Knowledge Discover Based On Law FrameNet Ontology Corpus System

With the continuous development of corpus technology and the corpus management system widely used, the amount of data stored in corpus increasing sharply. But the function of most corpus system is only to store the corpus, people get information through these corpus is only a small part of the whole information contained in corpus, because the current used for corpus analysis and processing tools are very little, and there are limitations. However, after the hidden data in the corpus, more important information has not been tapped, these information often has important reference value to linguists and natural language processing.At present, Shanxi University, department of Management build a Law FrameNet Ontology corpus system, this is a ontology-based corpus which is achieved on the Law FrameNet Ontology management, and store a large number of ontology instances——Raw material and annotated material. Therefore, the paper has research on knowledge mining of Law FrameNet Ontology corpus system.In this paper, the theory of knowledge discovery and knowledge discovery model are expound in the first and second chapter, we used a generally model to discover the knowledge of corpus data. The third chapter, describes the structure of law FrameNet corpus, put forward the principle of system construction, discusses the system model structure and database design and introduces five functions of the system. The fourth chapter is the key of this paper, explains knowledge discovery process and method based on raw materials, using three forms:the text feature extraction, text categorization and text similarity calculation to mining knowledge of raw corpus, and demonstrates the process and results of experiments. Chapter V is also the focal point of this paper, expounds knowledge mining of annotated materials, firstly raw materials are annotated into the annotated materials, then statistics the frames, frame elements, and semantic features of lexical finally gives the experimental result. ChapterⅥdoes a simple introduction with Law FrameNet Ontology corpus system. In the last chapter, makes a summary of the system and raises recommendations and prospects for future work.By research of knowledge discovery based on Law FrameNet Ontology corpus system will help law linguistics and natural language process doing more in-depth research, and making foundation for the future knowledge reasoning system and knowledge question-answering system.