| With the increasing focus on organic and healthy food,there is a rising demand for green and organic agricultural products Simultaneously,the government’s support for rural e-commerce has been steadily increasing,facilitating the rapid growth of agricultural product e-commerce platforms and expediting the development of new forms of rural e-commerce.This support has played a vital role in enhancing farmers’ income and contributing to rural revitalization and economic stability.Against this backdrop,the research focus has shifted towards how to recommend agricultural products that cater to user needs.However,traditional recommendation techniques are no longer sufficient in complex environments to meet user expectations.Consequently,advanced hybrid recommendation techniques have emerged,integrating multiple recommendation approaches in a synergistic manner to address the limitations of traditional methods.To address the practical requirements of agricultural product recommendation,this paper introduces a hybrid algorithm that combines the Apriori algorithm with item-based collaborative filtering.This hybrid algorithm leverages the Apriori algorithm to mine and analyze user behavior data,supplementing user behavior information and mitigating the challenges posed by data sparsity in the rating matrix.Consequently,it overcomes the issue of inaccurate recommendations caused by missing user behavior in collaborative filtering algorithms,thereby enhancing the accuracy of agricultural product recommendation systems.The algorithm is seamlessly integrated into the agricultural product recommendation system and showcased through the recommendation module.To tackle the cold-start problem faced by new users,who lack historical behavior data and struggle to define personalized preferences,the algorithm utilizes the popularity and sales volume of goods for recommendations,providing superior personalized services to new users.Backed by internet technology,the system incorporates a well-designed frontend shopping interface and a comprehensive backend admin system,encompassing functionalities such as user registration and login,agricultural product details,shopping cart,agricultural product recommendation,order generation,agricultural product management,order management,and evaluation,to facilitate seamless and efficient user purchasing experiences.Through experiments comparing the hybrid collaborative filtering algorithm with traditional methods,using the same number of neighbors,three widely-used recommendation metrics are employed to validate the results as the number of neighbors increases.The proposed hybrid collaborative filtering recommendation algorithm(APICF)demonstrates clear advantages and fulfills the requirements of the agricultural product recommendation system presented in this paper. |
