A semantic enhanced clothing recommendation system based on implicit trust graphs and ontology web language

In recent years, online clothing shopping has gained significant popularity due to its convenience and the vast range of products available to consumers. However, finding apparel that aligns with individual preferences remains a challenge for many shoppers, leading to time-consuming searches and unsatisfactory experiences. Traditional recommendation systems, such as collaborative filtering (CF), offer personalized clothing suggestions by analyzing user behavior and preferences. Despite their widespread use, these systems suffer from inherent limitations, including data sparsity, cold start issues, and an inability to accommodate users with multiple interests across diverse clothing categories effectively. To address these challenges, this article proposes a semantic-enhanced trust clothing recommendation system based on ant colony optimization, which overcomes the limitations of traditional recommendation systems when considering user context and trust. This system fully utilizes the collective intelligence of ant colony optimization algorithms, combined with the semantic information of clothing products and user trust models, to provide users with more personalized and reliable clothing recommendations. Specifically, the clothing products are initially modeled through semantic description, converting clothing features into semantic vectors to create a semantic similarity model. At the same time, a trust model is introduced to consider user historical behavior and semantic similarity, and to calculate the user’s trust in other users. Subsequently, the system simulates users with individual ants and uses ant colony optimization algorithms to simulate the collaborative behavior of ants in information search and transmission. The experimental results show that the system has significant advantages over traditional methods in providing personalized and reliable clothing recommendations. The system’s real-time updates and learning mechanism ensure continuous optimization of recommendation performance, making it better adapted to dynamic changes in user clothing interests.