An aspect sentiment of online review- based multi-criteria framework for technical attribute sustainable evaluation and purchase decision of sweeping robot

The rapid development of artificial intelligence and the Internet of Things has promoted the advent of sweeping robots, and the market competition has become increasingly fierce. In addition, e-commerce reviews now play a vital role in consumer decisions and product optimization. Therefore, this study proposes a decision model that combines the sentiment analysis method of comments with the Pythagorean multi-criteria decision-making (PMCDM) algorithm. The Term Frequency-Inverse Document Frequency (TF-IDF) algorithm is used to extract the technical attributes and features of products that consumers are concerned about. Using the the time exponential decay method to quantitatively calculate the sentimental tendencies of product technical attributes in online reviews, and transforming emotions into a Pythagorean evaluation matrix through proportional transformation. Subsequently, the Criteria Importance Though Intercrieria Correlation (CRITIC) method is used to calculate the weight of product technical attributes, and a PMCDM model was constructed based on the TOPSIS method to calculate the product ranking results. The results help provide decision support for consumers and guide manufacturers in optimizing their products.