I am using OpenAI and word embedding to research the Mi’kmaq language. I am currently evaluating vector similarity using cosine similarity. The intent is to find similarities in meaning and vector similarity. This is a passion project in an effort to preserve my native language by leveraging AI.
Any suggestions or comment is greatly appreciated.
https://github.com/AdieLaine/Mikmaq-Language-Embedding
Update
I wrote a long-form post on Twitter detailing the process.
3 Likes
Dent
2
I actually e-mailed my local tribal representative about using LLMs to help preserve the Mi’kmaq language!
They never got back to me… 
Definitely gonna follow your repository, though!
1 Like
Super cool and important work. Good luck. Let us know how you might benefit from additional support. Maybe someone in the community here can help.
1 Like
Thank you. I am currently compiling all the words in the Mi’kmaq language. I’ll certainly tap into the community for help. I appreciate the support.
1 Like
I appreciate the support. I added a Twitter link in the original post, detailing the process.
1 Like
