Resident-Based Store Recommendation Model for Community Commercial Planning.

The objective of community commercial planning is to identify appropriate stores to operate in a community shopping center, catering to the daily needs of residents and enhancing the appeal of the shopping center. However, obtaining data on the characteristics of all residents in the community is a major challenge, and practical methods for selecting suitable stores based on resident characteristics are unavailable. To address these issues, we propose a model that leverages mutual information maximization to learn representations of valuable residents in the shopping area and assess their value. Our key innovation is a value-ranking encoder-decoder that learns the characteristics of all residents in the community and recommends the most suitable store for each storefront. To balance the diversity and competition of businesses within the shopping center, we introduce a diversity loss function. Extensive experimental results show the effectiveness of our model.