The Intelligent Service Of Spatial Information Based On User Profile

Spatial information as an important basic resources, its applications and services has been increasingly deepening on economic, social, military and other aspects of life. The degree of dependence on the spatial information is also growing. Users were confronted with multi-sources and TB statistics operations of heterogeneous Earth observation data, unable to get the information which you need quickly and accurately. This makes the use-efficiency of space information decreasing with the increasing amount of data, the "information overload" phenomenon becomes more obvious. In order to solve the contradiction between the large amount, complex structure of geospatial information and the low efficiency of spatial data acquisition, analysis and processing, in the field of surveying and mapping science and remote sensing, the sharing and application services of spatial information has gradually become the focus of attention. So active and intelligent spatial information service is the key to solve this problem.Like most of the current remote sensing information publishing platform, it can provide a search interface to the user, then let the user input the regional scope, timing of demand, product types and other conditions. After that it return to the search results of remote sensing image which meet the requirements, and then view and download spatial data needed by the user. Traditional spatial information service methods have disadvantages such as identity, passivity and so on. Facing diverse types of mass space data, people urgent expect to establish a spatial intelligent service system, which can take a full consideration of the different users’ interest. Not only can it actively recommends data to the user, but it also can provides personalized search results for different user with different needs.In view of this, this paper would do some research about the intelligent service of spatial information based on user profile. Through the structural description about spatial data resources and users’ interest, to build user profile for intelligent distribution of spatial information, and based on user profile to achieve active recommendation and intelligent distribution of spatial information. Experiments show that the method can better realize active recommendation of spatial information for users’ interest, making the results more accurate. It can provide a certain theoretical reference for some related research in the future. Mainly includes the following contents:(1) Summarizing the research progress of domestic and international techniques and methods of intelligent service of spatial information. This paper focus on three kinds of personalized recommendation system and four kinds of user profile, expounding mainly on the existing research results. Analyzing the advantages and disadvantages of traditional information service methods and user profile. With the deficiency of existing user model and distribution method as the breakthrough point, to find out the key issues that can realize intelligent space distribution.(2) Establishing the user profile for intelligent distribution of spatial information. Giving full consideration to users’ interest and association rules of personalized recommendation. In view of the coverage characteristics of spatial information, this paper adopted the interval mathematics method into the representation of user profiles. Meanwhile, based on user profiles, the concepts of the correlation degree and interest degree are introduced. And designed correlation functions which is based on topological relation to quantitative calculate the correlation degree.(3) Presenting an intelligent distribution method of spatial information based on user profile. By the construction of decision matrixes, the solutions of the generations of intelligent services can be converted into the solutions of multiple attribute decision problems. Then prioritizing and ordering each alternative solution according to the utility level to determine the final results, improving the accuracy of recommendation of user profile.