User Semantic Portrayal And Visualization Based On Mobile Data

The study of the behavior of mobile user networks is very important for mobile operators to develop personalized network services, allocate cellular network traffic resources on demand, and launch services for users' preferences based on Internet. And a large number of mobile user network data is often collected inconsistently, not only hinders the validity of data processing, and can not be an effective way to analyze and present to the user. Therefore, this paper presents the analysis and visualization of user semantic portraits based on mobile data. By analyzing the semantic portrayal and data visualization of the base station data of the mobile users, the paper analyzes the network service preference rules of mobile Internet users.In this paper, we mainly combine the actual business needs, the user portrait design the following two themes: 1, business type transition law analysis visualization.2, the user preferences analysis of visualization. We will focus on these two topics.First of all, we study the current status and trends of mobile Internet,and the necessity of researching the visualization of user semantic portraits. Then, based on the Hadoop distributed computing platform, we use the FP-Growth association analysis algorithm to analyze the traffic data of the real operator using the jump rule of the network service type,cluster the city base station data with K-Means clustering algorithm, The city is divided into nine regions, which classify and analyze the network traffic used by users in each area, and analyze the user's preference. The user's movement trajectory can be visualized and mapped on the map through the base station. Then, we use data visualization technology to visualize the user semantic portraits through data visualization tools, and integrate the visualization applications, and analyze and summarize the visualization results of mobile user portraits data.Finally, we can get the business type jump rule of the user through portrait analysis and visualization: when most users use the network service, about 75% of the network service they will choose to use next is predictable, High preference areas are concentrated in the central area of the city, the most widely used business users focus on the Web network business, instant messaging services, streaming media business.