Study On The Kazakh Named Entity Recognition Method Based On N-gram Model

Named entity is a basic text information elements, it is the basis of correct understanding of an article. Named entity recognition plays an important role in machine translation, text retrieval and so on. Currently, English and Chinese Named entity recognition have made many results, Kazakh named entity recognition, however, is still in the research stage. Therefore, the research of Kazakh named entity recognition has great theoretical and practical significance.In this paper, analyzing the research on domestic and foreign named entity recognition, combining with the feature of Kazakh named entity recognition, we propose a Kazakh reliability Calculate method based on N-gram language model to research Kazakh named entity recognition, design and realize a Kazakh named entity recognition system. It can well complete the identification of Kazakh named entity and enable people to get effective information from the text promptly, the system has some application value. Finally, we test the system with one month's corpus of Xingjiang Daily of Kazakh language version and get satisfactory results. The results show that the accuracy, recall and F values were all more than 60%.