The retrieval of relevant cases plays a crucial role in case-based reasoning. There are three major methods for the retrieval of relevant cases: computational approaches (based upon measures of similarity), representational approaches (based upon indexing structures) and hybrid approaches. This paper looks at recent successful implementations of case retrieval with regard to this classification framework. Similarly it emphasises computational and representational models applied to feature-vector case representations.

Ontology has been recently proposed as an appropriate formalism to model software project management antipatterns, in order to encode antipatterns in a computer understandable form and introduce antipatterns to the Semantic Web. However, given two antipattern ontologies, the same entity can be described using different terminology. Therefore, the detection of similar antipattern ontologies is a difficult task. In this paper, we introduce a three-layered antipattern semantic social network, which involves the social network, the antipattern ontology network and the concept network. Social Network Analysis (SNA) techniques can be used to assist software project managers in finding similar antipattern ontologies. For this purpose, SNA measures are extracted from one layer of the semantic social network to another and this knowledge is used to infer new links between antipattern ontologies. The level of uncertainty associated with each new link is represented using Bayesian Networks (BNs). Furthermore, BNs address the issue of quantifying the uncertainty of the data collected regarding antipattern ontologies for the purposes of the conducted analysis. Finally, BNs are used to augment SNA by taking into account meta-information in their calculations. Hence, other knowledge not included in the social network can be used in order to search the social network for further inference. The benefits of using an antipattern semantic social network are illustrated using an example community of software project management antipattern ontologies.

Constraint satisfaction is a fundamental artificial intelligence technique offering a simple yet powerful representation. An increasing amount of attention has recently been paid to the development of constraint satisfaction techniques, and it has become clear that the original formulation of a static Constraint Satisfaction Problem (CSP) with hard, imperative constraints is insufficient to model many real problems. Two important extensions to the classical CSP framework which address some of these deficiencies are flexible and dynamic constraint satisfaction. This paper examines in detail classical, flexible and dynamic CSP. It reviews the motivations behind both extensions, and describes the techniques used to solve each type of problem. The paper employs a running example throughout to illustrate the ideas presented.

As the number and size of very large databases continues to grow rapidly, so does the need to make sense of them. This need is addressed by the field called knowledge Discovery in Databases (KDD), which combines approaches from machine learning, statistics, intelligent databases, and knowledge acquisition. KDD encompasses a number of different discovery methods, such as clustering, data summarization, learning classification rules, finding dependency networks, analysing changes, and detecting anomalies (Matheus et at., 1993).

The fast increase in the size and number of databases demands data mining approaches that are scalable to large amounts of data. This has led to the exploration of parallel computing technologies in order to perform data mining tasks concurrently using several processors. Parallelization seems to be a natural and cost-effective way to scale up data mining technologies. One of the most important of these data mining technologies is the classification of newly recorded data. This paper surveys advances in parallelization in the field of classification rule induction.

Artificial neural networks (ANNs) can be employed as controllers for robotic agents. Their structure is often complex, with many neurons and connections, especially when the robots have many sensors and actuators distributed across their bodies and/or when high expressive power is desirable. Pruning (removing neurons or connections) reduces the complexity of the ANN, thus increasing its energy efficiency, and has been reported to improve the generalization capability, in some cases. In addition, it is well-known that pruning in biological neural networks plays a fundamental role in the development of brains and their ability to learn. In this study, we consider the evolutionary optimization of neural controllers for the case study of Voxel-based soft robots, a kind of modular, bio-inspired soft robots, applying pruning during fitness evaluation. For a locomotion task, and for centralized as well as distributed controllers, we experimentally characterize the effect of different forms of pruning on after-pruning effectiveness, life-long effectiveness, adaptability to new terrains, and behavior. We find that incorporating some forms of pruning in neuroevolution leads to almost equally effective controllers as those evolved without pruning, with the benefit of higher robustness to pruning. We also observe occasional improvements in generalization ability.

Selecting pivot features that connect a source domain to a target domain is an important first step in unsupervised domain adaptation (UDA). Although different strategies such as the frequency of a feature in a domain, mutual (or pointwise mutual) information have been proposed in prior work in domain adaptation (DA) for selecting pivots, a comparative study into (a) how the pivots selected using existing strategies differ, and (b) how the pivot selection strategy affects the performance of a target DA task remain unknown. In this paper, we perform a comparative study covering different strategies that use both labelled (available for the source domain only) as well as unlabelled (available for both the source and target domains) data for selecting pivots for UDA. Our experiments show that in most cases pivot selection strategies that use labelled data outperform their unlabelled counterparts, emphasising the importance of the source domain labelled data for UDA. Moreover, pointwise mutual information and frequency-based pivot selection strategies obtain the best performances in two state-of-the-art UDA methods.

We survey the use and effect of decomposition-based techniques in qualitative spatial and temporal constraint-based reasoning, and clarify the notions of a tree decomposition, a chordal graph, and a partitioning graph, and their implication with a particular constraint property that has been extensively used in the literature, namely, patchwork. As a consequence, we prove that a recently proposed decomposition-based approach that was presented in the study by Nikolaou and Koubarakis for checking the satisfiability of qualitative spatial constraint networks lacks soundness. Therefore, the approach becomes quite controversial as it does not seem to offer any technical advance at all, while results of an experimental evaluation of it in a following work presented in the study by Sioutis become questionable. Finally, we present a particular tree decomposition that is based on the biconnected components of the constraint graph of a given large network, and show that it allows for cost-free utilization of parallelism for a qualitative constraint language that has patchwork for satisfiable atomic networks.

Multi-Robot System (MRS) is composed of a group of robots that work cooperatively. However, Multi-Agent System (MAS) is computational systems consisting of a group of agents that interact with each other to solve a problem. The central difference between MRS and MAS is that in the first case, the agent is a robot, and in the second, it is a software. Analyzing the scientific literature, it is possible to notice that few studies address the integration between MAS and MRS. In order to achieve the interdisciplinary integration, the theoretical background of these areas must be considered in this paper, so that the integration can be applied using a case study of decentralized MRS. The objective of this MRS is to track and surround a stationary target. Also, it has been implemented and validated in the robot simulator called Virtual Robot Experimentation Platform (V-REP). In the validation of the proposed MRS, a scenario with three robots and a stationary target were defined. In the tracking task, the robot can detect the target whose position is not known a priori. When the detection occurs, the V-REP informs the target position to the robot because the environment is discretized into a grid of rectangular cells. After that, all the robots are directed to the target, and the surround task is realized. In this task, a mathematical model with direct communication between the robots was used to keep the robots equidistant therefrom and from each other.

Agent communication languages (ACLs) are fundamental mechanisms that enable agents in multi-agent systems to talk, communicate with each other in order to satisfy their individual and social goals in a cooperative and competitive manner. Social approaches are advocated to overcome the shortcomings of ACL semantics delineated by using mental approaches in the figure of agents’ mental notions. Over the last two decades, social commitments have been the subject of considerable research in some of those social approaches as they provide a powerful representation for modeling and reasoning upon multi-agent interactions in the form of mutual contractual obligations. They particularly provide a declarative, flexible, verifiable, and social semantics for ACL messages while respecting agents’ autonomy, heterogeneity, and openness.

In this manuscript, we go through prominent and predominate proposals in the literature to explore the state of the art on how temporal logics can be devoted to define a formal semantics for ACL messages in terms of social commitments and associated actions. We explain each proposal and point out if and how it meets seven crucial criteria, four of them introduced by Munindar P. Singh to have a well-defined semantics for ACL messages. Far from deciding the best proposal, our aim is to present the advantages (strengths) and limitations of those proposals to designers and developers using a concrete running example and to compare between them, so that they can make the best choice with regard to their needs. We explore and evaluate current specification languages and different verification techniques that have been discussed within those proposals to, respectively, specify and verify commitment-based protocols. We also investigate logical languages of actions advocated to specify, model, and execute commitment-based protocols in other contributed proposals. Finally, we suggest some solutions that can contribute to address the identified limitations.

The promise of rule-based computing was to allow end-users to create, modify, and maintain applications without the need to engage programmers. But experience has shown that rule sets often interact in subtle ways, making them difficult to understand and reason about. This has impeded the widespread adoption of rule-based computing. This paper describes the design and implementation of XcelLog, a user-centered deductive spreadsheet system, to empower non-programmers to specify and manipulate rule-based systems. The driving idea underlying the system is to treat sets as the fundamental data type and rules as specifying relationships among sets, and use the spreadsheet metaphor to create and view the materialized sets. The fundamental feature that makes XcelLog suitable for non-programmers is that the user mainly sees the effect of the rules; when rules or basic facts change, the user sees the impact of the change immediately. This enables the user to gain confidence in the rules and their modification, and also experiment with what-if scenarios without any programming. Preliminary experience with using XcelLog indicates that it is indeed feasible to put the power of deductive spreadsheets for doing rule-based computing into the hands of end-users and do so without the requirement of programming or the constraints of canned application packages.

1 Believable interactive embodied agents

Among the goals of research on autonomous agents one important aim is to build believable interactive embodied agents that are apt to application to friendly interfaces in e-commerce, tourist and service query systems, entertainment (e.g. synthetic actors) and education (pedagogical agents, agents for help and instruction to the hearing impaired).

Potential-based reward shaping is a commonly used approach in reinforcement learning to direct exploration based on prior knowledge. Both in single and multi-agent settings this technique speeds up learning without losing any theoretical convergence guarantees. However, if speed ups through reward shaping are to be achieved in multi-agent environments, a different shaping signal should be used for each context in which agents have a different subgoal or when agents are involved in a different interaction situation.

This paper describes the use of context-aware potential functions in a multi-agent system in which the interactions between agents are sparse. This means that, unknown to the agents a priori, the interactions between the agents only occur sporadically in certain regions of the state space. During these interactions, agents need to coordinate in order to reach the global optimal solution.

We demonstrate how different reward shaping functions can be used on top of Future Coordinating Q-learning (FCQ-learning); an algorithm capable of automatically detecting when agents should take each other into consideration. Using FCQ-learning, coordination problems can even be anticipated before the actual problems occur, allowing the problems to be solved timely. We evaluate our approach on a range of gridworld problems, as well as a simulation of air traffic control.

The software and hardware applications are clearly on the way of becoming an integral tool of business, communication and popular culture in many parts of the world. People are interacting with the environment via the Internet to perform physical activities remotely. These applications are hosted in the public or private servers under the control of the server admin. The users’ online usage data can be stored in public or private cloud platforms, used for processing and monitoring users’ online behaviour and emotional factors and shared with third parties to facilitate making their business decisions. When users allow their data to be collected via software applications and mobile devices, users need to have some level of trust and control over their data. But, software applications or mobile devices connected to the cloud server using client–server architecture does not ensure the reliability, security and integrity among their data. To get over these kinds of limitations, we propose a database management system using blockchain technology that can be used by any software applications. The blockchain database connected to the cloud server can be used to increase the trustfulness of the application. Blockchain has the capability to provide decentralization, immutability and owner-controlled digital assets among software applications. Since users can save their data in a shared transaction repository with tamper-resistant records, it enables related parties to access and control users’ data without the need for a central control system.

A series of temporal reasoning tasks are identified which motivate the consideration and application of temporal logics in artificial intelligence. There follows a discussion of the broad issues involved in modelling time and constructing a temporal logic. The paper then presents a detailed review of the major approaches to temporal logics: first-order logic approaches, modal temporal logics and reified temporal logics. The review considers the most significant exemplars within the various approaches, including logics due to Russell, Hayes and McCarthy, Prior, McDermott, Allen, Kowalski and Sergot. The logics are compared and contrasted, particularly in their treatments of change and action, the roles they seek to fulfil and the underlying models of time on which they rest. The paper concludes with a brief consideration of the problem of granularity—a problem of considerable significance in temporal reasoning, which has yet to be satisfactorily treated in a temporal logic.

The success of a software system strongly depends on the ability of turning a precise domain analysis into a concrete architecture. Even if the domain model relies on sound ontological bases, there is often a wide semantic gap between the conceptual model and the concrete components that should reify it. To fill the semantic gap, relevant domain concepts should be engineered by identifying the corresponding architectural abstractions, which can be realized by concrete software components. Space plays a crucial role in many application domains, but surprisingly, related architectural abstractions have not emerged yet. This paper proposes space-related abstractions derived from the application of classical software engineering principles; in particular, the information hiding principle that leads to an operational definition of space. Basic abstractions are refined to deal with architectural aspects. As the underlying software engineering principles are close to principles that underlie the definition of space ontologies, the conjecture is that the proposed space architectural abstractions might be the basis for a formalization in ontological terms.

Reasoning about actions and plans is a vital aspect of the rational behaviour of intelligent agents, and hence represents a major research domain in artificial intelligence. Much work has been undertaken to develop logic-based formalisms and problem solving procedures for plan representation and plan synthesis. This paper consists of a survey of various paradigms for reasoning about actions and plans in artificial intelligence. Attention is focused on the logic-based theoretical frameworks which have built a formal foundation for the domain-independent approaches to the general principles of reasoning about actions and plans.

This paper focuses on coordination middleware for distributed applications based on active documents and XML technologies. First, the paper introduces the main concepts underlying active documents and XML, and identifies the strict relations between active documents and mobile agents (“document agents”). Then, the paper goes into details about the problem of defining a suitable middleware architecture to support coordination activities in applications including active documents and mobile agents, by specifically focusing on the role played by XML technologies in that context. A simple taxonomy is introduced to characterise coordination middleware architectures depending on the way they exploit XML documents in supporting coordination. The characteristics of several middleware infrastructures are then surveyed and evaluated, also with the help of a simple example scenario in the area of distributed workflow management. This analysis enables us to identify the advantages and the shortcomings of the different approaches, and the basic requirements of a middleware for XML-centric applications.

Biometric user authentication in large-scale distributed systems involves passive scanners and networked workstations and databases for user data acquisition, processing, and encryption. Unfortunately, traditional biometric authentication systems are prone to several attacks, such as Replay Attacks, Communication Attacks, and Database Attacks. Embedded biometric sensors overcome security limits of conventional software recognition systems, hiding its common attack points. The availability of mature reconfigurable hardware technology, such as field-programmable gate arrays, allows the developers to design and prototype the whole embedded biometric sensors. In this work, two strong and invasive biometric traits, such as fingerprint and iris, have been considered, analyzed, and combined in unimodal and multimodal biometric sensors. Biometric sensor performance has been evaluated using the well-known FVC2002, CASIA, and BATH databases.

Analogues to much of today's work in ontologies have existed for centuries in text retrieval. The use of controlled vocabularies, or thesauri, has been fundamental to document indexing in library science. Thesauri serve several purposes, including:

[bull ] Knowledge organisation A thesaurus provides a hierarchy of concepts that organises domain-specific knowledge.

[bull ] Terminology normalisation By selecting a unique word or phrase to represent each domain concept, then linking synonymous terms to it, a thesaurus enforces terminological consistency.

[bull ] Query expansion A thesaurus facilitates the addition of terms to a query by providing explicit hierarchical and lateral relationships among terms.

These properties serve to mediate the information flow from indexer to user. Thesauri thus serve many of the same functions for people that ontologies are designed to serve for software agents. As automated retrieval has developed over the decades since the inception of computer processing of text, many techniques have been introduced to apply this typically manual work to the automated arena (see Soergel (1985) for an introduction to library information systems, also Anderson and Pélrez-Carballo (2001a, 2001b) for a summary of the intersection of human and machine indexing).