Abstract

Shopping online for clothes is becoming very popular recently. But finding good clothes remains a difficult task. We face a wealth of clothes on offer, and without the possibility to fit or feel the product, making decisions is not easy. In this chapter, we present a use case of an online retailer that aims to improve the shopping experience of men. Differing from conventional online shops where the customers browse through various products and eventually add items to their shopping basket, this shopping service relies on the expertise of fashion advisers who, after getting in contact with the costumers, arrange a combination of different clothes and ship them to the customers. We present a case-based recommendation approach using the available user information entered explicitly, such as price constraints and preferred colors, and also learn a user model based on purchase histories. We evaluate and compare our case-based approach with standard recommendation approaches. The evaluation shows that even with little knowledge, a suitable user model can be learned and used for computing recommendations. The evaluation bases on real data of customers of an online shop. Based on the results, using a case-based recommendation approach could help to solve cold-start problems. But for computing good recommendations for all users, more information about explicit user preferences is needed, which is currently not available.

@inbook{sis_book_plumbaum3,
author = {Till Plumbaum, Benjamin Kille},
booktitle = {Smart Information Systems - Computational Intelligence for Real-Life Applications},
chapter = {Personalized Fashion Advice},
editor = {Frank Hopfgartner},
organization = {Springer},
publisher = {Springer}, 
series = {Advances in Computer Vision and Pattern Recognition},
volume = {...},
year = {2015},
ISBN = {978-3-319-14177-0},
ISSN = {...},
}