700 Journals and 15,000,000 Readers Each Journal is getting 25,000+ ReadersThis Readership is 10 times more when compared to other Subscription Journals (Source: Google Analytics)
Original Articles Open Access
As one of the important applications of Web2.0 technology, blog attracts more and more users. Writing and browsing blog has become a popular hotspot of network culture, which promotes the development of blog search service. But, the current blog search engines are mostly only based on matching query keywords; lack the ability of automatically extracting users’ interests and recommendation. Really Simple Syndication (RSS) is a format of describing website and keeping synchronization with website content. Using RSS to aggregate blog posts has the advantage of letting users get the latest update of blog posts. However, the posts collected by RSS don’t always attract users; users still need to browse every subscription post to find the interesting posts. To address this problem, the time spent by users on reading blog posts is viewed as a key factor to measure the users' interests. In this paper, we firstly used probabilistic latent semantic analysis (PLSA) to discovery the topics of blog posts, then adopted Naive Bayesian algorithm to classify the blog posts which was primarily connected with the users’ reading time, and lastly ranked and recommended the unread interesting posts to users. Experiments showed that our proposed method could recommend the favorite blog posts to users according to the users' browsing interests.
To read the full article Peer-reviewed Article PDF
Author(s): Lin Cui Caiyin Wang and Xiaoyin Wu
Blog Posts Recommendation, Probabilistic Latent Semantic Analysis, Naive Bayesian Classification Algorithm, Really Simple Syndication, Reading Time, development of blog search