A carregar...
Publicação
Developing a dynamic recommender system for the aftermarket car industry: dimensionality reduction techniques for recommendation system - the tips4y use case
| Nome: | Descrição: | Tamanho: | Formato: | |
|---|---|---|---|---|
| 542.74 KB | Adobe PDF |
Orientador(es)
Resumo(s)
Recommendation systems are a powerful tool for e-commerce businesses when applied properly. After extensive research on the various models used in the market, a recommendation system was successfully designed, with the purpose of helping TIPS 4Ysupport after-market distributors by optimizing the recommendation of auto parts. Adopting a CRIPS-DM framework, an Item-Based Collaborative Filtering was developed, with the data provided by TIPS 4Y, that performed far better than the other two baseline models. Attempts to improve the model, Dimensionality Reduction techniques were tested. A data dash board was also designed, allowing TIPS 4Yto make management decisions on working with data.
Descrição
Palavras-chave
Data Science Business analytics Recommendation system Tips4y Automotive aftermarket Dimensionality reduction Data dashboard