Research And Application Of Blog Recommendation Technology Based On Similarity

With applications of Web2.0, the propagation speed of the blog has obtained anunprecedented development, which has a huge information resource. In such a largenumber of blog system, it becomes very difficult for the user to find the most interestedblogs or blog articles and bloggers to make their own blog to gain a higher number ofvisits. The problem is solved due to the advent of the blog search engines, but it cannotreally meet the needs of users because of technology and requirements reasonsThe common recommendation algorithm has been studied and social informationof bloggers and blog articles have been analyzed. A blog recommendation algorithmbased on the similarity has been designed on existing technology, similarity of blog hasbeen researched from the aspects of the social information of the bloggers and blogarticles. the concepts of the blog article similarity of blogs and bloggers socialinformation similarity are introduced before the algorithm design, and the advantages ofthe similarity algorithm is described. The similarity weight value is determined usingthe linear combination method, and the contents of the references are combined todetermine its size. The experimental part uses the open source crawler tools (Heritrix) tograb the related blogs contents on Sina website as the experimental data, and crawl backto the data processing, and then the data stored in the database. Data crawled back fromthe website are processed and then stored in database.Two criteria are used to evaluate the improved algorithm: One method is used tocompare the accuracy with the text algorithm, which is suitable for automatic evaluationfor computer; Another method is to judge the similarities between the recommendedblog and target blog through human intervention. the comparison and analysis ofexperimental results prove the effectiveness of the improved algorithm which providetechnical support for the blog recommendation.