Healthcare recommender system (HRS) has shown the great potential of targeting medical experts or patients, and plays a key role in improving an individual's health by providing insightful recommendations. The HRSs generate recommendations based on a successful and widely applied method known as collaborative filtering (CF). Despite its success, the CF suffers from data sparsity and cold-start problem, which results in the poor quality of recommendations. In particular, it is a great challenge to seeking information relevant to patients' condition, and understanding the medical terms and relationships between them in HRSs. To address these problems, we design a novel collaborative variational deep learning model (CVDL) to exploit multi-sourced information for providing appropriate healthcare recommendations in primary care service. CVDL employs additional variational autoencoder (VAE) to learn deep latent representations for item contents (the description of primary care doctors) in latent space, instead of observation space through an inference network. Meanwhile, the CVDL extracts latent user (patient) features by incorporating user profile in a VAE neural network. Therefore, the CVDL can learn better implicit relationships between items and users from item content, user profile, and rating matrix. In addition, a Stochastic Gradient Variational Bayes (SGVB) approach is proposed to calculate the maximum posterior estimates for learning model parameters. The experiments conducted on three datasets have indicated that our method significantly outperforms the state-of-the-art hybrid CF methods.INDEX TERMS Collaborative topic regression, variational autoencoder, healthcare recommender system, side information, implicit feedback.

No abstract

Customer segmentation can enable company administrators to establish good customer relations and refine their marketing strategies to match customer expectations. To achieve optimal segmentation, a hybrid Artificial Bee Colony algorithm (ABC) is proposed to classify customers in mobile e-commerce environment, which is named KP-ABC. KP-ABC is based on three famous algorithms: the K-means, Particle Swarm Optimization (PSO), and ABC. The author first applied five clustering algorithms to a mobile customer segmentation problem using data collected from a well established chain restaurant which has operations throughout Japan. The results from the clustering were compared to the existing company customer segmentation data for verifications. Based on the initial analysis, special characteristics from those three algorithms were extracted and modified in our KP-ABC method which performed extremely well with mobile e-commerce applications. The result shows that KP-ABC is at least 2% higher than that of other three algorithms.

No abstract