On Tue, Aug 9, 2011 at 8:31 PM, Jörn Kottmann <[email protected]> wrote:

> On 8/9/11 6:58 PM, William Colen (JIRA) wrote:
>
>> What about the methods from Evaluator?
>>
>> void missclassified(Span references[], Span predictions[], String
>> referenceSample, String predictedSample, String sentence)
>> void missclassified(Span references[], Span predictions[], String
>> referenceSample, String predictedSample, String[] sentenceTokens)
>> void missclassified(String references[], String predictions[], String
>> referenceSample, String predictedSample, String[] sentenceTokens)
>>
>> Or do you think we should take advantage of some structure provided by
>> F-Measure classes? I can't see it yet.
>>
>
> Don't we have the samples?
>
> The evaluator knows that a sample was incorrectly classified.
>
> It could provide the original gold sample, and the predicted sample,
> this way a report tool can calculate the difference between the two samples
> and output/mark it,
> or compute statistics about mistakes.
>
> What do you think?
>
> Jörn
>

I think it would be much better, but we have different sample classes (one
for each tool) and no common parent. As far as I can see there is no way to
compare two samples without knowing the tool and it makes harder to
implement the monitor. That is way I avoided using the sample itself and
added 3 methods that covers different kinds of samples we have.

Reply via email to