development

Winners of the Doxa AI Matching Challenge

Winners of the Doxa AI Matching Challenge

It’s all over!

The Matching Challenge is now officially closed. Thank you to everyone who took part.

The wait is over! We have now closed the Matching Challenge which was hosted on DOXA AI. Over the course of the competition we saw a total of 26 participants with 14 finalists making it onto the scoreboard. The final days were tense with many participants improving on their scores, submitting different methods and swapping places at the top of the scoreboard.

The final scoreboard:

Final Scoreboard

..and the winners are…

1st Place 🏅
José Inés Martínez Berard ‘End’ @oreug, with an impressive MAE (mean absolute error) of 20.348
Prize: £500 in Amazon Vouchers
Congratulations! look out for an email from us with your prize

2nd Place 🥈
Rafi Ahmed Riyaz Ahmed Patel @rafa, with a very close 2nd place MAE of 20.544
Prize: £250 in Amazon Vouchers
Well done 👏👏👏 your prize will be emailed to you soon

A big thank you to everyone else that took part in the competition, if you weren’t successful this time round, fear not, as our next competition is already open and the prizes are even BIGGER!

Above: watch a video interview with José and Rafi, winners of the challenge.

Join our next challenge

Harmony: PDF Parsing Challenge. Develop an improved algorithm for identifying mental health survey questions.

Join our Discord

Join the Harmony Discord server. Check out the 🏅「parsing-challenge」 channel!

See other events

Sign up for our next competition on DOXA AI

Enter the competition on DOXA AI: Your challenge is to develop an improved algorithm for identifying mental health survey questions and selectable answers in plain text.

Related Posts

Introducing Harmony Meta: A Simpler Way to Discover Research Metadata

Introducing Harmony Meta: A Simpler Way to Discover Research Metadata

Introducing Harmony Meta: A Simpler Way to Discover Research Metadata Finding the right data is the foundation of good research, but it shouldn’t be the hardest part of the process. We are pleased to announce the beta release of Harmony Meta, a search tool built to help researchers find datasets and study information more efficiently. This first version is now live, and we are inviting the research community to test it and help us shape its future.
Improving Harmony's PDF extraction with user testing

Improving Harmony's PDF extraction with user testing

Since we built Harmony, a common complaint has been that it frequently identifies the wrong questions in PDFs. The original algorithm for finding questions in PDFs was a mixture of rule based heuristics and some hand coded logic to look for e.g. lines in the document which begin with numbers. This was very fragile and worked fine on short questionnaires such as the GAD-7, but failed on larger documents. We decided to run a competition with our partner DOXA AI where members of the public could train their own model to extract questions from PDFs.

Signup to our newsletter

The latest news on data harmonisation project.

Please select all the ways you would like to hear from Harmony project:

You can unsubscribe at any time by clicking the link in the footer of our emails. For information about our privacy practices, please visit our website. We use Mailchimp as our marketing platform. By clicking below to subscribe, you acknowledge that your information will be transferred to Mailchimp for processing. Learn more about Mailchimp's privacy practices.