A Text Analytics Framework for Performance Assessment and Weakness Detection From Online Reviews

Present research proposes a framework that integrates aspect-level sentiment analysis with multi-criteria decision making (TOPSIS) and control charts to uncover hidden quality patterns. While quantifies consumer opinions corresponding various product features, TOPSIS uses the scores rank manufacturers based on their relative performance. Finally, U P assist in discovering weak aspects attributes. To extract sentiments from reviews, we developed ontology of passenger cars designed heuristic connects opinion-bearing texts exact automobile attribute. The proposed was applied review dataset collected well-known car portal India. Considering five mid-size segment, identified weakest discovered attributes responsible for its perceived weakness.