Integrating Sentiment Analysis and Topic Modeling into QFD: A Data-Driven Approach for Product Design

Quality Function Deployment (QFD) turns customer needs into engineering goals, keeping product design aligned with what matters in the market. Traditionally, this relies on surveys, interviews, and manual feedback. This study offers QFD approach using natural language processing, combining sentiment analysis and topic modeling to pull insights from online reviews. With BERTopic, key customer concerns (WHATs) are identified, while BERT-based sentiment analysis adds emotional weight. Importance scores blend how often topics appear with how strongly customers feel about them. These scores link to engineering features (HOWs) through a simulated relationship matrix. A case study on biodegradable trash bags highlights compostability, handle strength, and leak resistance as top priorities, showing that automated analysis can offer clear, practical guidance for product development.