Citation analysis models
Goal of information retrieval is to find documents that are relevant to a user’s query. How is relevance defined? Basically Statistical Language Model have focused on content dependent models of relevance. Content dependent models has limitation since all documents are not created equal in the perspective of Credibility, Trust, and Authority. Can the link structure of the web be used to help infer the relevance of pages? → Citation analysis model (PageRank)
연결개수이용한 Google Search engine으로 Larry Page 등이 만들었다.
PageRank is a way of measuring the importance of website pages based on long-term visit-rate
Damping Factor 이용해서 수렴한다.
PageRank computation
simplest is power iteration