predict job title on the basis of skills

Question

I'm working on a project where I need to predict if a person should be frontend dev, backend dev, DevOps eng or data scientist, etc on the basis of his skills.

For example:

Sam has skills ['python','sql','machine learning','flask'] so he can be called as Data Scientist.

I've scrapped LinkedIn and got the job titles of people and their skills as well. I need guidance in a few things:

1. How I can normalize data using NLP
2. How I can predict a new user that he belongs into particular job or domain.

What should be the procedure to implement it using python and any reference articles or tutorials are most appreciated thanks in advance

Answer 1

I've solved this issue by using BERT it looked after vectorization and tokenization itself. I have a dataset that has particular skills and their domain (frontend, backend, etc). I used multi classifier to predict the domain and then that is how all the jobs are clustered through a particular domain