Research On Topic Map-Based Tag Semantic Mining

In recent years, with the popularity of flicker, Del.icio.us, watercress and other Web2.0sites, a new type of network information taxonomy-folksonomy began to be applied to more and more network systems. Folksonomy is described by the network user information for a spontaneous define a set of labels, and ultimately selected based on the frequency of high-frequency tag label is used as a method of network information to the class name of the classification of the information. Its advantage is that it has neither a strict classification, indexing language unrestricted. This freedom free, convenient and flexible classification despite the network of users, however, these features led to a diversity of labels, among ambiguity, non-hierarchical organization and lexical semantic relations lack of defects, these defects only use of network information resources to the organization brings some difficulties, but also difficult to adapt to the requirements of the Semantic Web.Topic Maps as a new digital information organization and knowledge representation technology to absorb the idea of ontology and semantic web. Thematic maps from the TAO three elements, T representatives Topic, A representative of Association, O behalf of Occurrence. The three elements constitute the subject layer and resource layer, and a subject having resource separation characteristics. Themes layer through thematic elements and associated elements reveal the relationship between topics, forming a semantic network, by linking themes and resources makes it easy for users to locate resources needed. Due to the characteristics and advantages of thematic maps of the network, its adaptability to the digital environment, Thematic map known as the world of information GPS. Because of thematic maps not only relative to other methods of information organization has a huge advantage, it will not only meet the needs of the label information organization also can effectively solve the problems of the label information resources exist.This paper is divided into six chapters, The first chapter introduces the background and significance of this topic, labels, tags and labels semantic mining research status topic map, while clearly the research content and research methods.The second chapter clarifies the concept of labels and thematic maps, basic characteristics, analyzes the semantic label study some shortcomings in the existing mass classification, while the other topic map technology and information organization by way of comparison, the advantageous features described topic map, for the use of topic maps label semantic mining technology for laying the groundwork.The third chapter first discusses the semantic tags used in mining analysis and thematic map labels semantic integration, feasibility semantic retrieval, knowledge navigation, analyze the feasibility of combining labels and thematic maps, combined with further proposed mechanism label next topic map.The fourth chapter presents the label semantic mining model based on topic maps, the major functional modules model are explained in detail. Introduction method of establishing the model label and theme topic map repository database, then explain the label topic maps used in information specific application knowledge in the service module.The fifth chapter of "watercress movie" label as an example, the use of topic maps Ontopia environment-related tools, label theme map generation, ontology editing, page views, visualization, navigation and other operations. Discussion topic map construct Chinese labels, to achieve aggregation associated label, the label will be hidden semantic relationships between dominance.Chapter VI summarizes the results obtained in this paper and shortcomings, propose future research ideas.