Abstract:

For intellectual activity in the time of information explosion, it is necessary to explore a lot of documents obtained from open sources of the Internet. The object of research is the interface and structure of the information system. This system allows to reduce the processed information flow by filtering documents. The filtration is based on the documents set clustering. This method is seldom used due to the complexity of the user interface. To solve this problem, it is proposed to use the mind map view for visualizing the clustering results. The cluster hierarchy automatically creates the initial graph of the map nodes. The binary graph of the clustering results will automatically transform to the n-ary graph tree. The n is no more than the Yngve-Miller's number and should be determined by the user. The user also controls the mapping of clusters to the mind map, using SQL-queries. The structure of the information system is determined. This system uses free software solutions as its integral parts. Neural network subsystem is required to adapt to the specific user needs. A prototype of the mind map user interface is developed. It is made in JavaScript and is represented as a web page. A list of the main use cases for implementation in the MVP (the minimum viable product) is given. Author Biography Alexey Dubinsky, State Establishment «Dnipropetrovsk Medical Academy», 9, Vladimir Vernadsky str., Dnipro, Ukraine, 49044 PhD, Associate Professor Department of Medical-biological Physics and Informatics

Keywords: