[
https://issues.apache.org/jira/browse/STANBOL-1418?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Rupert Westenthaler updated STANBOL-1418:
-----------------------------------------
Description:
Similar to STANBOL-1403 that added a PLAIN linking mode to the FST linking
engine. This will add an additional linking mode that only considers detected
Named Entities for linking.
When this is added the FST linking engine will support three modes:
* {{PLAIN}}: link every token
* {{LINKABLE_TOKENS}} (default): This used POS tags to define LINKABLE and
MATCHABLE tokens. Linkable tokens do trigger lockups in the controlled
vocabulary and matchable tokens are considered for matches. By default Proper
Nouns are considered as linkable and all nouns and numbers are considered as
matchable. For languages that use Upper Case letters to indicate Proper Nouns
this is also used for detecting linkable tokens
* {{NER}}: This one will only lookup Named Entities detected by some NER
component for linking.
This Issue will also introduce separate OSGI components for configuration of
FST linking engines for the different mode. The current component will still
allow to configured engines with any of the supported modes by setting the
{{enhancer.engines.linking.lucenefst.mode}} property to the correct mode.
was:Similar to STANBOL-1403 that added a PLAIN linking mode to the FST
linking engine. This will add an additional linking mode that only considers
detected Named Entities for linking.
> Add NAMED_ENTITY linking mode to FST linking engine
> ---------------------------------------------------
>
> Key: STANBOL-1418
> URL: https://issues.apache.org/jira/browse/STANBOL-1418
> Project: Stanbol
> Issue Type: Improvement
> Components: Enhancement Engines
> Reporter: Rupert Westenthaler
> Assignee: Rupert Westenthaler
>
> Similar to STANBOL-1403 that added a PLAIN linking mode to the FST linking
> engine. This will add an additional linking mode that only considers detected
> Named Entities for linking.
> When this is added the FST linking engine will support three modes:
> * {{PLAIN}}: link every token
> * {{LINKABLE_TOKENS}} (default): This used POS tags to define LINKABLE and
> MATCHABLE tokens. Linkable tokens do trigger lockups in the controlled
> vocabulary and matchable tokens are considered for matches. By default Proper
> Nouns are considered as linkable and all nouns and numbers are considered as
> matchable. For languages that use Upper Case letters to indicate Proper Nouns
> this is also used for detecting linkable tokens
> * {{NER}}: This one will only lookup Named Entities detected by some NER
> component for linking.
> This Issue will also introduce separate OSGI components for configuration of
> FST linking engines for the different mode. The current component will still
> allow to configured engines with any of the supported modes by setting the
> {{enhancer.engines.linking.lucenefst.mode}} property to the correct mode.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
