Automatic topic detection with an incremental clustering algorithm
WISM'10 Proceedings of the 2010 international conference on Web information systems and mining
Unsupervised topic detection model and its application in text categorization
Proceedings of the CUBE International Information Technology Conference
Hi-index | 0.00 |
In order to solve the two problems that the difficulty in distinguishing similar topics and topic center drift, a topic detection method based on bicharacteristic vectors is proposed. By this new method, named entity of information will be applied to the character representation and similar topics are differentiated by combining vectors of the named entities and keywords which can express the center vector of the topic more accurately. Then it deals with topic drift by single-pass clustering and continual modification of the topic center. The result of experiments shows that the new method can reduce the rate of missing and false reports, and improve the performance of topic detection effectively.