uima-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Yasen Kiprov <yasenkip...@yahoo.com>
Subject Re: f-score evaluation tool?
Date Tue, 06 Nov 2012 08:26:27 GMT
Hi Kameron, 

I'm not sure if I understand your question correctly. I'm writing a named entity recognition
system for text excerpts from the social/public domain: blogs, news, etc. I'm testing different
approaches with rules and ML and I need to evaluate annotations accuracy (in terms of f-score
against a gold corpus). My plan is to use the MASC corpus or build a custom one but the first
task is to find the right tools for evaluation.


P.S. I come from the GATE world and I believe UIMA will give me better performance and more
options for distribution and parallel processing.

 From: Kameron Cole <kameroncole@us.ibm.com>
To: user@uima.apache.org 
Cc: Peggy Zagelow <a2z@us.ibm.com>; William C Rollow <wcrollow@us.ibm.com> 
Sent: Monday, November 5, 2012 6:22 PM
Subject: Re: f-score evaluation tool?

what can you tell me about you f-score annotations? I'm assuming, of course.  are you writng
annotators to calculate f-scores from medical texts?

Best Regards,

Technical Solution Architect  United States 
IBM Content and Predictive Analytics for Healthcare  
IBM Global Business Services Center of Excellence        
IBM US Federal       
E-mail: kameroncole@us.ibm.com 
Work (cell): +1-305-389-8512 
Fax: +1-845-491-4052 
Twitter: @kameroncoleibm 
My Blog: Enterprise Linguistics 
Buy My Book  
Yasen Kiprov ---11/05/2012 09:27:49 AM---Hello,





I'm trying to setup a test environment where I can compare collections of annotated documents
in terms of precision, recall and f-scores. Is there any easy-to-use tool for comparing analysed
documents in the available UIMA xml formats?

I'm familiar with the GATE corpus evaluation tools so a CAS consumer which outputs documents
in the GATE xml format could also be a solution. Does anyone know about such an open-source

Thank you and all the best,
  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message