Finnish, Danish and Norwegian

Mon 01 July 2019

Today we're pleased to announce TextRazor support for three new languages - Finnish, Danish and Norwegian.

We've had a ton of interest in Nordic languages over the past year or so, both from local organisations and as part of a larger trend towards "international first" text analytics systems. Each of our languages needs to be individually modelled and tweaked, and these were our most challenging to date. We're particularly proud of Finnish - the complexity of the language together with the lack of available data means the usual "Deep Learning" methods can struggle.

These new languages are ready to try now, with support for our Entity Extraction and Linking, Topic Tagging and Classification functionality. As usual these new languages are in beta for now - if you spot any mistakes in the results or have any other feedback while we work out any remaining issues please get in touch.