Artificial Intelligence (AI), which is being taught in the classrooms, will soon move into the examination halls to evaluate answer scripts and grade the polytechnic students as well.
The State Board of Technical Education and Training (SBTET) is contemplating introducing the AI-based evaluation system for evaluating answer scripts of the diploma students of various engineering and pharmacy programmes.
The new computer-based evaluation system works as per the answer keys prepared by the Board. After examinations, the answer scripts are scanned and hosted on a dedicated server.
Further, the Board prepares answer keys for each subject after completion of exams. Using the new AI-based evaluation system, the answer key will be matched with content in the answer scripts. Based on the content, the system evaluates the answer scripts and awards marks.
However, the Board in its testing stage has encountered issues with respect to the students’ handwriting, evaluating mathematical problems and solutions, and engineering drawings. The AI could not perform optimally, leaving with errors in the evaluation. This was done
alongside evaluation by the faculty members.
As the results have not been up to the mark, the technology is now being trained to read handwritings, interpret answers as per the available information, diagrams and generate scores.
“There are issues in recognising some handwriting. There are also issues in evaluating solutions to the mathematical problems, with instances of the machine not understanding the concept written by the student and awarding zero marks. This will not be the case when the lecturers evaluate. Therefore, we are working to improve it. Once deployed, the new system will further speed up the evaluation process,” sources said.
Presently, diploma students’ answer scripts are evaluated using an on-screen evaluation system. As part of this, after completion of the exams, students’ answer scripts are scanned and hosted on a server.
Using a user ID and biometric authentication, polytechnic lecturers concerned evaluate the answer scripts on the computer and award marks. Before the introduction of onscreen evaluation, the answer scripts were evaluated manually in the colleges or evaluation camps.