Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Exploring Zero-Shot ACSA with Unified Meaning Representation in Chain-of-Thought Prompting

Filippos Ventirozos, Peter Appleby, Matthew Shardlow

TL;DR

This paper tackles zero-shot Aspect-Category Sentiment Analysis under data scarcity by introducing a Unified Meaning Representation (UMR) as an intermediate step in Chain-of-Thought prompting. It compares Baseline CoT and UMR-based CoT across three LLMs and four datasets, finding model- and dataset-dependent results, with mid-sized models showing the most promise. The work highlights that UMR can offer grounding benefits in specific configurations but is not universally superior, and it calls for deeper investigation into when structured reasoning aids zero-shot ACSA and how to embed it into model training. Overall, the findings motivate further exploration of architecture-dependent prompting strategies and broader applications to other fine-grained NLP tasks.

Abstract

Aspect-Category Sentiment Analysis (ACSA) provides granular insights by identifying specific themes within reviews and their associated sentiment. While supervised learning approaches dominate this field, the scarcity and high cost of annotated data for new domains present significant barriers. We argue that leveraging large language models (LLMs) in a zero-shot setting is a practical alternative where resources for data annotation are limited. In this work, we propose a novel Chain-of-Thought (CoT) prompting technique that utilises an intermediate Unified Meaning Representation (UMR) to structure the reasoning process for the ACSA task. We evaluate this UMR-based approach against a standard CoT baseline across three models (Qwen3-4B, Qwen3-8B, and Gemini-2.5-Pro) and four diverse datasets. Our findings suggest that UMR effectiveness may be model-dependent. Whilst preliminary results indicate comparable performance for mid-sized models such as Qwen3-8B, these observations warrant further investigation, particularly regarding the potential applicability to smaller model architectures. Further research is required to establish the generalisability of these findings across different model scales.

Exploring Zero-Shot ACSA with Unified Meaning Representation in Chain-of-Thought Prompting

TL;DR

This paper tackles zero-shot Aspect-Category Sentiment Analysis under data scarcity by introducing a Unified Meaning Representation (UMR) as an intermediate step in Chain-of-Thought prompting. It compares Baseline CoT and UMR-based CoT across three LLMs and four datasets, finding model- and dataset-dependent results, with mid-sized models showing the most promise. The work highlights that UMR can offer grounding benefits in specific configurations but is not universally superior, and it calls for deeper investigation into when structured reasoning aids zero-shot ACSA and how to embed it into model training. Overall, the findings motivate further exploration of architecture-dependent prompting strategies and broader applications to other fine-grained NLP tasks.

Abstract

Aspect-Category Sentiment Analysis (ACSA) provides granular insights by identifying specific themes within reviews and their associated sentiment. While supervised learning approaches dominate this field, the scarcity and high cost of annotated data for new domains present significant barriers. We argue that leveraging large language models (LLMs) in a zero-shot setting is a practical alternative where resources for data annotation are limited. In this work, we propose a novel Chain-of-Thought (CoT) prompting technique that utilises an intermediate Unified Meaning Representation (UMR) to structure the reasoning process for the ACSA task. We evaluate this UMR-based approach against a standard CoT baseline across three models (Qwen3-4B, Qwen3-8B, and Gemini-2.5-Pro) and four diverse datasets. Our findings suggest that UMR effectiveness may be model-dependent. Whilst preliminary results indicate comparable performance for mid-sized models such as Qwen3-8B, these observations warrant further investigation, particularly regarding the potential applicability to smaller model architectures. Further research is required to establish the generalisability of these findings across different model scales.
Paper Structure (19 sections, 1 equation, 3 figures, 3 tables)

This paper contains 19 sections, 1 equation, 3 figures, 3 tables.

Table of Contents

  1. Introduction
  2. Related Work
  3. Methodology
  4. Problem Statement
  5. Chain-of-Thought Approaches
  6. Baseline CoT
  7. CoT with Unified Meaning Representation (UMR)
  8. Prompt Crafting and Post-Processing
  9. Experiments
  10. Datasets
  11. Models and Evaluation
  12. Results and Discussion
  13. Model-Dependent UMR Effectiveness.
  14. Dataset-Specific Patterns.
  15. Model Size and Architecture.
  16. ...and 4 more sections

Figures (3)

  • Figure 1: An ACSA example where a review results in two category-polarity pairs.
  • Figure 2: Prompt used for predicting category–sentiment pairs without in-context examples.
  • Figure 3: Prompt of the CoT with UMR examples.