Design And Implementation Of A Context-aware Framework For The Intelligent Shopping Area

In recent years, as the core of the Internet of Things(IOT), smart devices collected data in all of the environment. Data mining theory and intelligent sensing systems have got a rapid development based on these data. IntelliSense technology used in various fields is expanding in both depth and breadth. As a result, pervasive computing is promoted to a new stage. It tends to be more open and consider more about the resource integration. There are more opportunities and challenges on unified management of these resources. Most of the academic research on IntelliSense is in the fields of medical, environmental monitoring and smart home. But there are few research about intelligent shopping area. Most of existing systems are based on consumer’s consumption data analysis, however, few are based on the study of their behavior.. What’s more, the recommender systems are lack of systematism, versatility and scalability.In this paper, we design a context-aware framework for the intelligent shopping area based recommender system, and model for customer behavior. We also study the efficient data storage and query and map index strategy in this framework. In the study of the framework of contextaware recommender system, we use a hierarchical and modular approach to separating the responsibilities of various parts of the system to achieve the effect of high cohesion, low coupling and enhance system’s scalability. The perception layer shielding smart device’s heterogeneity, gets all kinds of information in the intelligent shopping area. The modeling layer processes data cleaning and models customer’s behavior so that it is stored as a context model. The data analysis layer provides a unified data access interface for all types of services and data mining algorithms. The service layer will organize the recommended results and provides services for customers such as recommendation service. The application layer will present the service for customers and applies to user interaction. To support the unified management of data, this paper proposed the concept of behavior context. Customer behavior context is the collection of user behavior in their shopping path. A specific behavior point contains spatiotemporal data, behavior information and other data related. This paper describes the structure of the content and access interface of behavior context model in detail. Based on the context-aware framework, this paper also focuses on the efficient storage and extraction of map grid information.This paper first introduced the research background of intelligent sensing recommender system and context modeling, analyzed their characteristics and shortcomings and listed the research content and innovation of this paper. Next we put forward the corresponding design based on the challenges of context-aware framework. And we described the design of all levels and modules in the system architecture based on context-aware intelligent area. Then we presented and elaborated on the content and design of customer’s behavior context. We also proposed the difference grid grading strategy to extract map information and efficient storage method in data query. A prototype was provided to show the implement of the system and we tested the performance of this system. Finally, we concluded our research about the contextaware framework for intelligent shopping area and discussed the further work.