Aspect Extraction from E-Commerce Product and Service Reviews

TL;DR

This paper presents a comprehensive aspect extraction pipeline tailored for Taglish e-commerce reviews, combining rule-based, LLM, and fine-tuning methods, with the generative LLM outperforming others in implicit aspect detection.

Contribution

It introduces a novel AE framework for low-resource, code-switched languages, integrating multi-method topic modeling and dual-mode tagging for explicit and implicit aspects.

Findings

Generative LLM achieved highest Macro F1 score of 0.91.

Fine-tuned models showed limited performance due to dataset issues.

The framework enhances ABSA in linguistically diverse, low-resource settings.

Abstract

Aspect Extraction (AE) is a key task in Aspect-Based Sentiment Analysis (ABSA), yet it remains difficult to apply in low-resource and code-switched contexts like Taglish, a mix of Tagalog and English commonly used in Filipino e-commerce reviews. This paper introduces a comprehensive AE pipeline designed for Taglish, combining rule-based, large language model (LLM)-based, and fine-tuning techniques to address both aspect identification and extraction. A Hierarchical Aspect Framework (HAF) is developed through multi-method topic modeling, along with a dual-mode tagging scheme for explicit and implicit aspects. For aspect identification, four distinct models are evaluated: a Rule-Based system, a Generative LLM (Gemini 2.0 Flash), and two Fine-Tuned Gemma-3 1B models trained on different datasets (Rule-Based vs. LLM-Annotated). Results indicate that the Generative LLM achieved the highest performance across all tasks (Macro F1 0.91), demonstrating superior capability in handling implicit aspects. In contrast, the fine-tuned models exhibited limited performance due to dataset imbalance and architectural capacity constraints. This work contributes a scalable and linguistically adaptive framework for enhancing ABSA in diverse, code-switched environments.