Unsupervised Abstractive Summarization of Bengali Text Documents

TL;DR

This paper introduces a novel graph-based unsupervised abstractive summarization method for Bengali texts, overcoming the lack of parallel data by using only POS tagging and a pre-trained language model, and it outperforms existing extractive methods.

Contribution

The paper presents the first unsupervised abstractive summarization system for Bengali, utilizing minimal resources and providing a new dataset for evaluation.

Findings

Outperforms baseline extractive methods

Requires only POS tags and a pre-trained language model

Provides a human-annotated dataset for Bengali summarization

Abstract

Abstractive summarization systems generally rely on large collections of document-summary pairs. However, the performance of abstractive systems remains a challenge due to the unavailability of parallel data for low-resource languages like Bengali. To overcome this problem, we propose a graph-based unsupervised abstractive summarization system in the single-document setting for Bengali text documents, which requires only a Part-Of-Speech (POS) tagger and a pre-trained language model trained on Bengali texts. We also provide a human-annotated dataset with document-summary pairs to evaluate our abstractive model and to support the comparison of future abstractive summarization systems of the Bengali Language. We conduct experiments on this dataset and compare our system with several well-established unsupervised extractive summarization systems. Our unsupervised abstractive summarization model outperforms the baselines without being exposed to any human-annotated reference summaries.