Arff feature names
Nobody has claimed this yet.
- Dominant language
- PHP
- Stars
- 755
- Forks
- 128
- PR merge metrics
- No merged PRs in 30d
Description
Seems relevant as we had several discussions about what kind of arff we allow for our parsers. What kind of feature names do we allow on OpenML? Currently there is a check that checks whether a feature name is basic latin (first 128 characters of unicode). This check apparently fails on the following datasets (some of which are in the OpenML100/CC18): 1176, 4537, 41102, 41112, 41113, 41114, 41115, 41116.
Should we allow for more generic feature names in the arff files? My gut feeling is telling me that this is going to cause problems in future. My proposal would be to deactivate the datasets and replace them with a basic-latin complying variant. WDYT?
Contributor guide
First steps
- Read the whole issue, then the project's contributing guide.
- Comment on the issue to say you are picking it up — it saves two people doing the same work.
- Fork the repository and make your change on a branch.
- Open a pull request that references the issue number.
Research direction
Start by reviewing the existing basic-Latin feature-name check for ARFF parsers and the listed datasets: 1176, 4537, 41102, 41112, 41113, 41114, 41115, and 41116. Determine whether broader names should be allowed or whether affected datasets should be replaced, then document a decided policy and its effect on those datasets.
Written by the indexing model from the issue text.
Assessment
- Domain
- data, machine-learning
- Issue type
- Feature
- Difficulty
- 5/5
- Estimated time
- Over a week
- Activity status
- Stale
- Clarity
- Needs clarification
- Newbie friendliness
- 25/100