Academic Document Segregation Using Machine Learning

Abstract

Text Classification has proved to be one of the most popular problems in the Natural Language Processing domain. Document classification is one such problem which is an application of Text Classification. Convolutional Neural Network has been used in wide areas for image classification problems to achieve better accuracy. The R-CNN model is an advanced CNN model that focuses on each region in an image. It works on specific

regions thus lowering the problems caused by boundary interference. For selecting a spe- cific region it uses the selective search approach and then checks if the selected region

contains an object. The CNN than extracts important features. This paper focuses on the use of faster Region with Convolutional Neural network (R-CNN) to detect and later segregate the academic documents viz marksheet, leaving certificate, migration certificate and degree. For each document a number of classes have been defined and then applied R-CNN for detection.