OBO score - first draft

[matentzn]

measurements:

• dashboard score
• impact score (number of ontologies (or axioms?) reusing terms from O)
• re-use score (number of terms in O from other OBO ontologies)

[nataled]

1. I like impact score. The main problem with it is how this is determined. Every assessment of re-use I've seen simply counts the imports, not actual use. That is, terms can be imported (because, say, the entire ontology was imported) without actually appearing in any definitions or axioms in the importing ontology. Also, would there be some weighting factor built in? For example, an ontology might have very few terms that are used over and over, while another might have many terms used relatively few times. Which has greater impact? Or would they get similar scores...?

2. Undecided about dashboard score, pending the development of some mechanism to actually assign a score.

3. I'm not too keen on re-use score because it only tells anything about quality when there are actually external terms to use. That is, some ontologies won't require the re-use of external terms. To do it right, we'd need to know how many terms ARE reused vs how many terms CAN BE reused.

[matentzn]

sorry this ticket was more a note to self.. Nothing definite yet at all, just a brain dump. We will have a proper obo wide discussion about all this when the infrastructure gets nearer to finish. In the meantime, thanks for your take :)

[matentzn]

Alright, we have to start getting serious about this.

Lets do dashboard and "impact" score in these separate tickets:

• OBO Dashboard score: The OBO Dashboard score #64
• OBO Impact score: The OBO Impact score #65