In the context of “Internet+ healthcare” services, patient trust has become a key issue of concern. To address this, we propose a study on perception-based recommendation trust clustering using the GWO-K-means algorithm. First, we conduct data collection and preprocessing of perception-based recommendation trust text data. Then, we use the TFIDF_SP algorithm to select features from the perception-based recommendation trust text data. It was found that the K-means clustering algorithm has certain limitations. Therefore, the gray wolf optimization algorithm was used to optimize the K-means clustering algorithm, ultimately designing the GWO-K-means clustering algorithm. Using the GWO-K-means clustering algorithm, a clustering analysis of perceived recommendation trust in patient trust tendencies was conducted from the perspective of “Internet+ healthcare” services. Three categories of features were identified: the first category focuses on search recommendations and patient preferences, the second category emphasizes patient trust tendencies, and the third category primarily addresses patient trust and perceived value.
