Recommended System Based On UserProfile

With the continuous progress of Internet technology,the amount of data for the all society is growing.In the big data environment,the way of the user get the information what they want is also put forward higher requirements,Currently,the main way of the user to get information is to use a search engine,the search engine can indeed help the users to query what they want,but if there is a system can automatically give users the information what they want,then it can be very convenient the user,the recommendation engine has such a function.This paper constructs a recommendation engine based on userprofile to the user recommended,recommended to the user based on userprofile product is in line with the user's interests,but not to recommend the single,but also consider the new problem,so this paper also joined the collaborative filtering algorithm,the similarity calculation of users or items as a recommended basis,through cross calculation can be recommended for users of new products.In this paper,in the process of practice,through research to understand the production environment to the recommendation system is composed of real-time and off-line is composed of two parts,part of the Spark Streaming and Kafka combined with real-time Kafka technology,from the application to real-time access to user operation information,and then to the Spark Streaming filter,finally to obtain the recommendation list from the recommendation engine.The offline part adopts Mahout technology,and Mahout has a collaborative filtering algorithm,which can calculate the similarity between users and users and the similarity between users and objects offline.These calculations are time-consuming and are not suitable for online computing,so they are offline.The user's portrait data also needs to be calculated off-line and periodically updated to ensure that the recommended results can be changed according to the changes of the user's interest.During the implementation of the whole case,storage media such as Mysql,Redis and Hbase are used in the data storage.