Research On Mobile Context Aware Service Based On Multidimensional Context Convergence

In the era of mobile Internet,with the explosive growth of information and data,people can easily obtain the information which they want,but at the same time,massive data also make user worried about "information overload" problem.In order to solve people's information problems,personalized recommendation service emerged.The traditional recommendation service only considers two dimensions of users and projects,while in the mobile environment,users' complex and changeable situation will have a great impact on users' decision-making.In view of the existing problems,this thesis combines the research results of previous experts and scholars,trying to put forward its own solutions,building a multidimensional context aware service model,in order to solve the problem of user information overload,improve the accuracy and recall rate of personalized recommendation,and make the effect of personalized service better.This thesis first analyzes the research status of situation awareness at home and abroad,then describes the relevant concepts of situation and the steps of situation awareness in detail,introduces the data collection and preprocessing methods in situation awareness service,and also points out the principle and application scenarios of the currently commonly used personalized recommendation algorithm.Then it constructs a mobile context aware service model,briefly summarizes all aspects of the model,describes the classification of multidimensional context,describes the representation of context dimension,and describes the similarity measurement and combination of context in detail.Then,according to the steps of context awareness service,the detailed operation steps and algorithms are carried out,and the pseudo code is also given.After the context awareness system pushes the personalized recommendation list to the user,the user can have mobile user interaction,further modify and optimize the model,and improve the accuracy of recommendation.Finally,the author uses the open data set to simulate the algorithm proposed in this thesis,and makes a comparative experiment between the traditional recommendation algorithm(UCF)and the personalized recommendation algorithm(MCCF)based on multidimensional context convergence proposed in this thesis.The results show that the algorithm proposed in this thesis is effective.