Hot Topic Detection Based on Image Intelligent Understanding and Comment Mining

Micro-blog has become the main distribution center of public opinion. Therefore, how to mine hot topics from microblog has become a hot topic. However, microblogs are mostly short texts and contain a large number of invalid information, which significantly increases the difficulty of hot topic detection. On the other hand, micro-blog includes text, pictures and comment information, which greatly enriches the diversity of information. Therefore, this paper proposed a new method, which applies deep learning to image understanding. Image content is interpreted as analyzable text semantic information combined with comments to enrich short text information. Experimental results show the effectiveness of the proposed method, and better experimental results are obtained compared with other methods.