Automatic Evaluation of Scanned Multiple Choice Tests
Bogdan Ionete · 2010
Although on-line or off-line multiple choice test are used, in many cases, traditional on paper tests are still in use. If we deal with large amount of tests, their evaluation becomes a problem. The paper presents a technique of digitization of scanned multiple choice tests, including their evaluation. A three steps approach is considered: first, the scanned image is preprocessed in order to improve its quality, then character recognition phase assures the acquisition of the data that identifies the applicant and reads, and the final phase scores the test. A test by test processing or a batch procedure is possible. The method produces accurate results and has many applications in day by day activities.