[Announcement] Submission Updates & Reminders

[LLMs4OL Challenge]

Dear Participants,

I hope this email finds you well.

Here are a few points that we would like to share while you are doing the submissions:

1. Please remember to complete the registration using the Google Form provided (in addition to the CodaLab registration): https://forms.gle/EUYrViwD4HHqum8k6. If you’ve already filled it out, no further action is needed.
2. Please make sure to provide only a single file inside the zip file that you are submitting. Avoid adding directories such as `.ipynb_checkpoints` or `__MACOSX` that may accidentally be added to your zip file.
3. The training set for the SWEET ontology in Task B has been updated! Blank nodes, which were previously unintentionally included, have now been removed. You are free to use the training sets in any format you prefer.
4. The author guidelines and paper submission instructions have been added. You can find them here: https://sites.google.com/view/llms4ol2025/author-guidelines?authuser=0 

If you have any questions, please don't hesitate to reach out. We look forward to receiving your submissions and papers!

Warm Regards,

Hamed

On behalf of the LLMs4OL challenge organization team

[LLMs4OL Challenge]

Dear Participants,

I hope this email finds you well.

1. We noticed an important typo mistake in the submission page of the CodaLab regarding the submission format for Task B. Please use the "types" key in your submission file instead of "type". Previously, we mistakenly used the "type" key on the CodaLab submission page! So, your submission should be in the following format:

[
    {"id": "sample 1 ID from test", "types": ["type-1", "type-8"]},
    {"id": "sample 2 ID from test", "types": ["type-4", "type-8", "type-3"]},
    {"id": "sample 3 ID from test", "types": ["type-3"]},
    ...
]

2. A clarification on "Dataset Policy": I would like to clarify that the goal of this challenge is to tackle a series of tasks aimed at constructing a primitive ontology from scratch. By overcoming the specified tasks, the resulting ontology will consist of new terms, types, and both taxonomic and non-taxonomic relations. To stay aligned with this objective, the use of existing ontologies to support or guide any subtasks that have some overlap is not allowed, as it would undermine the vision of building ontologies autonomously using LLMs. That said, you are permitted to provide some external context to your model—such as generating type descriptions or definitions using LLMs or other external resources—so long as these are not drawn from existing ontologies related to the tasks. Using unrelated ontologies with no overlap is acceptable, but if there is any overlap with those used in the challenge, it may conflict with the intended goals.

If your submission timeline is affected by any of the above points, please don't hesitate to let us know. We would be happy to assist you with a smooth submission.

Warm Regards,

Hamed

On behalf of the LLMs4OL Challenge Organization Team