| Ontology is in the "definition layer" of vocabulary in Seven-layer model of Semantic Web,which is used to describe all kinds of concepts to the computer,to achieve the concept of human-computer interaction.In the Internet era,the search for the character information needs to be screened out from the mass of results for the specific character,and ontology can be used to solve the problem of ambiguity between different characters.This thesis designed a news ontology automatic construction model according to the demand of character information retrieval during news publishing,and analyzed the important and difficult points of automatic character news ontology construction by experimental study on the model.The main contents of this work are as follows:1.This work studied the character entity recognition,which is the essential point of automatic construction of character news ontology,and optimized the process of name recognition and name disambiguation.2.According to the understanding of the ontology concept and ontology construction,this work constructed the basic framework of simple character news ontology,combining the character-related terms in Baidu Encyclopedia and character-related news contents,and designed and implemented the automatic refinement of ontology construction based on the extracted content resources.This work applied the collected information of Baidu Encyclopedia to construct the basic character individuals of character-related news ontologies;constructed the news individuals of character-related news ontology after processing the acquired news data;and refine the basic information of individuals based on the news information.3.Based on the analysis of character-related terms in Baidu Encyclopedia,the content organization of character-related news,and content characteristics,this work designed and implement the extraction process of character-related terms in Baidu Encyclopedia and character-related news contents.This work summarized the key issues of the automatic construction of ontology on character news according to the preliminary validation and analysis on the results of the automatic construction process.This thesis constructed a simple ontology on character news and designed a system to implement the automatic construction of ontology.The preliminary experiments can associate the characters with the news.The constructed ontology on character news provides the Knowledge Service about character-related information in terms of single characters.It can also provide relatively accurate resources for the editing of character news and simplify the work of information collecting and processing.This work has certain practical value,and at the same time,we proposed a solution to further improve the system by the research on the automatic construction of ontology. |
PDF Full Text Request