User Interest Model For Active Service Of Geographic Information

Geographic information service, including geographic data service and information processing service of geographic information system, is the basic information means to solve the major social problems facing in sustainable development such as population, resources, environment and disaster. It not only has the public welfare, but also has the market value. Over the past decade, with the continuous development of the representative of the3S technology which is geographic information system, remote sensing and Global Positioning System, as the mapping results in social service application, the geographic information also achieves leapfrog development. But at the same time, because of the increased of data source means, geographic information data is growing exponentially, which leads to the problem of so-called "Rich Data Poor Information". In this paper, to meet the users’needs and requirements for geographical information personalization service, combined with Web mining, information recommendation and other methods, the construction and updating of user interest model is systematically studied, and is applied to a test area. The main research contents and results are as follows:(1) The calculation method of measure value in the process of constructing user interest modelBased on the study in expression of users’ preferences or interesting information, behavior and the in-depth research of user browsing behavior, this paper introduces the scoring mechanism from the browse information and behavior, and users’behavior, such as saving, printing, etc., is also included in the scope of the study, describes a clear representation of user interest model, and takes the study and improvement on the basis of the calculation method of the existing degree of users’interest, to get the information and behavior more comprehensive and effective which can reflect the user interest, which was to build a more accurate model of user interest, to provide users with personalized service.(2) Improved collaborative filtering recommendation algorithmBased on the large number of theoretical research, this paper improves the existing recommend algorithms, proposes the data filling method based on data attributes and geographic features. Then, it makes calculation of attribute similarity using the attributes of geographic information itself makes calculation of attribute similarity, and makes the filling to the sparse matrix with cloud model data filling technique to get the similarity of new information through dynamic weighted. And it gets the recommendation based on the algorithm of geographic information at last.(3) Application of active service model in the experimental areaAccording to proposed model of active service, by the research and exploration on the personalized service of the scholars at home and abroad, data mining function module, user interest model module, method and process of information recommendation module were systematically studied in this paper. It introduces the main functions of each module, and proposes the work flow based on active service mode. And then take the fusion data of geographic information and text information of experimentation area as the data source, to introduce the application of geographic information service.