Generating temporal semantic context of concepts using web search engines

In this paper, the problem of generating temporal semantic context for concepts is studied. The goal of the proposed problem is to annotate a concept with temporal, concise, and structured information, which can reflect the explicit and faceted meanings of the concept. The temporal semantic context can help users learn and understand unfamiliar or newly emerged concepts. The proposed temporal semantic context structure integrates the features from dictionary, Wikipedia, and LinkedIn web sites. A general method to generate temporal semantic context of a concept by constructing its associated words, associated concepts, context sentences, context graph, and context communities is proposed. Empirical experiments on three different datasets including Q–A dataset, LinkedIn dataset, and Wikipedia dataset show that the proposed algorithm is effective and accurate. Different from manually generated context repositories such as LinkedIn and Wikipedia, the proposed method can automatically generate the context and does not need any prior knowledge such as ontology or a hierarchical knowledge base. The proposed method is used on some applications such as trend analysis, faceted exploration, and query suggestion. These applications prove the effectiveness of the proposed temporal semantic context problem in many web mining tasks.