Design and implementation of sentiment analysis system based on rough set
Liangliang Zhao, Minling Zhu, Min Zhao · 2021
Aiming at the difficulty of sentence analysis caused by the fuzziness of words, this paper uses attribute reduction algorithm to generate emotion judgment rules based on rough set theory. Firstly, we extract features of the corpus sample that is the emotional words in the corpus, then through attribute reduction to reduce the extracted emotion feature words for text training and to generate rules for judging the emotion of the text. Secondly, we also extract the features of emotion words from the test corpus and compare with the trained rule text to achieve the classification of their emotions. Through experimental verification we know that different dimensions eigenvalues and boundary values have an important influence on the accuracy and recall rate of classification.