suggest the best fitting product based on some questions the user has to answer. Currently the questions and the corresponding products are stored in a ~3000 lines json file. In terms of cross selling etc I thought about migrating the algorithm to some magic ML stuff. Do you think that would be worth it? If yes, what are the terms I would have to search for, in order to learn how to implement this with ML?
Thanks a lot
You may build an expert system but i dont know about that much
You should use "recommender systems"
Is the question set static ?
Seems like it's item-based collaborative filtering. https://medium.com/@cfpinela/recommender-systems-user-based-and-item-based-collaborative-filtering-5d5f375a127f
Обсуждают сегодня