Recent progress in reinforcement learning (RL) using self-play has shown remarkable performance with several board games (e.g., Chess and Go) and video games (e.g., Atari games and Dota2). It is plausible to hypothesize that RL, starting from zero knowledge, might be able to gradually approach a winning strategy after a certain amount of training. In this paper, we explore neural Monte Carlo Tree Search (neural MCTS), an RL algorithm that has been applied successfully by DeepMind to play Go and Chess at a superhuman level. We try to leverage the computational power of neural MCTS to solve a class of combinatorial optimization problems. Following the idea of Hintikka’s Game-Theoretical Semantics, we propose the Zermelo Gamification to transform specific combinatorial optimization problems into Zermelo games whose winning strategies correspond to the solutions of the original optimization problems. A specially designed neural MCTS algorithm is then introduced to train Zermelo game agents. We use a prototype problem for which the ground-truth policy is efficiently computable to demonstrate that neural MCTS is promising.

This paper deals with the implications of showing the storyteller’s perspective when telling stories in virtual environments. The paper proposes an open and reusable architecture for the construction of virtual guides who tell stories about the worlds they inhabit from their own perspective. The system consists of two main components: the structure of a knowledge base to code the narrative knowledge about the worlds and a novel hybrid algorithm for the generation of narratives showing the storyteller’s viewpoint. The complete architecture has been developed. Furthermore, we carried out a study with users in order to both validate the implemented storytelling system and test whether the users consider it significantly better or worse (in terms of several aspects related to believability) adding the virtual guide perspective (as both text and animations showing emotions) to the guide discourse. The interesting results are discussed.

Repeated games are an important mathematical formalism to model and study long-term economic interactions between multiple self-interested parties (individuals or groups of individuals). They open attractive perspectives in modeling long-term multiagent interactions. This overview paper discusses the most important results that actually exist for repeated games. These results arise from both economics and computer science. Contrary to a number of existing surveys of repeated games, most of which originated from the economic research community, we are first to pay a special attention to a number of important distinctive features proper to artificial agents. More precisely, artificial agents, as opposed to the human agents mainly aimed by the economic research, are usually bounded whether in terms of memory or performance. Therefore, their decisions have to be based on the strategies defined using finite representations. Furthermore, these strategies have to be efficiently computed or approximated using a limited computational resource usually available to artificial agents.

An appropriately designed architecture of a neural network is essential to many realistic pattern-recognition tasks. A choice of just the right number of neurons, and their interconnections, can cut learning costs by orders of magnitude, and still warrant high classification accuracy. Surprisingly, textbooks often neglect this issue. A specialist seeking systematic information will soon realize that relevant material is scattered over diverse sources, each with a different perspective, terminology and goals. This brief survey attempts to rectify the situation by explaining the involved aspects, and by describing some of the fundamental techniques.

This paper discusses the development of Information Presentation Systems (IPS). It identifies that these systems have had two development paths; generic and application-specific. It discusses the main pieces of early work, and identifies the significant problems and shortcomings associated with them. The development of later systems is traced through to the present day, and outstanding research issues are identified.

Designing realistic multi-agent systems is a complex process, which involves specifying not only the functionality of individual agents, but also the authority relationships and lines of communication existing among them. In other words, designing a multi-agent system refers to designing an agent organisation. Existing methodologies follow a wide variety of approaches to designing agent organisations, but they do not provide adequate support for the decisions involved in moving from analysis to design. Instead, they require designers to make ad hoc design decisions while working at a low level of abstraction.

We have developed RAMASD (Role Algebraic Multi-Agent System Design), a method for semi-automatic design of agent organisations based on the concept of role models as first-class design constructs. Role models represent agent behaviour, and the design of the agent system is done by systematically allocating roles to agents. The core of this method is a formal model of basic relations between roles, which we call role algebra. The semantics of this role-relationships model are formally defined using a two-sorted algebra.

In this paper, we review existing agent system design methodologies to highlight areas where further work is required, describe how our method can address some of the outstanding issues and demonstrate its application to a case study involving telephone repair service teams.

The techniques of qualitative reasoning are now becoming sufficiently mature to be applied to real world problems. In order to better understand which techniques are being used successfully for real world applications, and which application areas can be suitably addressed using qualitative reasoning techniques, it is helpful to have a summary of what application oriented work has been done to date. This helps to provide a picture of the application areas in which the techniques are being applied, and who is working in each application domain. In this paper, we summarize over 40 relevant projects.

Regulatory information has a direct and significant effect on the safety, economics and quality of many industries' operations and end-products. However, standards documents are often voluminous and can be complex. Users of such standards need to invest considerable time and effort to become familiar with the structure and content of the documents. Therefore, it has long been recognized that computer handling of this information is a potentially useful application area. This paper provides an overview of the work that has been done to provide computer systems that will help both users and authors of regulations. The emphasis is on using advanced symbolic computing techniques that would help in accessing, interpreting and applying the information.

Drawing upon the theories of complexity and complex adaptive systems and the Singerian inquiring system from C. West Churchman's seminal work The Design of Inquiring Systems, the study herein develops a systems design theory for continuous auditing systems. The discussion focuses on the two foundational theories, development of the theory of Complex Adaptive Inquiring Organizations (CAIO) and associated design principles for a continuous auditing system supporting a CAIO, and instantiation of the CAIO theory. The instantiation consists of an agent-based model depicting the marketplace for Frontier Airlines that generates an anticipated market share used as an integral component in a mock auditor going concern opinion for the airline. As a whole, the study addresses the lack of an underlying system design theory and comprehensive view needed to build upon and advance the continuous assurance movement and addresses the question of how continuous auditing systems should be designed to produce knowledge – knowledge that benefits auditors, clients, and society as a whole.

This paper presents an approach for the enactment of policies in digital health based on our earlier work on the implementation of digital contracts in distributed systems. A formal policy model and an abstract policy language for the expression of healthcare policies are first proposed, leveraging the semantics of the ISO Reference Model for Open Distributed Processing enterprise language standard. Healthcare consent policies included in the HL7 Fast Health Interoperability Resource (FHIR®) standard are used to illustrate the modelling approach. Several distributed ledger and smart legal contract options were considered next as target platforms for implementation. Their benefits are highlighted along with considerations on their use reflecting business concerns of risk, trust and cost.

A logical spreadsheet is a spreadsheet in which the formula language is composed of logical expressions. Logical spreadsheets were invented shortly after traditional electronic spreadsheets were introduced, but since then logical spreadsheet research has been somewhat sparse. Recently, however, there has been a resurgence in the interest of logical spreadsheets in the research community. In this article, we summarize logical spreadsheet research up to this point.

In this paper, we present two broad styles of KBS reasoner: those based primarily on some general, explicit model of the knowledge of the domain (whether that model be expressed by heuristic rules or by a deep model of structure and function), which we term domain model-based reasoners; and those based primarily on a set of examples of events in the domain, which we term example-based reasoners (EBR), of which case-based reasoners are a subset. The aim of this paper is to guide developers in considering the trade-offs between these different styles of reasoning. We believe that this cannot be done in general, but may be possible for specific domains. Thus, the paper provides an example analysis of the usefulness of these reasoning styles. We assess the suitability of these styles against a series of requirements which we have identified that KBSs must fulfil if they are to support help desk operations. We conclude that EBR systems are more likely to meet those requirements (the analysis draws on our earlier work in Bridge & Dearden, 1992).

In this article I present and discuss some criteria to provide a diagrammatic classification. Such a classification is of use for exploring in detail the domain of diagrammatic reasoning. Diagrams can be classified in terms of the use we make of them—static or dynamic—and of the correspondence between their space and the space of the data they are intended to represent. The investigation is not guided by the opposition visual vs. non-visual, but by the idea that there is a continuous interaction between diagrams and language. Diagrammatic reasoning is characterized by a duality, since it refers both to an object, the diagram, having its spatial characteristics, and to a subject, the user, who interprets them. A particular place in the classification is occupied by constructional diagrams, which exhibit for the user instructions for the application of some procedures.

This paper presents our sketch drawing artist humanoid robot research. One of the limitations of the existing artist humanoid robot is the lack of feedback on the error that occurs during the drawing process. The contribution of this research is the development of a humanoid robot artist with drawing error correction capability. Based on our previous work with open-loop control pen-and-ink humanoid robot artist, we have implemented a closed-loop visual servoing approach to address this problem. Our experimental results show that this approach is sufficient to correct drawing errors that occur due to mechanical limitation of a robot.

Knowledge representation is a subarea of artificial intelligence concerned with using formal symbols to represent a set of facts within a knowledge domain. Two popular knowledge representation languages, namely Petri net and ontology, are promising knowledge sharing and reusing methods in knowledge engineering. The combination of Petri net and ontology can facilitate achieving complementary advantages. Currently, many efforts have been done on knowledge sharing between Petri nets and ontologies. To investigate these issues and more importantly serve as identifying the direction of knowledge sharing between Petri nets and ontologies, in this paper we give a comprehensive literature overview of knowledge sharing between Petri net models and ontology models to satisfy the obvious need. In detail, we discuss the knowledge sharing from two aspects: the different knowledge representation approaches of ontology to represent and reason Petri net and issues of constructing Petri net from ontology. In addition, other important issues on applications and directions for future research are discussed in detail.

This paper aims to describe the current state of knowledge systems technology and its commercialisation in the US. First, knowledge systems are defined and placed in a historical context. The introduction is concluded with a preview of major ideas. The paper will assess the technological state of the art and will survey the current state of commercialisation. Finally, some anticipated future trends will be discussed.

Algorithmic bias arises in machine learning when models that may have reasonable overall accuracy are biased in favor of ‘good’ outcomes for one side of a sensitive category, for example gender or race. The bias will manifest as an underestimation of good outcomes for the under-represented minority. In a sense, we should not be surprised that a model might be biased when it has not been ‘asked’ not to be; reasonable accuracy can be achieved by ignoring the under-represented minority. A common strategy to address this issue is to include fairness as a component in the learning objective. In this paper, we consider including fairness as an additional criterion in model training and propose a multi-objective optimization strategy using Pareto Simulated Annealing that optimizes for both accuracy and underestimation bias. Our experiments show that this strategy can identify families of models with members representing different accuracy/fairness tradeoffs. We demonstrate the effectiveness of this strategy on two synthetic and two real-world datasets.

This paper examines the claim that machine induction can alleviate the current knowledge engineering bottleneck in expert system construction. It presents a case study of the rule induction software tool known as Expert-Ease and proposes a set of criteria which might guide the selection of appropriate domains.

This paper presents a review of pattern matching techniques. The application areas for pattern matching are extensive, ranging from CAD systems to chemical analysis and from manufacturing to image processing. Published techniques and methods are classified and assessed within the context of three key issues: pattern classes, similarity types and matching methods. It has been shown that the techniques and approaches are as diverse and varied as the applications.

This paper surveys a variety of deductive database theories. Such theories differ from one another in the set of axioms and metarules that they allow and use. The following theories are discussed: relational, Horn, and stratified in the text; protected, disjunctive, typed, extended Horn, and normal in the appendix. Connections with programming in terms of the declarative, fixpoint, and procedural semantics are explained. Negation is treated in several different ways: closed world, completed database, and negation as failure. For each theory examples are given and implementation issues are considered.