Abstract Recommendation with Assistance of Interactive User Profile Extraction
Haodi Feng, Hong Liu, Shenpeng Lu · 2009
Abstraction and user-profile-based search are two important topics in information retrieval. In this paper, we combine these two problems together, and present a user-profile-based meta search system with which the user can read the abstracts of preferred documents. The user profile is represented as list of words, which are either given directly by the user or generated automatically by the system from the user's reading history.