Baidu Image User Behavior Research Method Based On Log Analysis System

As people ’s living standards improve, the Internet tends to be explosive growth. With thegrowing population of Internet users, network construction has improved steadily for thenetwork user behavior analysis has become the current network information systemsarchitecture analysis, performance optimization and improvement, and an important part ofsystem maintenance is an important research network information retrieval and knowledgediscovery one of the areas. A mature product, is not only able to meet the needs, the need toanalyze user behavior through representation, which we need to dig user behavior. Since thelog is an objective record of user behavior, so the log based on user behavior analysis enablesus to better determine the product-oriented.Although there are a lot of shaping user behavior analysis, but based on theconfidentiality of log information, Baidu picture playing figure has not yet been logging userbehavior analysis, in order to make the product more mature, better adapt to changingnetwork environment, Baidu Image user behavior analysis is necessary.In this context, this paper Baidu picture playing Figure logs, designed Baidu pictureplaying Figure user behavior analysis system is proposed to play diagram log analysis of userbehavior, analyze the relationship between the product and the user is proposed to improvethe user viscosity solutions.System for the user entry, each channel needs analysis, user retention data, user data,demand a return visit for design, analysis distributed by hadoop ’s map/reduce large user logdata, extract user data model and user behavior characteristics, analysis of user behavioralcharacteristics, behavior characteristics at the same time to get the user to increase pageviews, increase customer loyalty, and other products to improve the website user experiencepoint of view, put forward to improve the user portal page optimization, improve productback, improved picture quality channels, channel content optimization, etc. viscosity of thesolution, for the direction of the product presents an iterative forward solutions to make ourproducts more competitive in the future.