Research On Topic Classification And Ranking Method Optimization Based On Social Annotations

With the concept of Web2.0'impact on the information technology, the network pays more attention to interaction with the web users. The social annotation is the emerging new data arising from the user's interaction with the network. The main purpose of utilization of social annotation is to improve the efficiency of the classification, retrieval and discovery of network resources by describing the interesting resources. In academic field, the precision of social annotation is higher than the common field due to the professional of academic users. Therefore, in academic field, the topic classification of resource can guide the user's retrieval and help user find out their interested resources. Meanwhile, topic classification can also be integrated into the retrieval system by filtering the topic of initial results, which can improve the efficiency of ranking method. In any retrieval system, relevance ranking method is the key technology to improve the quality of search results.In this paper, academic retrieval system is used as the platform. The first proposed method in this paper is to make use of social annotations as a complementary data source, as well as the content of academic conferences to the construction of the representation of academic conferences. Based on the representation of academic conferences, the proposed method utilize Na?ve Bayes classification algorithm to classify the academic conferences. The second proposed method in this paper is to optimize academic retrieval ranking method by using the ranking strategy of integrating the query-annotation similarity into query-content similarity. Therefore, with the emergence and increase of social annotation in academic field, this paper proposes two methods in order to improve the performance of academic retrieval system. One is the optimization of academic conference classification using social annotations, and the other is the optimization of ranking method using social annotations. The aim of the two methods proposed in the paper is to making use of social annotation, as the new attribute of network resources, to improve the quality of search results in the academic retrieval system. Comparative experimental results validate that social annotation can optimize the academic conference classification and ranking method in academic retrieval system.