University of Crete (UOC)
Interactive exploration of multi-dimensional information spaces with preference support
Abstract
dc:descriptionUsers access large amounts of information resources (documents or data) mainly through search functions,where they type a few words and the system (web search engine, query engine) returns a linearlist of hits. While this is often satisfactory for focalized search, it does not provide enough support forrecall-oriented (exploratory) information needs, which constitute the majority according to various userstudies.The interaction of Faceted and Dynamic Taxonomies (FDT), is a highly prevalent model for exploratorysearch, which allows users to get an overview of the information space (e.g. search results) andoffer them various groupings of the results (based on their attributes, metadata, or other dynamicallymined information). These groupings enable users to restrict their focus gradually and in a simple way(through clicks, i.e. without having to formulate queries), enabling them to locate resources that wouldbe difficult to locate otherwise (especially the low ranked ones).The enrichment of search mechanisms with preferences could be proved useful for recall-orientedinformation needs. However, the current approaches for preference-based access (mainly from the areaof databases), seem to ignore the fact that users should be acquainted with the information space andthe available choices for describing effectively their preferences.In this dissertation we extend the interaction model of FDT with preference actions that allow usersto express their preferences interactively, gradually, and in a simple way.Initially, we introduce a preference framework appropriate for information spaces comprising resourcesdescribed by attributes whose values can be hierarchically valued and/or multi-valued. Wedefine the language, its semantics and the required algorithms. The framework supports preferenceinheritance in the hierarchies, automatic conflict resolution, as well as preference composition (prioritization,Pareto and their combination).Subsequently, we enrich the FDT model with preference actions and we propose logical optimizationsand methods for exploiting the intrinsic characteristics of the FDT-based interaction, aiming at making it applicable to large amounts of information. Then, we present the design and the implementationof the web-based system Hippalus, which realizes the extended interaction model.Regarding user benefits, at first we theoretically analyze user gain in terms of the number and difficultyof choices, and then we describe and analyze three user-based evaluations that we have conducted.The first investigates the degree of effectiveness of preferences (and the effort to express them) whenusers are not aware of the available choices. The results showed that only 20% of the users managed toexpress effective preferences without knowing the available choices.The second comparatively evaluates FDT and other exploratory models. The results showed that themajority of users preferred FDT, was more satisfied by FDT and achieved higher rates of task completionwith FDT.The last one concerns the evaluation of the preference-enriched FDT as realized by Hippalus. Theresults were impressive. Even in a very small dataset, with the preference-enriched FDT all users successfullycompleted all tasks in 1/3 of the time and with 1/3 of the actions in comparison to the plain FDT.Moreover all (100%) of the users (either plain or experts) preferred the preference-enriched interface.
Degree
thesis:*- Grantor dc:publisher
- University of Crete (UOC)
- Year dc:date
- 2013
Author and committee
dc:creator, dc:contributor.*- Authors dc:creator
-
- Papadakos, Panagiotis
- Παπαδάκος, Παναγιώτης
Subjects
dc:subject × 12Rights
- Language dc:language
- eng
Identifiers
dc:identifier.*- Identifier
- 10.12681/eadd/36568