In this PR I have created models for course recommendation as well as for skill recommendation. The skill rec model is trained on scrapped data which you can find at : skill_recommendation. For course recommendation I have used udacity course dataset. Also for each model I have employed primarily KNN and if the field is not present in the given dataset then we use cosine similarity algorithm. I have created classes of each model and modularized the code for better debugging and future changes. This will work as base for #30 issue where most probably we will be using custom pyresparser.
Note for Phase II:
Collect and preprocess the data made available to us by TnP cell and make appropriate changes in the model if required.
Description
In this PR I have created models for course recommendation as well as for skill recommendation. The skill rec model is trained on scrapped data which you can find at : skill_recommendation. For course recommendation I have used udacity course dataset. Also for each model I have employed primarily KNN and if the field is not present in the given dataset then we use cosine similarity algorithm. I have created classes of each model and modularized the code for better debugging and future changes. This will work as base for #30 issue where most probably we will be using custom pyresparser.
Note for Phase II:
Collect and preprocess the data made available to us by TnP cell and make appropriate changes in the model if required.