International Journal of Computational Intelligence and Applications

Web engines crawl hyperlinks to search for new documents; yet when they index discovered documents they basically revert to conventional information retrieval models and concentrate on the indexing of terms in a single document.

We propose to overcome such limits with an approach based on temporal logic. By modeling a web site as a finite state transition system we are able to define complex and selective queries over hyperlinks with the aid of Computation Tree Logic operators.

We deployed the proposed approach in a prototype system that allows users pose queries in natural language. Queries are automatically translated in Computation Tree Logic, and the answer returned by our system is a set of paths. Experiments carried out with the aid of human experts show improved retrieval effectiveness with respect to current search engines.

Clustering of data in a large dimension space is of great interest in many data mining applications. In this paper, we propose a method for clustering of web usage data in a high-dimensional space based on a concept hierarchy model. In this method, the relationship present in the web usage data are mapped into a fuzzy proximity relation of user transactions. We also described an approach to present the preference set of URLs to a new user transaction based on the match score with the clusters. The study demonstrates that our approach is general and effective for mining the web data for web personalization.

Topic distillation is the process of finding representative pages relevant to a given query. Well-known topic distillation approaches such as the HITS algorithm have shown to be useful for topic distillation. Many succeeding researchers focus on augmenting HITS with further content analysis to alleviate the steady deterioration of distillation quality suffered by HITS. In this paper, we attempt to revisit the behavior of HITS from a different point of view. Namely, a similarity-based analysis model is applied to observing the distillation procedure. By defining a generalized similarity, an algorithm is proposed, which can improve the quality of distillation only using the information of hyperlinks. The experimental results reveal that the new algorithm improves distillation quality without utilizing any content information of pages.

Due to the explosive growth of available information on the World Wide Web (WWW), users have suffered from the information overload. To alleviate this problem, there is a need for an intelligent tool to help the users screening and filtering for interesting and useful information. In this paper, a method of automatically identifying topics for Web documents via a classification technique is proposed. Topic identification can be applied as a filtering tool for recommender systems to prune down the number of documents to within some particular topics. We adopt the fuzzy association concept as a machine learning technique to classify the documents into some predefined categories or topics. Our approach is compared to the vector space model with the cosine coefficient using the data sets collected from three different Web portals: Yahoo!, Open Directory Project and Excite. The results show that our approach yields higher classification accuracy compared to the vector space model.

We present an automatic system that discovers competing companies from public information sources. The system extracts data from text, uses transformation-based learning to obtain appropriate data normalization, combines structured and unstructured information sources, uses probabilistic modelling to represent models of linked data, and succeeds in autonomously discovering competitors. We also introduce the iterative graph reconstruction process for inference in relational data, and show that it leads to improvements in performance. We validate system results and deploy it on the web as a powerful analytic tool for individual and institutional investors.

Web page categorization is an approach for improving precision and efficiency of information retrieval on the web by filtering out irrelevant pages. Current approaches to information filtering based on categorization assume the existence of a single classification hierarchy used for filtering. In this paper, we address the problem of filtering information categorized according to different classification hierarchies. We describe a method for approximating Boolean queries over class names across different class hierarchies.

Collaborative filtering techniques in the Internet are a means to make predictions about the behavior of a certain user based on the observation of former users. Frequently in literature the exploited information is contained in the access-log files of web servers storing requested data objects. However with additional effort on the server side it is possible to register, from which to which data object a client actually navigates. In this article the profile of a user in a distributed web environment will be modeled by the set of his navigation decisions between data objects. Such a set can be regarded as a graph with the nodes being the requested data objects and the edges being the decisions. A method is presented to learn the distribution of such graphs based on distance functions between graphs and the application of clustering techniques. The estimated distribution is used to predict future navigation decisions of new users. Results with randomly generated graphs show properties of the new algorithm. A measure to estimate the prediction quality for observed profiles is presented.

This paper presents an architecture of the e-Butler service, i.e. a personalized consumer-centric infomediary delivering a broad range of shopping-related services to consumers. In order to provide these services effectively, e-Butler should be "intelligent". This paper describes how a certain level of intelligence can be achieved through the utilization of a broad range of technologies, including data mining, personalization, profiling, recommender systems, data warehousing, Web proxy servers, data monitoring and XML-based technologies, that are integrated into a unique architecture of e-Butler.

XML's tree structure provides a rich background for complicated structural searches. In this paper we present a new system, called XML Query by Example (XML QBE) that allows the user to query XML documents exploiting their inherent tree structure. We present some interesting queries and describe the underlying query processing algorithms. We also describe the system's architecture and report its implementation status. Finally we conclude the paper by pointing out some future work.

We present a technique for simultaneously mining Web navigation patterns and maximally frequent context-sensitive itemsets (URL associations) from the historic user access data stored in Web server logs. A new hierarchical clustering technique that exploits the symbiosis between clusters in feature space and genetic biological niches in nature, called Hierarchical Unsupervised Niche Clustering (H-UNC) is presented. We use H-UNC as part of a complete system of knowledge discovery in Web usage data. Our approach does not necessitate fixing the number of clusters in advance, is insensitive to initialization, can handle noisy data, general non-differentiable similarity measures, and automatically provides profiles at multiple resolution levels. Our experiments show that our algorithm is not only capable of extracting meaningful user profiles on real Web sites, but also discovers associations between distinct URL pages on a site, with no additional cost. Unlike content based association methods, our approach discovers associations between different Web pages based only on the user access patterns and not on the page content. Also, unlike traditional context-blind association discovery methods, H-UNC discovers context-sensitive associations which are only meaningful within a limited context/user profile.

Efficient allocation of services to form a supply chain to solve complex tasks is a crucial problem. Optimal service allocation based on a single criterion is NP-Complete. Furthermore, complex tasks in general have multiple criteria that may be conflicting and non-commensurable. This paper presents a two-stage brokering algorithm for optimal anytime service allocation based on multiple criteria. In the first stage, a hierarchical task network planner is used to identify the services required to solve a task. In the second stage, a genetic algorithm (GA) determines service providers based on multiple criteria to provide the services identified by the planner. We present our algorithm and results from various experiments conducted to analyze the effect of various parameters that influence the complexity of the problem. In general, the results show the GA finds optimal solutions much quicker than a standard search algorithm. The empirical results also indicate the performance of the algorithm is sub-linear or polynomial time for various parameters. The algorithm has the ability to deal with any number of criteria. By addressing this problem, we expand the range of problems being addressed to any that require simultaneous optimization of multiple criteria and/or planning.