Research On Customer Service System Based On Big Data Machine Learning

With the development of mobile Internet,there has been a blowout growth in mobile communication data.The era of big data is coming.The huge amount of data,a variety of data types and complex data structures have brought great challenges to data storage and data processing.It has gradually exceeded the storage and processing capabilities of traditional relational databases.How to select a technology to deal with this mass data is the first topic we should study.At present,the work of the customer service system is usually as a sample of the existing artificial customer service corpus,and the whole question and answer process is completed through the training sample.With the development of the machine learning,the present intelligent customer service system can not only help the customers to search the answers to the problem,but also can dig out the interests and interests of the potential users,in the form or column of the pop-up window.In the form of a table,the user recommends their products or services.In the face of such recommendations,the user may be interested in clicking related recommendations.It may be ignored for the non interest.It is an increasingly important question for people to get what they need in a wide range of information.At this time,you need it.To be an automated tool,it can be based on your historical interest,from a large resource pool,to select the goods or requirements that conform to your taste.This tool is a personalized recommendation system,and how to implement the precise personalized recommendation is the second topic we have to study.This paper is to study the recommendation system in customer service system under big data environment.In this paper,the basic theoretical knowledge of Hadoop technology is introduced in detail.On this basis,the architecture diagram of the data analysis system is completed,and a mass data storage and analysis processing system based on the Hadoop cloud platform is designed and implemented.Hadoop HDFS distributed file system is used to deal with massive data storage and Hadoop-Yarn as resource manager to support various computing engines to improve resource utilization.The paper builds Hadoop cluster to enhance the stability of the system.At the same time,it introduces the Bayesian algorithm,the principle and workflow of collaborative filtering algorithm in detail.On the basis of this system,the Bayesian algorithm is used to analyze the user's access to interconnected behavior data,and the classification of items is realized.The collaborative filtering algorithm is used to realize the personalized recommendation for different users.The system combines the Bayesian algorithm with the collaborative filtering algorithm and is based on the distributed technology and is compared to the traditional one.Recommendation system can effectively deal with the storage and depth analysis of Internet data resources,and embody the perfect combination of big data and machine learning technology.