Text Information Processing Technology Based on Multi Feature Fusion and Keyword Extraction in the New Media of Ideological and Political Education

The improvement and popularization of new media technology have enriched people's channels for obtaining information. However, there are many uncertain factors in the impact of complex ideologies and ideological cultures, which have brought great challenges to ideological and political education work. Based on the current situation of ideological and political education in the era of new media, this study proposes to use text information processing technology to improve people's ability to screen and identify information, by introducing dual channel text similarity calculation and multi feature keyword extraction methods. The results show that the proposed analysis algorithm has good accuracy and recall in text analysis, with maximum values of 91.05% and 93.97%, significantly better than other comparative algorithms. And its mean square error and average absolute error value in analyzing ideological and political education content do not exceed 0.2, indicating good data processing ability. The text information processing technology proposed in the study can provide technical means for the application of new media technology in ideological and political education and classroom teaching, and improve the quality of teaching.