JevMade

Sign in
← Back to experiments

Benchmarks & research

Matching records with Jev: how many candidates to put in one question, and what goes wrong

Shaun Mataire compares Jev’s multiple-choice and yes-or-no questions for matching product listings and research papers.

Source screenshot of Matching records with Jev: how many candidates to put in one question, and what goes wrong
SOURCE SCREENSHOTFull screenshot ↗

What it does

The reported tests examine missed matches and sensitivity to candidate order. Choice finds one match; pairwise questions can find several duplicates. Only summary results, not raw API answers, are published.

How you can use it

To spot duplicates across product listings or paper archives, collect the records you need sorted. A developer can adapt this project's approach to ask the service either a yes-or-no question for each pair or a multiple-choice question from a short list of options.

Keep multiple-choice questions to ten or fewer items at a time. Longer lists or shifting item orders can cause the system to miss duplicates. If a record might match several entries at once, have your developer check pairs individually instead of using multiple-choice.

Keep this for later

Sign in to bookmark experiments, guides and videos, and keep notes only you can see.

Continue to sign in

We’ll bring you back to this listing.