Navigation Leads Selection Considering Navigational Value of Keywords
Róbert Móro, Mária Bieliková · 2015
Searching a vast information space such as the Web presents a challenging task and even more so, if the domain is unknown and the character of the task is thus exploratory in its nature. We have proposed a method of exploratory navigation based on navigation leads, i.e. terms that help users to filter the information space of a digital library. In this paper, we focus on the selection of the leads considering their navigational value. We employ clustering based on topic modeling using LDA (Latent Dirichlet Allocation). We present results of a preliminary evaluation on the Annota dataset containing more than 50,000 research papers.