uima-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Francisco Abad <francisco.a...@um.es>
Subject Newbie questions about UIMA modularity
Date Mon, 20 May 2019 12:56:04 GMT
Hi everyone!

My name is Fran and I am a new UIMA user, attracted by its capabilities 
in natural language processing.

I am trying to develop a component that should receive tokens as input 
in order to create its own annotations as output. I though that uima 
modularity permits the use of my component by adding it in any pipeline 
including a tokenizer before. However, I have to define what java 
implementation of token is expected by my annotator. In this way, my 
annotator would be dependent on the tokenizer so it only understands  
those tokenizers that produce a concrete implementation of token defined 
in my annotator. This does not permit the reuse of my annotator in 
different pipelines, for example, when a tokenizer produces TokenA.java 
and another one TokenB.java

My question is: is there any standard implementation of token so I can 
use it as input for my annotator? If not, what is the strategy for 
reusing uima components in different pipelines?

I hope that my question is clear. Thanks in advance.

View raw message