[
https://issues.apache.org/jira/browse/TIKA-3340?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17315541#comment-17315541
]
Tim Allison commented on TIKA-3340:
-----------------------------------
Nothing surprising in the results.
Optimaize is the off the shelf optimaize with no new models, so it only covers
half the languages and is scored only on the langs it knows.
OpenNLPLangDetector is the off the shelf OpenNLPLanguageDetector with the
langdetect-183.bin models
OpenNLPTikaEvalDetector is tika-eval's integration with opennlp using the
current model_20190626.bin model.
TikaDetector is the current homegrown langid model with new models built
freshly for all langs.
TikaOpenNLPDetector is the current Tika opennlp integration with the proposed
model.
> LanguageProfile for Myanmar
> ---------------------------
>
> Key: TIKA-3340
> URL: https://issues.apache.org/jira/browse/TIKA-3340
> Project: Tika
> Issue Type: Improvement
> Components: languageidentifier
> Reporter: Arky
> Priority: Major
> Attachments: table-summarized-truncated.txt.gz
>
>
> A language profile for detecting Myanmar/Burmese (my).
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
