Research Article
A Study on the Optimization of Text Emotional Classification Based on Class Representation Words and Naive Bayes
@INPROCEEDINGS{10.4108/eai.18-11-2022.2327117, author={Hongyu Liu}, title={A Study on the Optimization of Text Emotional Classification Based on Class Representation Words and Naive Bayes}, proceedings={Proceedings of the 4th International Conference on Economic Management and Model Engineering, ICEMME 2022, November 18-20, 2022, Nanjing, China}, publisher={EAI}, proceedings_a={ICEMME}, year={2023}, month={2}, keywords={text mining; naive bayes; classification}, doi={10.4108/eai.18-11-2022.2327117} }- Hongyu Liu
Year: 2023
A Study on the Optimization of Text Emotional Classification Based on Class Representation Words and Naive Bayes
ICEMME
EAI
DOI: 10.4108/eai.18-11-2022.2327117
Abstract
There is a certain similarity of the characteristic vectors of the comment statements between the comments between the commentary emotion category, bringing certain difficulties to text emotion 3 classification. This article proposes a specification of the specification. It is mainly from 8 aspects of fresh products to 8 aspects (emotional adjustment words + feature words). Add all kinds of representatives selected to a word-defined dictionary, by changing the word words in the comment statement, in turn increases the difference between the comment statement feature vector between the category during the quantization process, thereby enabling classification Easy to identify the emotional polarity of comment statements. After verification, this method achieved a good classification effect, the correct rate reached 86.14%, the recall rate reached 84.97%, which increased 4.24% and 4.42% compared to the method without dictionary.
