Causal Prompting for Implicit Sentiment Analysis with Large Language Models

TL;DR

This paper introduces CAPITAL, a causal prompting framework for implicit sentiment analysis with large language models, improving accuracy and robustness by integrating causal inference into reasoning processes.

Contribution

It proposes a novel causal prompting method that incorporates front-door adjustment into chain-of-thought reasoning for better sentiment inference.

Findings

CAPITAL outperforms baseline prompting methods in accuracy.

It enhances robustness under adversarial conditions.

The approach effectively reduces bias and spurious correlations.

Abstract

Implicit Sentiment Analysis (ISA) aims to infer sentiment that is implied rather than explicitly stated, requiring models to perform deeper reasoning over subtle contextual cues. While recent prompting-based methods using Large Language Models (LLMs) have shown promise in ISA, they often rely on majority voting over chain-of-thought (CoT) reasoning paths without evaluating their causal validity, making them susceptible to internal biases and spurious correlations. To address this challenge, we propose CAPITAL, a causal prompting framework that incorporates front-door adjustment into CoT reasoning. CAPITAL decomposes the overall causal effect into two components: the influence of the input prompt on the reasoning chains, and the impact of those chains on the final output. These components are estimated using encoder-based clustering and the NWGM approximation, with a contrastive learning objective used to better align the encoder's representation with the LLM's reasoning space. Experiments on benchmark ISA datasets with three LLMs demonstrate that CAPITAL consistently outperforms strong prompting baselines in both accuracy and robustness, particularly under adversarial conditions. This work offers a principled approach to integrating causal inference into LLM prompting and highlights its benefits for bias-aware sentiment reasoning. The source code and case study are available at: https://github.com/whZ62/CAPITAL.