Native Language Translation (not analyzers)

Elastic Stack Elasticsearch
1

Hello,

My name is Yossi. I'm pretty new to elastic. Just stated using it a few months.

I work a lot of different languages - English, German, Arabic (different dialects), French, Hebrew, Spanish and more. I'm using David Pilato's fscrawler to scan through a bunch of files (txt, docx, pptx, pdf, etc.) and send everything to my elasticsearch. I'm using an open source NLP plugin but I'm really trying to find a way to automatically translate any non-English text into English.

I have searched all over the web to see if there is a solution but couldn't find anything native. The only solution I found online was ditching fscrawler and just using python which I tried and it doesn't really work that well.

Any help / guidance / advise you can give me is much appreciated!

3

Welcome!

I have never heard about a translation plugin. And I don't think we have that out of the box.

I'm wondering if you could find a model on HuggingFace which could help doing that?

In which case you could upload this model to Elasticsearch (you will need a Platinum license I believe) and use the inference processor.

Hope this could help. I'd be very interested in knowing the outcome of your findings...