Representation Learning for Short Text Clustering

TL;DR

This paper introduces two autoencoder-based methods to enhance short text representations derived from pre-trained models, significantly improving clustering performance by exploiting structural information and clustering constraints.

Contribution

It proposes novel unsupervised autoencoder techniques, STN-GAE and SCA-AE, to fine-tune pre-trained models for better short text clustering.

Findings

SCA-AE improves clustering accuracy by up to 14% over BERT alone.

Pre-trained models like BERT outperform traditional methods in short text clustering.

Tuning pre-trained representations with proposed methods enhances clustering performance.

Abstract

Effective representation learning is critical for short text clustering due to the sparse, high-dimensional and noise attributes of short text corpus. Existing pre-trained models (e.g., Word2vec and BERT) have greatly improved the expressiveness for short text representations with more condensed, low-dimensional and continuous features compared to the traditional Bag-of-Words (BoW) model. However, these models are trained for general purposes and thus are suboptimal for the short text clustering task. In this paper, we propose two methods to exploit the unsupervised autoencoder (AE) framework to further tune the short text representations based on these pre-trained text models for optimal clustering performance. In our first method Structural Text Network Graph Autoencoder (STN-GAE), we exploit the structural text information among the corpus by constructing a text network, and then adopt graph convolutional network as encoder to fuse the structural features with the pre-trained text features for text representation learning. In our second method Soft Cluster Assignment Autoencoder (SCA-AE), we adopt an extra soft cluster assignment constraint on the latent space of autoencoder to encourage the learned text representations to be more clustering-friendly. We tested two methods on seven popular short text datasets, and the experimental results show that when only using the pre-trained model for short text clustering, BERT performs better than BoW and Word2vec. However, as long as we further tune the pre-trained representations, the proposed method like SCA-AE can greatly increase the clustering performance, and the accuracy improvement compared to use BERT alone could reach as much as 14\%.