Community analysis-based academic search engine ranking method

Abstract

The invention discloses a community analysis-based academic search engine ranking method. The method comprises the steps of establishing a two-dimensional complex graph model, based on work citation relationships and author cooperation relationships in academic circles, in academic search engines, converting the two-dimensional complex graph model into a one-dimensional graph model, performing community analysis in a weighted label propagation mode, and dividing work information into different communities; performing community relationship mapping on the basis of inputting content to be searched by a user; performing ranking on content in the communities by reference to text similarity and walk frequencies of graph nodes through a random walk process-based ranking policy; and finally obtaining a work set required by the user. According to the method, hidden related content that cannot be found by a conventional academic search engine ranking method can be found, the defect that the conventional method depends too much on the text similarity is overcome, and the required calculation time is relatively short; and therefore, the method is suitable for large academic search engine ranking scenes.

Description

technical field [0001] The invention relates to information retrieval and complex networks, in particular to a method for sorting content by search engines and performing community analysis in complex networks formed by academic circles, and in particular to a method for sorting academic search engines based on community analysis. Background technique [0002] With the rapid development of the Internet, Web services based on the HTTP protocol are becoming more and more popular, and the amount of resources and information on the Internet has increased dramatically. Users have the need to find resources distributed in various locations on the Internet based on their own personalized information. In July 1994, Lycos launched the data mining technology based on the robot protocol to support the sorting of search results, which is an important progress in the history of search engines. In 1995, the meta search engine was born. After a user submits a search request, the meta-sear...

AI Technical Summary

Problems solved by technology

[0005] Although traditional academic search engines such as Google Scholar, Microsoft Academic Search, Baidu Academic Search, etc. have a relatively complete function of matching and finding relevant content in the document index library according to the text content, because the index is mainly based on the text similarity of the retrieved content Therefore, for some documents that are closely related to the search content but do not directly have a high text similarity, they are ignored in this mechanism with a high probability, resulting in incomplete and inaccurate search results.

Images