past-exam-matcher

To detect duplicate questions in a past exam corpus, the tool applies semantic similarity scoring to compare new questions against a repository of past-exam items, flagging near-duplicates and related content for quality control.

Semantic similarity scoring for exam questions works by applying semantic search across multiple exams and question types to measure likeness between new and past items, optionally refining comparisons with concept or topic matching.

Yes, you can use semantic search to compare both MCQ and open-ended exam questions. The similarity check handles multiple question types and applies optional concept matching to quantify likeness.

To generate a similarity report for new exam questions, provide the new question and trigger the similarity check against the past-exams corpus. The tool surfaces related past items for reference and veto decisions.

The best way to prevent duplicate exam questions before publishing is to run a similarity check against a past-exams repository. This flags near-duplicates and related items to guide quality assurance workflows.