AutoLabel-ABSA: A Comparative Study of Transformer Embeddings and Clustering Algorithms for Automatic Text Labeling and Aspect-based Categorization

Marutho, Dhendra and Basuki, Ruri Suko and Muljono, Muljono and Giap, Yo Ceng (2026) AutoLabel-ABSA: A Comparative Study of Transformer Embeddings and Clustering Algorithms for Automatic Text Labeling and Aspect-based Categorization. International Journal of Intelligent Engineering and Systems, 19 (3). pp. 360-372.

[img] Text (Article)
Auto Label ABSA A Comparative Study of Transformer Embeddings and.pdf - Published Version
Download (1MB)
[img] Text (Turnitin)
Turnitin - AutoLabel ABSA A Comparative Study of Transformer Embeddings and.pdf - Published Version
Download (1MB)
[img] Text
korespondensi jurnal.pdf
Download (892kB)

Abstract

Automated text labeling is a promising approach to reduce annotation costs in natural language processing, particularly for aspect-based sentiment analysis and topic classification. This paper presents AutoLabel-ABSA, a systematic framework that combines state-of-the-art transformer embeddings BERT, RoBERTa, and BigBird with multiple clustering algorithms HDBSCAN, K-Means, Agglomerative, and Spectral Clustering to generate high-quality pseudo-labels from unlabeled text. We evaluate the pipeline on three benchmark datasets: IMDB, AG News, and 20 Newsgroupss. Results show that RoBERTa paired with HDBSCAN yields the most coherent clusters, achieving strong downstream classification accuracy (up to 94.5% on IMDB). Notably, BigBird offers minimal gains over RoBERTa despite higher computational cost, suggesting that long-context modeling is unnecessary for standard-length texts. Our findings provide practical guidance for unsupervised labeling in low-resource settings.

Item Type: Article
Uncontrolled Keywords: Automatic labeling, Clustering, Transformer embeddings, Pseudo-labeling, ABSA, HDBSCAN, RoBERTa, Text classification.
Subjects: 000 Karya Umum > 005 Pemograman > 005.5 Program Aplikasi dengan Kegunaan Khusus
Divisions: Fakultas Sains & Teknologi > Teknik Informatika
Depositing User: Hariyanto Rie
Date Deposited: 31 Aug 2026 04:03
Last Modified: 31 Aug 2026 04:03
URI: https://repositori.buddhidharma.ac.id//id/eprint/3483

Actions (login required)

View Item View Item