[Objective] This paper identifies the dimensions of customer interest in healthcare wearable devices and their impact on satisfaction, inspiring merchants to optimize their products and improve their services. [Methods] Using 11,349 online review data from the well-known shopping website Amazon as a corpus, the LDA model was used to identify customer satisfaction dimensions, and combined with machine learning algorithms to construct a satisfaction model. [Results] The satisfaction model constructed with the Multi-Layer Perceptron (MLP)has the best performance in predicting customer attention ( ) to products by focusing on 13 product dimensions with seven integrated attributes, such as quality, service, functionality, usefulness, social, value and ease of use. Functionality is the most important product feature for customer groups, while social, quality and service attributes have a negative impact on customer satisfaction that should be a priority for merchants for product optimization and service enhancement. [Limitations] We did not consider the authenticity of the reviews and in future will include cases of false and malicious reviews in the analysis process. [Conclusions] This paper obtains the dimensions of customer attention to products, the aspects of satisfaction influence and the priorities for improvement, providing rich management insights for businesses.