What are the main methods of text mining?

Text mining has always been a very important field of information processing, because we all need the power of text mining, whether it is recommendation system, search system or other extensive applications. The amount of information generated every day is increasing rapidly, and these information are basically unstructured massive texts, which are not easy to be processed and perceived by computers. Therefore, we need some effective techniques and algorithms to find useful patterns. Text mining has attracted extensive attention in recent years, and it is a task to extract effective information from text files.

In recent years, text mining (TM) has attracted extensive attention due to the amazing amount of text data in various forms (such as social networks, medical records, medical insurance data, news releases, etc.). ). IDC predicted in a report that by 2020, the amount of data will increase to 40 billion TB (4 * (10 22) bytes), that is, it will increase 50 times from 20 10 [50].

Text data is typical unstructured information, and it is one of the simplest data forms that can be generated in most cases. Humans can easily process and perceive unstructured texts, but machines are obviously difficult to understand. Needless to say, these texts must be a valuable source of information and knowledge. Therefore, it is urgent to design a method that can effectively deal with unstructured text in various applications. At present, the main text mining methods are: