[Phenoscape] Re: [go] evidence code ontology

[Larry Hunter]

On Feb 5, 2008, at 7:28 PM, Hilmar Lapp wrote:

> But aren't the act of creating evidence, and the act of using  
> evidence to make an inference two different things?

Yes, and that is one of the reasons for making the relationship  
between evidence and a hypothesis explicit -- to tease apart the  
evidence from the support (or contradiction) of some hypothesis.

> In my understanding of it, ECO would be aiming to conceptualize the  
> possible acts of making an inference. It seems to me that you could  
> use the same evidence to make different kinds of inferences (though  
> I admit that describing those would require further thought). At  
> present it also seems that the trust in an evidence is synonymous  
> with the ECO term - though in reality the same evidence could be  
> weakly or strongly supporting different assertions.

Yes!  While some quantitative degree of support is probably too hard  
to get right for now, I think the difference between outright  
demonstration (or contradiction) and some weaker relationship might be  
captured.   Although consider Racunas' contradiction-based logic (http://portal.acm.org/citation.cfm?id=938084 
) as one path toward quantitation.

Larry