Recommendation system for the fashion company - FORsystem
This study is proposes recommendation for the fashion company. We discuss the recomendation model using RNN units while considering many kinds of features including image, language and categorical data. We explore the possibility of the RNN based recommendation model for fashion company.
📁 Resources
Introdunction
In this project, we focused on fashion companies, especially in e-commerce. There are numerous choices are available so an efficient recommendation system is required in the fashion industry. But compared to other commodities, building effective recommendation systems for the fashion domain is challenging for a number of reasons.
- it is different from the conventional recommendation system method.
- The data generated by fashion companies are too diverse.
- Users’ preferences, product styles, and trends change over time.
So we set the fashion recommendation problem reflecting these characteristics. The problem is this. predict the items that a customer will buy in the next n-day period. N is the parameter. To solve this problem, we design a new recommendation architecture, and we call it FOR-system(fashion online recommendation system).
Model Design and Architecture
- Feature Extraction
- Image
- Language
-
Sequence mini-batches
-
Recommendation system with deep learning
- query + candidate
FOR-system(fashion online recommendation system)
Result
