A lightweight common webpage topic crawler method based on a search engine comprises the steps as follows: initializing seeds, namely extending a small amount of given related seeds describing specific topics to a series of seeds according to a certain rule; discovering websites, namely converting the initialized seeds into query words, and obtaining a plurality of related websites through an interface of the search engine; downloading the websites, namely downloading the related websites to a local machine and storing the related websites into a database; analyzing web pages, namely analyzing the downloaded websites to obtain link information in the websites; updating the seeds, namely analyzing crawled new websites, extracting topic related words from the crawled new websites, creating new seeds by the topic related words, and guaranteeing a crawling process to be continuously proceeded; updating the crawling process, namely, calculating a re-crawling cycle according to the update information of the crawled websites, and enabling the crawled websites to be automatically updated in a self-adaption way. The method has the characteristics of low cost, simplicity for realization, high efficiency, accuracy and the like.