[
https://issues.apache.org/jira/browse/STANBOL-1387?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14096941#comment-14096941
]
Chalitha_Perera commented on STANBOL-1387:
------------------------------------------
Aida-light internally uses Stanford NER to extract mentions to disambiguate
from a given text. But Stanbol already provides TextAnnotations which are
extracted by using various enhancement engines such as OpenNLP engines.
Therefore Aida-light is adapted to disambiguate a given text with detected
mentions (Stanbol TextAnnotations)
Code for Aida-light server is available in the following link
https://github.com/ChalithaUdara/Apache-Stanbol-Disambiguation-API/tree/master/stanbol-aidalight-server
> Adapt AIDA light to support disambiguation of mentions (text annotations)
> extracted by Stanbol
> ----------------------------------------------------------------------------------------------
>
> Key: STANBOL-1387
> URL: https://issues.apache.org/jira/browse/STANBOL-1387
> Project: Stanbol
> Issue Type: Sub-task
> Components: Enhancement Engines
> Reporter: Rafa Haro
> Fix For: 1.0.0
>
>
> AIDA-light original disambiguation service takes as input a text/plain
> content and returns as result a list of Mentions within the text. Each
> mention consists on the most prominent entity candidate in YAGO and a
> confidence value.
> In Stanbol, those mentions are already extracted by the NLP engines and
> converted to TextAnnotations. Therefore, AIDA-light should be extended with a
> new service that would receive TextAnnotations and the original content as
> parameters and would return YAGO's candidates ordered by their
> disambiguation's confidences values for each TextAnnotation. Those candidates
> will be mapped later to Entity Annotations back in Stanbol
--
This message was sent by Atlassian JIRA
(v6.2#6252)
