Framework for Classroom Student Grading with Open-Ended Questions: A Text-Mining Approach

The purpose of this paper is to present a framework based on text-mining techniques support teachers in their tasks grading texts, compositions, or essays, which form the answers open-ended questions (OEQ). approach assumes that OEQ must be used as learning and evaluation instrument with increasing frequency. Given time-consuming process for those questions, large-scale use only possible when computational tools can help teacher. This work decision entirely teacher’s task responsibility, not result an automatic process. In context, teacher author included tests, administration results assessment, entire cycle being noticeably short: few days at most. An attempt made address problem. method exploratory, descriptive data-driven, data assumed inputs texts essays compositions created by students answering single test specific occasion. Typically, involves exceedingly small volumes measured power current home computers, but big compared human capabilities. general idea software extract useful features from perform lengthy complex statistical analyses teacher, who, it believed, will combine information his her knowledge experience make decisions mark allocation. A generic path model formulated represent context kind should perform, estimated synthesised using graphic displays. illustrated analysing three corpora 126 originating different real contexts, time periods, educational levels disciplines.