International Journal of Pattern Recognition and Artificial Intelligence

This paper addresses the problem of transforming WordNet into a resource tailored to Information Retrieval (IR) applications. We address two of the major drawbacks pointed out in previous literature in relation to this semantic network. One is the fine granularity of senses defined in WordNet, which proves useless from an IR perspective. To solve this problem, we propose a set of methods that enable the automatic transformation of WordNet into a coarse grained dictionary. The other drawback is the encoding used in this resource, and the methods for accessing related words across the semantic net. Due to the high number of connections among concepts, the simple computation of a path in this net, or the generation of related concepts may become a computationally intensive process. This effect is highly undesirable in time sensitive applications such as IR applications. We propose a methodology for hierarchical encoding that enables increased efficiency in WordNet-based IR systems.

The World Wide Web provides an immense source of information. Accessing information of interest presents a challenge to scientists and analysts, particularly if the desired information is structural in nature. Our goal is to design a structural search engine that uses the hyperlink structure of the Web, in addition to textual information, to search for sites of interest. Our structural search engine, called WebSUBDUE, searches not only for particular words or topics but also for a desired hyperlink structure. Enhanced by WordNet text functions, our search engine retrieves sites corresponding to structures formed by graph-based user queries. We hypothesize that this system can form the heart of a structural query engine, and demonstrate the approach on a number of structural web queries.

Mutations are structural changes in DNA that can cause protein malfunction and genetic disease. This paper describes a machine learning approach to analyzing mutations associated to Osteogenesis Imperfecta (OI), also known as brittle bone disease. We apply SORCER, a second-order rule induction system to predict clinical phenotypes of OI from mutation and neighboring amino acid sequences in the COLIA1 gene. On the average, over ten 10-fold cross-validations, SORCER gives more accurate results than C4.5 with average accuracy of about 81.2%. The paper discusses the advantages and limitations of SORCER and demonstrates its use to provide initial exploration of biological sequences.

As part of the task of automated question answering from a large collection of text documents, the reduction of the search space to a smaller set of document passages that are actually searched for answers constitutes a difficult but rewarding research issue. We propose a set of precision-enhancing filters for passage retrieval based on semantic constraints detected in the submitted questions. The approach improves the performance of the underlying question answering system in terms of both answer accuracy and time performance.

Human and software agents exhibit regularities in their activities. We describe our ONASI algorithm, which derives a Markov model from such observed regularities, and dynamically scales the model through merging and splitting of states. ONASI uses this model to predict the agent's next action. We evaluate ONASI's predictive accuracy on a dataset of Wumpus World games, and demonstrate from this approach that the model can correctly predict the agent's next action with adjustable computation and memory resources. These predictions can be used to imitate, assist or obstruct an agent.

Many extensions to existing binary constraint satisfaction algorithms have been proposed that directly deal with nonbinary constraints. Another choice is to perform a structural transformation of the representation of the problem, so that the resulting problem is a binary CSP except that now the original constraints which were nonbinary are replaced by binary compatibility constraints between relations. A lot of recent work has focussed on comparing different levels of local consistency enforceable in the nonbinary representation with the dual representation. In this paper we present extensions to the standard dual encoding that can compactly represent the given CSP using an equivalent dual encoding that contains all the original solutions to the CSP, using constraint coverings. We show how enforcing arc consistency in these constraint covering based encodings, strictly dominates enforcement of generalized arc consistency (GAC) on the primal nonbinary encoding.

This paper reports our progress in applying multiagent systems analysis and design in the area of cooperative robotics. In this paper, we apply the Multiagent Systems Engineering (MaSE) methodology to design a team of autonomous, heterogeneous search and rescue robots. MaSE provides a top-down approach to building cooperative robotic systems instead of the bottom-up approach employed in most robotic implementations. We follow the MaSE steps and discuss various approaches and their impact on the final system design.

During the last two decades the scarcity of water resources is exponentially growing worldwide. Cyprus is one of the countries in the Mediterranean where the water shortage problem is evident and very severe. This paper describes an intelligent and flexible, hybrid decision support system for the management of water resources in the regions along the South Conveyor Project, the largest water development project in Cyprus. In particular, the paper presents the integrated system and its architecture, together with detailed descriptions of the hybrid components synthesizing it. Details on the implementation and preliminary evaluation of the system are also provided.

Recommender systems research combines techniques from user modeling and information filtering in order to build search systems that are better able to respond to the preferences of individual users during the search for a particular item or product. Collaborative recommenders leverage the preferences of communities of similar users in order to guide the search for relevant items. The success of collaborative recommendation has always been restrained by the so-called sparsity problem, in which a lack of available user similarity knowledge ultimately limits the formation of high-quality user communities and has a subsequent impact on recommender accuracy. This article presents an approach to addressing the sparsity problem by describing and evaluating how implicit similarity knowledge can be discovered and exploited using data-mining techniques and an approach to recommendation that is inspired by case-based reasoning research.

SUSTAIN (Supervised and Unsupervised STratified Adaptive Incremental Network) is a network model of human category learning. SUSTAIN initially assumes a simple category structure. If simple solutions prove inadequate and SUSTAIN is confronted with a surprising event (e.g. it is told that a bat is a mammal instead of a bird), SUSTAIN recruits an additional cluster to represent the surprising event. Newly recruited clusters are available to explain future events and can themselves evolve into prototypes/attractors/rules. SUSTAIN has expanded the scope of findings that models of human category learning can address. This paper extends SUSTAIN to account for both supervised and unsupervised learning data through a common mechanism. The modified model, uSUSTAIN (unified SUSTAIN), is successfully applied to human learning data that compares unsupervised and supervised learning performances.¹⁸