Recommending Diverse and Relevant Queries with A Manifold Ranking Based Approach
Xiaofei Zhu, Jiafeng Guo, Xueqi Cheng · 2010
Query recommendation has been considered as an e ective way to help search users in their information seeking activities. Traditional approaches mainly focused on recommending alternative queries with close search intent to the original query. However, to only take the relevance into account may generate redundant recommendations which provide almost the same information for users. Therefore, it is important to provide diverse as well as relevant query recommendations. In this way, we are able to cover multiple potential search intents of users and attract more clicks over recommendations. Besides, previous query recommendation approaches mostly relied on measuring the relevance or similarity between queries in the Euclidean space. However, there is no convincing evidence that the query space is Euclidean. Therefore, it is more natural and reasonable to assume that the query space is a manifold. In this paper, we aim to recommend diverse and relevant queries based on the intrinsic query manifold. We propose a unified model, named manifold ranking with stop points, for query recommendation. Specifically, by introducing stop points into query manifold, our approach can iteratively rank queries for recommendation by simultaneously considering both diversity and relevance between queries in an unified way. Empirical experimental results show that our approach can e ectively generate highly diverse as well as closely related query recommendations.