Chinese Terminology System Design And Implementation Based On Maximum Entropy

In reading academic information, people always hope to be able to quickly and accurately grasp the terms and definitions in the article, so that they can better understand and grasp the theme. In natural language processing work, the researchers also hope that through the term recognition for other deep research to provide support. In this article, we research the Chinese term extraction and Chinese term definition extraction and design a Chinese term extraction system based on maximum entropy, which includes Chinese term extraction subsystem and Chinese term definition extraction subsystem.The Chinese term extraction subsystem,which uses a large number of existing terms to study the characteristics of the internal structure of terms,is based on maximum entropy term extraction.With the aid of maximum entropy tools to train classification model, candidate terms are classified to determine whether they are terms or not through the model.Evaluation results show that the term extraction accuracy of the system can reach63.75%, and the recall rate can reach78.84%. The Chinese term definition extraction subsystem, which draws and inquires the precise definitions of the term word from the network, combines multiple network knowledge bases.At the same time, the maximum-entropy-based Chinese term extraction system in this paper is also based on AJAX Technology, and provides the user feedback function-the system upgrades itself through the errors submitted by the users.The system can extract the terms, and the definition of the terms accurately and quickly to Chinese text. The system interface is simple and beautiful, bringing great convenience to the user.