Industry-oriented topic search method

Abstract

The invention discloses an industry-oriented topic search method. The method comprises the steps that an initial to-be-crawled queue is initialized and established, whether crawling time of a crawler is due and whether the to-be-crawled queue is empty are judged, a Shark-Search-Advanced algorithm is adopted to calculate a relevancy value between a webpage and a topic, a PageRank-Advanced algorithm is adopted to calculate a webpage connection value and a webpage ranking value, and whether a secondary crawling time interval of the crawler is due is judged. Through the method, the accuracy and reliability of a search result can be effectively improved, therefore, a high-accuracy high-coverage retrieval result is acquired efficiently, and it is guaranteed that a search engine can respond to a search demand of a user for a specific industry with high efficiency, high accuracy and high coverage.

Description

technical field [0001] The invention belongs to the technical field of information retrieval, and in particular relates to an industry-oriented subject search method. Background technique [0002] The Internet has become the most important way for people to disseminate information and acquire content. The general search engines represented by Google, Baidu, and Bing provide great convenience for people to obtain information quickly and accurately on the Internet. However, general-purpose search engines need to build a huge search database, and the search content is oriented to the entire network. When users need to conduct vertical searches on specific industries, their accuracy rate is relatively low and resource consumption is large. At the same time, the vertical search engines represented by Qunar and Sogou Shopping have built their own databases for special fields, with heavy industry constraints, insufficient application flexibility, and unsatisfactory recall rates. ...

AI Technical Summary

Problems solved by technology

This method can establish a theme database with low redundancy, but if the correlation degree is used to sort, although the retrieval results are highly relevant to the theme, it reduces the overallness and cannot guarantee the reliability of the content; degree score ranking, the relevance between the search results and the topic cannot be guaranteed, resulting in "topic drift"

Images