Positron Emission Tomography (PET) is a powerful imaging modality for assessing biochemical processes within the body. However, accurate image reconstruction is challenged by photon attenuation, particularly in dense structures such as bones, leading to quantification errors and reduced diagnostic confidence. Computed Tomography (CT) based attenuation correction is the standard approach but introduces additional radiation exposure, longer imaging times, and patient inconvenience, as well as potential registration errors, motion artifacts, and energy scaling inaccuracies. In this study, we propose a 3D U-Net based deep learning framework that directly estimates attenuation information from PET sinograms, eliminating the need for additional imaging modalities. Our approach integrates PET physics and employs custom skip connections to enhance cross-domain learning. We evaluate our model on a simulated brain dataset derived from real patient templates, achieving a Dice coefficient of 0.650 and an accuracy of 0.486 for bone structures. The clinical applicability of our method is further assessed by reconstructing PET images with the generated attenuation maps, yielding an MSE of 0.007 and an SSIM of 0.956, demonstrating strong structural consistency with CT-based attenuation correction. These results highlight the feasibility of performing PET image attenuation correction using PET sinograms alone, offering a promising alternative that reduces imaging time, radiation exposure, and patient burden while enabling faster and more efficient PET reconstruction.

For an agent in a multi-agent environment, it is often beneficial to be able to predict what other agents will do next when deciding how to act. Previous work in multi-agent intention scheduling assumes a priori knowledge of the current goals of other agents. In this paper, we present a new approach to multi-agent intention scheduling in which an agent uses online goal recognition to identify the goals currently being pursued by other agents while acting in pursuit of its own goals. We show how online goal recognition can be incorporated into an MCTS-based intention scheduler, and evaluate our approach in a range of scenarios. The results demonstrate that our approach can rapidly recognise the goals of other agents even when they are pursuing multiple goals concurrently, and has similar performance to agents which know the goals of other agents a priori.

Recent work in multi-agent intention scheduling has shown that enabling agents to predict the actions of other agents when choosing their own actions can be beneficial. However existing approaches to 'intention-aware' scheduling assume that the programs of other agents are known, or are "similar" to that of the agent making the prediction. While this assumption is reasonable in some circumstances, it is less plausible when the agents are not co-designed. In this paper, we present a new approach to multi-agent intention scheduling in which agents predict the actions of other agents based on a high-level specification of the tasks performed by an agent in the form of a reward machine (RM) rather than on its (assumed) program. We show how a reward machine can be used to generate tree and rollout policies for an MCTS-based scheduler. We evaluate our approach in a range of multi-agent environments, and show that RM-based scheduling out-performs previous intention-aware scheduling approaches in settings where agents are not co-designed

Story sifting (or story recognition) allows for the exploration of events, stories, and patterns that emerge from simulated storyworlds. The goal of this work is to reduce the authoring burden for creating sifting queries. In this paper, we use the event traces of simulated storyworlds to create Dynamic Character Networks that track the changing relationship scores between characters in a simulation. These networks allow for the fortunes between any two characters to be plotted against time as a story arc. Similarity scores between story arcs from the simulation and a user’s query arc can be calculated using the Dynamic Time Warping algorithm. Events corresponding to the story arc that best matches the query arc can then be returned to the user, thus providing an intuitive means for users to sift a variety of stories without coding a search query. These components are implemented in our experimental prototype ARC SIFT. The results of a user study support our expectation that ARC SIFT is an intuitive and accurate tool that allows human users to sift stories out from a larger chronicle of events emerging from a simulated story world.

We propose a new approach to intention progression in multi-agent settings where other agents are effectively black boxes. That is, while their goals are known, the precise programs used to achieve these goals are not known. In our approach, agents use an abstraction of their own program called a partially-ordered goal-plan tree (pGPT) to schedule their intentions and predict the actions of other agents. We show how a pGPT can be derived from the program of a BDI agent, and present an approach based on Monte Carlo Tree Search (MCTS) for scheduling an agent's intentions using pGPTs. We evaluate our pGPT-based approach in cooperative, selfish and adversarial multi-agent settings, and show that it out-performs MCTS-based scheduling where agents assume that other agents have the same program as themselves.

The increasing capabilities of autonomous systems offer the potential for more effective teaming with humans. Effective human/agent teaming is facilitated by a mutual understanding of the team objective and how that objective is decomposed into team roles. This paper presents a framework for engineering human/agent teams that delineates the key human/agent teaming components, using TDF-T diagrams to design the agents/teams and then present contextualised team cognition to the human team members at runtime. Our hypothesis is that this facilitates effective human/agent teaming by enhancing the human's understanding of their role in the team and their coordination requirements. To evaluate this hypothesis we conducted a study with human participants using our user interface for the StarCraft strategy game, which presents pertinent, instantiated TDF-T diagrams to the human at runtime. The performance of human participants in the study indicates that their ability to work in concert with the non-player characters in the game is significantly enhanced by the timely presentation of a diagrammatic representation of team cognition.

A key problem in Belief-Desire-Intention agents is how an agent progresses its intentions, i.e., which plans should be selected and how the execution of these plans should be interleaved so as to achieve the agent’s goals. Previous approaches to the intention progression problem assume the agent has perfect information about the state of the environment. However, in many real-world applications, an agent may be uncertain about whether an environment condition holds, and hence whether a particular plan is applicable or an action is executable. In this paper, we propose SAU, a Monte-Carlo Tree Search (MCTS)-based scheduler for intention progression problems where the agent’s beliefs are uncertain. We evaluate the performance of our approach experimentally by varying the degree of uncertainty in the agent’s beliefs. The results suggest that SAU is able to successfully achieve the agent’s goals even in settings where there is significant uncertainty in the agent’s beliefs.

Sparse reward games, such as the infamous Montezuma’s Revenge, pose a significant challenge for Reinforcement Learning (RL) agents. Hierarchical RL, which promotes efficient exploration via subgoals, has shown promise in these games. However, existing agents rely either on human domain knowledge or slow autonomous methods to derive suitable subgoals. In this work, we describe a new, autonomous approach for deriving subgoals from raw pixels that is more efficient than competing methods. We propose a novel intrinsic reward scheme for exploiting the derived subgoals, applying it to three Atari games with sparse rewards. Our agent’s performance is comparable to that of state-of-the-art methods, demonstrating the usefulness of the subgoals found.

Aside from hand-coded bots, most of the videogame agents rely on experience in one way or another. Some agents improve over time by adjusting to real experience, while others make projections by leveraging simulated experience. In one sense, human play seems to resemble simulation, in that human players often try to visualize possible trajectories before deciding on a real course of action. However, most of the existing simulation-based agents require precise knowledge of the game's code, and their search depth is limited by the granular time scales encountered in videogames. Human players, on the other hand, appear to visualize plans in terms of abstract “skills,” such as running and jumping. These skills are learned from real experience, so,in this sense, human play resembles learning-based approaches. Motivated by these observations, we propose an approach that bridges the gap between skills, which are uncertain in outcome and duration, and traditional simulation-based planning, which is discrete. We apply this approach to maze-like navigation problems in Infinite Mario. After an initial skill acquisition phase, our agent is capable of navigating new levels without further training and also scales better with goal distance than a granular simulation method that exploits exact knowledge of the game's physics.

Applications of reinforcement learning to continuous control tasks often rely on a steady, informative reward signal. In videogames, however, tasks may be far easier to specify through a binary reward that indicates success or failure. In the absence of a steady, guiding reward, the agent may struggle to explore efficiently, especially if effective exploration requires strong coordination between actions. In this paper, we show empirically that this issue may be mitigated by exploring over an abstract action set, using hierarchically composed parameterized skills. We experiment in two tasks with sparse rewards in a continuous control environment based on the arcade game Asteroids. Compared to a flat learner that explores symmetrically over low-level actions, our agent explores a greater variety of useful actions, and its long-term performance on both tasks is superior.

In platform videogames, players are frequently tasked with solving medium-term navigation problems in order to gather items or powerups. Artificial agents must generally obtain some form of direct experience before they can solve such tasks. Experience is gained either through training runs, or by exploiting knowledge of the game's physics to generate detailed simulations. Human players, on the other hand, seem to look ahead in high-level, abstract steps. Motivated by human play, we introduce an approach that leverages not only abstract "skills", but also knowledge of what those skills can and cannot achieve. We apply this approach to Infinite Mario, where despite facing randomly generated, maze-like levels, our agent is capable of deriving complex plans in real-time, without relying on perfect knowledge of the game's physics.

In this work we present a novel approach to check the consistency of agent designs (prior to any implementation) with respect to the requirements specifications via automated planning. This checking is essentially a search problem which makes planning technology an appropriate solution. We focus our work on BDI agent systems and the Prometheus design methodology in order to directly compare our approach to previous work. Our experiments in more than 16K random instances prove that the approach is more effective than previous ones proposed: it achieves higher coverage, lower run-time, and importantly, can handle loops in the agent detailed design and unbounded subgoal reasoning.

A key advantage the reactive planning approach adopted by BDI-based agents is the ability to recover from plan execution failures, and almost all BDI agent programming languages and platforms provide some form of failure handling mechanism. In general, these consist of simply choosing an alternative plan for the failed subgoal (e.g., JACK, Jadex). In this paper, we propose an alternative approach to recovering from execution failures that relies on exploiting positive interactions between an agent's intentions. A positive interaction occurs when the execution of an action in one intention assists the execution of actions in other intentions (e.g., by (re)establishing their preconditions). We have implemented our approach in a scheduling algorithm for BDI agents which we call SP. The results of a preliminary empirical evaluation of SP suggest our approach out-performs existing failure handling mechanisms used by state-of-the-art BDI languages. Moreover, the computational overhead of SP is modest.

Intelligent agent systems built using the BDI model of agency have grown in popularity for implementing complex systems such as UAVs, military simulations, trading agents and intelligent games. The robust and flexible behaviours that these systems afford also makes testing the ‘correctness’ of these systems a non-trivial task. Whilst the main focus on existing work has been on checking the correctness of agent-programs, in this work we present an approach to formally verify agent-based designs for a particular BDI agent design methodology. The focus is on verifying whether the detailed design of the agents conform to the requirements specification. We present a sound and complete approach, formally verifiable properties, and an evaluation with respect to time and effectiveness.

No description available

Measures of semantic relatedness have been used in a variety of applications in information retrieval and language technology, such as measuring document similarity and cohesion of text. Definitions of such measures have ranged from using distance‐based calculations over WordNet or other taxonomies to statistical distributional metrics over document collections such as Wikipedia or the Web. Existing measures do not explicitly consider the domain associations of terms when calculating relatedness: This article demonstrates that domain matters. We construct a data set of pairs of terms with associated domain information and extract pairs that are scored nearly identical by a sample of existing semantic‐relatedness measures. We show that human judgments reliably score those pairs containing terms from the same domain as significantly more related than cross‐domain pairs, even though the semantic‐relatedness measures assign the pairs similar scores. We provide further evidence for this result using a machine learning setting by demonstrating that domain is an informative feature when learning a metric. We conclude that existing relatedness measures do not account for domain in the same way or to the same extent as do human judges.

Agent systems are used for a wide range of applications, and techniques to detect and avoid defects in such systems are valuable. In particular, it is desirable to detect issues as early as possible in the software development lifecycle. We describe a technique for checking the plan structures of a BDI agent design against the requirements models, specified in terms of scenarios and goals. This approach is applicable at design time, not requiring source code. A lightweight evaluation demonstrates that a range of defects can be found using this technique.

Given the current set of intentions an autonomous agent may have, intention selection is the agent's decision which intention it should focus on next. Often, in the presence of conflicts, the agent has to choose between multiple intentions. One factor that may play a role in this deliberation is the level of completeness of the intentions. To that end, this paper provides pragmatic but principled mechanisms for quantifying the level of completeness of goals in a BDI-style agent. Our approach leverages previous work on resource and effects summarization but we go beyond by accommodating both dynamic resource summaries and goal effects, while also allowing a non-binary quantification of goal completeness. We demonstrate the computational approach on an autonomous robot case study.

Intelligent agent systems are often used to implement complex software systems. A key aspect of agent systems is goals: a programmer or user defines a set of goals for an agent, and then the agent is left to determine how best to satisfy the goals assigned to it. Such goals include both achievement goals and maintenance goals. An achievement goal describes a particular state the agent would like to become true, such as being in a particular location or having a particular bank balance. A maintenance goal specifies a condition which is to be kept satisfied, such as ensuring that a vehicle stays below a certain speed, or that it has sufficient fuel. Current agent systems usually only utilize reactive maintenance goals, in that the agent only takes action after the maintenance condition has been violated. In this paper, we discuss methods by which maintenance goals can be made proactive, i.e., acting before a maintenance condition is violated, having predicted that the maintenance condition will be violated in the future. We provide a representation of proactive maintenance goals, reasoning algorithms, an operational semantics that realizes these algorithms and an experimental evaluation of our approach.

Software testing remains the most widely used approach to verification in industry today, consuming between 30-50 percent of the entire development cost. Test input selection for intelligent agents presents a problem due to the very fact that the agents are intended to operate robustly under conditions which developers did not consider and would therefore be unlikely to test. Using methods to automatically generate and execute tests is one way to provide coverage of many conditions without significantly increasing cost. However, one problem using automatic generation and execution of tests is the oracle problem: How can we automatically decide if observed program behavior is correct with respect to its specification? In this paper, we present a model-based oracle generation method for unit testing belief-desire-intention agents. We develop a fault model based on the features of the core units to capture the types of faults that may be encountered and define how to automatically generate a partial, passive oracle from the agent design models. We evaluate both the fault model and the oracle generation by testing 14 agent systems. Over 400 issues were raised, and these were analyzed to ascertain whether they represented genuine faults or were false positives. We found that over 70 percent of issues raised were indicative of problems in either the design or the code. Of the 19 checks performed by our oracle, faults were found by all but 5 of these checks. We also found that 8 out the 11 fault types identified in our fault model exhibited at least one fault. The evaluation indicates that the fault model is a productive conceptualization of the problems to be expected in agent unit testing and that the oracle is able to find a substantial number of such faults with relatively small overhead in terms of false positives.

One of the biggest bottlenecks for conversational systems is large-scale provision of suitable content. Our approach readily provides this without the need for custom-crafting. In this demonstration, we present the use of question-answer (QA) pairs mined from online question-and-answer websites to construct system utterances for a conversational agent. Our system uses QA pairs to formulate utterances that drive a conversation in addition to the answering of user questions as has been done in previous work. We use a collection of strategies that specify how and when the different parts of our question-answer pairs can be used and augmented with a small number of generic hand-crafted text snippets to generate natural and coherent system utterances.

A conversational agent for entertainment and engagement requires the ability to maintain coherent conversations. We describe the use of semantic relatedness to select the next conversational fragment that an agent utters, to maximise dialogue coherence or to possibly suggest new directions for a dialogue. We compare our approach, using a specfic semantic relatedness metric, to an existing nearest-context mechanism based on T F x IDF for selecting fragments to continue a conversation. Evaluation with human judges shows that use of semantic relatedness provides improved coherence across a sample collection of generated conversations.

Studies have shown that natural language interfaces such as question answering and conversational systems allow information to be accessed and understood more easily by users who are unfamiliar with the nuances of the delivery mechanisms (e.g., keyword‐based search engines) or have limited literacy in certain domains (e.g., unable to comprehend health‐related content due to terminology barrier). In particular, the increasing use of the web for health information prompts us to reexamine our existing delivery mechanisms. We present enquireMe, which is a contextual question answering system that provides lay users with the ability to obtain responses about a wide range of health topics by vaguely expressing at the start and gradually refining their information needs over the course of an interaction session using natural language. enquireMe allows the users to engage in “conversations” about their health concerns, a process that can be therapeutic in itself. The system uses community‐driven question–answer pairs from the web together with a decay model to deliver the top scoring answers as responses to the users' unrestricted inputs. We evaluated enquireMe using benchmark data from WebMD and TREC to assess the accuracy of system‐generated answers. Despite the absence of complex knowledge acquisition and deep language processing, enquireMe is comparable to the state‐of‐the‐art question answering systems such as START as well as those interactive systems from TREC.

More and more people are turning to the World Wide Web for learning and sharing information about their health using search engines, forums and question answering systems. In this demonstration, we look at a new way of delivering health information to the end-users via coherent conversations. The proposed conversational system allows the end-users to vaguely express and gradually refine their information needs using only natural language questions or statements as input. We provide example scenarios in this demonstration to illustrate the inadequacies of current delivery mechanisms and highlight the innovative aspects of the proposed conversational system.

Agent systems based on the BDI paradigm need to make decisions about which plans are used to achieve their goals. Usually the choice of which plan to use to achieve a particular goal is left up to the system to determine. In this paper we show how preferences, which can be set by the user of the system, can be incorporated into the BDI execution process and used to guide the choices made.

Introduction. Deliberation over courses of action to pursue is fundamental to agent systems. Agents designed to work in dynamic environments, such as a rescue robot or an online travel agent, must be able to reason about what actions they should take, incorporating deliberation into their execution cycle, reviewing decisions and taking corrective action with appropriate focus and frequency. Not only must agents reason about the effects of their courses of action, they must also consider the semantics of these corrective actions. Systems based on the well-known Belief-Desire-Intention (BDI) framework most often ascribe a set of goals to the agent, which is equipped with various techniques to deliberate over and manage this set. The centrality of reasoning over goals is seen in the techniques investigated in the literature, which include subgoaling and plan selection, detection and resolution of conflicts or opportunities for cooperation [9], checking goal properties to specification [10, 5], failure recovery and planning [2], and dropping, aborting, or suspending and resuming goals [7]. A variety of goals are described in the literature, including goals of performance of a task, achievement of a state, querying truth of a statement, testing veracity of beliefs, and maintenance of a condition [1, 11]. An agent must manage such a variety of goals, while incorporating pertinent sources of information into its decisions over them, such as preferences, quality goals, motivational goals, and advice [10]. The complexity of agent goal management stems from this combination of the variety of goals and the breadth of deliberation considerations. It is furthered because each goal can be dropped, aborted, suspended, or resumed (as illustrated in Figure 1) at arbitrary times. While goals themselves are static (i.e., they are specified at design time, and do not change during execution), their behaviour is dynamic: a goal may undergo a variety of changes of state during its execution cycle [5]. This evolution may include its initial adoption by the agent, being actively pursued, being suspended and then later resumed, and eventually succeeding (or failing). (Maintenance goals have a subtle life-cycle: the goal is retained even when the desired property is true; it is possible that such goals are never dropped.) Our work analyzes the behaviour of the above types of goals, including the behaviour when goals are aborted or suspended. We consider the complete life-cycle of goals, from their initial adoption by the agent to the time when they are no longer of interest, and all stages in between; we account for the dynamics of plan execution and sub-goaling. We develop a generic framework for goal states and transitions that captures the life-cycle of goals—shown in summary in Figure 1; the Active and Suspended states decomposed further [8]—