Though promising in healthcare consultation applications, large language models (LLMs) face critical limitations in retaining and utilizing long-term memory across multi-turn interactions. In particular, existing memory enhancing paradigms are constrained by limited context windows and embedding-based retrieval, often failing to maintain task relevance and still suffering from memory prototype collapse in multi-turn healthcare consultation. To address these challenges, we propose a cognitively-inspired memory framework named MemoryART, which is grounded in Adaptive Resonance Theory (ART)—a cognitive and learning theory of how humans and animals adapt to dynamic environments. MemoryART employs three memory modules—working memory, episodic memory, and semantic memory to support task-aware memory organization and dynamic retrieval. Specifically, episodic memory provides the storage of specific experiences along with contextual clues, which is crucial for managing patient-specific information and perfect for multi-turn healthcare consultation interactions. Building upon this concept, MemoryART leverages multi-channel competitive learning and resonance matching to enable efficient and interpretable episodic memory encoding, alleviating issues of prototype collapse and noisy memory associations. For evaluation, we construct a long-term medical dialogue benchmark called MediLongChat using a LLM-based generation pipeline. The resulting dataset features realistic, multi-disease chat histories, each exceeding 100K tokens across 20–30 dialogues, simulating real-world healthcare interaction patterns. Our experimental results show that MemoryART outperforms mainstream approaches in memory-intensive tasks, achieving SOTA results and significantly reducing token consumption across five popular LLMs, confirming its effectiveness and efficiency in providing scalable, reliable memory for LLMs in healthcare.

This paper introduces a novel system for in-home cognitive health assessment using ambient sensors and a machine learning technology that can robustly detect mild cognitive impairment (MCI) despite limited available data. The learned model can explain the aspects of individuals' daily lives led to the prediction, while reliably predicting MCI, providing more insights to healthcare workers for further clinical interventions. We developed the robust transparent machine learning model, based on fusion adaptive resonance theory (Fusion ART) neural network to learn individuals' daily patterns of activity from continuous sensor data in terms of a suite of digital biomarkers reflecting four key domains: physical, daily activity, cognitive engagement, and sleep patterns. Based on a longitudinal study of over one hundred participants, deployed with non-intrusive sensors in their homes to undergo parallel clinical evaluation across a period of five years, our model successfully identified individuals with MCI, achieving high predictive accuracy regardless the noisy and sparse availability of data. As a transparent neural network, the learned model can also be interpreted as classification rules to distinguish MCI from normal cognition (NC) cases based on the digital biomarkers. These results demonstrate that passively collected, sensor-derived digital biomarkers can be leveraged to indicate cognitive status and potentially providing clinically meaningful insights on the impairment conditions. We also discuss the practical challenges and lessons learned from this real-world deployment to inform future large-scale implementations of such AI-driven health monitoring systems.

Current large language models (LLMs) exhibit significant deficiencies in episodic memory tasks including encoding, storing, and retrieving specific information from temporally dependent events over a long period of time. Recent approaches to handle memory tasks in LLMs, such as in-context learning, retrieval-augmented generation (RAG), and fine-tuning, may resolve the long-term retention issues, but are still inadequate to handle tasks requiring chronological awareness of the stored information. We introduce Agentic Retrieval with Temporal-Episodic Memory (ARTEM), a hybrid LLM-based agent architecture integrating LLMs with a self-organizing neural network named Spatial-Temporal Episodic Memory (STEM), designed to handle episodic memory tasks. Our approach employs LLMs for event extraction from the inputs to represent temporal, spatial, entitative, and semantic information that may facilitate future retrieval, aside from generating outputs or direct responses. The extracted events can then be encoded vectorially and stored in a fast and stable manner in the episodic memory through an instance-based incremental learning in STEM. STEM supports precise episodes retrieval and helps reduce computational overhead in generating the appropriate responses by LLMs. Evaluation on standardized episodic memory benchmarks across four tasks—partial cue retrieval, epistemic uncertainty detection, recent event identification, and chronological recall—demonstrates superior performance of ARTEM compared to in-context learning, RAG, and fine-tuning in various popular LLMs.

In this work, we address the cooperation problem among large language model (LLM) based embodied agents, where agents must cooperate to achieve a common goal. Previous methods often execute actions extemporaneously and incoherently, without long-term strategic and cooperative planning, leading to redundant steps, failures, and even serious repercussions in complex tasks like search-and-rescue missions where discussion and cooperative plan are crucial. To solve this issue, we propose Cooperative Plan Optimization (CaPo) to enhance the cooperation efficiency of LLM-based embodied agents. Inspired by human cooperation schemes, CaPo improves cooperation efficiency with two phases: 1) meta plan generation, and 2) progress-adaptive meta plan and execution. In the first phase, all agents analyze the task, discuss, and cooperatively create a meta-plan that decomposes the task into subtasks with detailed steps, ensuring a long-term strategic and coherent plan for efficient coordination. In the second phase, agents execute tasks according to the meta-plan and dynamically adjust it based on their latest progress (e.g., discovering a target object) through multi-turn discussions. This progress-based adaptation eliminates redundant actions, improving the overall cooperation efficiency of agents. Experimental results on the ThreeDworld Multi-Agent Transport and Communicative Watch-And-Help tasks demonstrate CaPo's much higher task completion rate and efficiency compared with state-of-the-arts. The code is released at https://github.com/jliu4ai/CaPo.

Multi-agent reinforcement learning (MARL) has demonstrated remarkable success in collaborative tasks, yet faces significant challenges in scaling to complex scenarios requiring sustained planning and coordination across long horizons. While hierarchical approaches help decompose these tasks, they typically rely on hand-crafted subtasks and domain-specific knowledge, limiting their generalizability. We present L2M2, a novel hierarchical framework that leverages large language models (LLMs) for high-level strategic planning and MARL for low-level execution. L2M2 enables zero-shot planning that supports both end-to-end training and direct integration with pre-trained MARL models. Experiments in the VMAS environment demonstrate that L2M2's LLM-guided MARL achieves superior performance while requiring less than 20% of the training samples compared to baseline methods. In the MOSMAC environment, L2M2 demonstrates strong performance with pre-defined subgoals and maintains substantial effectiveness without subgoals - scenarios where baseline methods consistently fail. Analysis through kernel density estimation reveals L2M2's ability to automatically generate appropriate navigation plans, demonstrating its potential for addressing complex multi-agent coordination tasks.

One main challenge in constructing a Knowledge Graph (KG) is to deal with ambiguity. Specifically, an entity in the graph can be assigned with multiple meanings while two or more entities considered to have different meanings may actually be the same. Assigning an entity with the correct meaning may involve re-evaluation of its relevant contexts. This costly operation typically involves searching for other similar entities within the KG such that the context can be determined. In this article, a new model called DisambiguART is proposed leveraging multi-channel matching and inference in a self-organizing neural network for sense disambiguation in KGs. Unlike other disambiguation methods that rely on representation learning to identify the relevant contexts whereby similarities among entities are learned, DisambiguART extends the working principle of multi-channel Adaptive Resonance Theory (ART) to conduct inferences directly over the graph representation through bi-directional interactions of bottom-up activations and top-down matching to find similar entities and select the correct meaning according to the right context. The proposed method is evaluated on the tasks of entity sense disambiguation in three domain KGs (jet engine, biomedical, and kinship) and author name disambiguation in bibliographic KGs, demonstrating the effectiveness and efficiency of DisambiguART against the state-of-the-art methods.

Though reinforcement learning (RL) has shown an outstanding capability for solving complex computational problems, most RL algorithms lack an explicit method that would allow learning from contextual information. On the other hand, humans often use context to identify patterns and relations among elements in the environment, along with how to avoid making wrong actions. However, what may seem like an obviously wrong decision from a human perspective could take hundreds of steps for an RL agent to learn to avoid. This article proposes a framework for discrete environments called Iota explicit context representation (IECR). The framework involves representing each state using contextual key frames (CKFs), which can then be used to extract a function that represents the affordances of the state; in addition, two loss functions are introduced with respect to the affordances of the state. The novelty of the IECR framework lies in its capacity to extract contextual information from the environment and learn from the CKFs' representation. We validate the framework by developing four new algorithms that learn using context: Iota deep Q-network (IDQN), Iota double deep Q-network (IDDQN), Iota dueling deep Q-network (IDuDQN), and Iota dueling double deep Q-network (IDDDQN). Furthermore, we evaluate the framework and the new algorithms in five discrete environments. We show that all the algorithms, which use contextual information, converge in around 40000 training steps of the neural networks, significantly outperforming their state-of-the-art equivalents.

Learning of Activities of Daily Living (ADLs) provides insights into an individual’s habits, lifestyle, and well-being. However, it is crucial to address data privacy concerns in practical situations when learning the ADL routines of individuals. In this paper, we introduce FedSTEM-ADL, a federated spatial-temporal episodic memory model to address this privacy issue. FedSTEM-ADL utilizes a federation of Spatial-Temporal Episodic Memory for ADLs (STEM-ADL) for federated learning, wherein multiple local STEM-ADL models from individual users are combined into a global model while preserving the privacy of the original data. Specifically, each local model is designed to learn the spatio-temporal ADL routines of an individual user, representing them as ADL events and sequences of such events as episode patterns. The global model then integrates the local models without referring to the underlying individual data, thus addressing privacy concerns in multi-user ADL analysis. We conduct a series of experiments based on both pseudo and real-world multi-user ADL datasets. The results show that FedSTEM-ADL is able to learn global ADL models in an efficient manner and consistently outperforms the baseline models in the task of next ADL event prediction.

Learning to reach long-horizon goals in spatial traversal tasks is a significant challenge for autonomous agents. Recent subgoal graph-based planning methods address this challenge by decomposing a goal into a sequence of shorter-horizon subgoals. These methods, however, use arbitrary heuristics for sampling or discovering subgoals, which may not conform to the cumulative reward distribution. Moreover, they are prone to learning erroneous connections (edges) between subgoals, especially those lying across obstacles. To address these issues, this article proposes a novel subgoal graph-based planning method called learning subgoal graph using value-based subgoal discovery and automatic pruning (LSGVP). The proposed method uses a subgoal discovery heuristic that is based on a cumulative reward (value) measure and yields sparse subgoals, including those lying on the higher cumulative reward paths. Moreover, LSGVP guides the agent to automatically prune the learned subgoal graph to remove the erroneous edges. The combination of these novel features helps the LSGVP agent to achieve higher cumulative positive rewards than other subgoal sampling or discovery heuristics, as well as higher goal-reaching success rates than other state-of-the-art subgoal graph-based planning methods.

Video summarization is the process of selecting a subset of informative keyframes to expedite storytelling with limited loss of information. In this article, we propose an EEG-Video Emotion-based Summarization (EVES) model based on a multimodal deep reinforcement learning (DRL) architecture that leverages neural signals to learn visual interestingness to produce quantitatively and qualitatively better video summaries. As such, EVES does not learn from the expensive human annotations but the multimodal signals. Furthermore, to ensure the temporal alignment and minimize the modality gap between the visual and EEG modalities, we introduce a Time Synchronization Module (TSM) that uses an attention mechanism to transform the EEG representations onto the visual representation space. We evaluate the performance of EVES on the TVSum and SumMe datasets. Based on the rank order statistics benchmarks, the experimental results show that EVES outperforms the unsupervised models and narrows the performance gap with supervised models. Furthermore, the human evaluation scores show that EVES receives a higher rating than the state-of-the-art DRL model DR-DSN by 11.4% on the coherency of the content and 7.4% on the emotion-evoking content. Thus, our work demonstrates the potential of EVES in selecting interesting content that is both coherent and emotion-evoking.

Spatial mapping and navigation are critical cognitive functions of autonomous agents, enabling one to learn an internal representation of an environment and move through space with real-time sensory inputs, such as visual observations. Existing models for vision-based mapping and navigation, however, suffer from memory requirements that increase linearly with exploration duration and indirect path following behaviors. This article presents e -TM, a self-organizing neural network-based framework for incremental topological mapping and navigation. e -TM models the exploration trajectories explicitly as episodic memory, wherein salient landmarks are sequentially extracted as "events" from streaming observations. A memory consolidation procedure then performs a playback mechanism and transfers the embedded knowledge of the environmental layout into spatial memory, encoding topological relations between landmarks. Fusion adaptive resonance theory (ART) networks, as the building block of the two memory modules, can generalize multiple input patterns into memory templates and, therefore, provide a compact spatial representation and support the discovery of novel shortcuts through inferences. For navigation, e -TM applies a transfer learning paradigm to integrate human demonstrations into a pretrained locomotion network for smoother movements. Experimental results based on VizDoom, a simulated 3-D environment, have shown that, compared to semiparametric topological memory (SPTM), a state-of-the-art model, e -TM reduces the time costs of navigation significantly while learning much sparser topological graphs.

Hierarchical reinforcement learning (HRL) is a promising approach to perform long-horizon goal-reaching tasks by decomposing the goals into subgoals. In a holistic HRL paradigm, an agent must autonomously discover such subgoals and also learn a hierarchy of policies that uses them to reach the goals. Recently introduced end-to-end HRL methods accomplish this by using the higher-level policy in the hierarchy to directly search the useful subgoals in a continuous subgoal space. However, learning such a policy may be challenging when the subgoal space is large. We propose integrated discovery of salient subgoals (LIDOSS), an end-to-end HRL method with an integrated subgoal discovery heuristic that reduces the search space of the higher-level policy, by explicitly focusing on the subgoals that have a greater probability of occurrence on various state-transition trajectories leading to the goal. We evaluate LIDOSS on a set of continuous control tasks in the MuJoCo domain against hierarchical actor critic (HAC), a state-of-the-art end-to-end HRL method. The results show that LIDOSS attains better goal achievement rates than HAC in most of the tasks.

Stock market trend forecasting is a valuable and challenging research task for both industry and academia. In order to explore the influence of stock news information on the stock market trend, a textual embedding construction method is proposed to encode multiple textual features, including topic features, sentiment features, and semantic features extracted from stock news textual content. In addition, a deep learning method is designed by using financial data and multiple textual features obtained from multiple news textual embeddings for short-term stock market trend prediction. For evaluation, extensive experiments on real stock market data are conducted. The experimental results illustrate that the proposed method can enhance the performance of predicting stock market trend by obtaining effective information from stock news.

Goal recognition for path planning is an important task of intention identification and situation awareness, requiring an observer to predict the goal of an evader given observations of its movements. While existing models based on planning or Markov Decision Process (MDP) show superior performance over traditional library based methods, they require much effort in model design and can hardly provide legible decision rules for their users. To make the system more user-friendly while preserving accuracy of goal inference, this paper proposes a novel self-organizing neural network based inference model, which learns compact rule sets through generalizing the streaming observations of an evader. More critically, the system manifests a high level of interpretability with the linguistic if-then rule base, making it easily comprehensible for human decision makers. We conducted extensive experiments on a large-scale real-world road network. Results show that the proposed model produces accuracy comparable to those of two state-of-the-art methods while uniquely providing legible inference rules and strong robustness against multiple goals with missing data.

Activities of daily living (ADLs) refer to the activities performed by individuals on a daily basis and are the indicators of a person's habits, lifestyle, and wellbeing. Consequently, learning an individual's ADL daily routines has significant value in the healthcare domain. Specifically, ADL recognition and inter-ADL pattern learning problems have been studied extensively in the past couple of decades. However, discovering the patterns of ADLs performed in a day and clustering them into ADL daily routines has been a relatively unexplored research area. In this paper, a self-organizing neural network model, called the Spatiotemporal ADL Adaptive Resonance Theory (STADLART), is proposed for learning ADL daily routines. STADLART integrates multimodal contextual information that involves the time and space wherein the ADLs are performed. By encoding spatiotemporal information explicitly as input features, STADLART enables the learning of time-sensitive knowledge. Moreover, a STADLART variation named STADLART-NC is proposed to normalize and customize ADL weighting for daily routine learning. A weighting assignment scheme is presented that facilitates the assignment of weighting according to ADL importance in specific domains. Empirical experiments using both synthetic and real-world public data sets validate the performance of STADLART and STADLART-NC when compared with alternative pattern discovery methods. The results show that STADLART could cluster ADL routines with better performance than baseline algorithms.

Most of today's AI technologies are geared towards mastering specific tasks performance through learning from a huge volume of data. However, less attention has still been given to make the AI understand its own purposes or be responsible socially. In this paper, a new model of agent is presented with the capacity to represent itself as a distinct individual with identity, a mind of its own, unique experiences, and social lives. In this way, the agent can interact with its surroundings and other agents seamlessly and meaningfully. A practical framework for developing an agent architecture with this model of self and self-awareness is proposed allowing self to be ascribed to an existing intelligent agent architecture in general to enable its social ability, interactivity, and co-presence with others. Possible applications are discussed with some exemplifying cases based on an implementation of a conversational agent.

Removing the undesired reflections from images taken through the glass is of broad application to various computer vision tasks. Non-learning based methods utilize different handcrafted priors such as the separable sparse gradients caused by different levels of blurs, which often fail due to their limited description capability to the properties of real-world reflections. In this paper, we propose a network with the feature-sharing strategy to tackle this problem in a cooperative and unified framework, by integrating image context information and the multi-scale gradient information. To remove the strong reflections existed in some local regions, we propose a statistic loss by considering the gradient level statistics between the background and reflections. Our network is trained on a new dataset with 3250 reflection images taken under diverse real-world scenes. Experiments on a public benchmark dataset show that the proposed method performs favorably against state-of-the-art methods.

Neurocomputational modelling of long-term memory is a core topic in computational cognitive neuroscience, which is essential towards self-regulating brain-like AI systems. In this paper, we study how people generally lose their memories and emulate various memory loss phenomena using a neurocomputational autobiographical memory model. Specifically, based on prior neurocognitive and neuropsychology studies, we identify three neural processes, namely overload, decay and inhibition, which lead to memory loss in memory formation, storage and retrieval, respectively. For model validation, we collect a memory dataset comprising more than one thousand life events and emulate the three key memory loss processes with model parameters learnt from memory recall behavioural patterns found in human subjects of different age groups. The emulation results show high correlation with human memory recall performance across their life span, even with another population not being used for learning. To the best of our knowledge, this paper is the first research work on quantitative evaluations of autobiographical memory loss using a neurocomputational model.

To simulate the concept acquisition and binding of different senses in the brain, a biologically inspired neural network model named perception coordination network (PCN) is proposed. It is a hierarchical structure, which is functionally divided into the primary sensory area (PSA), the primary sensory association area (SAA), and the higher order association area (HAA). The PSA contains feature neurons which respond to many elementary features, e.g., colors, shapes, syllables, and basic flavors. The SAA contains primary concept neurons which combine the elementary features in the PSA to represent unimodal concept of objects, e.g., the image of an apple, the Chinese word "[píng guǒ]" which names the apple, and the taste of the apple. The HAA contains associated neurons which connect the primary concept neurons of several PSA, e.g., connects the image, the taste, and the name of an apple. It means that the associated neurons have a multimodal response mode. Therefore, this area executes multisensory integration. PCN is an online incremental learning system, it is able to continuously acquire and bind multimodality concepts in an online way. The experimental results suggest that PCN is able to handle the multimodal concept acquisition and binding effectively.

A biologically plausible neural network model named Perception Coordination Network (PCN) is proposed for online multi-modal concept acquisition and binding. It is a hierarchical structure inspired by the structure of the brain, and functionally divided into the primary sensory area (PSA), the primary sensory association area (SAA), and the higher order association area (HAA). The PSA processes many elementary features, e.g., colors, shapes, syllables, and basic flavors, etc. The SAA combines these elementary features to represent the unimodal concept of an object, e.g., the image, name and taste of an apple, etc. The HAA connects several primary sensory association areas like a function of synaesthesia, which means associating the image, name and taste of an object. PCN is able to continuously acquire and bind multi-modal concepts in an online way. Experimental results suggest that PCN can handle the multi-modal concept acquisition and binding problem effectively.

Removing the undesired reflections from images taken through the glass is of broad application to various computer vision tasks. Non-learning based methods utilize different handcrafted priors such as the separable sparse gradients caused by different levels of blurs, which often fail due to their limited description capability to the properties of real-world reflections. In this paper, we propose the Concurrent Reflection Removal Network (CRRN) to tackle this problem in a unified framework. Our proposed network integrates image appearance information and multi-scale gradient information with human perception inspired loss function, and is trained on a new dataset with 3250 reflection images taken under diverse real-world scenes. Extensive experiments on a public benchmark dataset show that the proposed method performs favorably against state-of-the-art methods.

Segmenting video content into events provides semantic structures for indexing, retrieval, and summarization. Since motion cues are not available in continuous photo-streams, and annotations in lifelogging are scarce and costly, the frames are usually clustered into events by comparing the visual features between them in an unsupervised way. However, such methodologies are ineffective to deal with heterogeneous events, e.g. taking a walk, and temporary changes in the sight direction, e.g. at a meeting. To address these limitations, we propose Contextual Event Segmentation (CES), a novel segmentation paradigm that uses an LSTM-based generative network to model the photo-stream sequences, predict their visual context, and track their evolution. CES decides whether a frame is an event boundary by comparing the visual context generated from the frames in the past, to the visual context predicted from the future. We implemented CES on a new and massive lifelogging dataset consisting of more than 1.5 million images spanning over 1,723 days. Experiments on the popular EDUB-Seg dataset show that our model outperforms the state-of-the-art by over 16% in f-measure. Furthermore, CES' performance is only 3 points below that of human annotators.

Removing the undesired reflections in images taken through the glass is of broad application to various image processing and computer vision tasks. Existing single image based solutions heavily rely on scene priors such as separable sparse gradients caused by different levels of blur, and they are fragile when such priors are not observed. In this paper, we notice that strong reflections usually dominant a limited region in the whole image, and propose a Region-aware Reflection Removal (R3) approach by automatically detecting and heterogeneously processing regions with and without reflections. We integrate content and gradient priors to jointly achieve missing contents restoration as well as background and reflection separation in a unified optimization framework. Extensive validation using 50 sets of real data shows that the proposed method outperforms state-of-the-art on both quantitative metrics and visual qualities.

To facilitate the browsing of long videos, automatic video summarization provides an excerpt that represents its content. In the case of egocentric and consumer videos, due to their personal nature, adapting the summary to specific user's preferences is desirable. Current approaches to customizable video summarization obtain the user's preferences prior to the summarization process. As a result, the user needs to manually modify the summary to further meet the preferences. In this paper, we introduce Active Video Summarization (AVS), an interactive approach to gather the user's preferences while creating the summary. AVS asks questions about the summary to update it on-line until the user is satisfied. To minimize the interaction, the best segment to inquire next is inferred from the previous feedback. We evaluate AVS in the commonly used UTEgo dataset. We also introduce a new dataset for customized video summarization (CSumm) recorded with a Google Glass. The results show that AVS achieves an excellent compromise between usability and quality. In 41% of the videos, AVS is considered the best over all tested baselines, including summaries manually generated. Also, when looking for specific events in the video, AVS provides an average level of satisfaction higher than those of all other baselines after only six questions to the user.

Self-awareness is a crucial feature for a sociable agent or robot to better interact with humans. In a futuristic scenario, a conversational agent may occasionally be asked for its own opinion or suggestion based on its own thought, feelings, or experiences as if it is an individual with identity, personality, and social life. In moving towards that direction, in this paper, a brain inspired model of self-awareness is presented that allows an agent to learn to attend to different aspects of self as an individual with identity, physical embodiment, mental states, experiences, and reflections on how others may think about oneself. The model is built and realized on a NAO humanoid robotic platform to investigate the role of this capacity of self-awareness on the robot's learning and interactivity.

Removing undesired reflections from a photo taken in front of a glass is of great importance for enhancing the efficiency of visual computing systems. Various approaches have been proposed and shown to be visually plausible on small datasets collected by their authors. A quantitative comparison of existing approaches using the same dataset has never been conducted due to the lack of suitable benchmark data with ground truth. This paper presents the first captured Single-image Reflection Removal dataset ‘SIR2’ with 40 controlled and 100 wild scenes, ground truth of background and reflection. For each controlled scene, we further provide ten sets of images under varying aperture settings and glass thicknesses. We perform quantitative and visual quality comparisons for four state-of-the-art single-image reflection removal algorithms using four error metrics. Open problems for improving reflection removal algorithms are discussed at the end.

Reflection removal aims at separating the mixture of the desired background scenes and the undesired reflections, when the photos are taken through the glass. It has both aesthetic and practical applications which can largely improve the performance of many multimedia tasks. Existing reflection removal approaches heavily rely on scene priors such as separable sparse gradients brought by different levels of blur, and they easily fail when such priors are not observed in many real scenes. Sparse representation models and nonlocal image priors have shown their effectiveness in image restoration with self similarity. In this work, we propose a reflection removal method benefited from the sparsity and nonlocal image prior as a unified optimization framework. We leverage the retrieved image patch from an external database to overcome the limited prior information in the input mixture image and self similarity search. The experimental results show that our proposed model performs better than the existing state-of-the-art reflection removal method for both objective and subjective image qualities.

Episodic memory enables a cognitive system to improve its performance by reflecting upon past events. In this paper, we propose a computational model called STEM for encoding and recall of episodic events together with the associated contextual information in real time. Based on a class of self-organizing neural networks, STEM is designed to learn memory chunks or cognitive nodes, each encoding a set of co-occurring multi-modal activity patterns across multiple pattern channels. We present algorithms for recall of events based on partial and inexact input patterns. Our empirical results based on a public domain data set show that STEM displays a high level of efficiency and robustness in encoding and retrieval with both partial and noisy search cues when compared with a state-of-the-art associative memory model.

The introduction of wearable video cameras (e.g., GoPro) in the consumer market has promoted video life-logging, motivating users to generate large amounts of video data. This increasing flow of first-person video has led to a growing need for automatic video summarization adapted to the characteristics and applications of egocentric video. With this paper, we provide the first comprehensive survey of the techniques used specifically to summarize egocentric videos. We present a framework for first-person view summarization and compare the segmentation methods and selection algorithms used by the related work in the literature. Next, we describe the existing egocentric video datasets suitable for summarization and, then, the various evaluation methods. Finally, we analyze the challenges and opportunities in the field and propose new lines of research.

Semantic memory plays a critical role in reasoning and decision making. It enables an agent to abstract useful knowledge learned from its past experience. Based on an extension of fusion adaptive resonance theory network, this paper presents a novel self-organizing memory model to represent and learn various types of semantic knowledge in a unified manner. The proposed model, called fusion adaptive resonance theory for multimemory learning, incorporates a set of neural processes, through which it may transfer knowledge and cooperate with other long-term memory systems, including episodic memory and procedural memory. Specifically, we present a generic learning process, under which various types of semantic knowledge can be consolidated and transferred from the specific experience encoded in episodic memory. We also identify and formalize two forms of memory interactions between semantic memory and procedural memory, through which more effective decision making can be achieved. We present experimental studies, wherein the proposed model is used to encode various types of semantic knowledge in different domains, including a first-person shooting game called Unreal Tournament, the Toads and Frogs puzzle, and a strategic game known as StarCraft Broodwar. Our experiments show that the proposed knowledge transfer process from episodic memory to semantic memory is able to extract useful knowledge to enhance the performance of decision making. In addition, cooperative interaction between semantic knowledge and procedural skills can lead to a significant improvement in both learning efficiency and performance of the learning agents.

Humans are not rational beings. Deviations from rationality in human thinking are currently well documented [25] as non-reducible to rational pursuit of egoistic benefit or its occasional distortion with temporary emotional excitation, as it is often assumed. This occurs not only outside conceptual reasoning or rational goal realization but also subconsciously and often in certainty that they did not and could not take place 'in my case'. Non-rationality can no longer be perceived as a rare affective abnormality in otherwise rational thinking, but as a systemic, permanent quality, 'a design feature' of human cognition. While social psychology has systematically addressed non-rationality of human cognition (including its non-emotional aspects) for decades [63]. It is not the case for computer science, despite obvious relevance for individual and group behavior modeling. This paper proposes brief survey of work in computational disciplines related to human-like non-rationality modeling including: Social Signal Processing, Cognitive Architectures, Affective Computing, Human-Like Agents and Normative Multi-agent Systems. It attempts to establish a common terminology and conceptual frame for this extremely interdisciplinary issue, reveal assumptions about non-rationality underlying the discussed models and disciplines, their current limitations and potential in contributing to solution. Finally, it also presents ideas concerning possible directions of development, hopefully contributing to solution of this challenging issue.

Reflection removal aims at separating the mixture of the desired scene and the undesired reflections. Locating reflection and background edges is a key step for reflection removal. In this paper, we present a visual depth guided method to remove reflections. Our idea is to use Depth of Field (DoF) to label the background and reflection edges. We propose a DoF confidence map where pixels with higher DoF values are assumed to belong to the desired background components. Moreover, we observe that images with different resolutions show different properties in the DoF map. Thus, we introduce a multi-scale DoF computing strategy to classify edge pixels more efficiently. Based on the results of edge classification, the background and reflection layers can be separated. Experimental results validate the effectiveness of our method using real-world photos.

Due to their online learning nature, incremental clustering techniques can handle a continuous stream of data. In particular, various incremental clustering techniques based on Adaptive Resonance Theory (ART) have been shown to have low computational complexity in adaptive learning and are less sensitive to noisy information. However, parameter regularization in existing ART clustering techniques is applied either on different features or on different clusters exclusively. In this paper, we introduce Interest-Focused Clustering based on Adaptive Resonance Theory (IFC-ART), which self-regulates the vigilance parameter associated with each feature and each cluster. As such, we can incorporate the domain knowledge of the data set into IFC-ART to focus on certain preferences during the self-regulated clustering process. For performance evaluation, we use a real-world data set, named American Time Use Survey (ATUS), which records nearly 160,000 telephone interviews conducted with U.S. residents from 2003 to 2014. Specifically, we conduct case studies to explore three types of interesting relationship, focusing on the wage, age, and provision of elderly care, respectively. Experimental results show that the performance of IFC-ART is highly competitive and stable when compared with two well-established clustering techniques and three ART models. In addition, we highlight the important and unexpected findings observed from the clusters discovered.

Abstract This paper presents a semi‐automatic method of parameterizing an existing social context cognition model. It discusses benefits of the social context cognition models for example in personality modeling and their key issue that is parametrization. It briefly introduces social context cognition model and describes a new method of its crowd‐sourcing‐based parametrization. Later, validation is provided, and ability to recreate social context cognition in the provided samples is presented with good generalization for the unknown cases. Finally, model's stability for the continuous stream of dynamic social context input data is shown. Presented system contributes to the believable agent modeling and social simulations by making much needed applications of social context cognition models easier by addressing the so far unsolved troublesome parametrization issues. Copyright © 2016 John Wiley & Sons, Ltd.

The large scale and complex nature of social media data raises the need to scale clustering techniques to big data and make them capable of automatically identifying data clusters with few empirical settings. In this paper, we present our investigation and three algorithms based on the fuzzy adaptive resonance theory (Fuzzy ART) that have linear computational complexity, use a single parameter, i.e., the vigilance parameter to identify data clusters, and are robust to modest parameter settings. The contribution of this paper lies in two aspects. First, we theoretically demonstrate how complement coding, commonly known as a normalization method, changes the clustering mechanism of Fuzzy ART, and discover the vigilance region (VR) that essentially determines how a cluster in the Fuzzy ART system recognizes similar patterns in the feature space. The VR gives an intrinsic interpretation of the clustering mechanism and limitations of Fuzzy ART. Second, we introduce the idea of allowing different clusters in the Fuzzy ART system to have different vigilance levels in order to meet the diverse nature of the pattern distribution of social media data. To this end, we propose three vigilance adaptation methods, namely, the activation maximization (AM) rule, the confliction minimization (CM) rule, and the hybrid integration (HI) rule. With an initial vigilance value, the resulting clustering algorithms, namely, the AM-ART, CM-ART, and HI-ART, can automatically adapt the vigilance values of all clusters during the learning epochs in order to produce better cluster boundaries. Experiments on four social media data sets show that AM-ART, CM-ART, and HI-ART are more robust than Fuzzy ART to the initial vigilance value, and they usually achieve better or comparable performance and much faster speed than the state-of-the-art clustering algorithms that also do not require a predefined number of clusters.

Developing a complex intelligent system by abstracting their behaviors, functionalities, and reasoning mechanisms can be tedious and time consuming. In this paper, we present a framework for developing an application or software system based on smart autonomous components that collaborate with the developer or user to realize the entire system. Inspired by teachable approaches and programming-by-demonstration methods in robotics and end-user development, we treat intelligent agents as teachable components that make up the system to be built. Each agent serves different functionalities and may have prebuilt operations to accomplish its own design objectives. However, each agent may also be equipped with in-built social-cognitive traits to interact with the user or other agents in order to adapt its own operations, objectives, and relationships with others. The results of adaptation can be in the form of groups or multiagent systems as new aggregated components. This approach is made to tackle the difficulties in completely programming the entire system by allowing the user to teach the components toward the desired behaviors in the situated context of the application. We exemplify this novel method with cases in the domains of human-like agents in virtual environment and agents for in-house caregiving.

This paper analyzes advanced reinforcement learning techniques and compares some of them to motivated learning. Motivated learning is briefly discussed indicating its relation to reinforcement learning. A black box scenario for comparative analysis of learning efficiency in autonomous agents is developed and described. This is used to analyze selected algorithms. Reported results demonstrate that in the selected category of problems, motivated learning outperformed all reinforcement learning algorithms we compared with.

Weak supervisory information of web images, such as captions, tags, and descriptions, make it possible to better understand images at the semantic level. In this paper, we propose a novel online multimodal co-indexing algorithm based on Adaptive Resonance Theory, named OMC-ART, for the automatic co-indexing and retrieval of images using their multimodal information. Compared with existing studies, OMC-ART has several distinct characteristics. First, OMC-ART is able to perform online learning of sequential data. Second, OMC-ART builds a two-layer indexing structure, in which the first layer co-indexes the images by the key visual and textual features based on the generalized distributions of clusters they belong to; while in the second layer, images are co-indexed by their own feature distributions. Third, OMC-ART enables flexible multimodal search by using either visual features, keywords, or a combination of both. Fourth, OMC-ART employs a ranking algorithm that does not need to go through the whole indexing system when only a limited number of images need to be retrieved. Experiments on two published data sets demonstrate the efficiency and effectiveness of our proposed approach.

Abstract Expanding the spectrum of agent social capabilities is an important challenge in agent‐based simulation and other domains. While human‐like emotionality has been vastly explored in the last 20years, little research addresses explicit, psychologically believable social situation modeling. Recently, some important elements have been underlined: hybrid connectionist models outside formal ontologies; complex subjective representations linking culture, personality and norms, and so on, but proposed solutions do not provide a formalized structure of a social experience, expressive and well‐grounded in psychology. In this paper, we develop a new approach to social situation modeling based on the dramaturgical and dissonance theories. A new component (Dramaturgical Module) is described with implementation used to generate example behavior depicting new social modeling capabilities and a believable representation of the relevant psychological theories. We present a case scenario with a dramaturgical interpretation of dynamic social situations and related cognitive dissonances resulting in a simple and flexible classification. Easily usable in reasoning, planning or affect generation, dramaturgical interpretation is additionally presented here as basis of social affect generation. Copyright © 2015 John Wiley & Sons, Ltd.

Games are good test-beds to evaluate AI methodologies. In recent years, there has been a vast amount of research dealing with real-time computer games other than the traditional board games or card games. This paper illustrates how we create agents by employing FALCON, a self-organizing neural network that performs reinforcement learning, to play a well-known first-person shooter computer game called Unreal Tournament. Rewards used for learning are either obtained from the game environment or estimated using the temporal difference learning scheme. In this way, the agents are able to acquire proper strategies and discover the effectiveness of different weapons without any guidance or intervention. The experimental results show that our agents learn effectively and appropriately from scratch while playing the game in real-time. Moreover, with the previously learned knowledge retained, our agent is able to adapt to a different opponent in a different map within a relatively short period of time.

The use of domain knowledge in learning systems is expected to improve learning efficiency and reduce model complexity. However, due to the incompatibility with knowledge structure of the learning systems and real-time exploratory nature of reinforcement learning (RL), domain knowledge cannot be inserted directly. In this paper, we show how self-organizing neural networks designed for online and incremental adaptation can integrate domain knowledge and RL. Specifically, symbol-based domain knowledge is translated into numeric patterns before inserting into the self-organizing neural networks. To ensure effective use of domain knowledge, we present an analysis of how the inserted knowledge is used by the self-organizing neural networks during RL. To this end, we propose a vigilance adaptation and greedy exploitation strategy to maximize exploitation of the inserted domain knowledge while retaining the plasticity of learning and using new knowledge. Our experimental results based on the pursuit-evasion and minefield navigation problem domains show that such self-organizing neural network can make effective use of domain knowledge to improve learning efficiency and reduce model complexity.

In this paper, we propose a multi-memory model, ADLART model, to discover the daily activity pattern of a sensor monitored user from his/her activities of daily living (ADL). The proposed model mimics the human multiple memory system which comprises a working memory, an episodic memory, and a semantic memory. Through encoding user's daily activities patterns in episodic memory and extracting the regularities of activity routines in semantic memory, the ADLART system is able to learn, recognize, compare, and retrieve daily ADL patterns of the user. Experiments are presented to show the performance of the ADLART model using different parameter settings and its performance is discussed in details.

Most non-trivial problems require the coordinated performance of multiple goal-oriented and time-critical tasks. Coordinating the performance of the tasks is required due to the dependencies among the tasks and the sharing of resources. In this work, an agent learns to perform a task using reinforcement learning with a self-organizing neural network as the function approximator. We propose a novel coordination strategy integrating Motivated Learning (ML) and a self-organizing neural network for multi-agent reinforcement learning (MARL). Specifically, we adapt the ML idea of using pain signal to overcome the resource competition issue. Dependency among the agents is resolved using domain knowledge of their dependence. To avoid domineering agents, the task goals are staggered over multiple stages. A stage is completed by attaining a particular combination of task goals. Results from our experiments conducted using a popular PC-based game known as Starcraft Broodwar show goals of multiple tasks can be attained efficiently using our proposed coordination strategy.

Human, especially elderly, require frequent attention, continuous companionship, and deep understanding from the others. To provide more specific and appropriate tender care to the elderly, knowing their affective states is a great advantage. Recent work on human emotion recognition shows promising results that the expressive emotion can be successfully captured through visual, audio, and keyboard or touchpad stroke pattern signals. Furthermore, human activities are shown to be accurately recognizable with context by non-intrusive sensors within or connected to the smartphones. In this paper, we propose a computational model to characterize the affective states of the elderly based on the recognizable daily activities. Therefore, by integrating such an understanding module into a humanoid agent residing in the smartphone platform, we make the mobile agent more human-like. The initial knowledge of the activity-affect associations is taken from published work in psychology and gerontology. Based on the provided training signals, our model adapts the activity-affect knowledge accordingly. Consequently, by modeling mood awareness of the elderly, our agent can carry out more specific task and provide more appropriate tender care.

Discovering social communities of web users through clustering analysis of heterogeneous link associations has drawn much attention. However, existing approaches typically require the number of clusters a prior, do not address the weighting problem for fusing heterogeneous types of links and have a heavy computational cost. In this paper, we explore the feasibility of a newly proposed heterogeneous data clustering algorithm, called Generalized Heterogeneous Fusion Adaptive Resonance Theory (GHF-ART), for discovering user communities in social networks. Different from existing algorithms, GHF-ART performs real-time matching of patterns and one-pass learning which guarantee its low computational cost. With a vigilance parameter to restrain the intra-cluster similarity, GHF-ART does not need the number of clusters a prior. To achieve a better fusion of multiple types of links, GHF-ART employs a weighting function to incrementally assess the importance of all the feature channels. Extensive experiments have been conducted to analyze the performance of GHF-ART on two heterogeneous social network data sets. The promising results comparing with existing methods demonstrate the effectiveness and efficiency of GHF-ART.

Co-clustering is a commonly used technique for tapping the rich meta-information of multimedia web documents, including category, annotation, and description, for associative discovery. However, most co-clustering methods proposed for heterogeneous data do not consider the representation problem of short and noisy text and their performance is limited by the empirical weighting of the multi-modal features. In this paper, we propose a generalized form of Heterogeneous Fusion Adaptive Resonance Theory, called GHF-ART, for co-clustering of large-scale web multimedia documents. By extending the two-channel Heterogeneous Fusion ART (HF-ART) to multiple channels, GHF-ART is designed to handle multimedia data with an arbitrarily rich level of meta-information. For handling short and noisy text, GHF-ART does not learn directly from the textual features. Instead, it identifies key tags by learning the probabilistic distribution of tag occurrences. More importantly, GHF-ART incorporates an adaptive method for effective fusion of multi-modal features, which weights the features of multiple data sources by incrementally measuring the importance of feature modalities through the intra-cluster scatters. Extensive experiments on two web image data sets and one text document set have shown that GHF-ART achieves significantly better clustering performance and is much faster than many existing state-of-the-art algorithms.

Though not a fundamental pre-requisite to efficient machine learning, insertion of domain knowledge into adaptive virtual agent is nonetheless known to improve learning efficiency and reduce model complexity. Conventionally, domain knowledge is inserted prior to learning. Despite being effective, such approach may not always be feasible. Firstly, the effect of domain knowledge is assumed and can be inaccurate. Also, domain knowledge may not be available prior to learning. In addition, the insertion of domain knowledge can frame learning and hamper the discovery of more effective knowledge. Therefore, this work advances the use of domain knowledge by proposing to delay the insertion and moderate the effect of domain knowledge to reduce the framing effect while still benefiting from the use of domain knowledge. Using a non-trivial pursuit-evasion problem domain, experiments are first conducted to illustrate the impact of domain knowledge with different degrees of truth. The next set of experiments illustrates how delayed insertion of such domain knowledge can impact learning. The final set of experiments is conducted to illustrate how delaying the insertion and moderating the assumed effect of domain knowledge can ensure the robustness and versatility of reinforcement learning.

Despite the advantages of fast and stable learning, Adaptive Resonance Theory (ART) still relies on an empirically fixed vigilance parameter value to determine the vigilance regions of all of the clusters in the category field (F2), causing its performance to depend on the vigilance value. It would be desirable to use different values of vigilance for different category field nodes, in order to fit the data with a smaller number of categories. We therefore introduce two methods, the Activation Maximization Rule (AMR) and the Confliction Minimization Rule (CMR). Despite their differences, both ART with AMR (AM-ART) and with CMR (CM-ART) allow different vigilance levels for different clusters, which are incrementally adapted during the clustering process. Specifically, AMR works by increasing the vigilance value of the winner cluster when a resonance occurs and decreasing it when a reset occurs, which aims to maximize the participation of clusters for activation. On the other hand, after receiving an input pattern, CMR first identifies all of the winner candidates that satisfy the vigilance criteria and then tunes their vigilance values to minimize conflicts in the vigilance regions. In this paper, we chose Fuzzy ART to demonstrate these concepts, but they will clearly carry over to other ART architectures. Our comparative experiments show that both AM-ART and CM-ART improve the robust performance of Fuzzy ART to the vigilance parameter and usually produce better cluster quality.

A major drawback of evolutionary optimization approaches in the literature is the apparent lack of automated knowledge transfers and reuse across problems. Particularly, evolutionary optimization methods generally start a search from scratch or ground zero state, independent of how similar the given new problem of interest is to those optimized previously. In this paper, we present a study on the transfer of knowledge in the form of useful structured knowledge or latent patterns that are captured from previous experiences of problem-solving to enhance future evolutionary search. The essential contributions of our present study include the meme learning and meme selection processes. In contrast to existing methods, which directly store and reuse specific problem solutions or problem sub-components, the proposed approach models the structured knowledge of the strategy behind solving problems belonging to similar domain, i.e., via learning the mapping from problem to its corresponding solution, which is encoded in the form of identified knowledge representation. In this manner, knowledge transfer can be conducted across problems, from differing problem size, structure to representation, etc. A demonstrating case study on the capacitated arc routing problem (CARP) is presented. Experiments on benchmark instances of CARP verified the effectiveness of the proposed new paradigm.

Existing efforts on web image organization usually transform the task into surrounding text clustering. However, Current text clustering algorithms do not address the problem of insufficient statistical information for image representation and noisy tags which greatly decreases the clustering performance while increases the computational cost. In this paper, we propose a two-step semi-supervised hierarchical clustering algorithm, Personalized Hierarchical Theme-based Clustering (PHTC), for web image organization. In the first step, the Probabilistic Fusion ART (PF-ART) is proposed for grouping semantically similar images and simultaneously learning the probabilistic distribution of tag occurrence for mining the key tags/topics of clusters. In this way, the side-effect of noisy tags can be largely eliminated. Moreover, PF-ART can incorporate user preference for semi-supervised learning and provide users a direct control of clustering results. In the second step, a novel agglomerative merging strategy based on Cluster Semantic Relevance, proposed for measuring the semantic similarity between clusters, is employed for associating the clusters by generating a semantic hierarchy. Different from existing hierarchical clustering algorithms, the proposed merging strategy can provide a multi-branch tree structure which is more systematic and clearer than traditional binary tree structure. Extensive experiments on two real world web image data sets, namely NUS-WIDE and Flickr, demonstrate the effectiveness of our algorithm for large web image data sets.

Although various emotion models have been proposed based on appraisal theories, most of them focus on designing specific appraisal rules and there is no unified framework for emotional appraisal. Moreover, few existing emotion models are biologically-inspired and are inadequate in imitating emotion process of human brain. This paper proposes a bio-inspired computational model called Cognitive Regulated Affective Architecture (CRAA), inspired by the cognitive regulated emotion theory and the network theory of emotion. This architecture is proposed by taking the following positions: (1) Cognition and emotion are not separated but interacted systems; (2) The appraisal of emotion depends on and should be regulated through cognitive system; and (3) Emotion is generated though numerous neural computations and networks of brain regions. This model contributes to an integrated system which combines emotional appraisal with the cognitive decision making in a multi-layered structure. Specifically, a self-organizing neural model called Emotional Appraisal Network (EAN) is proposed based on the Adaptive Resonance Theory (ART), to learn the associations from appraisal components involving expectation, reward, power, and match to emotion. An appraisal module is positioned within EAN contributing to translate cognitive information to emotion appraisal. The above model has been evaluated in a first person shooting game known as Unreal Tournament. Comparing with non-emotional NPC, emotional NPC obtains a higher evaluation in improving game playability and interest. Moreover, comparing with existing emotion models, our CRAA model obtains a higher accuracy in determining emotion expressions. .

This paper reports on an agent-oriented approach for the modeling of adaptive doctrine-equipped computer generated force (CGF) using a commercial-grade simulation platform known as CAE STRIVE®CGF. A self-organizing neural network is used for the adaptive CGF to learn and generalize knowledge in an online manner during the simulation. The challenge of defining the state space and action space and the lack of domain knowledge to initialize the adaptive CGF are addressed using the doctrine used to drive the non-adaptive CGF. The doctrine contains a set of specialized knowledge for conducting 1-v-1 dogfights. The hierarchical structure and symbol representation of the propositional rules are incompatible to the self-organizing neural network. Therefore, it has to be flattened and then translated to vector pattern before it can inserted into the self-organizing neural network. The state space and action space are automatically extracted using the flattened doctrine as well. Experiments are conducted using several initial conditions in round robin fashions. The experimental results show that the selforganizing neural network is able to make good use of the domain knowledge with complex knowledge structure to discover the knowledge to out-maneuver the doctrine-driven CGF consistently in an efficient manner.

This paper presents a self-organizing approach to the learning of procedural and declarative knowledge in parallel using independent but interconnected memory models. The proposed system, employing fusion Adaptive Resonance Theory (fusion ART) network as a building block, consists of a declarative memory module, that learns both episodic traces and semantic knowledge in real time, as well as a procedural memory module that learns reactive responses to its environment through reinforcement learning. More importantly, the proposed multi-memory system demonstrates how the various memory modules transfer knowledge and cooperate with each other for a higher overall performance. We present experimental studies, wherein the proposed system is tasked to learn the procedural and declarative knowledge for an autonomous agent playing in a first person game environment called Unreal Tournament. Our experimental results show that the multi-memory system is able to enhance the performance of the agent in a real time environment by utilizing both its procedural and declarative knowledge.

This paper presents a neural model that learns episodic traces in response to a continuous stream of sensory input and feedback received from the environment. The proposed model, based on fusion adaptive resonance theory (ART) network, extracts key events and encodes spatio-temporal relations between events by creating cognitive nodes dynamically. The model further incorporates a novel memory search procedure, which performs a continuous parallel search of stored episodic traces. Combined with a mechanism of gradual forgetting, the model is able to achieve a high level of memory performance and robustness, while controlling memory consumption over time. We present experimental studies, where the proposed episodic memory model is evaluated based on the memory consumption for encoding events and episodes as well as recall accuracy using partial and erroneous cues. Our experimental results show that: 1) the model produces highly robust performance in encoding and recalling events and episodes even with incomplete and noisy cues; 2) the model provides enhanced performance in a noisy environment due to the process of forgetting; and 3) compared with prior models of spatio-temporal memory, our model shows a higher tolerance toward noise and errors in the retrieval cues.

Memetics is a new science that has attracted in creasing attentions in the recent decades. Beyond the formalism of simple hybrids, adaptive hybrids and memetic algorithms, the notion of memetic automaton as an adaptive entity that is self-contained and uses memes as building blocks of information is recently conceptualized in the context of computational intelligence as potential tools for effective problem-solving. Taking this cue, this paper embarks a study on Memetic Multi agent system (MeM) towards human-like social agents with memetic automaton. Particularly, we introduce a potentially rich meme-inspired design and operational model, with Darwin's theory of natural selections and Dawkins' notion of a meme as the principal driving forces behind interactions among agents, whereby memes formed the fundamental building blocks of the agents' mind universe. Experimental studies on a Mine Navigation Task indicates the modeling of memetic agents that resemble the natural way of human interaction can lead to greater level of adaptivity and effective problem-solving.

The presence of noise or errors in the stated feature values of biomedical data can lead to incorrect prediction. We introduce a Bayesian Network-based Noise Correction framework named BN-NC. After data preprocessing, a Bayesian Network (BN) is learned to capture the feature dependencies. Using the BN to predict each feature in turn, BN-NC estimates a feature's error rate as the deviation between its predicted and stated values in the training data, and allocates the appropriate uncertainty to its subsequent findings during prediction. BN-NC automatically generates a probabilistic rule to explain BN prediction on the class variable using the feature values in its Markov blanket, and this is reapplied as necessary to explain the noise correction on those features. Using three real-life benchmark biomedical data sets (on HIV-1 drug resistance prediction and leukemia subtype classification), we demonstrate that BN-NC (1) accurately detects the errors in biomedical feature values, (2) automatically corrects for the errors to maintain higher prediction accuracy over competing methods including Decision Trees, Naive Bayes and Support Vector Machines, and (3) generates probabilistic rules that concisely explain the prediction and noise correction decisions. In addition to achieving more robust biomedical prediction in the presence of feature noise, by highlighting erroneous features and explaining their corrections, BN-NC provides medical researchers with high utility insights to biomedical data not found in other methods.

Capacitated arc routing problem (CARP) has attracted much attention due to its generality to many real world problems. Memetic algorithm (MA), among other meta-heuristic search methods, has been shown to achieve competitive performances in solving CARP ranging from small to medium size. In this paper we propose a formal probabilistic memetic algorithm for CARP that is equipped with an adaptation mechanism to control the degree of global exploration against local exploitation while the search progresses. Experimental study on benchmark instances of CARP showed that the proposed probabilistic scheme led to improved search performances when introduced into a recently proposed state-of-the-art MA. The results obtained on 24 instances of the capacitated arc routing problems highlighted the efficacy of the probabilistic scheme with 9 new best known solutions established.

We have designed and implemented a 3D virtual world based on the Co-Space concept encompasses the Youth Olympic Village (YOV) and several sports competition venues. It is a massively multiplayer online (MMO) virtual world built according to the actual, physical locations of the YOV and sports competition venues. On top of that, the Co-Space is being populated with human-like avatars, which are created according to the actual human size and appearance, they perform their activities and interact with the users in real-world context. In addition, autonomous intelligent agents are integrated into the Co-Space to provide context-aware and personalized services to the users. These agents, which are in the form of human-like avatars, act as the users' personal tour guides, they interact with the users and gradually learn the users' preferences. Then, the agents are able to suggest suitable locations for the users to visit. The intention of deploying these agents is to enrich the Co-Space with a variety of interactions tailored to each user's preferences instead of the same contents for everyone. Besides, our Co-Space also provides an avenue for people from different locations throughout the world to explore the YOV and experience the atmosphere of the sports events in Singapore.

Because of the inherent ambiguity in user queries, an important task of modern retrieval systems is faceted topic retrieval (FTR), which relates to the goal of returning diverse or novel information elucidating the wide range of topics or facets of the query need. We introduce a generative model for hypothesizing facets in the (news) video domain by combining the complementary information in the visual keyframes and the speech transcripts. We evaluate the efficacy of our multimodal model on the standard TRECVID-2005 video corpus annotated with facets. We find that: (1) the joint modeling of the visual and text (speech transcripts) information can achieve significant F-score improvement over a text-alone system; (2) our model compares favorably with standard diverse ranking algorithms such as the MMR. Our FTR model has been implemented on a news search prototype that is undergoing commercial trial.

This paper proposes self-organizing neural networks for modeling behavior of non-player characters (NPC) in first person shooting games. Specifically, two classes of self-organizing neural models, namely Self-Generating Neural Networks (SGNN) and Fusion Architecture for Learning and Cognition (FALCON) are used to learn non-player characters' behavior rules according to recorded patterns. Behavior learning abilities of these two models are investigated by learning specific sample Bots in the Unreal Tournament game in a supervised manner. Our empirical experiments demonstrate that both SGNN and FALCON are able to recognize important behavior patterns and learn the necessary knowledge to operate in the Unreal environment. Comparing with SGNN, FALCON is more effective in behavior learning, in terms of lower complexity and higher fighting competency.

We present a self-organizing neural model for creating intelligent learning agents in virtual worlds. As agents in a virtual world roam, interact and socialize with users and other agents as in real world without explicit goals and teachers, learning in virtual world presents many challenges not found in typical machine learning benchmarks. In this paper, we highlight the unique issues and challenges of building learning agents in virtual world using reinforcement learning. Specifically, a self-organizing neural model, named TD-FALCON (Temporal Difference - Fusion Architecture for Learning and Cognition), is deployed, which enables an autonomous agent to adapt and function in a dynamic environment with immediate as well as delayed evaluative feedback signals. We have implemented and evaluated TD-FALCON agents as virtual tour guides in a virtual world environment. Our experimental results show that the agents are able to adapt and improve their performance in real time. To the best of our knowledge, this is one of the few in-depth works on building complete learning agents that adapt their behaviors through real time reinforcement learning in virtual world.

Motivated learning is a new machine learning approach that extends reinforcement learning idea to dynamically changing, and highly structured environments. In this approach a machine is capable of defining its own objectives and learns to satisfy them though an internal reward system. The machine is forced to explore the environment in response to externally applied negative (pain) signals that it must minimize. In doing so, it discovers relationships between objects observed through its sensory inputs and actions it performs on the observed objects. Observed concepts are not predefined but are emerging as a result of successful operations. For the optimum development of concepts and related skills, the machine operates in the protective environment that gradually increases its complexity. Simulation illustrates the advantage of this gradual increase in environment complexity for machine development. Comparison to reinforcement learning indicates weakness of the later method in learning proper behavior, even in such protective environments with gradually increasing complexity. The method shows how mental development stimulates learning of new concepts and at the same time benefits from this learning. Thus the method addresses a well know problem of merging connectionist (bottom-up) and symbolic (top down) approaches for intelligent autonomous machine operation in developmental robotics.

This paper presents a neural model that learns episodic traces in response to a continual stream of sensory input and feedback received from the environment. The proposed model, based on fusion Adaptive Resonance Theory (fusion ART) network, extracts key events and encodes spatio-temporal relations between events by creating cognitive nodes dynamically. The model further incorporates a novel memory search procedure, which performs parallel search of stored episodic traces continuously. Comparing with prior systems, the proposed episodic memory model presents a robust approach to encoding key events and episodes and recalling them using partial and erroneous cues. We present experimental studies, wherein the model is used to learn episodic memory of an agent's experience in a first person game environment called Unreal Tournament. Our experimental results show that the model produces highly robust performance in encoding and recalling events and episodes even with incomplete and noisy cues.

Abstract Domain ontologies play an important role in supporting knowledge‐based applications in the Semantic Web. To facilitate the building of ontologies, text mining techniques have been used to perform ontology learning from texts. However, traditional systems employ shallow natural language processing techniques and focus only on concept and taxonomic relation extraction. In this paper we present a system, known as Concept‐Relation‐Concept Tuple‐based Ontology Learning (CRCTOL), for mining ontologies automatically from domain‐specific documents. Specifically, CRCTOL adopts a full text parsing technique and employs a combination of statistical and lexico‐syntactic methods, including a statistical algorithm that extracts key concepts from a document collection, a word sense disambiguation algorithm that disambiguates words in the key concepts, a rule‐based algorithm that extracts relations between the key concepts, and a modified generalized association rule mining algorithm that prunes unimportant relations for ontology learning. As a result, the ontologies learned by CRCTOL are more concise and contain a richer semantics in terms of the range and number of semantic relations compared with alternative systems. We present two case studies where CRCTOL is used to build a terrorism domain ontology and a sport event domain ontology. At the component level, quantitative evaluation by comparing with Text‐To‐Onto and its successor Text2Onto has shown that CRCTOL is able to extract concepts and semantic relations with a significantly higher level of accuracy. At the ontology level, the quality of the learned ontologies is evaluated by either employing a set of quantitative and qualitative methods including analyzing the graph structural property, comparison to WordNet, and expert rating, or directly comparing with a human‐edited benchmark ontology, demonstrating the high quality of the ontologies learned.

The problem of polysemy in keyword-based image search arises mainly from the inherent ambiguity in user queries. We propose a latent model based approach that resolves user search ambiguity by allowing sense specific diversity in search results. Given a query keyword and the images retrieved by issuing the query to an image search engine, we first learn a latent visual sense model of these polysemous images. Next, we use Wikipedia to disambiguate the word sense of the original query, and issue these Wiki-senses as new queries to retrieve sense specific images. A sense-specific image classifier is then learnt by combining information from the latent visual sense model, and used to cluster and re-rank the polysemous images from the original query keyword into its specific senses. Results on a ground truth of 17K image set returned by 10 keyword searches and their 62 word senses provides empirical indications that our method can improve upon existing keyword based search engines. Our method learns the visual word sense models in a totally unsupervised manner, effectively filters out irrelevant images, and is able to mine the long tail of image search.

In content-based image retrieval, the ldquosemantic gaprdquo between visual image features and user semantics makes it hard to predict abstract image categories from low-level features. We present a hybrid system that integrates global features (G-features) and region features (R-features) for predicting image semantics. As an intermediary between image features and categories, we introduce the notion of mid-level concepts, which enables us to predict an image's category in three steps. First, a G-prediction system uses G-features to predict the probability of each category for an image. Simultaneously, a R-prediction system analyzes R-features to identify the probabilities of mid-level concepts in that image. Finally, our hybrid H-prediction system based on a Bayesian network reconciles the predictions from both R-prediction and G-prediction to produce the final classifications. Results of experimental validations show that this hybrid system outperforms both G-prediction and R-prediction significantly.

Web information fusion can be defined as the problem of collating and tracking information related to specific topics on the World Wide Web. Whereas most existing work on Web information fusion has focused on text-based multidocument summarization, this paper concerns the topic of image and text association, a cornerstone of cross-media Web information fusion. Specifically, we present two learning methods for discovering the underlying associations between images and texts based on small training data sets. The first method based on vague transformation measures the information similarity between the visual features and the textual features through a set of predefined domain-specific information categories. Another method uses a neural network to learn direct mapping between the visual and textual features by automatically and incrementally summarizing the associated features into a set of information templates. Despite their distinct approaches, our experimental results on a terrorist domain document set show that both methods are capable of learning associations between images and texts from a small training data set.

This paper presents an efficient and effective solution for retrieving image near-duplicate (IND) from image database. We introduce the coherent phrase model which incorporates the coherency of local regions to reduce the quantization error of the bag-of-words (BoW) model. In this model, local regions are characterized byvisual phraseof multiple descriptors instead of visual word of single descriptor. We propose two types of visual phrase to encode the coherency in feature and spatial domain, respectively. The proposed model reduces the number of false matches by using this coherency and generates sparse representations of images. Compared to other method, the local coherencies among multiple descriptors of every region improve the performance and preserve the efficiency for IND retrieval. The proposed method is evaluated on several benchmark datasets for IND retrieval. Compared to the state-of-the-art methods, our proposed model has been shown to significantly improve the accuracy of IND retrieval while maintaining the efficiency of the standard bag-of-words model. The proposed method can be integrated with other extensions of BoW.

To generate a believable and dynamic virtual world is a great challenge in interactive storytelling. In this paper, we propose a model, namely Evolutionary Fuzzy Cognitive Map (E-FCM), to model the dynamic causal relationships among different context variables. As an extension to conventional FCM, E-FCM models not only the fuzzy causal relationships among the variables, but also the probabilistic property of causal relationships, and asynchronous activity update of the concepts. With this model, the context variables evolve in a dynamic and uncertain manner with the according evolving time. As a result, the virtual world is presented more realistically and dynamically.

Traditional approaches to integrating knowledge into neural network are concerned mainly about supervised learning. This paper presents how a family of self-organizing neural models known as Fusion Architecture for Learning, COgnition and Navigation(FALCON) can incorporate a priori knowledge and perform knowledge refinement and expansion through reinforcement learning. Symbolic rules are formulated based on pre-existing know-how and inserted into FALCON as a priori knowledge. The availability of knowledge enables FALCON to start performing earlier in the initial learning trials. Through a temporal-difference (TD) learning method, the inserted rules can be refined and expanded according to the evaluative feedback signals received from the environment. Our experimental results based on a minefield navigation task have shown that FALCON is able to learn much faster and attain a higher level of performance earlier when inserted with the appropriate a priori knowledge.

This paper presents a neural architecture for learning category nodes encoding mappings across multimodal patterns involving sensory inputs, actions, and rewards. By integrating adaptive resonance theory (ART) and temporal difference (TD) methods, the proposed neural model, called TD fusion architecture for learning, cognition, and navigation (TD-FALCON), enables an autonomous agent to adapt and function in a dynamic environment with immediate as well as delayed evaluative feedback (reinforcement) signals. TD-FALCON learns the value functions of the state-action space estimated through on-policy and off-policy TD learning methods, specifically state-action-reward-state-action (SARSA) and Q-learning. The learned value functions are then used to determine the optimal actions based on an action selection policy. We have developed TD-FALCON systems using various TD learning strategies and compared their performance in terms of task completion, learning speed, as well as time and space efficiency. Experiments based on a minefield navigation task have shown that TD-FALCON systems are able to learn effectively with both immediate and delayed reinforcement and achieve a stable performance in a pace much faster than those of standard gradient-descent-based reinforcement learning systems.

Traditional text mining techniques transform free text into flat bags of words representation, which does not preserve sufficient semantics for the purpose of knowledge discovery. In this paper, we present a two-step procedure to mine generalized associations of semantic relations conveyed by the textual content of Web documents. First, RDF (resource description framework) metadata representing semantic relations are extracted from raw text using a myriad of natural language processing techniques. The relation extraction process also creates a term taxonomy in the form of a sense hierarchy inferred from WordNet. Then, a novel generalized association pattern mining algorithm (GP-Close) is applied to discover the underlying relation association patterns on RDF metadata. For pruning the large number of redundant overgeneralized patterns in relation pattern search space, the GP-Close algorithm adopts the notion of generalization closure for systematic overgeneralization reduction. The efficacy of our approach is demonstrated through empirical experiments conducted on an online database of terrorist activities

Acquisition of context poses unique challenges to mobile context-aware recommender systems. The limited resources in these systems make minimizing their context acquisition a practical need, and the uncertainty in the mobile environment makes missing and erroneous context inputs a major concern. In this paper, we propose an approach based on Bayesian networks (BNs) for building recommender systems that minimize context acquisition. Our learning approach iteratively trims the BN-based context model until it contains only the minimal set of context parameters that are important to a user. In addition, we show that a two-tiered context model can effectively capture the causal dependencies among context parameters, enabling a recommender system to compensate for missing and erroneous context inputs. We have validated our proposed techniques on a restaurant recommendation data set and a Web page recommendation data set. In both benchmark problems, the minimal sets of context can be reliably discovered for the specific users. Furthermore, the learned Bayesian network consistently outperforms the J4.8 decision tree in overcoming both missing and erroneous context inputs to generate significantly more accurate predictions.

Temporal-Difference-Fusion Architecture for Learning, Cognition, and Navigation (TD-FALCON) is a generalization of adaptive resonance theory (a class of self-organizing neural networks) that incorporates TD methods for real-time reinforcement learning. In this paper, we investigate how a team of TD-FALCON networks may cooperate to learn and function in a dynamic multiagent environment based on minefield navigation and a predator/prey pursuit tasks. Experiments on the navigation task demonstrate that TD-FALCON agent teams are able to adapt and function well in a multiagent environment without an explicit mechanism of collaboration. In comparison, traditional Q-learning agents using gradient-descent-based feedforward neural networks, trained with the standard backpropagation and the resilient-propagation (RPROP) algorithms, produce a significantly poorer level of performance. For the predator/prey pursuit task, we experiment with various cooperative strategies and find that a combination of a high-level compressed state representation and a hybrid reward function produces the best results. Using the same cooperative strategy, the TD-FALCON team also outperforms the RPROP-based reinforcement learners in terms of both task completion rate and learning efficiency.

Mobile context-aware recommender systems face unique challenges in acquiring context. Resource limitations make minimizing context acquisition a practical need, while the uncertainty inherent to the mobile environment makes missing context values a major concern. This paper introduces a scalable mechanism based on Bayesian network learning in a tiered context model to overcome both of these challenges. Extensive experiments on a restaurant recommender system showed that our mechanism can accurately discover causal dependencies among context, thereby enabling the effective identification of the minimal set of important context for a specific user and task, as well as providing highly accurate recommendations even when context values are missing.

With the rapid development and wide deployment of wireless local area networks (WLANs), WLAN-based positioning system employing signal-strength-based technique has become an attractive solution for location estimation in indoor environment. In recent years, a number of such systems has been presented, and most of the systems use the common nearest neighbor in signal space (NNSS) algorithm. In this paper, we propose an enhancement to the NNSS algorithm. We analyze the enhancement to show its effectiveness. The performance of the enhanced NNSS algorithm is evaluated with different values of the parameters. Based on the performance evaluation and analysis, we recommend some guidelines on optimizing the parameters of our proposed enhanced algorithm.

In this paper, we present a novel frequent generalized pattern mining algorithm, called GP-Close, for mining generalized associations from RDF metadata. To solve the over-generalization problem encountered by existing methods, GP-Close employs the notion of emphgeneralization closure for systematic over-generalization reduction.

Domain ontology has been used in many Semantic Web applications. However, few applications explore the use of ontology for personalized services. This paper proposes an ontology based user model consisting of both concepts and semantic relations to represent users' interests. Specifically, we adopt a statistical approach to learning a semantic-based user ontology model from domain ontology and a spreading activation procedure for inferencing in the user ontology model. We apply the methods of learning and exploiting user ontology to a semantic search engine for finding academic publications. Our experimental results support the efficacy of user ontology and spreading activation theory (SAT) for providing personalized semantic services.

Traditional text mining systems employ shallow parsing techniques and focus on concept extraction and taxonomic relation extraction. This paper presents a novel system called CRCTOL for mining rich semantic knowledge in the form of ontology from domain-specific text documents. By using a full text parsing technique and incorporating both statistical and lexico-syntactic methods, the knowledge extracted by our system is more concise and contains a richer semantics compared with alternative systems. We conduct a case study wherein CRCTOL extracts ontological knowledge, specifically key concepts and semantic relations, from a terrorism domain text collection. Quantitative evaluation, by comparing with a state-of-the-art ontology learning system known as text-to-onto, has shown that CRCTOL produces much better precision and recall for both concept and relation extraction, especially from sentences with complex structures.

Traditional approaches to recommender systems have not taken into account situational information when making recommendations, and this seriously limits the relevance of the results. This paper advocates context-awareness as a promising approach to enhance the performance of recommenders, and introduces a mechanism to realize this approach. We present a framework that separates the contextual concerns from the actual recommendation module, so that contexts can be readily shared across applications. More importantly, we devise a learning algorithm to dynamically identify the optimal set of contexts for a specific recommendation task and user. An extensive series of experiments has validated that our system is indeed able to learn both quickly and accurately.

This work presents a natural extension of self-organizing neural network architecture for learning cognitive codes across multi-modal patterns involving sensory input, actions, and rewards. The proposed cognitive model, called FALCON, enables an autonomous agent to adapt and function in a dynamic environment. Simulations based on a minefield navigation task indicate that the system is able to adapt amazingly well and learns rapidly through it's interaction with the environment in an online and incremental manner. The scalability and robustness of the system is further enhanced by an online code evaluation and pruning procedure, that maintains the number of cognitive codes at a manageable size without degradation of system performance.

This paper introduces the Adaptive Resonance Theory under Constraint (ART-C 2A) learning paradigm based on ART 2A, which is capable of generating a user-defined number of recognition nodes through online estimation of an appropriate vigilance threshold. Empirical experiments compare the cluster validity and the learning efficiency of ART-C 2A with those of ART 2A, as well as three closely related clustering methods, namely online K-Means, batch K-Means, and SOM, in a quantitative manner. Besides retaining the online cluster creation capability of ART 2A, ART-C 2A gives the alternative clustering solution, which allows a direct control on the number of output clusters generated by the self-organizing process.

We address the problem of mining a class of semantic networks, called concept frame graphs (CFG's), for knowledge discovery from text. This new representation is motivated by the need to capture richer text content so that nontrivial mining tasks can be performed. We first define the CFG representation and then describe a rule-based algorithm for constructing a CFG from text documents. Treating the CFG as a networked knowledge base, we propose new methods for text mining. On a specific task of discovering the top companies in an area, we observe that our approach leads to simpler content mining algorithms, once the CFG has been constructed. Moreover, exploiting the network structure of CFG results in significant improvements in precision and recall.

Clustering of gene expression patterns is of great value for the understanding of the various molecular biological processes. While a number of algorithms have been applied to gene clustering, there are relatively few studies on the application of neural networks to this task. In addition, there is a lack of quantitative evaluation of the gene clustering results. This paper proposes Adaptive Resonance Theory under Constraint (ART-C) for efficient clustering of gene expression data. We illustrate that ART-C can effectively identify gene functional groupings through a case study on rat CNS data. Based on a set of quantitative evaluation measures, we compare the performance of ART-C with those of K-Means, SOM, and conventional ART. Our comparative studies on the yeast cell cycle and the human hematopoietic differentiation data sets show that ART-C produces reasonably good quantitative performance. More importantly, compared with K-Means and SOM, ART-C shows a significantly higher learning efficiency, which is crucial for knowledge discovery from large scale biological databases.

We address the text content mining problem through a concept based framework by constructing a conceptual knowledge base and discovering knowledge therefrom. Defining a novel representation called the Concept Frame Graph (CFG), we propose a learning algorithm for constructing a CFG knowledge base from text documents. An interactive concept map visualization technique is presented for user-guided knowledge discovery from the knowledge base. Through experimental studies on real life documents, we observe that the proposed approach is promising for mining deeper knowledge.

Web intelligence can be defined as the process of scanning and tracking information on the World Wide Web so as to gain competitive advantages. This paper describes a system known as Flexible Organizer for Competitive Intelligence (FOCI) that transforms raw URLs returned by internet search engines into personalized information portfolios. FOCI builds information portfolios by gathering and organizing online information according to a user's needs and preferences. Through a novel method called User-Configurable Clustering, a user can personalize his/her portfolios in terms of the content as well as the information structure. The personalized portfolios can then be used to track new information and organize them into appropriate folders accordingly. We show a sample session of FOCI which illustrates how a user may create and personalize an information portfolio according to his/her preferences and how the system discovers novel information groupings while organizing familiar information according to the user-defined themes.

This paper describes how an integrated web-based application, code-named FOCI (Flexible Organizer for Competitive Intelligence), can help the knowledge worker in the gathering, organizing, tracking, and dissemination of competitive intelligence or knowledge bases on the web. It shows how text mining techniques including a novel user-configurable clustering, trend analysis and visualization techniques can be used synergistically to address the problem of managing information gathered from the web. FOCI allows a user to define and personalize the organization of the information clusters according to their needs and preferences into portfolios. Predefined sections for organizing information in specific domains is also supported. The personalized portfolios created can be saved and subsequently tracked and shared with other users. In addition, FOCI is designed to handle multilingual documents.

This paper introduces a hybrid system termed cascade adaptive resonance theory mapping (ARTMAP) that incorporates symbolic knowledge into neural-network learning and recognition. Cascade ARTMAP, a generalization of fuzzy ARTMAP, represents intermediate attributes and rule cascades of rule-based knowledge explicitly and performs multistep inferencing. A rule insertion algorithm translates if-then symbolic rules into cascade ARTMAP architecture. Besides that initializing networks with prior knowledge can improve predictive accuracy and learning efficiency, the inserted symbolic knowledge can be refined and enhanced by the cascade ARTMAP learning algorithm. By preserving symbolic rule form during learning, the rules extracted from cascade ARTMAP can be compared directly with the originally inserted rules. Simulations on an animal identification problem indicate that a priori symbolic knowledge always improves system performance, especially with a small training set. Benchmark study on a DNA promoter recognition problem shows that with the added advantage of fast learning, cascade ARTMAP rule insertion and refinement algorithms produce performance superior to those of other machine learning systems and an alternative hybrid system known as knowledge-based artificial neural network (KBANN). Also, the rules extracted from cascade ARTMAP are more accurate and much cleaner than the NofM rules extracted from KBANN.

This article introduces a neural network based cognitive architecture termed Concept Hierarchy Memory Model (CHMM) for conceptual knowledge representation and commonsense reasoning. CHMM is composed of two subnetworks: a Concept Formation Network (CFN), that acquires concepts based on their sensory representations; and a Concept Hierarchy Network (CHN), that encodes hierarchical relationships between concepts. Based on Adaptive Resonance Associative Map (ARAM), a supervised Adaptive Resonance Theory (ART) model, CHMM provides a systematic treatment for concept formation and organization of a concept hierarchy. Specifically, a concept can be learned by sampling activities across multiple sensory fields. By chunking relations between concepts as cognitive codes, a concept hierarchy can be learned/modified through experience. Also, fuzzy relations between concepts can now be represented in terms of the weights on the links connecting them. Using a unified inferencing mechanism based on code firing, CHMM performs an important class of commonsense reasoning, including concept recognition and property inheritance.

A neural network expert system called adaptive connectionist expert system (ACES) which will learn adaptively from past experience is described. ACES is based on the neural logic network, which is capable of doing both pattern processing and logical inferencing. The authors discuss two strategies, pattern matching ACES and rule inferencing ACES. The pattern matching ACES makes use of past examples to construct its neural logic network and fine-tunes itself adaptively during its use by further examples supplied. The rule inferencing ACES conceptualizes new rules based on the frequencies of use on the rule-based neural logic network. A new rule could be considered as a pattern matching example and be incorporated into pattern matching ACES.>

An inertial navigation system interactive diagnostic expert (INSIDE) was developed for troubleshooting an avionic line-replaceable unit, the inertial navigation system. INSIDE was designed based on a neural network model called neural-logic network. The knowledge base can be constructed using a neural-logic network by learning from past cases recorded in the workshop log book. To complement the connectionist knowledge base, a flowchart module which captures the knowledge of troubleshooting flowcharts was also implemented as part of the system. During operation, if the connectionist module fails to derive the solution, the user will be directed to the flowchart module for guidance. After the case is solved, it can be captured as a new example to be acquired by the connectionist module. Besides providing an economical way for developing fault diagnostic systems in general, the learning process of the system highly resembles the way an expert acquires knowledge through experience