Research On Central Red Wowo E-commerce User Interest Recommendation Based On Context-Awareness

With the development of the times and the advancement of technology,e-commerce has gradually become a trading method for people,enabling people to live in the environment of big data.Faced with a huge amount of information,users need to spend a lot of time and cost to select and identify products that are suitable for them,and even have no choice.The emergence of this phenomenon is inevitable and a situation that needs urgent solution.A recommendation system is needed to intelligently help the user to filter out the information he or she needs,and the evaluation recommendation system is determined by the recommendation algorithm.Because the current collaborative filtering algorithm has a single data value,a small amount of data,poor real-time performance.large amount of calculation,and low recommendation efficiency,the paper conducts further research based on the research of Central Red Wowo e-commerce enterprise collaborative filtering recomimendation algorithm.as follows:(1)Aimina at the problems of scalability,low recommendation efficiency and large computational complexity in traditional collaborative filtering algorithms,a K-means clusteriim collaborative recommendation algorithm based on Improved Artificial Bee Colony Algorithm(IABC)is proposed.The paper improves the artificial bee colony algorithm(ABC)through two aspects of initialization and fitness function,and combines with K-means iteration to obtain more accurate clusterins effect,and then incorporates collaborative filtering algorithm to complete the recommendation.The experimental analysis proves that the K-means clusterine collaborative filtering recmmendation effect based on IABC algorithm is better than the traditional collaborative filtering algorithm and the K-means clustering based collaborative recommendation algorithm,can be effectively applied to the recommendation system of the Central Red Wowo e-commerce platform.(2)Aiming at the problem of data sparsity,real-time difference and single data value in traditional collaborative filterina algorithm,a context-awareness user interest model(CA-UI)is proposed.The paper records the data through the user's explicit behavior and implicit behavior.which enriches the data volume,relieves the problem of data sparsity to a certain extent,and introduces the user's real-time behavior and situational factors,so that the data is no longer a single Boolean quantization.The value,but accurately represents the user's attention to the project,get the user-project attention matrix,solve the problem of real-time difference and single data value,Effectively alleviated the problem of scarcity of user data on Central Red Wo wo e-commerce platforms.(3)The K-means clustering collaborative recommendation algorithm based on IABC algorithm is used to recommend users to ensure the superiority of personalized recommendation in recommendation performance and recommendation accuracy.The experimental analysis proves that the K-means clustering collaborative filtering recommendation algorithm based on IABC algorithm has better effect in context-based user interest model(CA-UI)recommendation,can effectively solve the problems of the Central Red Wowo e-commerce platform recommendation system.Aiming at the shortcomings of collaborative filtering algorithm,this paper proposes a context-aware user interest model(CA-UI)and a user clustering collaborative recommendation algorithm.The K-means clustering collaborative recommendation algorithm based on the IABC algorithm is used to recommend users in the attention matrix.Experiments show that the recommendation effect based on CA-UI model and user clustering collaborative recommendation algorithm is better than traditional collaborative filtering recommendation,and can provide Central Red Wowo users with personalized recommendations.