[
https://issues.apache.org/jira/browse/NLPCRAFT-385?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17391080#comment-17391080
]
Aaron Radzinski edited comment on NLPCRAFT-385 at 8/1/21, 12:52 AM:
--------------------------------------------------------------------
We also need the following IDL functions:
* tok_is_before_id(...)
* tok_is_before_group(...)
* tok_is_before_parent(...)
* tok_is_after_id(...)
* tok_is_after_group(...)
* tok_is_after_parent(...)
So that token can form relative order relationship in the intent.
was (Author: aradzinski):
We also need the following IDL functions:
* tok_is_before_id(...)
* tok_is_before_group(...)
* tok_is_before_parent(...)
* tok_is_after_id(...)
* tok_is_after_group(...)
* tok_is_after_parent(...)
> Add tok_index(), tok_is_first() and tok_is_last() functions to IDL.
> -------------------------------------------------------------------
>
> Key: NLPCRAFT-385
> URL: https://issues.apache.org/jira/browse/NLPCRAFT-385
> Project: NLPCraft
> Issue Type: Task
> Components: model
> Reporter: Aaron Radzinski
> Assignee: Aaron Radzinski
> Priority: Major
> Fix For: 0.9.1
>
>
> These functions should allows to specify the desired location (index) of the
> token in the original sentence. Right now, we don't directly support
> something like "Hey, Siri" that should be at the beginning of the sentence
> requiring that entire intent should be either ordered or not.
> These functions will allow a single term to be pinned to a specific location
> to be matched.
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
