/AutoGrader

Auto Grading Software to grade short subjective questions.

Primary LanguagePython

AutoGrader

Automatic Short Answer Grading (ASAG) is the process of grading the student answers by computational approaches given a question and the desired answer.

Objective

  • Remove human intervention while grading students’ answers
  • Compare with other models on the basis of evaluation metrics like Pearson correlation coefficient, RMSE

Dataset used

  • Consists of 2273 answers with 80 questions collected from 31 students of University of North Texas
  • Graded from 0 to 5. The grades are allocated by two evaluators.

Features implemented

  • Preprocessing
  • Feature Extraction
  • Similarity measurement
  • Spelling correction

Performance