How to Participate

The challenge is open and free to join. Anyone, from any institution or background, can take part.

1. Register your team through the registration form. Registration helps us organise participation and communication.
2. Choose a task and download its definition, dataset, and gold standard from the challenge-catalog repository [coming soon].
3. Run your system on the task inputs and produce outputs in the required format.
4. Prepare the artifacts you want to submit (see the Submission section for details).
5. Submit via the process described in the Submission section.

FAQ

Can I participate in only one task?

Yes. You can enter one or both tasks, and each is evaluated independently.

Is there a participation fee?

No. The challenge is open and free to join.

Can I use any LLM or external tools?

Yes, as long as you comply with the size limits and the model is open.

How are my outputs evaluated?

We have automated tools for this. Refer to the corresponding benchmark to see which metrics are used.

What if my system cannot be reproduced?

If a reviewer cannot reproduce your results from the public artifacts, the submission does not receive a Green status and is not included in the ranked leaderboard. Under the traffic-light system, it may receive a Yellow status if only minor clarifications or fixes are needed, or a Red status if the code is unavailable and only the outputs were provided.

Will the reports be published?

Yes. Accepted papers will be published in a high-ranking Semantic Web journal.

Are there prizes?

Yes. Information about the award for the best systems will be published soon.

How to Join Us

Task Contributor and Reviewer roles are open to the community. To get involved, open a pull request to add yourself under the Contributors section, subject to approval by the Organizers. Governance and decision-making follow the rules in governance.md.