uima-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Peter Klügl (JIRA) <...@uima.apache.org>
Subject [jira] [Reopened] (UIMA-3071) Break up sequential matching in Ruta rules
Date Thu, 01 Aug 2013 10:57:56 GMT

     [ https://issues.apache.org/jira/browse/UIMA-3071?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel

Peter Klügl reopened UIMA-3071:

forgot explanation of rule inference
> Break up sequential matching in Ruta rules
> ------------------------------------------
>                 Key: UIMA-3071
>                 URL: https://issues.apache.org/jira/browse/UIMA-3071
>             Project: UIMA
>          Issue Type: New Feature
>          Components: ruta
>    Affects Versions: 2.0.2ruta
>            Reporter: Peter Klügl
>            Assignee: Peter Klügl
>             Fix For: 2.0.2ruta
> Break up sequential matching in Ruta rules: Right now a list of rule elements specify
a sequential pattern. In some use cases, however, the sequence of annoations is not as important
as the their existence. An example: A rule should fire, if some complex annotations patterns
occur within a sentence whereas the location or order is not important. Two use cases can
be distinghuished: "and" and "or". A disjunctive matcher is already implemented, but only
suppport simple matching conditions, but not complex patterns (e.g., additional conditions)
for the alternatives. I am still thinking about the best syntax for this. Right now, my favorite
is a special character that separates the rule elements. An example:
> {noformat}
> BLOCK(b) Sentence{}{
> }
> {noformat}
> ... a sentence that contains a capitalized word followed by a period AND a small written
word followed by a colon, regardless of where they occur in the sentence.
> Maybe also something like the follwoing is then possible:
> {noformat}
> NUM (CW{REGEXP("A") -> MARK(NUMA,1,2)} | CW{REGEXP("B") -> MARK(NUMB,1,2)});
> {noformat}

This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira

View raw message