Personalized Diversity Search Based on User’s Social Relationships

Ming Li,Juanzi Li,Lei Hou,Hai-Tao Zheng

DOI: https://doi.org/10.1007/978-3-642-35527-1_55

2012-01-01

Abstract:Keyword-based web search is nowadays the most popular and convenient means for people to access information, whereas traditional search methods usually disappoint people with inaccurate, insufficient or redundant results, because those methods often fail to understand user’s search intents and interest preference. Diversity search is an effective approach to present various kinds of results so that average people may satisfy with as least one result, however, most existing diversity search methods are uniformly applied to all users and queries, the returned results generally reflect the masses’ needs, individual users’ requirements are not fully considered. To deal with this issue, we present a systematic method named personalized diversity search based on user’s social relationships (PDSSR), this method is a combination of personalization and diversification, which enables computer better understand user’s search intents and interests, consequently returns a personalized and reduced diversified result set. Besides, we introduce social relationships into the personalization, which helps to avoid “cold start” and “data sparsity” problems. Empirical experiments conducted show that the proposed method outperforms the baseline in terms of nDCG, which proves the effectiveness of our method.