Hi All,
We've built a prototype recommendations engine using the taste item similarity 
algorithms and have wrapped this around a Rest API.  I have also written code 
to basically take the mathematical intersection of the recommendations returned 
from taste with the actual user preferences both of which are stored in our 
MySQL instance.  This is somewhat tedious so I was wondering if it makes sense 
to build a configurable plugin that takes in a set of user preferences and have 
that influence the item similarity recommendations that are being returned.


Thoughts, let me know if I was unclear in anything above.

Sent from my iPhone

Reply via email to