Feature-based informal short text information analysis and research method

Abstract

The invention relates to the technical field of intelligence analysis and research, and discloses a feature-based informal short text intelligence analysis and research method.The output end of an original data source is electrically connected with a data center, and the output end of the data center is electrically connected with a feature extraction module; the output end of the feature extraction module is electrically connected with an information acquisition module, the output end of the information acquisition module is electrically connected with an information classification and retrieval module, and the output end of the information classification and retrieval module is electrically connected with an emotion analysis and research module. Short texts in the text data are extracted through a paragraph extraction and dispersion module, a short text extraction module and an entry extraction module, and obtained paragraph data, short text data and entry data enter an information acquisition module; the short texts and entries in a large amount of text data are classified by the classification module through the short text distinguishing module and the entry distinguishing module according to different emotions, and the short texts and the entries in the large amount of text data are rapidly extracted and numbered.

Description

technical field [0001] The invention relates to the technical field of information analysis and research, in particular to a characteristic-based informal short text information analysis and research method. Background technique [0002] Text analysis research is to quantify the emotional information expressed in the text, and study the positive, negative or neutral polarity expressed in a piece of text to analyze the emotional tendency of the text. In recent years, there have been a lot of research on text sentiment analysis. Generally, it can be divided into two categories according to the nature, one is based on supervised sentiment analysis research, the other is based on unsupervised sentiment analysis research, and there are also some semi-supervised sentiment analysis research, but the existing text intelligence analysis research methods are still There are some shortcomings, for example, the existing intelligence analysis research method has weak ability to distingui...

AI Technical Summary

Problems solved by technology

[0004] Aiming at the deficiencies of the prior art, the present invention provides a feature-based informal short text intelligence analysis research method, which can quickly extract and number short texts and entries in a large amount of text data, and at the same time pass sentiment analysis according to comparative documents. It saves the time of material collection, facilitates follow-up analysis and research and can quickly search and analyze related entries according to needs, improves work efficiency, etc., and solves the problem of existing information analysis and research methods. The ability to discriminate the emotional polarity of statements in short texts is weak, and the extraction efficiency of overall emotional tuples is not high

Images