KDSTM: Neural Semi-supervised Topic Modeling with Knowledge Distillation

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Xu, Weijie, Jiang, Xiaoyu, Desai, Jay, Han, Bin, Yan, Fuqin, Iannacci, Francis
Format: Preprint
Published: 2023
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866909101114523648
author Xu, Weijie
Jiang, Xiaoyu
Desai, Jay
Han, Bin
Yan, Fuqin
Iannacci, Francis
author_facet Xu, Weijie
Jiang, Xiaoyu
Desai, Jay
Han, Bin
Yan, Fuqin
Iannacci, Francis
contents In text classification tasks, fine tuning pretrained language models like BERT and GPT-3 yields competitive accuracy; however, both methods require pretraining on large text datasets. In contrast, general topic modeling methods possess the advantage of analyzing documents to extract meaningful patterns of words without the need of pretraining. To leverage topic modeling's unsupervised insights extraction on text classification tasks, we develop the Knowledge Distillation Semi-supervised Topic Modeling (KDSTM). KDSTM requires no pretrained embeddings, few labeled documents and is efficient to train, making it ideal under resource constrained settings. Across a variety of datasets, our method outperforms existing supervised topic modeling methods in classification accuracy, robustness and efficiency and achieves similar performance compare to state of the art weakly supervised text classification methods.
format Preprint
id arxiv_https___arxiv_org_abs_2307_01878
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle KDSTM: Neural Semi-supervised Topic Modeling with Knowledge Distillation
Xu, Weijie
Jiang, Xiaoyu
Desai, Jay
Han, Bin
Yan, Fuqin
Iannacci, Francis
Computation and Language
Artificial Intelligence
68T50
I.2.6
In text classification tasks, fine tuning pretrained language models like BERT and GPT-3 yields competitive accuracy; however, both methods require pretraining on large text datasets. In contrast, general topic modeling methods possess the advantage of analyzing documents to extract meaningful patterns of words without the need of pretraining. To leverage topic modeling's unsupervised insights extraction on text classification tasks, we develop the Knowledge Distillation Semi-supervised Topic Modeling (KDSTM). KDSTM requires no pretrained embeddings, few labeled documents and is efficient to train, making it ideal under resource constrained settings. Across a variety of datasets, our method outperforms existing supervised topic modeling methods in classification accuracy, robustness and efficiency and achieves similar performance compare to state of the art weakly supervised text classification methods.
title KDSTM: Neural Semi-supervised Topic Modeling with Knowledge Distillation
topic Computation and Language
Artificial Intelligence
68T50
I.2.6
url https://arxiv.org/abs/2307.01878