Automatic online subjective text evaluation using text mining

Abstract

Semantic based text mining is essential in high dimensional data processing in today’s environment. In educational sector Question Answer (QA) evaluation has done using semantic as well as semantic analysis in many approaches. Numerous existing mechanisms have implemented using different machine learning algorithms. Semantic and semantic both works for evaluate the text data, but semantic approach should done same work with low time complexity. In this work system carried out automated text evaluation for online examination system with semi structured dataset. The system has categorized into two phases, NLP and Features base evaluation. Natural Language Processing (NLP) has used for preprocessing of data using tokenization, stop word removal, porter stemmer etc. Similarity technique has used for generate similarity score between test answer and train answer data. Artificial Neural Network (ANN) has used to generate the similarity score between two features vectors. Experimental analysis shows the how proposed system is better than some traditional approaches for semantic text evaluation.