Identification of Navigation Lead Candidates Using Citation and Co-Citation Analysis

摘要

Query refinement is an integral part of search, especially for the exploratory search scenarios, which assume that the users start with ill-defined information needs that change over time. In order to support exploratory search and navigation, we have proposed an approach of exploratory navigation in digital libraries using navigation leads. In this paper, we focus specifically on the identification of the navigation lead candidates using keyword extraction. For this purpose, we utilize the citation sentences as well as the co-citations. We hypothesize that they can improve the quality of the extracted keywords in terms of finding new keywords (that would not be otherwise discovered) as well as promoting the important keywords by increasing their relevance. We have quantitatively evaluated our method in the domain of digital libraries using experts' judgement on the relevance of the extracted keywords. Based on our results, we can conclude that using the citations and the co-citations improves the results of extraction of the most relevant terms over the TF-IDF baseline.