Design Of Automatic Term Extraction System And Study Of Key Techniques

Automatic Term Extraction is an important issue in Chinese Information Processing.Term represents the core knowledge in one scientific field. The development of term reflects the development of this field to some extent. Term Extraction not only plays an important role when doing research in Information Retrieval, Information Extraction, Data Mining, Machine Translation, Ontology and other parts of Nature Language Processing, but also shows its theoretical and real significance of keeping up with the pace of world in this field.Nowadays, research on Chinese Information Processing focuses on Chinese Automatic Word Segmentation, Parsing, but seldom in Automatic Term Extraction. The extraction and retrieval of term usually rely on foreign research achievements based on western languages, but it is not quite suitable for research based on Chinese, so to develop suitable way for Chinese Term Extraction with the help of western methods is very important for Chinese term standardization as well as Chinese Information Processing.Aiming at the present situation, this paper pays much attention to the research of Automatic Term Extraction techniques, the main work are as follows:1 .Analysis and compare all kinds of Natural Language Processing Model.2.Design an automatic term extraction system, and make the system be an optional one considering different purposes.3.Propose a novel Automatic Term Extraction Model based on Linguistically-Oriented and Statistically-Oriented Approach, many disadvantages are avoided in our Model.4. Propose the method using two statistic parameters mi_f and logL_f in automatic term extraction system, then implement the key part of the system by using this algorithm. The results of the evaluation indicate that this algorithm ismore efficient and robust than previous methods.