Introduction to Information Filtering and Retrieval: Soft Computing-Based Approaches Minitrack

The problem of automatically accessing information/data/services relevant to specific user’s needs on the World Wide Web still poses a big challenge. The Web is, in fact, a continuously evolving information repository that needs to be accessed with the support of automatic systems to help users in locating and retrieving relevant information. These systems include Search Engines and Information Filtering systems; the former, also known as Information Retrieval Systems (IRSs), offer a pull technology, which produces an answer when solicited by an explicit user input (usually a keyword based query). The latter are based on the knowledge of some form of user preferences to push to users the information estimated relevant to them; they belong to the category of push technologies. The task of providing a fast and effective localisation of the information relevant to specific users' needs is very hard and complex, since it is pervaded with subjectivity, vagueness, and uncertainty, and it also requires context adaptation. The huge amount of research in this scientific domain witnesses that we still need advanced and user-centric systems that allow to effectively access various kind of information on the Web that can be relevant to address specific needs and tasks. To overcome the limitations of current search engines, in recent years a great deal of research has been deployed to the aim of modelling factors of subjectivity, uncertainty, vagueness and the context adaptation being intrinsic to the process of locating relevant information. For this purpose, the role of Soft Computing (SC) techniques has been advocated and demonstrated through various studies as a conceptual and algorithmic vehicle to obtain a greater flexibility in designing systems for Information Access. This mini-track on Information Filtering and Retrieval: Soft Computing Based Approaches is focused on the theory and applications of fuzzy logic and Soft Computing to support the development of techniques that support users to easily access information relevant to their needs. The paper authored by Endang Djuana, Yue Xu and Yuefeng Li is entitled A Combined Method for Mitigating Sparsity Problem in Tag Recommendation. The authors address the problem of tag recommendation, a specific recommendation task for recommending metadata (tags) for a web resource (item) during the user annotation process. In particular, the authors propose a combined method for mitigating the sparsity problem in tag recommendation by mainly expanding and ranking candidate tags based on similar items’ tag and tag ontology. The approach is evaluated using two publicly available social bookmarking datasets. The second paper by Anwar Alhenshiri, Carolyn Watters, Michael Shepherd, and Jack Duffy entitled Investigating Features in Support of Web Tools for Information Gathering brings an interesting study that investigates features to improve how users perform information gathering tasks on the Web by 1) keeping track of references to information as it is gathered for the task 2) keeping the task information integrated between sessions; and 3) providing integrated search, browsing, and editing capabilities. Finally, the paper entitled Soft Retrieval and Fuzzy Databases is authored by Ronald Yager and Rachel Yager. The paper addresses the problem of retrieving information from databases; in particular, the paper reviews mechanisms of flexible querying in databases, and discusses a method for aggregating the individual satisfaction levels that can model a lexicographic relation between the individual requirements. Moreover querying database that store information affected by probabilistic uncertainty is considered. 2014 47th Hawaii International Conference on System Science