Multi-Label Text Classification with Transfer Learning
Publication Date
1-1-2024
Document Type
Conference Proceeding
Publication Title
9th IEEE/ACIS International Conference on Big Data, Cloud Computing, and Data Science, BCD 2024
DOI
10.1109/BCD61269.2024.10743077
First Page
21
Last Page
26
Abstract
Multi-label text categorization is a crucial task in Natural Language Processing, where each text instance can be simultaneously assigned to numerous labels. In this research, our goal is to assess how well several deep learning models perform on a real-world dataset for multi-label text classification. We employed data augmentation techniques to address the problem of data imbalance and evaluated the effectiveness of several deep learning architectures, as well as fine-tuned pretrained models. We also discovered that models performed better when data augmentation approaches were used. Our study shows that pretrained models are effective for multi-label text classification, with good performance across various metrics. However, the performance may vary on different datasets and require fine-tuning to achieve optimal results. This study also highlights the importance of addressing data imbalance and the potential benefits of using pretrained language models for this task.
Keywords
data augmentation, deep learning, multi-label text classification, Natural Language Processing, pretrained models, random word substitution, synonym replacement, transfer learning
Department
Computer Science
Recommended Citation
Likhitha Yelamanchili, Ching Seh Mike Wu, Chris Pollett, and Robert Chun. "Multi-Label Text Classification with Transfer Learning" 9th IEEE/ACIS International Conference on Big Data, Cloud Computing, and Data Science, BCD 2024 (2024): 21-26. https://doi.org/10.1109/BCD61269.2024.10743077
