• We propose DASP, a novel domain-aware prompting framework for cross-domain sequential recommendations with natural explanations. • Addresses domain gaps through contrastive learning and meta-learned adapters for efficient domain adaptation. • DASP outperforms baselines with 10.7 % HR@10 and 10.5 % NDCG@10 improvements with 17.3 % better explanation quality. • DASP reduces training time by 54 % using a lightweight architecture and frozen large language models. Cross-domain sequential recommendation faces persistent challenges in addressing domain shift, data sparsity, and the trade-off between performance, efficiency, and explainability. Existing methods often struggle with inefficient cross-domain adaptation or fail to generate coherent explanations that bridge user preferences across domains. To overcome these limitations, we propose Domain-Aware Self-Prompting (DASP) , a novel framework that integrates cross-domain recommendation with natural language explanation generation. DASP introduces three key innovations: (1) a domain-invariant self-prompt generator that captures shared user preferences via contrastive alignment across domains; (2) lightweight domain adapters with meta-learned initialization for parameter-efficient adaptation to target domains; and (3) a cross-domain explanation generator that grounds recommendations in semantically aligned multi-domain prompts using large language models. Extensive experiments on Amazon Movie-Book and Food-Kitchen datasets demonstrate DASP’s superiority, achieving 10.7 % and 10.5 % improvements in HR@10 and NDCG@10 over state-of-the-art baselines on the Movie-Book dataset, while reducing training time by 54 % compared to full large language models fine-tuning approaches. Qualitative and quantitative analyses validate DASP’s ability to generate interpretable explanations that link cross-domain preferences, offering a scalable and trustworthy solution for cross-domain sequential recommendation. Our work bridges critical gaps in efficiency, adaptability, and explainability for real-world multi-domain recommendation systems.

Generating realistic two-person interaction motions from text holds immense potential in computer vision and animations. While existing latent motion diffusion models offer compact and efficient representations, they often fail to produce physically plausible contacts and are typically constrained to a single canonical body shape. As a result, the generated motion sequences exhibit substantial mesh penetrations and lack interaction realism. To address these limitations, we propose a contact and shape-aware latent motion representation and diffusion model (CoShMDM) for generating realistic two-person interactions from text. Our framework begins by constructing contact-compatible motion using SMPL-based meshes and a normal alignment-based mesh contact matrix to capture fine-grained mesh-level contacts. To account for shape diversity, we incorporate SMPL shape parameters and iteratively learn contact dynamics across different body shapes. Additionally, a reinforcement learning-based mesh penetration avoidance policy network, guided by signed distance fields, is introduced to minimize mesh penetrations while preserving contact fidelity and shape-aware motion. We further employ a dual-encoder VQ-VAE to learn disentangled latent representations for motion and contacts, which are then utilized in a text- and body-shape-conditioned diffusion model. To ensure spatial, temporal, and semantic coherence, we integrate a novel contact and motion consistency module into the diffusion transformer. Extensive evaluations on the InterHuman and InterX datasets demonstrate that our method outperforms state-of-the-art approaches achieving lowest FID scores (4.801 and 0.013), with 19% and 17.3% reductions in mesh penetrations, and 17.8% and 33.2% gains in contact similarity, respectively.

We propose an energy amplification technique to address the issue that existing models easily overlook low-energy components in time series forecasting. This technique comprises an energy amplification block and an energy restoration block. The energy amplification block enhances the energy of low-energy components to improve the model's learning efficiency for these components, while the energy restoration block returns the energy to its original level. Moreover, considering that the energy-amplified data typically displays two distinct energy peaks in the frequency spectrum, we integrate the energy amplification technique with a seasonal-trend forecaster to model the temporal relationships of these two peaks independently, serving as the backbone for our proposed model, Amplifier. Additionally, we propose a semi-channel interaction temporal relationship enhancement block for Amplifier, which enhances the model's ability to capture temporal relationships from the perspective of the commonality and specificity of each channel in the data. Extensive experiments on eight time series forecasting benchmarks consistently demonstrate our model's superiority in both effectiveness and efficiency compared to state-of-the-art methods.

The significance of Temporal Knowledge Graphs (TKGs) in Artificial Intelligence (AI) lies in their capacity to incorporate time-dimensional information, support complex reasoning and prediction, optimize decision-making processes, enhance the accuracy of recommendation systems, promote multimodal data integration, and strengthen knowledge management and updates. This provides a robust foundation for various AI applications. To effectively learn and apply both static and dynamic temporal patterns for reasoning, a range of embedding methods and large language models (LLMs) have been proposed in the literature. However, these methods often rely on a single underlying embedding space, whose geometric properties severely limit their ability to model intricate temporal patterns, such as hierarchical and ring structures. To address this limitation, this paper proposes embedding TKGs into projective geometric space and leverages LLMs technology to extract crucial temporal node information, thereby constructing the 5EL model. By embedding TKGs into projective geometric space and utilizing Möbius Group transformations, we effectively model various temporal patterns. Subsequently, LLMs technology is employed to process the trained TKGs. We adopt a parameter-efficient fine-tuning strategy to align LLMs with specific task requirements, thereby enhancing the model's ability to recognize structural information of key nodes in historical chains and enriching the representation of central entities. Experimental results on five advanced TKG datasets demonstrate that our proposed 5EL model significantly outperforms existing models.

The recent boom of large pre-trained models witnesses remarkable success in developing foundation models (FMs) for time series forecasting. Despite impressive performance across diverse downstream forecasting tasks, existing time series FMs possess massive network architectures and require substantial pre-training on large-scale datasets, which significantly hinders their deployment in resource-constrained environments. In response to this growing tension between versatility and affordability, we propose **SEMPO**, a novel lightweight foundation model that requires pretraining on relatively small-scale data, yet exhibits strong general time series forecasting. Concretely, SEMPO comprises two key modules: 1) _energy-aware **S**p**E**ctral decomposition module_, that substantially improves the utilization of pre-training data by modeling not only the high-energy frequency signals but also the low-energy yet informative frequency signals that are ignored in current methods; and 2) _**M**ixture-of-**P**r**O**mpts enabled Transformer_, that learns heterogeneous temporal patterns through small dataset-specific prompts and adaptively routes time series tokens to prompt-based experts for parameter-efficient model adaptation across different datasets and domains. Equipped with these modules, SEMPO significantly reduces both pre-training data scale and model size, while achieving strong generalization. Extensive experiments on two large-scale benchmarks covering 16 datasets demonstrate the superior performance of SEMPO in both zero-shot and few-shot forecasting scenarios compared with state-of-the-art methods. Code and data are available at https://github.com/mala-lab/SEMPO.

Understanding users’ satisfaction is fundamental for enhancing the effectiveness and usability of e-learning platforms. The existing approaches for analyzing users’ satisfaction leverage word embedding vectors to represent sentiment information, but they often fail to fully address the complex relationship between emotional and semantic information. Additionally, several emotional and semantic word embedding models are proposed, but they require sentiment information. In this study, we propose a novel multi-task deep neural model, called Sentiment-Emotion-Semantic Network (SES-Net), capable of learning sentiment, emotion, and semantic information simultaneously. The proposed model comprises three main sub-neural tasks: Bidirectional Long Short-Term Memory (BiLSTM) to capture sentiment, BiLSTM to extract semantics, and Convolutional Neural Networks (CNN) to learn emotional features. Experimental results reveal that, SES-Net outperforms the previous approaches by achieving an average F1-score of 90.59%.

Deep graph clustering is a highly significant tool for community detection, enabling the identification of strongly connected groups of nodes within a graph. This technology is crucial in various fields such as education and E-learning. However, deep graph clustering can be more misled by the graph topology, disregarding node information. For example, an excessive number of intercommunity edges or insufficient intracommunity edges can lead to inaccurate community distinction by the model. In this article, we propose a novel model, deep Graph Clustering with Triple Fusion Autoencoder (GC-TriFA) for community detection, which utilizes a triple encoding fusion mechanism to balance the incorporation of node and topological information, thereby mitigating this issue. Specifically, GC-TriFA employs a shallow linear coding fusion and a deep coding fusion method within an autoencoder structure. This approach enables the model to simultaneously learn and capture the embedding of cross-modality information and later utilizes weight fusion to equalize the two modalities. Furthermore, GC-TriFA also reconstructs the graph structure, learns relaxed$k$-means, and undergoes self-supervised training to enhance the quality of the graph embedding. The experimental results of GC-TriFA, when evaluated as an end-to-end model on publicly available datasets, demonstrate its superiority compared to the baseline models.

In the globalized academic landscape, university academic evaluation plays a crucial role. However, predicting university academic evaluation in time series prediction poses significant challenges due to its nonlinearity and nonstationarity. In this article, we propose a novel method to address the prediction problem in scenarios characterized by limited data volume. By leveraging the properties of Gaussian processes, our method constructs nonstationary kernels that are fully trainable. The time series data is decomposed into trend and seasonal components, which are then inputted into an independent Gaussian process model equipped with trainable parameters. The core of our method lies in the trend Gaussian process model, which consists of classical kernel functions and multilayer perceptron stacks. This combination enhances the model’s capacity to capture long-term dependencies and improves prediction accuracy. To evaluate the effectiveness of our novel method, experiments were conducted on both the M4 dataset and an academic indicator dataset. The experimental results demonstrate significant improvements in prediction accuracy and stability compared with traditional time series prediction methods. Additionally, comparative experiments were performed to contrast our method against other popular time series prediction methods, consistently affirming the superior predictive performance of our approach.

Due to the nonstationary nature, the distribution of real-world multivariate time series (MTS) changes over time, which is known as distribution drift. Most existing MTS forecasting models greatly suffer from distribution drift and degrade the forecasting performance over time. Existing methods address distribution drift via adapting to the latest arrived data or self-correcting per the meta knowledge derived from future data. Despite their great success in MTS forecasting, these methods hardly capture the intrinsic distribution changes, especially from a distributional perspective. Accordingly, we propose a novel framework temporal conditional variational autoencoder (TCVAE) to model the dynamic distributional dependencies over time between historical observations and future data in MTSs and infer the dependencies as a temporal conditional distribution to leverage latent variables. Specifically, a novel temporal Hawkes attention (THA) mechanism represents temporal factors that subsequently fed into feedforward networks to estimate the prior Gaussian distribution of latent variables. The representation of temporal factors further dynamically adjusts the structures of Transformer-based encoder and decoder to distribution changes by leveraging a gated attention mechanism (GAM). Moreover, we introduce conditional continuous normalization flow (CCNF) to transform the prior Gaussian to a complex and form-free distribution to facilitate flexible inference of the temporal conditional distribution. Extensive experiments conducted on six real-world MTS datasets demonstrate the TCVAE's superior robustness and effectiveness over the state-of-the-art MTS forecasting baselines. We further illustrate the TCVAE applicability through multifaceted case studies and visualization in real-world scenarios.

Few-shot named entity recognition (NER) aims to recognize novel named entities in low-resource domains utilizing existing knowledge. However, the present few-shot NER models assume that the labeled data are all clean without noise or outliers, and there are few works focusing on the robustness of the cross-domain transfer learning ability to textual adversarial attacks in Few-shot NER. In this work, we comprehensively explore and assess the robustness of few-shot NER models under textual adversarial attack scenario, and found the vulnerability of existing few-shot NER models. Furthermore, we propose a robust two-stage few-shot NER method with Boundary Discrimination and Correlation Purification (BDCP). Specifically, in the span detection stage, the entity boundary discriminative module is introduced to provide a highly distinguishing boundary representation space to detect entity spans. In the entity typing stage, the correlations between entities and contexts are purified by minimizing the interference information and facilitating correlation generalization to alleviate the perturbations caused by textual adversarial attacks. In addition, we construct adversarial examples for few-shot NER based on public datasets Few-NERD and Cross-Dataset. Comprehensive evaluations on those two groups of few-shot NER datasets containing adversarial examples demonstrate the robustness and superiority of the proposed method.

Currently, the increasing information overload on Massive Open Online Courses(MOOCs) inhibits the appropriate choice of learning objects by learners, leading to low efficiency and high dropout rates. However, in MOOC platforms, recommendation network structures that can selectively extract implicit features such as heterogeneous learning preference and knowledge organization of learning objects are still not comprehensively studied. To this end, we propose a learning object recommendation model namely ACGCN based on heterogeneous learning behavior and knowledge graph. By introducing an attention mechanism, information is amplified when updating the representation of the heterogeneous graph, which eliminates the impact of noise and improves the robustness of ACGCN. Experimental results using a real-world dataset revealed that our proposed model has the best performance compared to those of several existing baselines.

BACKGROUND: COVID-19 is a disease that caused a contagious respiratory ailment that killed and infected hundreds of millions. It is necessary to develop a computer-based tool that is fast, precise, and inexpensive to detect COVID-19 efficiently. Recent studies revealed that machine learning and deep learning models accurately detect COVID-19 using chest X-ray (CXR) images. However, they exhibit notable limitations, such as a large amount of data to train, larger feature vector sizes, enormous trainable parameters, expensive computational resources (GPUs), and longer run-time. RESULTS: In this study, we proposed a new approach to address some of the above-mentioned limitations. The proposed model involves the following steps: First, we use contrast limited adaptive histogram equalization (CLAHE) to enhance the contrast of CXR images. The resulting images are converted from CLAHE to YCrCb color space. We estimate reflectance from chrominance using the Illumination-Reflectance model. Finally, we use a normalized local binary patterns histogram generated from reflectance (Cr) and YCb as the classification feature vector. Decision tree, Naive Bayes, support vector machine, K-nearest neighbor, and logistic regression were used as the classification algorithms. The performance evaluation on the test set indicates that the proposed approach is superior, with accuracy rates of 99.01%, 100%, and 98.46% across three different datasets, respectively. Naive Bayes, a probabilistic machine learning algorithm, emerged as the most resilient. CONCLUSION: Our proposed method uses fewer handcrafted features, affordable computational resources, and less runtime than existing state-of-the-art approaches. Emerging nations where radiologists are in short supply can adopt this prototype. We made both coding materials and datasets accessible to the general public for further improvement. Check the manuscript's availability of the data and materials under the declaration section for access.

Abstract As an extension of conventional top-K item recommendation solution, bundle recommendation has aroused increasingly attention. However, because of the extreme sparsity of user-bundle (UB) interactions, the existing top-K item recommendation methods suffer from poor performance when applied to bundle recommendation. While some graph-based approaches have been proposed for bundle recommendation, these approaches primarily leverage the bipartite graph to model the UB interactions, resulting in suboptimal performance. In this paper, a dual hypergraph contrastive learning model is proposed for bundle recommendation. First, we model the direct and indirect UB interactions as hypergraphs to represent the higher-order UB relations. Second, we utilize the hypergraph convolution networks to learn the user and bundle embeddings from the hypergraphs, and improve the learned embeddings through a bidirectional contrastive learning strategy. Finally, we adopt a joint loss that combines the InfoBPR loss supporting multiple negative samples and the contrastive losses to optimize model parameters for prediction. Experiments on the real-world datasets indicate that our model performs better than the state-of-the-art baseline methods.

Analyzing lecturers’ and students’ satisfaction with using e-learning is important to improve the teaching-learning processes. The existing approaches have been widely employing machine learning algorithms, usage-based, and System Usability Scale (SUS) metrics based on users’ opinions, activities, and usability testing, respectively. However, the usage-based and SUS metrics fail to cover users’ opinions about e-learning systems and they involve manual features engineering. Whereas, the machine learning classifiers do not analyze satisfaction based on activities and usability. Toward this end, we propose a machine learning model that employs CNN and BiLSTM algorithms to concatenate the features extracted from users’ activities, usability testing, and users’ opinions. The proposed model is coined as E-learning Users’ Satisfaction Detection (El-USD). Experimental results suggest that there is a significant correlation between satisfaction analysis by achieving an average r = 0.778. The evaluation results further suggest that our proposed approach can analyze users’ satisfaction accurately.

Frequent meteorological disasters present new challenges for decision-making in disaster response. As a timely and effective source of intelligent information, social media plays a vital role in detecting and monitoring these situations. Meteorological social briefings summarize valuable information from numerous social media posts, providing essential decision-support services. This article proposes a multi-knowledge-enhanced summarization (MKES) model for automatically generating meteorological social briefing content from multiple Sina Weibo posts. The MKES model consists of a summary generation module and a knowledge enhancement module. The knowledge enhancement module guides and constrains the summary generation process using meteorological events and geographical location knowledge, resulting in summaries that focus on describing specific knowledge from the source text. The MKES model outperforms baseline models in content evaluation, as measured by$\text {ROUGE-1}$,$\text {ROUGE-2}$, and$\text {ROUGE-L}$scores, and in sentiment evaluation, as measured by$F_{1}$scores. Based on the MKES model, a framework for generating meteorological social briefings is developed, providing decision support services for the China Meteorological Administration (CMA).

Multivariate time series (MTS) forecasting penetrates various aspects of our economy and society, whose roles become increasingly recognized. However, often MTS forecasting is unfair, not only degrading their practical benefits but even incurring potential risk. Unfair MTS forecasting may be attributed to disparities relating to advantaged and disadvantaged variables, which has rarely been studied in the MTS forecasting. In this work, we formulate the MTS fairness modeling problem as learning informative representations attending to both advantaged and disadvantaged variables. Accordingly, we propose a novel framework, namedFairFor, for fairness-aware MTS forecasting, i.e.,fair MTS forecasting.FairForuses adversarial learning to generate both group-irrelevant and -relevant representations for downstream forecasting.FairForfirst adopts recurrent graph convolution to capture spatio-temporal variable correlations and to group variables by leveraging a spectral relaxation of the K-means objective. Then, it utilizes a novel filtering$\&$fusion module to filter group-relevant information and generate group-irrelevant representations by orthogonality regularization. The group-irrelevant and -relevant representations form highly informative representations, facilitating to share the knowledge from advantaged variables to disadvantaged variables and guarantee the fairness of forecasting. Extensive experiments on four public datasets demonstrate theFairForeffectiveness for fair forecasting and significant performance improvement.

Multivariate time series (MTS) forecasting is crucial in many real-world applications. To achieve accurate MTS forecasting, it is essential to simultaneously consider both intra- and inter-series relationships among time series data. However, previous work has typically modeled intra- and inter-series relationships separately and has disregarded multi-order interactions present within and between time series data, which can seriously degrade forecasting accuracy. In this article, we reexamine intra- and inter-series relationships from the perspective of mutual information and accordingly construct a comprehensive relationship learning mechanism tailored to simultaneously capture the intricate multi-order intra- and inter-series couplings. Based on the mechanism, we propose a novel deep coupling network for MTS forecasting, named DeepCN, which consists of a coupling mechanism dedicated to explicitly exploring the multi-order intra- and inter-series relationships among time series data concurrently, a coupled variable representation module aimed at encoding diverse variable patterns, and an inference module facilitating predictions through one forward step. Extensive experiments conducted on seven real-world datasets demonstrate that our proposed DeepCN achieves superior performance compared with the state-of-the-art baselines.

Tuples extraction is a fundamental task for information extraction and knowledge graph construction.The extracted tuples are usually represented as knowledge triples consisting of subject, relation, and object.In practice, however, the validity of knowledge triples is associated with and changes with the spatial, temporal, or other kinds of constraints.Motivated by this observation, this paper proposes a constrained tuple extraction (CTE) task to guarantee the validity of knowledge tuples.Formally, the CTE task is to extract constrained tuples from unstructured text, which adds constraints to conventional triples.To this end, we propose an interaction-aware network.Combinatorial interactions among context-specific external features and distinct-granularity internal features are exploited to effectively mine the potential constraints.Moreover, we have built a new dataset containing totally 1,748,826 constrained tuples for training and 3656 ones for evaluation.Experiments on our dataset and the public CaRB dataset demonstrate the superiority of the proposed model.The constructed dataset and the codes are publicly available.

Multivariate time series (MTS) forecasting has shown great importance in numerous industries. Current state-of-the-art graph neural network (GNN)-based forecasting methods usually require both graph networks (e.g., GCN) and temporal networks (e.g., LSTM) to capture inter-series (spatial) dynamics and intra-series (temporal) dependencies, respectively. However, the uncertain compatibility of the two networks puts an extra burden on handcrafted model designs. Moreover, the separate spatial and temporal modeling naturally violates the unified spatiotemporal inter-dependencies in real world, which largely hinders the forecasting performance. To overcome these problems, we explore an interesting direction of directly applying graph networks and rethink MTS forecasting from a pure graph perspective. We first define a novel data structure, hypervariate graph, which regards each series value (regardless of variates or timestamps) as a graph node, and represents sliding windows as space-time fully-connected graphs. This perspective considers spatiotemporal dynamics unitedly and reformulates classic MTS forecasting into the predictions on hypervariate graphs. Then, we propose a novel architecture Fourier Graph Neural Network (FourierGNN) by stacking our proposed Fourier Graph Operator (FGO) to perform matrix multiplications in Fourier space. FourierGNN accommodates adequate expressiveness and achieves much lower complexity, which can effectively and efficiently accomplish {the forecasting}. Besides, our theoretical analysis reveals FGO's equivalence to graph convolutions in the time domain, which further verifies the validity of FourierGNN. Extensive experiments on seven datasets have demonstrated our superior performance with higher efficiency and fewer parameters compared with state-of-the-art methods. Code is available at this repository: https://github.com/aikunyi/FourierGNN.

Time series forecasting has played the key role in different industrial, including finance, traffic, energy, and healthcare domains. While existing literatures have designed many sophisticated architectures based on RNNs, GNNs, or Transformers, another kind of approaches based on multi-layer perceptrons (MLPs) are proposed with simple structure, low complexity, and superior performance. However, most MLP-based forecasting methods suffer from the point-wise mappings and information bottleneck, which largely hinders the forecasting performance. To overcome this problem, we explore a novel direction of applying MLPs in the frequency domain for time series forecasting. We investigate the learned patterns of frequency-domain MLPs and discover their two inherent characteristic benefiting forecasting, (i) global view: frequency spectrum makes MLPs own a complete view for signals and learn global dependencies more easily, and (ii) energy compaction: frequency-domain MLPs concentrate on smaller key part of frequency components with compact signal energy. Then, we propose FreTS, a simple yet effective architecture built upon Frequency-domain MLPs for Time Series forecasting. FreTS mainly involves two stages, (i) Domain Conversion, that transforms time-domain signals into complex numbers of frequency domain; (ii) Frequency Learning, that performs our redesigned MLPs for the learning of real and imaginary part of frequency components. The above stages operated on both inter-series and intra-series scales further contribute to channel-wise and time-wise dependency learning. Extensive experiments on 13 real-world benchmarks (including 7 benchmarks for short-term forecasting and 6 benchmarks for long-term forecasting) demonstrate our consistent superiority over state-of-the-art methods. Code is available at this repository: https://github.com/aikunyi/FreTS.

Proper noun recognition is a sub-task in named entity recognition.However, few methods have been specifically applied to the Chinese.The reason is that most of the existing deep clustering methods rely on manually labeled training sets, which take a long time in the learning process.And due to the wide and large-scale nature of the proprietary domain and the lack of word boundaries, recognizing Chinese specialized terms from unstructured text remains challenging.In this paper, we design an unsupervised method to improve Chinese proper noun recognition.The first step is to implement the word separation for Chinese, followed by a BERT-based improved word characterization method to obtain word vectors.Finally, we use the autoencoder-based deep clustering method to complete the extraction of proper nouns from books.We have done comparison experiments on the public dataset and our selected professional book data respectively, and the result is an improvement of our method in both the accuracy and F1 values.

Multivariate time-series (MTS) forecasting plays a crucial role in various real-world applications, but the complex dependencies between time-series variables (i.e., inter-series dependencies) make this task extremely challenging. While most existing studies focus on modeling intra-series (temporal) dependencies by capturing long- and short-term patterns, they fail to explore and exploit the inter-series dependencies to enhance MTS forecasting. In this paper, we propose a Cluster-aware Attentive Convolutional Recurrent Network (CACRN) to capture both inter-series and intra-series dependencies in MTS data. Specifically, CACRN first introduces a cluster-aware variable representation module that separates irrelevant variables and captures the interaction between relevant variables to learn cluster-aware variable representations. Then, CACRN feeds these representations into parallel convolutional recurrent neural networks (CRNNs) to capture the short- and long-term temporal dependencies in a cluster-wise manner. Next, a cluster-aware attention mechanism is introduced to attend to temporal information in each cluster and co-attend all cluster information jointly to capture intra-cluster and inter-cluster dependencies for the downstream forecasting task. Our extensive experiments on six real-world datasets demonstrate that CACRN is effective and outperforms representative and state-of-the-art baselines. Our proposed method is suitable for a wide range of real-world data collections, especially those with clear dependencies of variables.

Current evidence indicates that the semantic representation of question and answer sentences is better generated by deep neural network-based sentence models than traditional methods in community answer selection tasks. In particular, as a widely recognized language model, the self-attention model computes the similarity between the specific word and the whole sets of words in the same sentence and generates new semantic representation through the similarity-weighted summation of semantic representations of the whole words. However, the self-attention operation entirely considers all the signals with a weighted sum operation, which disperses the distribution of attention, which may result in overlooking the relation of neighboring signals. This issue becomes serious when applying the self-attention model to online community question answering platforms because of the varied length of the user-generated questions and answers. To address this problem, we introduce an attention mechanism enhanced local self-attention (LSA), which restricts the range of original self-attention by a local window mechanism, thereby scaling linearly when increasing the sequence length. Furthermore, we propose stacking multiple LSA layers to model the relationship of multiscale$n$-gram features. It captures the word-to-word relationship in the first layer and then captures the chunk-to-chunk (such as lexical$n$-gram phrases) relationship in its deeper layers. We also test the effectiveness of the proposed model by applying the learned representation through the LSA model to a Siamese and a classification network in community question answer selection tasks. Experiments on the public datasets show that the proposed LSA achieves a good performance.

Natural disasters are public emergencies characterized by suddenness, universality, and nonconventionality. Realizing the early warning, monitoring, and intervention of natural disasters and their derivative social impacts is significant for reducing the disasters’ damage and benefits the maintenance of social stability. Social sensors are ubiquitous sensors based on social network platforms. It uses the concepts and methods of physical space to mine social signals that integrate human perception and intelligence in cyberspace. Compared with traditional physical sensors, social sensors represent a crucial data acquisition channel in the emergency management of natural disasters and have the advantages of real time, comprehensive coverage, low cost, and flexible deployment. This article reviews the application of social sensors in natural disasters emergency management. We summarize the application functions of social sensors into three categories: natural disaster situation awareness and event detection, disaster information dissemination and communication, and disaster sentiment analysis and public opinion mining. Based on the above functions, this article analyzes the research status, data, technical methods, and application systems. Finally, this article proposes a research trend of applying social sensors in natural disaster emergency management according to the requirements of real scenarios.

Multimodal sentiment analysis (MSA) is a very challenging task due to its complex and complementary interactions between multiple modalities, which can be widely applied into areas of product marketing, public opinion monitoring, and so on. However, previous works directly utilized the features extracted from multimodal data, in which the noise reduction within and among multiple modalities has been largely ignored before multimodal fusion. This paper proposes a multi-level attention map network (MAMN) to filter noise before multimodal fusion and capture the consistent and heterogeneous correlations among multi-granularity features for multimodal sentiment analysis. Architecturally, MAMN is comprised of three modules: multi-granularity feature extraction module, multi-level attention map generation module, and attention map fusion module. The first module is designed to sufficiently extract multi-granularity features from multimodal data. The second module is constructed to filter noise and enhance the representation ability for multi-granularity features before multimodal fusion. And the third module is built to extensibly mine the interactions among multi-level attention maps by the proposed extensible co-attention fusion method. Extensive experimental results on three public datasets show the proposed model is significantly superior to the state-of-the-art methods, and demonstrate its effectiveness on two tasks of document-based and aspect-based MSA tasks.

Massive Open Online Courses (MOOCs) have received unprecedented attention, in which learners can obtain a large number of learning objects anytime and anywhere. However, the increasing information overload on MOOCs inhibits the appropriate choice of learning objects by learners, leading to a low efficiency and high dropout rates in the learning process of this human-computer interaction scenario. E-learning recommendation systems have been studied to present learning objects directly to learners, thereby relieving such problem. However, in MOOC platforms, recommendation network structures which can selectively extract implicit feature such as heterogeneous learning preference and knowledge organization of learning objects are still not comprehensively studied. To this end, we propose a learning object recommendation model based on heterogeneous learning behavior and knowledge graph. To generate a unified representation of each entity and relation, we first propose an Attentive Composition based Graph Convolutional Network (ACGCN). By introducing an attention mechanism, information is amplified when updating the representation of the heterogeneous graph, which eliminates the impact of noise and improves the robustness of the model. Then, a Dense Feature based Operation-Aware Network (DFOAN) is utilized to capture implicit and complex learners’ interactive behaviors, and to further provide a recommendation. Experimental results using two real-world datasets revealed that our proposed model has the best precision, recall, F1, and accuracy scores compared to those of several existing models.

Modeling complex hierarchical and grouped feature interaction in the multivariate time series data is indispensable to comprehend the data dynamics and predicting the future condition. The implicit feature interaction and high-dimensional data make multivariate forecasting very challenging. Many existing works did not put more emphasis on exploring explicit correlation among multiple time series data, and complicated models are designed to capture long- and short-range pattern with the aid of attention mechanism. In this work, we think that pre-defined graph or general learning method is difficult due to their irregular structure. Hence, we present CATN, an end-to-end model of Cross Attentive Tree-aware Network to jointly capture the inter-series correlation and intra-series temporal pattern. We first construct a tree structure to learn hierarchical and grouped correlation and design an embedding approach that can pass dynamic message to generalize implicit but interpretable cross features among multiple time series. Next in temporal aspect, we propose a multi-level dependency learning mechanism including global&local learning and cross attention mechanism, which can combine long-range dependencies, short-range dependencies as well as cross dependencies at different time steps. The extensive experiments on different datasets from real world show the effectiveness and robustness of the method we proposed when compared with existing state-of-the-art methods.

The widespread communication of academic ideas and research achievements through literature and media has generated massive academic big data. Analyzing such academic big data and discovering knowledge can discover comprehensive and predictive academic intelligence and provide corresponding services, which are valuable for scholars, journals, institutions and governments for career assessment, topic selection, funding management and resource allocation. This paper proposes a heterogeneous knowledge-learning method for understanding and predicting academic impact in the transportation field. The proposed method is illustrated on the academic big data collected from the papers published in 34 transportation journals from 2008 to 2018. We extract four types of features including bibliometric, altmetric, network and semantic features, and build hybrid feature embedding via TransR and Doc2vec that involving domain knowledge. Further, an academic impact prediction model for articles named as Hy-LSTM-Att is proposed, which weighs the hybrid features by the attention mechanism and predicts academic impact with the bi-LSTM recurrent neural networks. Experimental results demonstrate that the proposed Hy-LSTM-Att model outperforms competing shallow and deep learning models. Additionally, the feature ablation experiments illustrate that the four types of features positively contribute to the performance of impact prediction.

In recommendation, both stationary and dynamic user preferences on items are embedded in the interactions between users and items (e.g., rating or clicking) within their contexts. Sequential recommender systems (SRSs) need to jointly involve such context-aware user-item interactions in terms of the couplings between the user and item features and sequential user actions on items over time. However, such joint modeling is non-trivial and significantly challenges the existing work on preference modeling, which either only models user-item interactions by latent factorization models but ignores user preference dynamics or only captures sequential user action patterns without involving user/item features and context factors and their coupling and influence on user actions. We propose a neural time-aware recommendation network (TARN) with a temporal context to jointly model 1) stationary user preferences by a feature interaction network and 2) user preference dynamics by a tailored convolutional network. The feature interaction network factorizes the pairwise couplings between non-zero features of users, items, and temporal context by the inner product of their feature embeddings while alleviating data sparsity issues. In the convolutional network, we introduce a convolutional layer with multiple filter widths to capture multi-fold sequential patterns, where an attentive average pooling (AAP) obtains significant and large-span feature combinations. To learn the preference dynamics, a novel temporal action embedding represents user actions by incorporating the embeddings of items and temporal context as the inputs of the convolutional network. The experiments on typical public data sets demonstrate that TARN outperforms state-of-the-art methods and show the necessity and contribution of involving time-aware preference dynamics and explicit user/item feature couplings in modeling and interpreting evolving user preferences.

Urban transportation systems are shaped by factors that include people, vehicles, roads, and the environment, forming a complex and giant system with dynamics, diversity, and uncertainty. Physical signal-driven intelligent transportation systems (ITSs) typically lack the ability to capture social behaviors or crowd willingness, and they achieve only information automation for transportation decision support. The crowdsourcing social signals consist of timely, extensive, comprehensive, and rich intelligence that concern urban dynamics, social behaviors, and traffic environments. Such social signals provide a new paradigm for operating ITS with unstructured semantic data, making knowledge automation for decision intelligence a possibility. This article reviews the knowledge automation paradigms for cyber-physical-social systems (CPSSs) compared with traditional information automation paradigms for cyber-physical systems (CPSs) in ITS, from the perspective of data-driven, modeling space, analytical methodologies, and decision support services. To investigate the key methodology in social spaces that enhance information automation into knowledge automation, we summarize the current research into a multisource heterogeneous social signal-based traffic decision knowledge automation framework and further exploit the computational paradigm and applications scenarios of this framework. Finally, we discuss future challenges for designing and realizing knowledge automation on CPSS in transportation.

Effective connectivity analysis has been extensively used in recent years by scientists studying brain functional networks because it can illuminate the brain's mechanisms and functional logic. Dynamic causal modeling is the most widely used method for effective connectivity analysis, and it is often performed with statistical parametric mapping, the most popular statistical analysis procedure for brain function data. To solve problems related to the complexity of statistical parametric mapping and the limitations of existing meta-analytic databases, we designed an auxiliary system for calculating effective brain connectivity based on co-activation networks. This system improves the efficiency of dynamic causal modeling calculations and simplifies the pre-processing steps of statistical parametric mapping. To improve the execution efficiency of dynamic causal modeling, we used connectivity information for the co-activation network as a priori knowledge to assist the selection of regions of interest based on our previously proposed co-activation network method.During dynamic causal modeling calculations, we decreased the dimensionality and runtime of the analysis by applying association rule algorithms to exclude some impossible connectivity relationships in the co-activation network. We automatically extracted information from tables to establish an a priori knowledge database on published literature for greater convenience in meta-analyses and co-activation network mining. We decreased the risk of human errors in statistical parametric mapping by optimizing its preprocessing steps to modularize its processes and by simplifying unnecessary parameter settings. We have verified that our system increases the efficiency and accuracy of dynamic causal modeling. This will greatly aid both new and experienced users.

Video captioning is a fundamental task for visual understanding. Previous works employ end-to-end networks to learn from the low-level vision feature and generate descriptive captions, which are hard to recognize fine-grained objects and lacks the understanding of crucial semantic concepts. According to DPC [19], these concepts generally present in the narrative transcripts of the instructional videos. The incorporation of transcript and video can improve the captioning performance. However, DPC directly concatenates the embedding of transcript with video features, which is incapable of fusing language and vision features effectively and leads to the temporal mis-alignment between transcript and video. This motivates us to 1) learn the semantic concepts explicitly and 2) design a temporal alignment mechanism to better align the video and transcript for the captioning task. In this paper, we start with an encoder-decoder backbone using transformer models. Firstly, we design a semantic concept prediction module as a multi-task to train the encoder in a supervised way. Then, we develop an attention based cross-modality temporal alignment method that combines the sequential video frames and transcript sentences. Finally, we adopt a copy mechanism to enable the decoder(generation) module to copy important concepts from source transcript directly. The extensive experimental results demonstrate the effectiveness of our model, which achieves state-of-the-art results on YouCookII dataset.

Complementarity between activities reveals that doing any one of them increases the returns to doing the others. In other words, complementarity leads to the synergistic effect that the whole is greater than the sum of its parts. Identifying and exploiting complementarity can benefit many cybernetic activities, where human-machine interactions are inherent and dominant. One such activity is requirements tracing that helps stakeholders to track the status of their goals. Although various kinds of support for human analysts in requirements tracing have been proposed, little is known about the nature of complementarity when different tracing practices are involved. In this paper, we explore the role of complementarity by considering together the tagging-to-trace (T2T) and learning-to-trace (L2T) activities. We present a novel approach to examining which T2T and L2T practices enhance the qualities of each other. Our approach also uncovers the environmental factors which the complementarity is sensitive to. Applying our approach to the logs of 140 analyst-tracing units offers operational insights into the rigorous detection of complementarity and shows the importance of understanding the cybernetic conditions under which the requirements tracing practices may in fact be complementary.

In e-learning recommender systems, interpersonal information between learners is very scarce, which makes it difficult to apply collaborative filtering (CF) techniques to achieve recommendations. In this study, we propose a hybrid filtering recommendation approach (SI - IFL) combining learner influence model (LIM), self-organization based (SOB) recommendation strategy, and sequential pattern mining (SPM) together for recommending learning objects (LOs) to learners. The method works as follows: (i) LIM is applied to acquire the interpersonal information by computing the influence that a learner exerts on others. LIM consists of learner similarity, knowledge credibility, and learner aggregation, meanwhile, LIM is independent of ratings. Furthermore, to address the uncertainty and fuzzy natures of learners, intuitionistic fuzzy logic (IFL) is applied to optimize the LIM. (ii) A SOB recommendation strategy is applied to recommend the optimal learner cliques for active learners by simulating the influence propagation among learners. Influence propagation means that a learner can move towards active learners, and such behaviors can stimulate the moving behaviors of his/her neighbors. This SOB recommendation approach achieves a stable structure based on distributed and bottom-up behaviors of individuals. (iii) SPM is applied to decide the final learning objects (LOs) and navigational paths based on the recommended learner cliques. The experimental results demonstrate that SI - IFL can provide personalized and diversified recommendations, and it shows promising efficiency and adaptability in e-learning scenarios.

Understanding narrated instructional videos is important for both research and real-world web applications.Motivated by video dense captioning, we propose a model to generate procedure captions from narrated instructional videos which are a sequence of stepwise clips with description.Previous works on video dense captioning learn video segments and generate captions without considering transcripts.We argue that transcripts in narrated instructional videos can enhance video representation by providing fine-grained complimentary and semantic textual information.In this paper, we introduce a framework to ( 1) extract procedures by a cross-modality module, which fuses video content with the entire transcript; and (2) generate captions by encoding video frames as well as a snippet of transcripts within each extracted procedure.Experiments show that our model can achieve state-of-the-art performance in procedure extraction and captioning, and the ablation studies demonstrate that both the video frames and the transcripts are important for the task.

Image-text matching is a vital cross-modality task in artificial intelligence and has attracted increasing attention in recent years. Existing works have shown that learning semantic concepts is useful to enhance image representation and can significantly improve the performance of both image-to-text and text-to-image retrieval. However, existing models simply detect semantic concepts from a given image, which are less likely to deal with long-tail and occlusion concepts. Frequently co-occurred concepts in the same scene, e.g. bedroom and bed, can provide common-sense knowledge to discover other semantic-related concepts. In this paper, we develop a Scene Concept Graph (SCG) by aggregating image scene graphs and extracting frequently co-occurred concept pairs as scene common-sense knowledge. Moreover, we propose a novel model to incorporate this knowledge to improve image-text matching. Specifically, semantic concepts are detected from images and then expanded by the SCG. After learning to select relevant contextual concepts, we fuse their representations with the image embedding feature to feed into the matching module. Extensive experiments are conducted on Flickr30K and MSCOCO datasets, and prove that our model achieves state-of-the-art results due to the effectiveness of incorporating the external SCG.

Electronic health records contain a wealth of information on a patients healthcare over many visits, such as diagnoses, treatments, drugs administered, and so on. The untapped potential of these data in healthcare analytics is vast. However, given that much of medical information is a cause and effect science, new embedding methods are required to ensure the learning representations reflect the comprehensive interplays between medical concepts and their relationships over time. Unlike one-hot encoding, a distributed representation should preserve these complex interactions as high-quality inputs for machine learning-based healthcare analytics tasks. Therefore, we propose a novel attentive dual embedding method called MC2Vec. MC2Vec captures the proximity relationships between medical concepts through a two-step optimization framework that recursively refines the embedding for superior output. The framework comprises a Skip-gram model to generate the initial embedding and an attentive CBOW model to fine-tune the embedding with temporal information gleaned from sequences of patient visits. Experiments with two public datasets demonstrate that MC2Vecs produces embeddings of higher quality than five state-of-the-art methods.

During software evolution, developers often receive new requirements expressed as feature requests. To implement the requested features, developers have to perform necessary modifications (refactorings) to prepare for new adaptation that accommodates the new requirements. Software refactoring is a well-known technique that has been extensively used to improve software quality such as maintainability and extensibility. However, it is often challenging to determine which kind of refactorings should be applied. Consequently, several approaches based on various heuristics have been proposed to recommend refactorings. However, there is still lack of automated support to recommend refactorings given a feature request. To this end, in this paper, we propose a novel approach that recommends refactorings based on the history of the previously requested features and applied refactorings. First, we exploit the stateof-the-art refactoring detection tools to identify the previous refactorings applied to implement the past feature requests. Second, we train a machine classifier with the history data of the feature requests and refactorings applied on the commits that implemented the corresponding feature requests. The machine classifier is then used to predict refactorings for new feature requests. We evaluate the proposed approach on the dataset of 43 open source Java projects and the results suggest that the proposed approach can accurately recommend refactorings (average precision 73%).

The structured case representation improves case-based reasoning (CBR) by exploring structures in the case base and the relevance of case structures. Recent CBR classifiers have mostly been built upon the attribute-value case representation rather than structured case representation, in which the structural relations embodied in their representation structure are accordingly overlooked in improving the similarity measure. This results in retrieval inefficiency and limitations on the performance of CBR classifiers. This paper proposes a hierarchical case-based classifier, HCBC, which introduces a concept lattice to hierarchically organize cases. By exploiting structural case relations in the concept lattice, a novel dynamic weighting model is proposed to enhance the concept similarity measure. Based on this similarity measure, HCBC retrieves the top-K concepts that are most similar to a new case by using a bottom-up pruning-based recursive retrieval (PRR) algorithm. The concepts extracted in this way are applied to suggest a class label for the case by a weighted majority voting. Experimental results show that HCBC outperforms other classifiers in terms of classification performance and robustness on categorical data, and also works confidently well on numeric datasets. In addition, PRR effectively reduces the search space and greatly improves the retrieval efficiency of HCBC.

Requirements traceability provides critical support throughout all phases of software engineering. Automated tracing based on information retrieval (IR) reduces the effort required to perform a manual trace. Unfortunately, IR-based trace recovery suffers from low precision due to polysemy, which refers to the coexistence of multiple meanings for a term appearing in different requirements. Latent semantic indexing (LSI) has been introduced as a method to tackle polysemy, as well as synonymy. However, little is known about the scope and significance of polysemous terms in requirements tracing. While quantifying the effect, we present a novel method based on artificial neural networks (ANN) to enhance the capability of automatically resolving polysemous terms. The core idea is to build an ANN model which leverages a term's highest-scoring coreferences in different requirements to learn whether this term has the same meaning in those requirements. Experimental results based on 2 benchmark datasets and 6 long-lived open-source software projects show that our approach outperforms LSI on identifying polysemous terms and hence increasing the precision of automated tracing.

Automated classification of citation function in scientific text is a new emerging research topic inspired by traditional citation analysis in applied linguistic and scientometric fields.The aim is to classify citations in scholarly publication in order to identify author's purpose or motivation for quoting or citing a particular paper.Several citation schemes have been proposed to classify the citations into different functions.However, it is extremely challenging to find standard scheme to classify citations, and some of the proposed schemes have similar functions.Moreover, most of previous studies mainly used classical machine learning methods such as support vector machine and neural networks with a number of manually created features.These features are incomplete and suffer from time-consuming and error prone weakness.To address these problems, we present a new citation scheme with less functions and propose a deep learning model for classification.The citation sentences and author's information were fed to convolutional neural networks to build citation and author representations.A corpus was built using the proposed scheme and a number of experiments were carried out to assess the model.Experimental results have shown that the proposed approach outperforms the existing methods in term of accuracy, precision and recall.

In many critical industrial information systems, tracking a dependability requirement is instrumental to the verification and validation (V&V) of security, privacy, and other dependability concerns. Automated traceability tools employ information retrieval methods to recover candidate links, which saves much manual effort. Integrating relevance feedback (RF) could potentially improve the retrieval effectiveness by soliciting the relevance judgments on a subset of the retrieval results and then incorporating the feedback into subsequent retrieval. However, little is known about how to use RF to trace dependability requirements. In this paper, we propose a novel term-based RF algorithm that leverages the term usage context to recommend positive and negative feedback. Experiments on two software datasets show that our algorithm significantly outperforms the contemporary link-based RF tracing method. Our work not only contributes a new solution to dependability requirements' V&V, but also enables further automation to reduce the manual effort in the development life cycle of dependable industrial systems.

Currently, dynamic causal modeling (DCM) is one of the most widely used models for an effective brain connectivity network, but it also has some disadvantages (e.g., researchers' selection of cerebral regions of interest [ROIs] is subjective, a substantial time is required for computation, etc.). Statistical Parametric Mapping (SPM) is the most popular statistical data analysis software for brain function, but its settings cumbersome, especially the data preprocessing section. In response to these disadvantages of DCM and SPM, we designed and created a computer-aided system for an effective brain connectivity network, modularized the data preprocessing section of SPM, and we explored the cerebral ROIs and possible co-activation network based on our proposed approach. The co-activation network has as a prior interconnection relationship, and it is used to assist in the selection of ROIs in similar cognitive experiments; thus, the testing of meaningless noise connection modes by the DCM is prevented, the number of models DMC is decreased, and the accuracy of the conclusions and computational efficiency of the DCM are improved.

We propose a novel Hierarchical CBC model (HCBC) based on Formal Concept Analysis (FCA). Firstly, Concept Lattice (CL), the hierarchical and conceptual structure in FCA, is adopted to represent cases. Thus a novel dynamic weight model is proposed from CL to measure similarities between cases and concepts. Then the similarity metric is applied to retrieve the top-K similar concepts which are used to vote for adaptive solutions for new cases by majority voting in case adaption. Experiments show our model shows good performance in terms of accuracy and outperforms the other classification methods.

Many efficient deployments of large-scale wireless sensor networks based on the tree-based community rise into view recently. Sensor nodes are severely resource constrained, and lack sophisticated defense mechanisms to fight virus attacks. Cyber viruses spread through node populations over the networks, and a number of results about the prevalence have been derived in recent years by exploiting epidemic behaviors and the percolation processes on networks. A network model based on the Cayley tree is proposed to depict the underlying tree-based architectures of the network and the community. The percolation thresholds are calculated and analyzed in two cases. Due to random links in the communities, the sensor virus extends drastically on the network. The analysis and evaluation shows that the percolation threshold keeps decreasing with the increase of the shortcut probability. There is the smallest percolation threshold in a random network, where the virus easily attacks the network from one side to another. The conclusions can further our understanding of epidemic dynamics on tree-based communities of wireless sensor networks.

Developers often spend valuable time navigating and seeking relevant code in software maintenance. Currently, there is a lack of theoretical foundations to guide tool design and evaluation to best shape the code base to developers. This paper contributes a unified code navigation theory in light of the optimal food-foraging principles. We further develop a novel framework for automatically assessing the foraging mechanisms in the context of program investigation. We use the framework to examine to what extent the clustering of software entities affects code foraging. Our quantitative analysis of long-lived open-source projects suggests that clustering enriches the software environment and improves foraging efficiency. Our qualitative inquiry reveals concrete insights into real developer's behavior. Our research opens the avenue toward building a new set of ecologically valid code navigation tools.

Most code smell detection tools expose thresholds to engineers for customization because code smell detection is essentially subjective and application specific. Another reason why engineers should customize these thresholds is that they have different working schedules and different requirements on software quality. They have their own unique need on precision and recall in smell detection. This unique need should be fulfilled by adjusting thresholds of smell detection tools. However, it is difficult for software engineers, especially inexperienced ones, to adjust often contradicting and related thresholds manually. One of the possible reasons is that engineers do not know the exact quantitative relation between threshold values and performance, e.g., precision. In this paper, we propose an approach to adapting thresholds automatically and dynamically. Engineers set a target precision manually according to their working schedules and quality requirements. With feedback from engineers, the proposed approach then automatically searches for a threshold setting to maximize recall while having precision close to the target precision. The proposed approach has been evaluated on open-source applications. Evaluation results suggest that the proposed approach is effective.

Meta-analyses aim to mine reliable research findings from individual studies and can be usefully applied to an increasing number of published functional brain imaging studies. In this paper we use a new meta-analysis method based on the Apriori algorithm of association rules that can be used to find the possible connection network of the brain. We analyzed 13 studies that investigated how the brain processes Chinese phonology, compared the results of activation likelihood estimation (ALE) meta-analysis, and generated a potential core connection network model of the brain during this processing. The identified regions included the left inferior frontal gyrus (IFG), left middle frontal gyrus (MFG), bilateral supplementary motor areas (SMA), and left precentral gyrus. The described connection network model of brain regions was consistent with previous research. Furthermore, the network can facilitate better understanding of how the brain processes the Chinese language.

Software refactoring is to improve readability, maintainability and expansibility of software by adjusting its internal structure, whereas the external behaviors of software are not changed. Although software refactoring should not change the external behaviors of software systems, they might make a regression test case obsolete (with syntax and runtime errors) or fail. People have investigated which refactorings had an influence on regression test case. However, how test cases are influenced by refactorings and what kind of errors might be introduced remain unknown. To this end, in this paper, we proposed an approach to analyze refactorings' impact on regression test cases. On one hand, we analyzed why regression test cases failed. On the other hand, we analyzed the influence of refactorings on software interfaces. Based on the analysis, we built up a mapping between refactorings and test case failure. Such a mapping can be used to guide test case repair automation where test cases are made obsolete by refactorings. The approach was evaluated on five open-source applications. Evaluation results suggest that the precision of the approach is greater than 80%.

With the development of computer science and multimedia technology, Web-based learning becomes increasingly popular. User evaluation plays a significant role in the process of guided learning. In recent years, there is great progress in the development of evaluation technology. However, few evaluation methods fully take online learning activity analysis and individual differences into account. This paper proposes a personalized user evaluation model for Web-based learning systems. The model is utilized to record and analyze various learning activities throughout the entire learning process. Considering individual differences, learners are clustered and specific evaluation standards are set for different clusters. Comprehensive evaluation is achieved by combining Analytic Hierarchy Process, Fuzzy C-Means clustering and normalization algorithm. Through the comparison with several other common evaluation methods, experimental results show that the proposed method outperforms existing ones on the accuracy of learner evaluation.

Just-in-time requirements management, characterized by lightweight representation and continuous refinement of requirements, fits many iterative and incremental development projects. Being lightweight and flexible, however, can cause wasteful and procrastinated implementation, leaving certain stakeholder goals not satisfied. This paper proposes traceability-enabled refactoring aimed at fulfilling more requirements fully. We make a novel use of requirements traceability to accurately locate where the software should be refactored, and develop a new scheme to precisely determine what refactorings should be applied to the identified places. Our approach is evaluated through an industrial study. The results show that our approach recommends refactorings more appropriately than a contemporary recommender.

Recommender systems are often based on collaborative filtering. Previous researches on collaborative filtering mainly focus on one single recommender or formulating hybrid with different approaches. In consideration of the problems of sparsity, recommender error rate, sample weight update, and potential, we adapt AdaBoost and propose two novel boosting frameworks for collaborative filtering. Each of the frameworks combines multiple homogeneous recommenders, which are based on the same collaborative filtering algorithm with different sample weights. We use seven popular collaborative filtering algorithms to evaluate the two frameworks with two MovieLens datasets of different scale. Experimental result shows the proposed frameworks improve the performance of collaborative filtering.

Software refactoring might be done in two different tactics. The first one is XP-style small-step refactoring, also called floss refactoring. The other tactic, called root canal refactoring, is to set aside an extended period specially for refactoring. Floss refactoring, as one of the corner stones of XP, is well acknowledged. In contrast, root canal refactoring is doubted, especially by XP advocators. Despite the doubts, however, no large scale empirical study on refactoring tactics has been reported. In contrast to the doubts, cases of root canal refactoring have been reported from industry, e.g., Microsoft. Researchers from academe have also proposed various approaches to facilitating root canal refactoring. To this end, this paper would investigate the following questions. (1)How often are the two different tactics employed, respectively? (2) Is there any correlation between refactoring tactics and categories of refactorings? In other words, are some kinds of refactorings more likely than others to be done as floss refactorings or root canal refactorings? To answer these questions, we analyze refactoring histories collected by Eclipse Usage Data Collector (UDC). The data are collected from 753,367 engineers worldwide. Analysis results suggest that about 11.5 percent of refactorings collected by UDC are root canal refactorings, whereas others (88.5 percent) are floss refactorings. We also find that some kinds of refactorings, e.g., Introduce Parameter, are more likely than others to be performed as root canal refactorings.

Bad smells are signs of potential problems in code. Detecting and resolving bad smells, however, remain time-consuming for software engineers despite proposals on bad smell detection and refactoring tools. Numerous bad smells have been recognized, yet the sequences in which the detection and resolution of different kinds of bad smells are performed are rarely discussed because software engineers do not know how to optimize sequences or determine the benefits of an optimal sequence. To this end, we propose a detection and resolution sequence for different kinds of bad smells to simplify their detection and resolution. We highlight the necessity of managing bad smell resolution sequences with a motivating example, and recommend a suitable sequence for commonly occurring bad smells. We evaluate this recommendation on two nontrivial open source applications, and the evaluation results suggest that a significant reduction in effort ranging from 17.64 to 20 percent can be achieved when bad smells are detected and resolved using the proposed sequence.

Opinion mining is a challenging task to identify the opinions or sentiments underlying user generated contents, such as online product reviews, blogs, discussion forums, etc. Previous studies that adopt machine learning algorithms mainly focus on designing effective features for this complex task. This paper presents our approach based on tree kernels for opinion mining of online product reviews. Tree kernels alleviate the complexity of feature selection and generate effective features to satisfy the special requirements in opinion mining. In this paper, we define several tree kernels for sentiment expression extraction and sentiment classification, which are subtasks of opinion mining. Our proposed tree kernels encode not only syntactic structure information, but also sentiment related information, such as sentiment boundary and sentiment polarity, which are important features to opinion mining. Experimental results on a benchmark data set indicate that tree kernels can significantly improve the performance of both sentiment expression extraction and sentiment classification. Besides, a linear combination of our proposed tree kernels and traditional feature vector kernel achieves the best performances using the benchmark data set.

Long and complex methods are hard to read or maintain, and thus usually treated as bad smells, known as Long Method. On the contrary, short and well-named methods are much easier to read, maintain, and extend. In order to divide long methods into short ones, refactoring Extract Method was proposed and has been widely used. However, extracting methods manually is time consuming and error prone. Though existing refactoring tools can automatically extract a selected fragment from its inclosing method, which fragment within a long method should be extracted has to be determined manually. In order to facilitate the decision-making, we propose an approach to recommend fragments within long methods for extraction. The approach is implemented as a prototype, called AutoMeD. With the tool, we evaluate the approach on a nontrivial open source project. The evaluation results suggest that refactoring cost of long methods can be reduced by nearly 40%. The main contribution of this paper is an approach to recommending fragments within long methods to be extracted, as well as an initial evaluation of the approach.

Searching in unstructured peer-to-peer network with scalability and effectiveness is still a challenging problem. Recently in many applications, one-hop index replication becomes a fundamental technique for improving the search performance. However, when we extend one-hop index replication to multiple hops for locating both popular contents and rare objects with high lookup success rate, index storage cost and index data transfer overload become serious obstacles. In this paper, we propose a novel technique to exploit the advantage of multi-hop index replication and significantly reduce the cost of index storage. Our approach is to separate the peers and contents into groups and associating the nodes with shared files in group space. Each node works as a delegate node to storing the same group file indices which come from its neighbors. Search process is restricted in a group of peers according to the query's hash value. We study the search performance and index storage cost of random walk with group based index replication (GBIR) through theoretical model and give the mathematical relationship between GBIR and random walk with normal index replication (IR). We further evaluate the GBIR technique by simulator-based experiments, and the results show that GBIR can achieve the similar lookup success rate to IR and reduce the index storage cost to only 1/g of IR's (g is the number of groups in the system). Finally, we significantly improve the GBIR's search performance using GBIR with the highest degree neighbor routing (GBIR+).

In this paper, a novel similarity measuring method based on Rough Formal Concept Analysis (RFCA) and information content similarity(ICS) is proposed which evaluates the similarity degree between the concepts. We use the information content approach to automatically obtain part of similarity scores of two concepts which makes up the normal featural and structural evaluating model. Thus the similarity of two concepts can be directly calculated from the lower object approximations and lower attribute approximations based on the RFCA and ICS. Consequently the proposed method combines semantic, featural and structural information into decision which can be viewed as the development of Tverskyąrs similarity model.

Bad smell is a key concept in software refactoring. We have a bunch of bad smells, refactoring rules, and refactoring tools, but we do not know which kind of bad smells should be resolved first. The resolution of one kind of bad smells may have impact on the resolution of other bad smells. Con-sequently, different resolution orders of the same set of bad smells may require different effort, and/or lead to differ-ent quality improvement. In order to ease the work and maximize the effect of refactoring, we try to analyze the re-lationships among different kinds of bad smells, and their impact on resolution orders of these bad smells. With the analysis, we recommend a resolution order of common bad smells. The main contribution of this paper is to motivate the necessity to arrange resolution orders of bad smells, and recommend a resolution order of common bad smells.

Due to the heightened concern about bioterrorism and emerging/reemerging infectious diseases, a flood of molecular data about human pathogens has been generated and maintained in disparate databases. However, scientific findings regarding these pathogens and their host responses are buried in the growing volume of biomedical literature and there is an urgent need to mine information pertaining to pathogenesis-related proteins especially host-pathogen protein-protein interactions from literature. In this paper, we report our exploration of developing an automated system to identify MEDLINE abstracts referring to host-pathogen protein-protein interactions. An annotated corpus consisting of 1,360 MEDLINE abstracts was generated. With this corpus, we developed and evaluated document classification systems using support vector machines (SVMs). We also investigated the effects of feature selection using the information gain (IG) measure. Document classification systems were designed at two levels, abstract-level and sentence-level. We observed that feature selection was effective not only in reducing the dimensionality of features to build a compact system, but also in improving document classification performance. We also observed abstract-level systems and sentence-level systems yielded different classification of MEDLINE abstracts, and the combination of these systems could improve the overall document classification.

Digital Libraries have emerged as an important research and application field, and been facilitated by advances in information technology, such as the World Wide Web. The scale and diversity of documents and information of Digital Libraries are immense, which are often stored with different formats. Therefore research in Digital Libraries has generally focused on transparent access to heterogeneous data sources. For the enormous, heterogeneous, dynamic characteristic of Digital Libraries resource, the integrated retrieval system framework of heterogeneous information in Digital Libraries was designed in the paper with the structure of wrapper and mediator. It includes user interface agent, wrapper agent. Mediator agent, knowledge retrieval agent and knowledge base search agent. The agents communicate each other and cooperate together to carry out integrated retrieval function of heterogeneous information in Digital libraries.