| As the information data presents the geometric series expansion, obtain valuable knowledge from large data has become a hot research direction in the field of data mining. Researchers had been launched a lot of knowledge extraction and analysis of research at home and abroad, at the same time, knowledge acquisition research based on Chinese data has also been rapid development.This article will analyz review data of the film and television works, study a knowledge acquisition method based on grammar and knowledge ontology, the method is a kind of knowledge acquisition method for unstructured data in Chinese, from a lot of important knowledge in film and television areas which is stored in unstructured formats, we can acquire valuable and typical knowledge in the field of film and television. This method is also called analysis method and system of film’s reviews based on EKEL. Which aim to automatically acquire basic information and evaluative information which are useful in the film. Analysis method and system of film’s reviews based on EKEL is a text processing method that combine with grammar, the combination of film reviews grammar with EKEL form a text knowledge mining system, which can effectively and accurately acquire commentary information and knowledge in the field of film and television, we can acquire valuable and typical knowledge. Give a further analysis for the acquired knowledge, which can acquire more representative of knowledge in the field of film and television, then processing knowledge in the field of film and television, through transforming them into the forms of RDF and OWL, so that can be used in some specific areas, such as movie recommendation system and question answering system. In this paper, we firstly introduce the definition method of initial film review ontology, and its role in film and television works tags extraction. Then details the four parts of film reviews grammar system, respectively, load the initial film reviews ontology file, constants of film and television works, film reviews grammar patterns, processing subject of film reviews knowledge. Also details the definition method of each part and its role in the film reviews grammar system. After the operation mechanism and function of EKEL system are introduced, and explains how to use film reviews grammar system that combining EKEL system for extracting tags of film and television works. And then studied how to completely and systematically design film reviews grammar system, and illustrates the technical route of designing the film reviews grammar system in detail. Through designing of specific film reviews grammar and summarize the designing principles and methods of it. After using completed film reviews grammar system reviews, at the same time, combining EKEL and proceeding mining the tags of film and television works from the extracted text corpus of film and television works, this procedure can produce the parsing trees of EKEL. Afterwards, elaborately analyze and have a post processing on these parsing trees of EKEL, which can perfect film’s reviews grammar, optimized film grammar system with clear structure, reasonable description, a strong ability of generalization and extract tags. In order to verify the validity of the research on the method of knowledge extraction presented in this paper, design extraction tags experiments of film and television works. In the end, then research on analysis method of tags which had extracted from studied the film and television works. In the process of extracting experiment, study an analysis method of tags extraction effect of film and TV works, we make a concrete analysis of the precision rate and recall rate through a kind of tags, analysis of the experimental result and summarizes the advantages and in need of improvement that use the knowledge extraction method in this article research. Finally summarizes the significance of knowledge acquisition method based on film reviews grammar and knowledge ontology in film and TV areas, and the next research of this knowledge acquisition method is discussed. |
PDF Full Text Request