AiXAM - AI assisted Online MCQ Generation Platform using Google T5 and Sense2Vec

Multiple Choice Question (MCQ) tests provide a quick way to assess knowledge objectively. This pandemic scenario has transformed majority of exams as MCQ based. However, the manual creation of quality MCQ tests is a tedious and time-consuming process. This paper proposes a method to automatically build such tests from a given piece of text. Thus, it is expected to reduce the time and effort required to formulate a MCQ test. This will aid schoolteachers, college faculties and students (for self-assessment). This paper first introduces the readers to the components of an MCQ, then presents some background on such systems with relevant examples. Moreover, it discusses the list of techniques and a few approaches adopted from literature, later shows the implementation of the proposed system - aiXAM MCQ generation platform. Towards the end of this paper, survey results and inferences pertaining to the system have been presented.