Research on Text Extraction and Analysis Based on Social Media

The onlineization of social networks is a typical feature of the big data era and one of the important reasons for the emergence of big data. Social media data contains rich information, which is carried by human language and contains a large amount of causal analysis and multi-dimensional description of events. It can provide powerful supplements to traditional information collection methods and has become a source of information for public opinion monitoring in recent years. This article takes Facebook as the main data source and uses Latent Semantic Analysis (LSA) method to automatically extract and express knowledge from large data corpora. Based on this, various methods are combined to conduct in-depth analysis of latent semantic features. The method established in this article can provide information supplementation for traditional public opinion monitoring.