[ https://issues.apache.org/jira/browse/TIKA-2790?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16839904#comment-16839904 ]
Tim Allison commented on TIKA-2790: ----------------------------------- [^langid_20190514_plus_minus_1.zip] I implemented the noisification of randomly adding or subtracting 1 from the codepoint as an alternative to randomly picking between 0 and 1,000,000. Yalder isn't nearly as brittle in the plus_minus_1 noisification as in the 0-1000000. The conclusions are still basically the same, I think... > Consider switching lang-detection in tika-eval to open-nlp > ---------------------------------------------------------- > > Key: TIKA-2790 > URL: https://issues.apache.org/jira/browse/TIKA-2790 > Project: Tika > Issue Type: Improvement > Reporter: Tim Allison > Priority: Major > Attachments: fra_mixed_100000_0.0_0.txt, langid_20190509.zip, > langid_20190510.zip, langid_20190514.zip, langid_20190514_plus_minus_1.zip > > -- This message was sent by Atlassian JIRA (v7.6.3#76005)