The hot topics evaluation and key issues in online news about emergency events

This paper concentrates on the evaluation system of hot topics in the online news about the emergency events. First, the overall technology framework of the system was established and description was formulated on the key issues need to be resolved prior to the evaluation of the hot topics. Focusing on the vector representation of the news documents and clustering algorithm of the news topics, a modified TF-IDF text representation model was formulated and an improved selection method of initial topic clustering center was established. The evaluation model of the hot topics was then established to evaluate the heat value of the clustered topics based on extracting websites feature parameters, such as the time properties of the news reports, reporting properties, and user attention. Finally, the paper takes news of 2011 Japan Earthquake as data source for algorism and model evaluation. The results showed that the identification and evaluation system of hot topics in online news about the emergency events was valid, and the evaluation result of hot topics by the systems agreed with the expected results. It lays a foundation for the subsequent studying about the tracking and evolution of the hot topics.