Knowledge Discovery of News Text Based on Artificial Intelligence

Abstract

The explosion of news text and the development of artificial intelligence provide a new opportunity and challenge to provide high-quality media monitoring service. In this article, we propose a semantic analysis approach based on the Latent Dirichlet Allocation (LDA) and Apriori algorithm, and we realize application to improve media monitoring reports by mining large-scale news text. First, we propose to use LDA model to mine news text topic words and reducing news dimensionality. Then, we propose to use Apriori algorithm to discovering the relationship of topic words. Finally, we discovery the relevance of news text topic words and show the intensity and dependency among topic words through drawing. This application can realize to extract the news topics and discover the correlation and dependency among news topics in mass news text. The results show that the method based on LDA and Apriori can help the media monitoring staff to better understand the hidden knowledge in the news text and improve the media analysis report.