Design And Implementation Of Query Recommmandation System For E-commerce Products Based On Cloud Computing

In the current Internet era,Data is given a new era of new energy;Data is growing explosively,and customers in the E-Commerce industry are paying more and more attention to interactions with websites.Through the analysis and mining of data,and the rational use of these analysis results,we can provide the best service for customers.In the existing query recommendation methods,the accuracy of query intent of users and the sparseness of data are important factors affecting the effectiveness of query recommendation.Most of the existing methods are mainly based on user query and clinck characterization,and ignore the user's interest behavior information.At the same time,the sparseness between query clinck data cannot be solved well.By analyzing the query clinck behavior of users on the E-Commerce website,users can be provided with more personalized product query term recommendations when inquiring.Under the recommendation system,it still faces many challenges and problems,such as the consistency and stability of transmission with high concurrency of massive data,the problem of massive data storage calculation and scalability,the problem of native data sparsity,and the timeliness of user analysis behavior Questions and more.In order to address the problems in the above query recommendation method and recommendation system,and in the context of the E-Commerce field,a query and recommendation system for E-Commerce products is proposed in the context of cloud computing technology.In the existing query recommendation methods,how to combine the user's interest behavior with the query behavior is a problem.In order to optimize the system query recommendation effect,In the E-Commerce scenario,users are rich in interest behaviors and sparse query clinck behaviors.A query recommendation method based on queries and interests is proposed.While incorporating user query and clinck behavior representations,it also introduces the behavior of users when they interact with E-Commerce websites to solve the problem of low circulation of user behavior nodes forming subgraphs.In the recommendation system,message queue is a component that is used in many scenarios such as log collection,message distribution,and log processing.In order to improve the timeliness of the query recommendation system,Under the problem of imbalanced throughput performance at both ends of a message queue Kafka task and the problem of repeated message consumption after a task failure / downtime,a resource allocation method based on the message queue throughput ratio was proposed.This method is based on a multi-source topic publish-subscribe message model,which performs task management and message distribution.Finally,this article designs and implements an E-Commerce product query recommendation system based on cloud computing technology,including a data source subsystem,a task management subsystem,a non-relational data storage subsystem,and a service monitoring subsystem.According to the proposed query and interest-based E-Commerce query recommendation method,the existing query recommendation method is improved,and the message queue model under the recommendation system is optimized,and it is added to the multi-source topic consumption and management and buffer layers.Realize the management and resource allocation of message queue tasks in the system.The test results show that the cloud computing-based e-commerce product query recommendation system constructed in this paper can respond to personalized,low-latency query requirements,and reflects the parallel processing capability and reliability of the system in the cloud environment.