Research On Key Technology Of Geologic Information Management And Building The Topic Model

With the development and popularity of the information technology and computer science, information technology has been in fully use of geology science. Its applications are involved in not only daily work of geology, such as data mining, processing, management and outputting, but in information sharing and focusing as well. Nowadays, in the field of geology, built-in single-chip microcomputers, handheld computer, personal computer, workstation and giant-scale computers are being used. More and more technology has been used by geological data service.In the conventional management of geologic data, it’s difficult to complete co-working and sharing in the situation of keeping synchronization and independence. Based on the logic and data analysis of Geologic Information Management and Business Platform, it is urgent to improve the whole platform. First, according to the survey of the characteristic of Geologic Information Management’s data, I analyzed the users and business and functions. Overall design of Geologic Data Management was given. Second, the topic model GIC-LDA was given. With layered-design and analysis of data, the topics recommendation of Geologic Information Management and Business Platform were completed.Geologic Information Management and Business Platform has the function of fundamental support of geologic information management. In this thesis, Geologic Information Management and Business Platform was designed. The overall design was given and different function parts were included. Such as multi-data management, running support, data processing, business application, business interfaces, business analysis, unified analysis, cloud management and security management.In this thesis, based on the existing topic model LDA, the topic model GIC-LDA that based on geologic information characteristic was given. First, after designing the data of geologic information management business, analysis of two different kinds of data’s reports was completed. Second, topics were recommended by using the topic model GIC-LDA. At last, comparing with the LDA, we had the conclusion that the topic model GIC-LDA had a better result of topic recommendations in the experiment.