MohammadSadegh Zahedi - Department of Electrical and Computer Engineering, University of Tehran, Tehran. Iran
Abolfazl Aleahmad - Department of Electrical and Computer Engineering, University of Tehran, Tehran. Iran
Masoud Rahgozar - Department of Electrical and Computer Engineering, University of Tehran, Tehran. Iran
Farhad Oroumchian - Department of Computer Science and Engineering, University of Wollongong in Dubai, Dubai. UAE

Recently, user generated content is growing rapidly and becoming one of the most important sources of information in the web. Blogosphere (the collection of blogs on the web) is one of the main sources of information in this category. User’s information needs in blogosphere are different from those of general web users. So, it is necessary to present retrieval algorithms to the meet information need of blog users. In this paper, we focus on the blog feed search task. The goal of blog feed search is to rank blogs regarding their recurrent relevance to the topic of the query. In this paper, the state-of-the-art blog retrieval methods are surveyed and then they are evaluated and compared in Persian blogosphere. Then we introduce our proposed method which is an extension of the voting model that one of the best blog retrieval methods. We have addressed the excessive dependency of the voting model on a list of initially retrieved top N relevant posts by using data fusion methods. Evaluation of the proposed algorithm is carried out based on a standard Persian weblogs dataset with 45 diverse queries. Our comparisons show considerable improvement over existing blog retrieval algorithms. The results show 67% increase in MAP score for the CompVote data fusion method.

Blog Feed Search; Blog Retrieval; Persian Blogosphere; Voting model