Maurice Pagnucco

AAAI Conference 2026 System Paper

Principles2Plan: LLM-Guided System for Operationalising Ethical Principles into Plans

• Tammy Zhong
• Yang Song
• Maurice Pagnucco

Ethical awareness is critical for robots operating in human environments, yet existing automated planning tools provide little support. Manually specifying ethical rules is labour-intensive and highly context-specific. We present Principles2Plan, an interactive research prototype demonstrating how a human and a Large Language Model (LLM) can collaborate to produce context-sensitive ethical rules and guide automated planning. A domain expert provides the planning domain, problem details, and relevant high-level principles such as beneficence and privacy. The system generates operationalisable ethical rules consistent with these principles, which the user can review, prioritise, and supply to a planner to produce ethically-informed plans. To our knowledge, no prior system supports users in generating principle-grounded rules for classical planning contexts. Principles2Plan showcases the potential of human-LLM collaboration for making ethical automated planning more practical and feasible.

KR Conference 2025 Conference Paper

Belief Revision in a Probabilistic Setting

• James P. Delgrande
• Gerhard Lakemeyer
• Maurice Pagnucco
• Joshua Sack

This work develops an approach to qualitative belief revision in a fully probabilistic setting. We begin with a logic where possible worlds are assigned probabilities. In this logic an agent may believe a formula is true even though the subjective probability of the formula is less than 1. 0. Similarly, after revision by a formula ϕ, the agent will believe ϕ is true, even though the agent’s subjective probability of ϕ may be less than 1. 0. We establish a correspondence with the hallmark AGM postulates for belief revision. Moreover, we use Jeffrey Conditionalisation to establish a link with iterated belief change. To this end, we develop an approach that satisfies appropriately modified Darwiche-Pearl postulates (with clear justification). Thus, we provide a connection between quantitative probabilistic approaches on the one hand and the qualitative formulation of belief change, on the other. This work holds potential for the development of practical belief revision systems by applying a (qualitative) approach to belief change in probabilistic, uncertain domains.

JAIR Journal 2025 Journal Article

Computational Machine Ethics: A Survey

• Tammy Zhong
• Yang Song
• Raynaldio Limarga
• Maurice Pagnucco

Computational Machine Ethics (CME) is an interdisciplinary field that integrates moral philosophy into an agent’s decision-making process, contributing to the broader domain of Artificial Intelligence Ethics. Technological advancements have transformed the world, where technology has become an integral part of society, progressively given more autonomy in making judgments within various domains in our lives. Inevitably, issues of ethics come into play in these judgments, making ethical decision-making in machines an increasingly critical problem to solve. This survey provides an overview of CME, highlighting the breadth of directions and the use of techniques within the field. We also provide some background on the ethical dimension before introducing our taxonomy used to categorise and detail the variety of existing approaches from a more technical perspective. Finally, we identify limitations in the research and suggest potential open challenges for future work.

ICRA Conference 2025 Conference Paper

SynerGuard: A Robust Framework for Point Cloud Classification via Local Geometry and Spatial Topology

• Haonan Zhong
• Wei Song
• Maurice Pagnucco
• Yang Song 0001

Point cloud recognition models are known to be vulnerable to adversarial attacks. The state-of-the-art defense solutions either focus on partial features of the point cloud, limiting their effectiveness, or rely heavily on known adversarial examples, reducing their generalizability, while others, like point cloud reconstruction, will degrade the classifier's accuracy on clean examples. To address this, we introduce SynerGuard, a novel robust point cloud classification framework mitigating adversarial attacks by considering comprehensive geometric and topological attributes of the point cloud, without relying on known adversarial examples while attaining classification accuracies on clean examples. We comprehensively test SynerGuard against seven attack types from three leading adversarial attack approaches on two widely used datasets, ModelNet40 and ShapeNetPart. The results demonstrate SynERGUARD's superiority against existing defenses in mitigating adversarial attacks, as well as managing clean examples.

AAAI Conference 2024 Conference Paper

Decoupled Optimisation for Long-Tailed Visual Recognition

• Cong Cong
• Shiyu Xuan
• Sidong Liu
• Shiliang Zhang
• Maurice Pagnucco
• Yang Song

When training on a long-tailed dataset, conventional learning algorithms tend to exhibit a bias towards classes with a larger sample size. Our investigation has revealed that this biased learning tendency originates from the model parameters, which are trained to disproportionately contribute to the classes characterised by their sample size (e.g., many, medium, and few classes). To balance the overall parameter contribution across all classes, we investigate the importance of each model parameter to the learning of different class groups, and propose a multistage parameter Decouple and Optimisation (DO) framework that decouples parameters into different groups with each group learning a specific portion of classes. To optimise the parameter learning, we apply different training objectives with a collaborative optimisation step to learn complementary information about each class group. Extensive experiments on long-tailed datasets, including CIFAR100, Places-LT, ImageNet-LT, and iNaturaList 2018, show that our framework achieves competitive performance compared to the state-of-the-art.

ECAI Conference 2024 Conference Paper

Domain Generalised Cell Nuclei Segmentation in Histopathology Images Using Domain-Aware Curriculum Learning and Colour-Perceived Meta Learning

• Kunzi Xie
• Ruoyu Guo
• Cong Cong 0001
• Maurice Pagnucco
• Yang Song 0001

Cell nuclei segmentation in histopathology images is critical in computer-aided diagnosis and treatment planning. However, this task is challenging due to inherent heterogeneity in histopathology images especially when originating from different domains, caused by variations in imaging protocols, staining techniques, and tissue preparation methods. Such domain shifts can significantly affect segmentation performance when the segmentation model is trained and tested on different domains. In this work, we present a novel gradient-based meta-learning approach for domain generalisation in histopathology cell nuclei segmentation. Specifically, we propose a domain-aware regularisation to correct each pixel’s classification based on the specific domain. We also embed a novel network module to preserve the colour features in histopathology images via an enhanced feature extraction procedure. We demonstrate that our proposed framework can achieve consistent and accurate segmentation performance across domains through extensive experiments on multiple histopathology datasets from diverse sources. Our code is available at: https: //github. com/winnie172026/DG.

IJCAI Conference 2024 Conference Paper

Formalisation and Evaluation of Properties for Consequentialist Machine Ethics

• Raynaldio Limarga
• Yang Song
• Abhaya Nayak
• David Rajaratnam
• Maurice Pagnucco

As artificial intelligence (AI) technologies continue to influence our daily lives, there has been a growing need to ensure that AI enabled decision making systems adhere to principles expected of human decision makers. This need has given rise to the area of Machine Ethics. We formalise several ethical principles from the philosophical literature in the situation calculus framework to verify the ethical permissibility of a plan. Moreover, we propose several important properties, including some of our own that are intuitively appealing, and a number derived from the social choice literature that would appear to be relevant in evaluating the various approaches. Finally we provide an assessment of how our various situation calculus models of Machine Ethics that we examine satisfy the important properties we have identified.

NeurIPS Conference 2024 Conference Paper

Fully Distributed, Flexible Compositional Visual Representations via Soft Tensor Products

• Bethia Sun
• Maurice Pagnucco
• Yang Song

Since the inception of the classicalist vs. connectionist debate, it has been argued that the ability to systematically combine symbol-like entities into compositional representations is crucial for human intelligence. In connectionist systems, the field of disentanglement has gained prominence for its ability to produce explicitly compositional representations; however, it relies on a fundamentally symbolic, concatenative representation of compositional structure that clashes with the continuous, distributed foundations of deep learning. To resolve this tension, we extend Smolensky's Tensor Product Representation (TPR) and introduce Soft TPR, a representational form that encodes compositional structure in an inherently distributed, flexible manner, along with Soft TPR Autoencoder, a theoretically-principled architecture designed specifically to learn Soft TPRs. Comprehensive evaluations in the visual representation learning domain demonstrate that the Soft TPR framework consistently outperforms conventional disentanglement alternatives -- achieving state-of-the-art disentanglement, boosting representation learner convergence, and delivering superior sample efficiency and low-sample regime performance in downstream tasks. These findings highlight the promise of a distributed and flexible approach to representing compositional structure by potentially enhancing alignment with the core principles of deep learning over the conventional symbolic approach.

IROS Conference 2023 Conference Paper

HyperTraj: Towards Simple and Fast Scene-Compliant Endpoint Conditioned Trajectory Prediction

• Renhao Huang
• Maurice Pagnucco
• Yang Song 0001

An important task in trajectory prediction is to model the uncertainty of agents' motions, which requires the system to propose multiple plausible future trajectories for agents based on their past mo vement s. Recently, many approaches have been developed following an endpointconditioned deep learning framework by firstly predicting the distribution of endpoints, then sampling endpoints from it and finally completing their waypoints. However, this framework suffers a severe efficiency issue as it needs to repeatedly execute a separate decoder conditioned on multiple sampled endpoints. In this work, we propose a simple and fast endpoint conditioned fully convolutional trajectory prediction framework, called HyperTraj, by using dynamic convolutions to generate multiple trajectories, with the main benefits that (1) our prediction is conditioned on endpoint but takes almost constant time when the number of goals increases and (2) our model benefits from convolutional based predictions, such as the acceptance of various scene sizes and better modeling of agent-scene interactions. In our experiment, our model shows comparable or even better accuracy than our state-of-the-art baselines on SDD and VIRAT datasets with around 84% of acceleration and 90% model weight reduction for waypoint decoding.

ECAI Conference 2023 Conference Paper

Identifying the Defective: Detecting Damaged Grains for Cereal Appearance Inspection

• Lei Fan 0007
• Yiwen Ding
• Dongdong Fan
• Yong Wu
• Maurice Pagnucco
• Yang Song 0001

Cereal grain plays a crucial role in the human diet as a major source of essential nutrients. Grain Appearance Inspection (GAI) serves as an essential process to determine grain quality and facilitate grain circulation and processing. However, GAI is routinely performed manually by inspectors with cumbersome procedures, which poses a significant bottleneck in smart agriculture. In this paper, we endeavor to develop an automated GAI system: AI4GrainInsp. By analyzing the distinctive characteristics of grain kernels, we formulate GAI as a ubiquitous problem: Anomaly Detection (AD), in which healthy and edible kernels are considered normal samples while damaged grains or unknown objects are regarded as anomalies. We further propose an AD model, called AD-GAI, which is trained using only normal samples yet can identify anomalies during inference. Moreover, we customize a prototype device for data acquisition and create a large-scale dataset including 220K high-quality images of wheat and maize kernels. Through extensive experiments, AD-GAI achieves considerable performance in comparison with advanced AD methods, and AI4GrainInsp has highly consistent performance compared to human experts and excels at inspection efficiency over 20× speedup. The dataset, code and models will be released at https: //github. com/hellodfan/AI4GrainInsp.

NeurIPS Conference 2023 Conference Paper

LoRA: A Logical Reasoning Augmented Dataset for Visual Question Answering

• Jingying Gao
• Qi Wu
• Alan Blair
• Maurice Pagnucco

The capacity to reason logically is a hallmark of human cognition. Humans excel at integrating multimodal information for locigal reasoning, as exemplified by the Visual Question Answering (VQA) task, which is a challenging multimodal task. VQA tasks and large vision-and-language models aim to tackle reasoning problems, but the accuracy, consistency and fabrication of the generated answers is hard to evaluate in the absence of a VQA dataset that can offer formal, comprehensive and systematic complex logical reasoning questions. To address this gap, we present LoRA, a novel Logical Reasoning Augmented VQA dataset that requires formal and complex description logic reasoning based on a food-and-kitchen knowledge base. Our main objective in creating LoRA is to enhance the complex and formal logical reasoning capabilities of VQA models, which are not adequately measured by existing VQA datasets. We devise strong and flexible programs to automatically generate 200, 000 diverse description logic reasoning questions based on the SROIQ Description Logic, along with realistic kitchen scenes and ground truth answers. We fine-tune the latest transformer VQA models and evaluate the zero-shot performance of the state-of-the-art large vision-and-language models on LoRA. The results reveal that LoRA presents a unique challenge in logical reasoning, setting a systematic and comprehensive evaluation standard.

ICLR Conference 2023 Conference Paper

Maximizing Spatio-Temporal Entropy of Deep 3D CNNs for Efficient Video Recognition

• Junyan Wang 0001
• Zhenhong Sun
• Yichen Qian
• Dong Gong
• Xiuyu Sun
• Ming Lin 0002
• Maurice Pagnucco
• Yang Song 0001

3D convolution neural networks (CNNs) have been the prevailing option for video recognition. To capture the temporal information, 3D convolutions are computed along the sequences, leading to cubically growing and expensive computations. To reduce the computational cost, previous methods resort to manually designed 3D/2D CNN structures with approximations or automatic search, which sacrifice the modeling ability or make training time-consuming. In this work, we propose to automatically design efficient 3D CNN architectures via a novel training-free neural architecture search approach tailored for 3D CNNs considering the model complexity. To measure the expressiveness of 3D CNNs efficiently, we formulate a 3D CNN as an information system and derive an analytic entropy score, based on the Maximum Entropy Principle. Specifically, we propose a spatio-temporal entropy score (STEntr-Score) with a refinement factor to handle the discrepancy of visual information in spatial and temporal dimensions, through dynamically leveraging the correlation between the feature map size and kernel size depth-wisely. Highly efficient and expressive 3D CNN architectures, i.e., entropy-based 3D CNNs (E3D family), can then be efficiently searched by maximizing the STEntr-Score under a given computational budget, via an evolutionary algorithm without training the network parameters. Extensive experiments on Something-Something V1&V2 and Kinetics400 demonstrate that the E3D family achieves state-of-the-art performance with higher computational efficiency.

IJCAI Conference 2022 Conference Paper

Epistemic Logic of Likelihood and Belief

• James P. Delgrande
• Joshua Sack
• Gerhard Lakemeyer
• Maurice Pagnucco

A major challenge in AI is dealing with uncertain information. While probabilistic approaches have been employed to address this issue, in many situations probabilities may not be available or may be unsuitable. As an alternative, qualitative approaches have been introduced to express that one event is no more probable than another. We provide an approach where an agent may reason deductively about notions of likelihood, and may hold beliefs where the subjective probability for a belief is less than 1. Thus, an agent can believe that p holds (with probability <1); and if the agent believes that q is more likely than p, then the agent will also believe q. Our language allows for arbitrary nesting of beliefs and qualitative likelihoods. We provide a sound and complete proof system for the logic with respect to an underlying probabilistic semantics, and show that the language is equivalent to a sublanguage with no nested modalities.

AAMAS Conference 2019 Conference Paper

A Representation Theorem for Reasoning in First-Order Multi-Agent Knowledge Bases

• Christoph Schwering
• Maurice Pagnucco

Levesque’s notion of only-knowing provides a natural formalisation of a knowledge base: it precisely captures the beliefs and non-beliefs that follow from the knowledge base, including introspection and de dicto versus de re distinctions in a first-order setting. Apart from its attractive properties in terms of specification, a major result about only-knowing is Levesque’s representation theorem, which shows how reasoning in (single-agent) knowledge bases can be Turingreduced to ordinary first-order logic. While numerous proposals have been made to lift the logic of only-knowing to the multiagent case, generalising the representation theorem has remained an open problem. In this paper, we develop a Turing reduction from reasoning in multi-agent knowledge bases to ordinary, nonepistemic first-order logic and thus obtain a new representation theorem for the multi-agent case.

AIJ Journal 2017 Journal Article

Belief revision and projection in the epistemic situation calculus

• Christoph Schwering
• Gerhard Lakemeyer
• Maurice Pagnucco

This article considers defeasible beliefs in dynamic settings. In particular, we examine the belief projection problem: what is believed after performing an action and/or receiving new information? The approach is based on an epistemic variant of Reiter's situation calculus, where actions not only have physical effects but may also provide new information to the agent. The preferential belief structure is initially determined using conditional statements. New information is then incorporated using two popular belief revision schemes, namely natural and lexicographic revision. The projection problem is solved twofold in this formalism: by goal regression and by knowledge base progression.

IJCAI Conference 2016 Conference Paper

A Framework for Integrating Symbolic and Sub-Symbolic Representations

• Keith Clark
• Bernhard Hengst
• Maurice Pagnucco
• David Rajaratnam
• Peter Robinson
• Claude Sammut
• Michael Thielscher

This paper establishes a framework that hierarchically integrates symbolic and sub-symbolic representations in an architecture for cognitive robotics. It is formalised abstractly as nodes in a hierarchy, with each node a sub-task that maintains its own belief-state and generates behaviour. An instantiation is developed for a real robot building towers of blocks, subject to human interference; this hierarchy uses a node with a concurrent multitasking teleo-reactive program, a node embedding a physics simulator to provide spatial knowledge, and nodes for sensor processing and robot control.

IJCAI Conference 2015 Conference Paper

Belief Revision and Progression of Knowledge Bases in the Epistemic Situation Calculus

• Christoph Schwering
• Gerhard Lakemeyer
• Maurice Pagnucco

Fundamental to reasoning about actions and beliefs is the projection problem: to decide what is believed after a sequence of actions is performed. Progression is one widely applied technique to solve this problem. In this paper we propose a novel framework for computing progression in the epistemic situation calculus. In particular, we model an agent’s preferential belief structure using conditional statements and provide a technique for updating these conditional statements as actions are performed and sensing information is received. Moreover, we show, by using the concepts of natural revision and only-believing, that the progression of a conditional knowledge base can be represented by only-believing the revised set of conditional statements. These results lay the foundations for feasible belief progression due to the uniquemodel property of only-believing.

AAAI Conference 2014 Conference Paper

A Framework for Task Planning in Heterogeneous Multi Robot Systems Based on Robot Capabilities

• Jennifer Buehler
• Maurice Pagnucco

KR Conference 2014 Conference Paper

Forgetting in Action

• David Rajaratnam
• Hector Levesque
• Maurice Pagnucco
• Michael Thielscher

A further motivation for the concept of forgetting is evident in the formal analysis of security and cryptographic protocols. Cryptographic protocols have previously been analysed using the Situation Calculus equipped with a notion of the knowledge of agents (Delgrande, Hunter, and Grote 2010). However, such an encoding cannot always represent the class of nonmonotonic cryptographic protocols (Rubin and Honeyman 1994), where a notion of forgetting can be important. For example, in the analysis of credit card protocols it is a critical requirement to model vendors that forget (i. e., not retain) customer credit card details. In order to illustrate our approach to forgetting we consider the following running example. A robot needs to enter a room that is protected by a closed door with a keypad lock. When the robot senses that the door is closed, it needs to download the key combination from an external data source (e. g., the cloud or a database). It can then use this key combination to open the door and enter the room. Furthermore, since the door will now be open, the robot no longer needs the key combination, so is free to forget this information. The rest of the paper proceeds as follows. First we introduce the Situation Calculus (McCarthy 1963; Reiter 2001) and its epistemic extension (Lesperance et al. 1995) that allows for knowledge acquisition without forgetting. We then present our approach that handles both knowledge acquisition and forgetting and perform an extensive analysis of its properties and the conditions under which both knowledge acquisition and forgetting can occur. Having established our approach to forgetting, we then place it within the broader context of two of the main models that have been developed within the literature: AGM belief revision (Alchourrón, Gärdenfors, and Makinson 1985) and logical forgetting (Lin and Reiter 1994). In particular, we show that our approach is well-behaved with respect to the AGM belief contraction postulates, but is distinct from that of logical forgetting since knowledge forgetting can provide for more fine-grained control over what is forgotten. Finally, we provide some concluding remarks and discuss directions for future research. In this paper we develop a general framework that allows for both knowledge acquisition and forgetting in the Situation Calculus. Based on the Scherl and Levesque (Scherl and Levesque 1993) possible worlds approach to knowledge in the Situation Calculus, we allow for both sensing as well as explicit forgetting actions. This model of forgetting is then compared to existing frameworks. In particular we show that forgetting is well-behaved with respect to the contraction operator of the well-known AGM theory of belief revision (Alchourrón, Gärdenfors, and Makinson 1985) but that knowledge forgetting is distinct from the more commonly known notion of logical forgetting (Lin and Reiter 1994).

AAAI Conference 2014 Conference Paper

Minimising Undesired Task Costs in Multi-Robot Task Allocation Problems with In-Schedule Dependencies

• Bradford Heap
• Maurice Pagnucco

In multi-robot task allocation problems with inschedule dependencies, tasks with high costs have a large influence on the total time required for a team of robots to complete all tasks. We reduce this influence by calculating a novel task cost dispersion value that measures robots’ collective preference for each task. By modifying the winner determination phase of sequential single-item auctions, our approach inspects the bids for every task to identify tasks which robots collectively consider to be high cost and ensures these tasks are allocated prior to other tasks. Our empirical results show this method provides a significant reduction in the total time required to complete all tasks.

ECAI Conference 2014 Conference Paper

Planning and Execution of Robot Tasks Based on a Platform-Independent Model of Robot Capabilities

• Jennifer Elisabeth Buehler
• Maurice Pagnucco

The diversity of robotic architectures is a major factor in developing platform-independent algorithms. There is a need for a widely usable model of robot capabilities which can help to describe and reason about the diversity of robotic systems. We propose such a model and present an integrated framework for task planning and task execution using this model. Existing planning techniques need to be extended to support this model, as it requires 1) generating new objects during planning time; and, 2) establishing concurrency based on data flow within the robotic system. We present results on planning and execution of an object transportation task in simulation.

IJCAI Conference 2013 Conference Paper

Definability of Horn Revision from Horn Contraction

• Zhiqiang Zhuang
• Maurice Pagnucco
• Yan Zhang

In the AGM framework [Alchourrón and Makinson, 1985], a revision function can be defined directly through constructions like systems of spheres, epistemic entrenchment, etc. , or indirectly through a contraction operation via the Levi identity. A recent trend is to construct AGM style contraction and revision functions that operate under Horn logic. A direct construction of Horn revision is given in [Delgrande and Peppas, 2011]. However, it is unknown whether Horn revision can be defined indirectly from Horn contraction. In this paper, we address this problem by obtaining a model-based Horn revision through the model-based Horn contraction studied in [Zhuang and Pagnucco, 2012]. Our result shows that, under proper restrictions, Horn revision is definable through Horn contraction via the Levi identity.

KR Conference 2012 Conference Paper

Model Based Horn Contraction

• Zhiqiang Zhuang
• Maurice Pagnucco

2010; Delgrande and Wassermann 2010; Zhuang and Pagnucco 2010a; 2010b; 2011; Delgrande and Peppas 2011). The topic is interesting for several reasons. Horn logic is an important subset of propositional logic which has found use in many artificial intelligence and database applications. The study of belief change under Horn logic broadens the practical applicability of the AGM framework and in particular it provides a key step towards applying the AGM framework to non-classical logics with less expressive and reasoning power than propositional logic. The classic construction for AGM contraction is based on the notion of remainder sets. Remainder sets of a belief set K with respect to a sentence φ are the maximal1 subsets of K that fail to imply φ, denoted by K ↓ φ. In this construction, the resulting belief set is obtained by intersecting the most desirable remainder sets chosen by a selection function. If the selection function is transitive and relational, then the contraction constructed is called transitively relational partial meet contraction (TRPMC). Levi suggests that revision can be defined from contraction via the identity:.. K ∗ φ = (K −¬φ) ∪ {φ} for ∗ a revision operator and − a contraction operator. The revision obtained from TRPMC via the Levi identity is called transitively relational partial meet revision (TRPMR). AGM revision can be constructed directly without referring to a contraction. Due to its logical closure, a belief set can be identified by its set of models which makes it possible to study change operations in terms of the models involved. Katsuno and Mendelzon (Katsuno and Mendelzon 1992) gave a model based approach for constructing revision directly. In this approach the input to the revision are models of the belief set and those of the new belief and the output is a set of revised models from which the resulting belief set is obtained. A preorder over those models is used to determine the revision. According to (Katsuno and Mendelzon 1992), if the preorder is total and it is faithful with respect to the original belief set, then the determined revision performs identically to TRPMR. So essentially it gives a model theoretic account for TRPMR. Although (Katsuno and Mendelzon 1992) deals with revision, its contraction counterpart can be easily derived which gives a model theoretic account Following the recent trend of adapting the AGM (Alchourrón and Makinson 1985) framework to propositional Horn logic, Delgrande and Peppas (Delgrande and Peppas 2011) give a model theoretic account for revision in the Horn logic setting. The current paper complements their work by studying the model theoretic approach for contraction. A model based Horn contraction is constructed and shown to give a model theoretic account to the transitively relational partial meet Horn contraction studied in (Zhuang and Pagnucco 2011). Significantly however, in contrast to (Delgrande and Peppas 2011), our model-based characterisation of Horn contraction does not require the property of Horn compliance and totality over preorders. The model based contraction, upon proper restriction, also gives a model theoretic account for the epistemic entrenchment based Horn contraction studied in (Zhuang and Pagnucco 2010a).

AAAI Conference 2012 Conference Paper

Repeated Sequential Auctions with Dynamic Task Clusters

• Bradford Heap
• Maurice Pagnucco

Sequential auctions can be used to provide solutions to the multi-robot task-allocation problem. In this paper we extend previous work on sequential auctions and propose an algorithm that clusters and auctions uninitiated task clusters repeatedly upon the completion of individual tasks. We demonstrate empirically that our algorithm results in lower overall team costs than other sequential auction algorithms that only assign tasks once.

AIJ Journal 2011 Journal Article

Iterated belief change in the situation calculus

• Steven Shapiro
• Maurice Pagnucco
• Yves Lespérance
• Hector J. Levesque

John McCarthy's situation calculus has left an enduring mark on artificial intelligence research. This simple yet elegant formalism for modelling and reasoning about dynamic systems is still in common use more than forty years since it was first proposed. The ability to reason about action and change has long been considered a necessary component for any intelligent system. The situation calculus and its numerous extensions as well as the many competing proposals that it has inspired deal with this problem to some extent. In this paper, we offer a new approach to belief change associated with performing actions that addresses some of the shortcomings of these approaches. In particular, our approach is based on a well-developed theory of action in the situation calculus extended to deal with belief. Moreover, by augmenting this approach with a notion of plausibility over situations, our account handles nested belief, belief introspection, mistaken belief, and handles belief revision and belief update together with iterated belief change.

IJCAI Conference 2011 Conference Paper

Transitively Relational Partial Meet Horn Contraction

• Zhiqiang Zhuang
• Maurice Pagnucco

Following the recent trend of studying the theory of belief revision under the Horn fragment of propo- sitional logic this paper develops a fully charac- terised Horn contraction which is analogous to the traditional transitively relational partial meet contraction [Alchourron et al. , 1985]. This Horn con- traction extends the partial meet Horn contraction studied in [Delgrande and Wassermann, 2010] so that it is guided by a transitive relation that models the ordering of plausibility over sets of beliefs.

JELIA Conference 2010 Conference Paper

Horn Contraction via Epistemic Entrenchment

• Zhiqiang Zhuang
• Maurice Pagnucco

Abstract Belief change studies the way in which a reasoner should maintain its beliefs in the face of newly acquired information. The AGM account of belief change assumes an underlying logic containing classical propositional logic. Recently, there has been interest in studying belief change, specifically contraction, under the Horn fragment of propositional logic (i. e. , Horn logic ). In this paper we continue this line of research, and propose a Horn contraction that is based on the Epistemic Entrenchment (EE) construction of AGM contraction. The standard EE construction refers to arbitrary disjunctions which are not available in Horn logic. Therefore, we make use of a Horn approximation technique called Horn strengthening. An ideal Horn contraction should be as plausible as an AGM contraction. In other words it should performs identically with AGM contractions when restricted to Horn logic. We demonstrate that no EE based Horn contraction satisfies this criterion unless we apply certain restrictions to the AGM contraction. A representation theorem is proved which identifies the characterising postulates for our Horn contraction.

IJCAI Conference 2009 Conference Paper

• Thomas Ströder
• Maurice Pagnucco

IJCAI Conference 2005 Conference Paper

Inverse Resolution as Belief Change

• Maurice Pagnucco
• David

ECAI Conference 2004 Conference Paper

Iterated Belief Change and Exogeneous Actions in the Situation Calculus

• Steven Shapiro
• Maurice Pagnucco

ECAI Conference 2004 Conference Paper

Simplicity in Solving the Frame Problem

• Victor Jauregui
• Maurice Pagnucco
• Norman Y. Foo

AIJ Journal 2003 Journal Article

Dynamic belief revision operators

• Abhaya C. Nayak
• Maurice Pagnucco
• Pavlos Peppas

The AGM approach to belief change is not geared to provide a decent account of iterated belief change. Darwiche and Pearl have sought to extend the AGM proposal in an interesting way to deal with this problem. We show that the original Darwiche–Pearl approach is, on the one hand excessively strong and, on the other rather limited in scope. The later Darwiche–Pearl approach, we argue, although it addresses the first problem, still remains rather permissive. We address both these issues by (1) assuming a dynamic revision operator that changes to a new revision operator after each instance of belief change, and (2) strengthening the Darwiche–Pearl proposal. Moreover, we provide constructions of this dynamic revision operator via entrenchment kinematics as well as a simple form of lexicographic revision, and prove representation results connecting these accounts.

IJCAI Conference 2003 Conference Paper

Prolegomenon to a Theory of Conservative Belief Revision

• James P Delgrande
• Abhaya C. Nayak
• Maurice Pagnucco

IJCAI Conference 1999 Conference Paper

Diagrammatic Proofs

• Norman Y. Foo
• Maurice Pagnucco
• Abhaya C. Nayak

Diagrammatic reasoning comprises phenomena that range from the so-called "free-rides" (e. g. almost immediate understanding of visually perceived relationships) to conventions about tokens. Such reasoning must involve cognitive processes that are highly perceptual in content. In the domain of mathematical proofs where diagrams have had a long history, we have an opportunity to investigate in detail and in a controlled setting the various perceptual devices and cognitive processes that facilitate diagrammatically based arguments. This paper continues recent work by examining two kinds of diagrammatic proofs, called Categories 1 and 3 by Jamnik, et. al. 97], the first being one in which generalization of a diagram instance is implied, and the second being one in which an infinite completion is represented by an ellipsis. We provide explanations of why these proofs work, a semantics for ellipses, and conjectures about the underlying cognitive processes that seem to resonate with such proofs.

IJCAI Conference 1999 Conference Paper

Preferential Semantics for Causal Systems

• Pavlos Peppas
• Maurice Pagnucco
• Mikhail Prokopenko
• Norman Y. Foo
• Abhaya Nayak

In the present work we examine the causal theory of actions put forward by McCain and Turner [Mc- Cain and Turner, 1995] for determining ramifications. Our principal aim is to provide a characterisation of this causal theory of actions in terms of a Shoham-like preferential semantics [Shoham, 19881. This would have a twofold advantage: it would place McCain and Turner's theory in perspective, allowing a comparison with other logics of action; and, it would allow us to glean further insights into the nature of causality underlying their work. We begin by showing that our aim is not attainable by a preferential mechanism alone. At this point we do not abandon preferential semantics altogether but augment it in order to arrive at the desired result. We draw f he following moral which is at the heart of our paper: two components — minimal change under a preferential structure and causality — are required to provide a concise solution to the frame and ramification problems.

TARK Conference 1996 Conference Paper

Changing Conditional Belief Unconditionally

• Abhaya C. Nayak
• Norman Y. Foo
• Maurice Pagnucco
• Abdul Sattar 0001

Although the AGM account of beliefchange tells us how to change unconditional beliefs, it fails to guide us in changing our conditional beliefs. That explains why, in the AGM account of belief change proper, a decent account of iterated belief change is not forthcoming. Darwiche and Pearl provide an account of changing conditional beliefs. We argue that the Darwiche-Pearl postulates are, on the one hand, limited in scope and, on the other, excessively strong and suggest how they should be supplemented. We show, contrary to the generally held view, that the revision operation changes to a new (though in some cases the same) revision operation after each instance of belief change. Finally, we prove what may be viewed as representation results connecting the acount offered in this paper with entrenchment kinematics.

ECAI Conference 1996 Conference Paper

Definitional Constraints

• Norman Y. Foo
• Abhaya C. Nayak
• Maurice Pagnucco

ECAI Conference 1996 Conference Paper

Learning From Conditionals: Judy Benjamin's Other Problems

• Abhaya C. Nayak
• Maurice Pagnucco
• Norman Y. Foo
• Pavlos Peppas

ECAI Conference 1996 Conference Paper

Revision vs. Update: Taking a Closer Look

• Pavlos Peppas
• Abhaya C. Nayak
• Maurice Pagnucco
• Norman Y. Foo
• Rex Bing Hung Kwok
• Mikhail Prokopenko

IJCAI Conference 1995 Conference Paper

Determining Explanations using Transmutations

• Mary-Anne Williams
• Maurice Pagnucco
• Norman Foo
• Brailey Sims

Intelligent Information systems do not usually possess complete information about the world with which they interact. The AGM paradigm has become one of the standard frameworks for modeling changes to repositories of information. Its principal constructions for change operators rely on some form of underlying preference relation. The process of changing such a preference relation is known as a transmutation. Spohn's conditionalization can be interpreted as a transmutation that imposes a relative minimal change. A transmutation based on an absolute minimal change is an adjustment. In this paper we develop a notion of explanation using transmutations of information systems. Following Gardenfors lead we recast Spohn's notion of reason for within the general setting of transmutations and extend this to characterize most plausible explanations. We also investigate the relationship between explanation based on abduction and Spohnian reasons based on adjustments. Finally, and rather surprisingly, we identify explicit conditions that characterize the various forms of explanations identified by Boutilier and Becher using Spohnian reasons.