The Design And Implementation Of The Hot Education News Topic Detection System

With the development and application popularization of Internet technology, the network has become a kind of important information channel for public to participate in social life.Education-related issue has been public concern. The implementation of each policy will lead to a wide range of comments and conjecture, which creats huge pressure on the publicity and Journalism.To be able to manage and analyse the development of network hot spots,education management introduce the hot education news analysis system.Hot topic detection system, which this paper studies, is the subsystem of the hot education news analysis system.The system crawling from the social media data, using data mining related technologies, finding the intrinsic link between the mass press, and making cluster news into topics, implements a real-time hot topic detection.This paper which analyse the requirment of topic detection system, studies the features of education news field, designs a modular hot education news topic detection system.By discussing and comparing the text extraction, text categorization, text clustering and other related technologys, analyzing their advantages, disadvantages and scope, and analyzing network data features of education news, this paper designs a modular and configurable system of the hot topic detection and correction, and designs the based on word frequency Feature extraction strategy, the based on word group match text classification strategy and the based on threshold 1NN text clustering strategy. By using the Java language to implement the whole system, and adjusts parameters to test the quality of data, this paper achieves a satisfactory of topic set results and system evaluation results, completes the system integration work, and achieves a real-times hot education topic detection. This paper is the important basis of the hot education news analysis system, is a important complement of automated network information management, and is a certain reference to the network hot spots analysis.