I still think this is a confusing request.
It sounds like you are interested in testing an entity extraction annotator that
is implemented as a UIMA annotator, and the "performance" measure you're looking
for is % of entities found, or something like that.
This is not a test of "UIMA", but rather a test of some particular annotators.
So, the first task is to find an annotator you might be interested in testing.
Here's one that's available on the internet (you have to sign up for a license,
I think): http://www.opencalais.com/
And, there is a UIMA annotator which, in turn, makes use of this service:
http://uima.apache.org/sandbox.html#opencalais.annotator
There may be many other Annotators that could be available, some of which might
be better suited to your particular task; that will be up to you to discover.
Or, you may decide to take some NLP technology (like Apache OpenNLP) which
supports "training", and train one.
HTH -Marshall
On 5/23/2012 3:03 PM, Burt Buchman wrote:
We are interested installing and using UIMA as an entity extraction
engine. Before we do that, I'd like to test the system to see the
results. Could I send 25-100 documents , name your format, to someone to
put through their UIMA engine to test. Results sent back to me to
review. It would be greatly appreciated. Thanks!
Burt Buchman
Solution Architect - Editorial
EBSCO Publishing
Phone: 978-356-6500, ext 2346
[email protected]
