Crawling Strategy of Focused Crawler Based on Niche Genetic Algorithm

摘要

In order to improve the search efficiency of focused crawler, we design a new crawling strategy based on the niche genetic algorithm. Rather than colleting and indexing all accessible hypertext documents to be able to answer all possible ad-hoc queries, the new crawling strategy, combined the advantages of hyperlinks structure and web content strategies, uses hyperlink as genetic individual and topic-keywords based VSM is used to evaluate individual fitness, and imports new URLs to implement crossover and mutation, and the URLs that have the same prefix are regarded as niche. Guide the crawl direction by niche genetic algorithm to selectively seek out pages that are likely to be most relevant to a pre-defined set of topics. Compared with the other algorithms, experiments show that the strategy has higher precision and recall in searching the topic pages.