Online short answer evaluation system a model for descriptive answer assessment using semantic relational features

Abstract

The rapid increase in the usage of internet based technologies in elearning systems enables knowledge acquisition of an individual at various levels Every day the process of knowledge acquisition renews as many innovations are possible that ease the effort of mankind as a result of meticulous learning The World Wide Web is the unsurpassed library where learning quest happens without any hindrance E learning and online courses are a promising platform for aspirants who engage in all time learning The essential aspects of learning process must also include a suitable assessment technique which guarantees the acquaintance of learning There arises an obligatory need to develop standard evaluation techniques such as Automatic Question Answering and evaluation systems in educational organizations and forums Even though automatic evaluation systems are already available it effectively handles only objective type answers There is a wide scope to develop automated assessment system to evaluate descriptive type answers The proposed automatic evaluation and grading system focuses on evaluating close ended questions where the challenge lies in fetching the students level of learning precisely at a perfect scale for the descriptive type of answers presented during online exams Various architectures proposed earlier have different approaches that dealt with multiple choice questions single word as answers and polar answers that have adopted string matching sequence analysis and quantitative analysis As the volume of online learning increases there arises a need for automated evaluation of descriptive type answers This automation is achieved with the effective utilization of Natural Language Processing NLP techniques that renders efficiency and quality of grading These automatic graders are highly essential in educational institutions to adopt uniform and unbiased evaluation newline