Third Workshop on Recommender Systems in Fashion

15th ACM Conference on Recommender Systems,
Amsterdam, Netherlands, 27th September-1st October 2021

Online Fashion retailers have significantly increased in popularity over the last decade, making it possible for customers to explore hundreds of thousands of products without the need to visit multiple stores or stand in long queues for checkout. However, the customers still face several hurdles with current online shopping solutions. For example, customers often feel overwhelmed with the large selection of the assortment and brands. In addition, there is still a lack of effective suggestions capable of satisfying customers’ style preferences, or size and fit needs, necessary to enable them in their decision-making process. Moreover, in recent years social shopping in fashion has surfaced, thanks to platforms such as Instagram, providing a very interesting opportunity that allows to explore fashion in radically new ways. Such recent developments provides exciting challenges for Recommender Systems and Machine Learning research communities.

This workshop aims to bring together researchers and practitioners in the fashion, recommendations and machine learning domains to discuss open problems in the aforementioned areas. This involves addressing interdisciplinary problems with all of the challenges it entails. Within this workshop we aim to start the conversation among professionals in the fashion and e-commerce industries and recommender systems scientists, and create a new space for collaboration between these communities necessary for tackling these deep problems. To provide rich opportunities to share opinions and experience in such an emerging field, we will accept paper submissions on established and novel ideas, as well as new interactive participation formats.

List of Accepted Contributions

Invited Industry Talk: H&M, A Journey into Recommender Systems, by Bjorn Hertzberg.
[preprint] A critical analysis of offline evaluation decisions against online results: A real-time recommendations case study, by Pedro Nogueira, Diogo Goncalves, Vanessa Queiroz Marinho, Ana Rita Magalhães and João Sá.
[preprint] Attentive Hierarchical Label Sharing for Enhanced Garment and Attribute Classification of Fashion Imagery, by Stefanos-Iordanis Papadopoulos, Christos Koutlis, Martina Pugliese, Manjunath Sudheer, Delphine Rabiller, Symeon Papadopoulos and Ioannis Kompatsiaris.
[ preprint ] Knowing When You Don’t Know in Online Fashion: An Uncertainty Aware Size Recommendation Framework, by Hareesh Bahuleyan, Julia Lasserre, Leonidas Lefakis and Reza Shirvany.
[preprint, video] Style-based Interactive Eyewear Recommendations, by Michiel Braat and Jelle Stienstra.
[preprint] SkillSF: In the sizing game, your size is your skill, by Hamid Zafar, by Julia Lasserre and Reza Shirvany.
[preprint] Using Relational Graph Convolutional Networks to assign Fashion Communities to Users, by Amar Budhiraja, Mohak Sukhwani, Manasvi Aggarwal, Shirish Shevade, Ravindra Babu Tallamraju and Girish Sathyanarayana.
[preprint] End-to-End Image-Based Fashion Recommendation, by Shereen Elsayed, Lukas Brinkmeyer and Lars Schmidt-Thieme.
[preprint] What Users Want? WARHOL: A Generative Model for Recommendation, by Jules Samaran, Ugo Tanielian, Romain Beaumont and Flavian Vasile.

Workshop Program

This year the workshop will run in a hybrid mode. On September 27 we will host the in-person component of the workshop, with lighting talks and topical round tables aimed to foster interesting discussions within the fashion recommender systems community. On October 2, we will host the virtual component of the workshop, with invited talks, keynotes, demos and oeral paper presentations.

Shatha Jaradat

KTH Royal Institute of Technology, Sweden.

Nima Dokoohaki

Accenture AI, Sweden.

Humberto Corona

Spotify, The Netherlands.

Reza Shirvany

Zalando, Germany

First Workshop on Recommender Systems in Fashion, 2019: Held in Copenhagen (Denmark)

Second Workshop on Recommender Systems in Fashion, 2020: Held in Online (Worldwide)

Julian McAuley Profile Picture

Keynote Speaker, Julian McAuley

"Fashion Recommender Systems: Some old tasks, some new tasks, and some emerging challenges"

Abstract: In this talk I'll give a high-level overview of the history and some of the emerging problems at the intersection of recommendation and fashion. We'll start by introducing some of the "traditional" approaches to fashion recommendation, and discuss the main challenges involved in handling complex visual data, dealing with temporal signals, collecting reliable ground-truth (etc.). Next we'll talk about some of the unique problems in fashion that don't fall into usual recommendation paradigms, such as wardrobe recommendation, size estimation (etc.). Finally, we'll discuss some of the emerging trends in this area, especially focusing on fairness and bias issues in fashion recommendation data.

Bio: Julian McAuley has been a professor in the Computer Science Department at the University of California, San Diego since 2014. Previously he was a postdoctoral scholar at Stanford University after receiving his PhD from the Australian National University in 2011. His research is concerned with developing predictive models of human behavior using large volumes of online activity data.

Sharon Chiarella profile picture

Keynote Speaker, Sharon Chiarella

Sharon Chiarella is the Chief Product Officer for Stitch Fix, the leading online personal styling company. In this role, Ms. Chiarella leads the Product, Design and Data Science teams responsible for product innovations that leverage machine learning and a team of 5,600+ expert Stylists to deliver a personalized shopping experience.

Prior to Stitch Fix, Ms Chiarella held engineering and product leadership positions, most recently as VP of Technology at Amazon. During her 13 year tenure at Amazon, she led iconic experiences including Customer Reviews, Deals, Shoppable Content, and AWS Mechanical Turk. She has 25+ years of experience developing and managing internet based products at Amazon, Yahoo, Microsoft/WebTV, and Kodak. Ms. Chiarella holds a BS in Computer Science from Manhattan College and an MBA from Harvard Business School.

Call for Contributions

Suggested topics for submissions are (but not limited to):

  • Computer vision in Fashion (image classification, semantic segmentation, object detection.)
  • Deep learning in recommendation systems for Fashion.
  • Learning and application of fashion style (personalized style, implicit and explicit preferences, budget, social behaviour, etc.)
  • Size and Fit recommendations through mining customers implicit and explicit size and fit preferences.
  • Modelling articles and brands size and fit similarity.
  • Usage of ontologies and article metadata in fashion and retail (NLP, social mining, search.)
  • Addressing cold-start problem both for items and users in fashion recommendation.
  • Knowledge transfer in multi-domain fashion recommendation systems.
  • Hybrid recommendations on customers’ history and on-line behavior.
  • Multi- or Cross- domain recommendations (social media and online shops)
  • Privacy preserving techniques for customer’s preferences tracing.
  • Understanding social and psychological factors and impacts of influence on users’ fashion choices (such as Instagram, influencers, etc.)

In order to encourage the reproducibility of research work presented in the workshop, we put together a list of open datasets in the fashionXrecsys website. All submissions present work evaluated in at least one of the described open datasets will be considered for the best paper, best student paper and best demo awards, which will be given by our sponsors.

Mentorship

This year we will continue to offer mentorship opportunities to students who would like to get initial feedback on their work by industry colleagues. We aim to increase the chances of innovative student’s work being published, as well as to foster an early exchange across academia and industry. As a mentee, you should expect at least one round of review of your work p r to the submission deadline. If your work is accepted, you should also expect at least one feedback session regarding your demo, poster or oral presentation.

If you want to participate in the mentorship program, please get in touch via e-mail.

Paper Submission Instructions

  • All submissions and reviews will be handled electronically via EasyChair
  • Submissions should be prepared according to the a single-column ACM RecSys format. Long papers should report on substantial contributions of lasting value. The maximum length is 14 pages (excluding references) in the new single-column format. For short papers, the maximum length is 7 pages (excluding references) in the new single-column format.
  • The peer review process is double-blind (i.e. anonymised). This means that all submissions must not include information identifying the authors or their organisation. Specifically, do not include the authors’ names and affiliations, anonymise citations to your previous work and avoid providing any other information that would allow to identify the authors, such as acknowledgments and funding. However, that it is acceptable to explicitly refer in the paper to the companies or organizations that provided datasets, hosted experiments or deployed solutions, if specifically necessary for understanding the work described in the paper.
  • Submitted work should be original. However, technical reports or ArXiv disclosure prior to or simultaneous with the workshop submission, is allowed, provided they are not peer-reviewed. The organizers also encourage authors to make their code and datasets publicly available.
  • Accepted contributions are given either an oral or poster presentation slot at the workshop. At least one author of every accepted contribution must be registered and attend the workshop to present their work, either in-person or virtually. Please contact the workshop organization if none of the authors will be able to attend.
  • All accepted papers will be available through the program website. Moreover, we are currently in conversations with Springer in order to publish the workshop papers in a special issue journal.

Additional Submission Instructions for Demos

Submissions for extended abstract of demos should be prepared according to the standard double-column ACM SIG proceedings format as described in the RecSys 2021 Demos Call for Contributions. The submission should include:

  • An overview of the algorithm or system that is the core of the demo, including citations to any publications that support the work..
  • A discussion of the purpose and the novelty of the demo.
  • A link to a narrated screen capture of your system in action, ideally a video. The maximum duration of screen capture is 10 minutes. (This section will be removed for the camera-ready version of accepted contributions but might be included in the virtual platform used to host the online part of the workshop.)
  • A link to a narrated screen capture of your system in action, ideally a video (This section will be removed for the camera-ready version of accepted contributions)
  • We also highly encourage any external material related to the demo (e.g., shared code on GitHub). Please, provide a link to the shared code in the extended abstract accompanying the demo.

Datasets

The following is a non-exhaustive list of datasets that are relevant for the fashionXrecsys workshop. Participants presenting work in any of these datasets will automatically be part of the workshop's challenge track. If there is a public dataset that you think should be added to the list, please contact the organizing comittee.

  • Amazon Reviews dataset:introduced by McAuley et al. contains product reviews and metadata from Amazon, including products in the Clothing, Shoes and Jewelry category. It includes 142.8 million reviews, ratings, product metadata as well as bought and viewed actions at product level. This dataset can be used in the tasks of single item recommendation or similarity calculation.

  • Large-scale Fashion (DeepFashion) is a fashion dataset built by Lui et al, which contains a set of 800,000 diverse fashion images that can be used for dif- ferent tasks such as fashion similarity, or fashion item recognition.

  • DeepFashion2 dataset is a dataset collected by Ge et al, which contains 491,000 images, mapping to 801,000 clothing items and more detailed metadata associated with them. It has useful metadata information that can be used for tasks such as fashion similarity, fashion item recognition and single item recom- mendation.

  • Clothing Fit Dataset for Size Recommendation are datasets collected by by Misra et al. from sever fashion e-commerce sites ModCloth and Rent The Runway. It contains size and self-reported fit information, as well as reviews and ratings for 1,738 items and 47,958 users. The dataset can be used for research in size and fit related recommendation problems.

  • ViBE (Dressing for Diverse Body Shapes) is a Dataset collected by [167] Hsiao et al. from the fashion site Birdsnest [168] which contains photography of gar- ments worn by a variety of models of different body shapes. The dataset also contains medatada for both the item and the model. Thus, this dataset is very interesting for solving size and fit related recommendation problems

  • Street2Shop is a dataset collected by M. Hadi Kiapour et al., which con- tains 20,357 labeled images of clothing worn by people in the real world, and 404,683 images of clothing from shopping websites. The dataset contains 39,479 pairs of exactly matching items worn in street photos and shown in shop images, which can be used for the task of similar item detection from images, complementary item recommendations, or outfit recommendations.

  • Pinterest’s Shop The Look Dataset collected by Kang et al. [140]. The dataset contains 47,739 scenes of people wearing fashion, which are labeled and linked to the corresponding 38,111 items. The dataset also has style labels, which makes it useful for the task of outfit recommendations or complementary item recom- mendations.

  • Alibaba iFashion is a dataset collected by Xu et al. The dataset contains 1.01 million outfits that are created by fashion experts. It also contains the clicks behaviors from 3.57 million users over a period of three months. This dataset can be useful for the task of outfits’ personalised recommendations.

  • PolyVore Outfits isis a dataset collected by Vasileva et al. It contains 68,306 outfits and 365,054 clothing items with their descriptions and types. This dataset is suitable for outfit recommendations or complementary item recommendations.

Important Dates

All deadlines refer to 23:59 (11:59pm) in the AoE (Anywhere on Earth) time zone.

  • Mentorship deadline: June 11, 2021
  • Submission deadline: August 10, 2021
  • Review deadline: August 24, 2021
  • Author notification: August 27, 2021
  • Camera-ready version deadline: preprints only, workshop proceedings will be published late 2021
  • Workshop: September 27 2021 & October 2 2021
  • Ana Peleteiro (Zalando)
  • Diogo Goncalves (Farfetch)
  • Eder Martins (Farfetch)
  • Evgenii Koriagin (Zalando)
  • Gordon Blackadde (ASOS)
  • Hamid Zafar (Zalando)
  • Hareesh Bahuleyan (Zalando)
  • Hosna Sattar (Zalando)
  • Jelle Stienstra (PTTRNS.AI)
  • Julia Laserre (Zalando)
  • Katrien Laenen (KU Leuven)
  • Leonidas Lefakis (Zalando)
  • Lucy Dunne (University of Minnesota)
  • Luís Baía (Farfetch)
  • Marjan Celikik (Zalando)
  • Mirela Riveni (University of Groeningen)
  • Nour Karessli (Zalando)
  • Pedro Nogueira (Farfetch)
  • Ricardo Savii (UNIFESP)
  • Roberto Roverso (Zalando)
  • Sofiede Cnudde (Asos)
  • Steven Bourke (Zalando)
  • Sergio Gonzalez Sanz (Zalando)