Dynamic Sentiment Analysis with Local Large Language Models using Majority Voting: A Study on Factors Affecting Restaurant Evaluation

TL;DR

This study explores how local large language models can be used for dynamic sentiment analysis of restaurant reviews, emphasizing the importance of hyper-parameters, variability, and the effectiveness of majority voting for more reliable results.

Contribution

It introduces a majority voting mechanism with local LLMs for sentiment analysis, demonstrating improved robustness over single attempts with larger models.

Findings

Majority voting with multiple attempts yields more consistent results.

Medium-sized models with voting outperform large models with single attempts.

Analysis of review aspects influences overall sentiment evaluation.

Abstract

User-generated contents (UGCs) on online platforms allow marketing researchers to understand consumer preferences for products and services. With the advance of large language models (LLMs), some studies utilized the models for annotation and sentiment analysis. However, the relationship between the accuracy and the hyper-parameters of LLMs is yet to be thoroughly examined. In addition, the issues of variability and reproducibility of results from each trial of LLMs have rarely been considered in existing literature. Since actual human annotation uses majority voting to resolve disagreements among annotators, this study introduces a majority voting mechanism to a sentiment analysis model using local LLMs. By a series of three analyses of online reviews on restaurant evaluations, we demonstrate that majority voting with multiple attempts using a medium-sized model produces more robust results than using a large model with a single attempt. Furthermore, we conducted further analysis to investigate the effect of each aspect on the overall evaluation.