The Research And Implementation Of Text Information Mining In Police Business Field

Text mining technology is one of the key technologies in the computer data mining. At present, text mining technology has achieved great development in many fields, however, due to the date privacy and its unique complexity, text mining technology develops slowly in public security information field. With the continuous development of network information, public security information department received more and more text data, this made it urgently needs a kind of intelligent way to deal with the collected text data. So based on the public security business text mining become an important issue for public security information technology department.At present, problems which need to be solved are on several main technical points,such as word segmentation of Chinese, text clustering, text classification and so on in public security text data mining, I analyzes the characteristics of text mining in public security,then made a job focusing on text clustering as follow:We study the text clustering algorithm. We analyzes some common text clustering algorithm and points out its insufficiency in public security text clustering.We studied the structure characteristics of public security text. We analyses the sources of public security. text date and the commonly used date structure characteristics. Aimed at the characteristic of date structure, we proposed satisfy text clustering must comply with some conditions.We mainly studied neural network in the application of data clustering. We analyses two kinds of model which we call them SOM and GSOM, and we proposed an improved neural network model based on both the advantages and disadvantages between the SOM and GSOM. Meanwhile we do a simulated clustering experiment based on the new proposed neural network model. Experiments have proved that this model has many advantages in clustering algorithm.Finally, according to the characteristics of public security text data, we apply the new neural network model to the public security of text mining, and this method solve the topic which text mining used in public security system.