Hi Team,
I have list of employees from entire globe with their skillsets, trainings they have done, country, work-experience ,nationality, visa, passport, immunization etc. We have job requirement page where request for some kind of job comes and we select the person after looking into their skillsets, trainings, availability and so on.
I want to create a recommendation system for this. I have seen many movie and e-commerce recommendation systems and what I found is all are based on cosine similarity. And generally they create a ‘soup column’ which consists of all appended text together which is then used for comparing similarity.
In my case, I have huge list of trainings and skillset required for each country depending upon the job location . So, using text similarities I think won’t be a good choice. Plus I don’t know how to tune and improve the model accuracy.
Any idea team, how to proceed on this.
Thanks,
Birendra Singh