Text visualization method for retaining unstructured text semantics

Abstract

The invention provides a social media text visualization method for retaining unstructured text semantics, which comprises the following steps of: S101, performing word segmentation, filtering and part-of-speech tagging on an input text to obtain a dependency relationship; S102, constructing a syntax binary tree based on the dependency relationship between the part-of-speech tagging and the vocabularies, calculating the emotional polarity of each text, and dividing the text set into a positive class and a negative class; S103, for the positive text and the negative text, generating a vocabulary sequence mode based on the co-occurrence relationship between the word frequency and the vocabularies in each text, and reserving semantics; S104, based on the weights occupied by the positive and negative text sets, distributing a visual space, and designing visual fonts and colors; S105, adopting a layout algorithm to display semantic relationships in the sequence modes and between the sequence modes; and S106, introducing an interaction design, so that the user can pay attention to local details. By the adoption of the method, visualization of the social media text is achieved, emotion orientation, viewpoint semantics and public opinion support degree of the text are clearly presented, text information is effectively presented, and text analysis is facilitated.

Description

technical field [0001] The invention relates to the technical field of data visualization, in particular to a text visualization method for preserving unstructured text semantics. Background technique [0002] The traditional disciplines involved in data visualization technology include scientific visualization and information visualization, the purpose of which is to extract information from big data, gain insights into knowledge, and present them in an intuitive way. In visualization technology, the visualization of text information is an important research branch. Text information visualization is to visually and intuitively display the semantic features contained in a large number of texts (such as the frequency of words, the importance of words, the logical structure of the text, the topic clustering of multiple texts, and the dynamic change trend of the topic, etc.). [0003] Typical visualization techniques in this article include word clouds (or tag clouds), sorting...

AI Technical Summary

Problems solved by technology

[0004] Existing visual models have achieved good results to a certain extent, but they either focus on perceiving the overall content, or focus on reflecting the semantics of the text itself, which limits the ability of text analysis

Images