Ranking Documents with Query-Topic Sensitivity
Shota Hatakenaka, Satoshi Shimada, Takao Miura · 2012 IEEE/WIC/ACM International Conferences on Web Intelligence and Intelligent Agent Technology · 2012
In this work, we discuss Query-Topic Sensitive Ranking algorithm, called Topic-Driven PageRank (TDPR), to inquire general documents based on a notion of importance. The main idea is that we extract knowledge from training data for multiple classification and build characteristic feature for each topic. By this approach, we get documents reflecting queries and topics within so that we can improve query results and to avoid topic drift problems.