Research Article
A Weakly Supervised Text Classification Method Based on Vocabulary Construction
@INPROCEEDINGS{10.1007/978-3-031-30237-4_12, author={Peidong Li and Di Lin and Zijian Li}, title={A Weakly Supervised Text Classification Method Based on Vocabulary Construction}, proceedings={Machine Learning and Intelligent Communication. 7th EAI International Conference, MLICOM 2022, Virtual Event, October 23-24, 2022, Proceedings}, proceedings_a={MLICOM}, year={2023}, month={4}, keywords={Text Classification Weak Supervision Word Vocabulary}, doi={10.1007/978-3-031-30237-4_12} }- Peidong Li
Di Lin
Zijian Li
Year: 2023
A Weakly Supervised Text Classification Method Based on Vocabulary Construction
MLICOM
Springer
DOI: 10.1007/978-3-031-30237-4_12
Abstract
Text classification is an important research direction in natural language processing. The computer can automatically classify and label texts according to certain classification standards through text classification technology. Traditional text classification tasks require a large amount of labeled data. However, human-labeled data is not only expensive, but also susceptible to the subjective consciousness of the labelers. Therefore, unsupervised text classification using computers becomes relevant. In most cases, the label name of each category is instructive for the classification task. In this paper, we design a weakly supervised text classification method. This method only needs to provide the label names that guide the classification to complete the automated text classification. Our method was tested on several publicly available datasets and performed well.
