Research of PageRank Algorithm Based on Transition Probability

摘要

In order to improve the veracity of the web search, this paper studies the PageRank algorithm, proposes a new method PBTP Algorithm (PageRank Based on Transition Probability), that is an improvement for the classical PageRank method. As forwarding links in a web page are different, the transition probability of a link to be clicked is different too. For the classical PageRank value, should assign more authority value to the page according to its clicking probability with high authority value, effectively to focus the authority value on more meaningful web page, finally extracts meaningful page with high authority value. This paper takes advantage of Web link structure, proposes an unequal way to treat the different pages when distributing authorities. And the experiment shows that PBTP can improve PageRank effectively.