Evaluating LLM Usage for Efficient and Explainable Numerical and Classified Implicit Sentiment Analysis of Product Desirability

This paper presents a framework leveraging large language models (LLMs) for implicit sentiment analysis of product desirability, utilizing two datasets with 106 respondent groupings. The framework achieved Pearson correlations up to 0.97 and classification accuracy of 94%, with GPT-4o-mini demonstrating comparable performance to larger models at 94% lower cost. The approach enhances interpretability through model confidence ratings and human-readable explanations, making it a valuable tool for practitioners in product evaluation and development.