FashionBrain

Summary

A core business of Europe’s fashion industry is to acquire a deep understanding of customer needs and to predict next trends. Search engines and social networks are often used as a bridge between the customer’s potential purchase decision and the retailer. In order to reinforce Europe’s position in the fashion industry and better exploit its distinctive characteristics e.g., multiple languages, fashion and cultural differences, it is pivotal to reduce its dependence to search engines. This goal can be achieved by harnessing various data channels that retailers can leverage in order to gain more insight about potential buyers, and on the industry trends as a whole. The main outcome of the FashionBrain project is the improvement of the fashion industry value chain obtained thanks to the creation of novel on-line shopping experiences, the detection of influencers, and the prediction of upcoming fashion trends. Tangible outcomes will include software, demonstrators, and novel algorithms for a data-driven fashion industry.

The FashionBrain project targets the two main actors of the fashion industry, i.e., the retailers and the customers. We propose to gather and combine the sheer amount of data generated by (emanating from) different fashion industry multisectorial players, starting from manufacturers and distribution networks, to online shops, large retailers, and value-added services companies (e.g., social media analysis, market observers, call centers, press/magazines etc). The gathered data will be curated, analyzed and used as input for machines learning algorithms. The outcome of the project will benefit retailers by providing novel services to their customers in order to improve their shopping experience and boost their loyalty. For example, a customer will be able to receive personalized recommendations and perform advanced fashion items search by image, complex textual description, etc. Also, a retailer will be able to compose a marketing story about a product that fits the customer’s taste, instead of merely showing an item and its price.

At a technical level, we propose ‘Big Data’ techniques to i) automatically process text and image data and extract key features such as fashion items, ii) capture customers’ preferences, iii) build a fashion-based recommendation engine and iv) predict new fashion trends as they emerge on social media and the blogosphere. In order to achieve these technical objectives, we propose to investigate deep learning methods and their execution on GPU-based hardware. Our research portfolio also covers interactive data sampling and curating methods, such as crowdsourcing or in-database-text mining, to deliver the fashion industry with large-scale curated data. Additionally, we will investigate the trend prediction by analyzing the similarity between fashion time series data, e.g., purchase quantity of a specific item, social media activity, etc.

The output of FashionBrain project will benefit EU’s retailers as they can offer exclusive data services to their customers; can anticipate customer needs, and predict short and long term trends or frauds. The main industry participants are i) Zalando SE is the Europe’s largest online retailer that collects data from 135 million customer transactions per month from 15 European countries, ii) Fashwell AG provides insight gathered from social web images and iii) MonetDB Solutions is the Europe’s pioneer in ultra-fast in-memory data processing techniques that will provide the infrastructure to store and query the data. The leading academic partners are i) UNIFR with semantic data integration and time series expertise, ii) USFD with hybrid human-machine computation experience and iii) Beuth-HS Berlin specializing in in-database text mining. The academic partners will transfer activities to industry, disseminate results at international conferences and train Europe’s next generation of data scientists. Moreover, both MonetDB Solutions and Zalando have a strong record in producing world-class open source code.

Grant Details

FashionBrain is a €2.8M (University of Fribourg’s share: 701’800 CHF) three-year project funded by the European Commission’s Horizon 2020 programme. The project is being led by Dr Alessandro Checco as coordinator and Dr. Mourad Khayati as technical lead, and involves several partners from across Europe, including the Beuth University of Applied Sciences, Université de Fribourg, Zalando SE, Fashwell AG and MonetDB Solutions B.V.

The FashionBrain project aims at combining data from different sources to support different fashion industry players by predicting upcoming fashion trends from social media as well as by providing personalised recommendations and advanced fashion item search to customers.

This project has received funding from the European Union’s Horizon 2020 research and innovation programme under grant agreement No 732328.