C4.5 Decision Tree Machine Learning Algorithm Based GIS Route Identification

Advancement of Geographic Information System (GIS) technologies and Artificial Intelligence (AI) supports development of Decision Support System (DSS) and Spatial Decision Support System (SDSS) to solve complex real life problems. Its areas of implementation encompasses diverse fields from defense to civil departments, from meteorology to disaster management, from traffic analysis to network planning etc. Centre for Development of Telematics (CDOT) has developed a software system for analyzing and identifying optimized GIS routes for laying optical fiber who's planning has been done by Bharat Broadband Network Limited (BBNL). BBNL has done planning for connecting Customer Premise Equipment (CPE) to the Central Office (CO). These routes have been planned by doing manual foot survey. This software system analyses whether the planned GIS routes are optimal or not and suggests optimized fiber routes for planning. In this paper we have presented a solution to the problem that we faced during above mentioned project execution. We have suggested an algorithmic solution based on C4.5 Decision Tree Machine Learning Algorithm. This software system first learns based on user input and then creates rules. Using these rules software system provides fiber planning of optical network over GIS. This algorithm, using machine learning, finds best possible route from a set of routes calculated using different GIS data sources.