The Scanii Changelog 08/2018

NSFW Language identification improvements

Two large announcements, first, Improvements to language identification for short text leading to more effective identification of NSFW language for corpus with fewer than 15 words in it. Second, we kicked off a large refactoring of our open source NSFW language database, stating with the English language, aimed at making it easier to identify which exact word triggered our finding as well as improved plural identification - details here.

Omni Search

Omni search was extended to audit logs ( so you can now search and filter them blazingly fast be it by user or type (such as filtering just authentication events)

Tags: news

Last updated on 08/27/2018.