A Framework for FCA-based Exploratory Web Search

Authors: Peter Butka, Thomas Low, Michael Kotzyba, Stefan Haun, Andreas Nürnberger
Year: 2015
Venue: 1st International Symposium on Companion-Technology (ISCT 2015) - September 23rd-25th, Ulm University, Germany, 131-136
Product of the Action: Yes

Keystone Members Authors:

In this paper we present a conceptual framework for exploratory search by structuring web results based on concept lattices, which are created using methods from Formal Concept Analysis (FCA). The approach is to organize query based search engine results (web documents) by a hierarchy of clusters that are composed of documents with similar attributes. The resulting concept lattice provides a structured view on query-related domains and hence can improve the user’s understanding of document attributes and shared features. Instead of using the traditional approaches, we applied a fuzzy extension of FCA in order to support the usage of different types of attributes within the analyzed result set. Afterward the result set is presented using the graph visualization toolkit CET.