Adaptive ranking of search results by considering user's comprehension
Makoto Nakatani, Adam Jatowt, Katsumi Tanaka · 2010
Given a search query, conventional Web search engines provide users with the same ranking although users' comprehension levels can be different. It is often difficult especially for non-expert users to find comprehensible Web pages from the list of search results. In this paper, we propose the method of adaptively ranking search results by considering user's comprehension level. The main issues are (a) estimating the comprehensibility of Web pages and (b) estimating the user's comprehension level. In our method, the com-prehensibility of each search result is computed by using the readability index and technical terms extracted from Wikipedia. User's comprehension level is estimated by the users' feedback about the difficulty of search results that they have viewed. We implement a prototype system and evaluate the usefulness of our approach by user experiments.