[
https://issues.apache.org/jira/browse/CTAKES-373?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14643268#comment-14643268
]
Sean Finan commented on CTAKES-373:
-----------------------------------
Checked in a change to ignore a sentence if it does not contain letters or
digits. This change improved performance to an acceptable (seconds rather than
minutes per such note) level. Marking as fixed - resolved.
> MaxentParserWrapper can't handle section dividers: "=========="
> ---------------------------------------------------------------
>
> Key: CTAKES-373
> URL: https://issues.apache.org/jira/browse/CTAKES-373
> Project: cTAKES
> Issue Type: Improvement
> Components: ctakes-constituency-parser
> Affects Versions: 3.2.0, 3.1.1, 3.2.1, 3.2.2, 3.2.3
> Reporter: Sean Finan
> Assignee: Sean Finan
> Labels: performance
> Fix For: 3.2.3
>
>
> Notes often contain section "dividers" of a single text character, such as:
> "============================================"
> When the Constituency Parser hits these [sentences], it can churn for 30
> seconds (in my runs). For 60 notes containing two of these lines, that is a
> solid hour of useless processing.
> There shouldn't be any downstream dependencies on such lines, so they
> shouldn't be parsed.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
