On content-driven search-keyword suggesters for literature digital libraries
Proceedings of the 8th ACM/IEEE-CS joint conference on Digital libraries
| Hi-index | 0.00 |
Organizing and ranking online Vertical (domain-specific) Literature Digital Library (VLDL) results into categories enables easy and efficient browsing of search results. Studies show that publications in any scientific discipline are ‘naturally’ categorized into groups called ‘research pyramids’, where a research pyramid refers to publications that belong to a most-specific research topic. In this paper, we present elGiza, a research-pyramid based search tool for VLDLs. elGiza is equipped with (i) a research-pyramid-based Content-Based Search-Keyword Suggester that helps user develop search terms to reduce search failures, and (ii) a Topic-Sensitive Ranking Tool to effectively order search results via research pyramids. Such a tool significantly reduces topic diffusion. Furthermore, elGiza can successfully group and label search results based on research pyramids. Research-pyramid-based grouping supports pruned-search in which the user ‘omits’ unrelated topics, and safely focuses on his/her search. Evaluation experiments show that research-pyramid-based VLDLs are highly effective.