Research And Improvement Of Ranking Algorithms On Search Engine

The21st century is the century of information technology,and the search engine is aneffective tool for dealing with vast amouts of information.Ranking of the technology for searchengine is a top priority in a variety of technical difficulties,and it is becoming one of the hottesttopics.Google search engine stands in the majority of the top search engines by the excellentperformance of PageRank algorithm.However,with the increasing of the quantitative informationand the developing of intelligence,it is hard to meet people’s requirements only relying onexisting ranking technology.It is urgent for a more in line with the times of ranking algorithm toappear.This paper is based on the current situation.First,it introduces PageRank algorithm which isbased on link ranking and a text reference ranking algorithm which is based on contentranking,and then proposes revised PageRank algorithm which considering time-based feedbackby the knowledge of Markov chain and introduces the idea of text reference to the method ofLatent Semantic Indexing which is based on Singular Value Decomposition.Then by the study ofBaidu links it combine the link-based ranking method and the content-based rankingmethod.Finally it proposes an evaluation criteria based on multi-object decision making,and thengives the final evaluation value to the algorithm which is discussed in this paper.