Machine learning meets fashion

Data, algorithms and analytics for the fashion industry


Fashion is a multi-billion dollar industry with social and economic implications worldwide. The fashion industry has traditionally placed high value on human creativity and has been slower to realize the potential of data analytics. With the advent of modern cognitive computing technologies (data mining and knowledge discovery, machine learning, deep learning, computer vision, natural language understanding etc.) and vast amounts of (structured and unstructured) fashion data the impact on fashion industry could be transformational. Already fashion e-commerce portals are using data to be branded as not just an online warehouse, but also as a fashion destination. Luxury fashion houses are planning to recreate physical in-store experience for their virtual channels, and a slew of technology startups are providing trending, forecasting, and styling services to fashion industry.

The first international workshop on fashion and KDD will be hosted at KDD 2016 in San Francisco, California on 14th August, 2016. The goal of this workshop is to gather people from academia, industry, and startups working at the intersection of fashion and data mining and knowledge discovery to further the technology and its adoption.

Workshop Schedule

14th August, 2016 afternoon (1pm - 5pm) @ Franciscan C/D

1:00 - 1:30 pm - Introduction
1:30 - 2:30 pm - Oral Paper Presentations
  • 1:30 - 1:50 pm - Fashion DNA: Merging Content and Sales Data for Recommendation and Article Mapping
  • 1:50 - 2:10 pm - Joint multi-modal representations for e-commerce catalog search driven by visual attributes
  • 2:10 - 2:30 pm - Visual Product Discovery

2:30 - 3:30 pm - Invited Talks
3:30 - 4:10 pm - Coffee break & Poster Session

4:10 - 4:50 pm - Oral Paper Presentations
  • 4:10 - 4:30 pm - Decoding Fashion Contexts Using Word Embeddings
  • 4:30 - 4:50 pm - Detection of fashion trends and seasonal cycles through the analysis of implicit and explicit client feedback

4:50 - 5:00 pm - Open house (5 min spot-light talks/demos inviting interested participants to present)

Invited Speakers

Elena Eberhard
Public Relations & Special Events Manager, School of Fashion, Academy of Art University, San Francisco

Elena Eberhard is a Public Relations and Special Events Manager in the Academy of Art University School of Fashion. She started as a Public Relations Director for Parfionova, a leading fashion house in Russia. After moving to San Francisco from Paris in 2013, Elena managed international relations and expansion on US and Russian market for German-based fashion industry tradeshow Premium Berlin and consulted fashion startups on international business development, as well as gave talks on Fashion Wearable Tech at Silicon Valley conferences. Throughout her international career of 18 years she attended numerous fashion weeks and tradeshows all over Europe and USA, as well as organized fashion shows and events, and contributed as a free-lance journalist for fashion and culture medias in Russia and France.

Jinah Oh
Director of Fashion Merchandising, School of Fashion

Director of Fashion Merchandising, earned her M.B.A. in Marketing at Golden Gate University in San Francisco, as well as a B.S. in Clothing and Textiles and a B.A. in Philosophy from EWHA Women’s University in Seoul, Korea. She has extensive experience in international fashion and luxury brands. With Escada Asia, she developed business and market strategies, managed merchandising and buying for multiple brands, pioneered and initiated brick and mortar and e-commerce channel development. Later with Cartier, Richemont Korea Ltd., she oversaw retail operations and all aspects of retail marketing activities for the Korean market. In 2010, she joined Savannah College of Art and Design (SCAD) and was the first appointed chair of Fashion Marketing and Management (B.F.A.), and Luxury and Fashion Management (M.A. / M.F.A.). To this day, both are the fastest growing programs at SCAD. She has developed and led multiple industry sponsored projects with partners in various sectors including technology, fashion and luxury, and trade associations such as Microsoft, Kohl’s, Sonoma Brand, Silver Promotion, JC Penny, HSBC, Benetton, and others.

Brad Klingenberg
Director of Data Science at StitchFix, San Francisco

Brad Klingenberg is the Director of Data Science at Stitch Fix in San Francisco. His team uses data and algorithms to improve the selection of merchandise sent to clients. Prior to joining Stitch Fix Brad worked with data and predictive analytics at financial and technology companies. He studied applied mathematics at the University of Colorado at Boulder and earned his PhD in Statistics at Stanford University in 2012.

Julian McAuley
Assistant Professor, Computer Science Department,University of California, San Diego

Dr. McAuley has been an Assistant 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.

Invited Talks

Challenges of quantifying fashion data: creativity, art and emotions
Elena Eberhard and Jinah Oh

Fashion is a field at the border of art and industry, combining elements of creative spontaneity in a unexpected ways, based on various sources of inspiration. It takes a human to create a clothing and a celebrity to make it fashionable. Real fashion world, designers and creative consumers (street fashion) provide an eclectic ever-changing content that science and technology are trying to optimize in order to increase sales and decrease the waste of over-production. In this talk we provide an overview of fashion big data problems: forecasting fashion trends, influencer analytics, visual search, natural language processing, style recommendation algorithms and the need to understand the natural life-cycle of a fashion garment before applying science in order to accelerate or alter it. Also, we will share some examples of collaboration projects between giants of technology and academics exploring the potential of quantifying fashion data.

Making fashion recommendations with human-in-the-loop machine learning
Brad Klingenberg

Most recommendation algorithms produce results without human intervention. Especially in hard-to-quantify domains like fashion combining algorithms with expert human curation can make recommendations more effective. But it can also complicate traditional approaches to training and evaluating algorithms. In this talk I will share lessons from making personalized fashion recommendations with humans in the loop at Stitch Fix, where we commit to our recommendations through the physical delivery of merchandise to clients.

Recommendation and Opinion Mining with Visual Signals
Julian McAuley

Building personalized systems for fashion recommendation presents several challenges due to the complicated semantics of people's preferences and styles. One challenge is simply the need to deal with sparse, long-tailed datasets, where new content is constantly introduced and recommendation is inherently a cold-start problem. Another challenge is the need to model visual signals, where the semantics of what makes items "attractive" are incredibly subtle. Finally, there is the need to model temporal dynamics that account for how fashion continually (and rapidly) evolves. In this talk we'll see how traditional recommendation approaches can be extended to explicitly account for the visual appearance of the items being recommended, in order to overcome these challenges and make visually- and stylistically-aware recommendations.

Accepted Papers/Posters

Topics of Interest

This is a new emerging area for the KDD community and we hope this workshop will bring together all the researchers, practitioners, and interested audiences to explore the open problems, applications, and future directions in this field. We believe that the fashion industry introduces a number of interesting data analytics problems that are either not studied or scarcely studied in the past and can attract great interest in the general KDD community given their practical implications. Suggested topics include (but not limited to):

  • Detect and forecast fashion trends and cycles
  • Big data for fast fashion (the like of Zara, H&M, and Primark)
  • Analyzing fashion blogs, articles, and images
  • Visual search for fashion e-commerce
  • Fashion image understanding and auto-tagging of apparel
  • Novel search mechanisms for large fashion catalogs
  • Virtual personal fashion assistants
  • Recommendation engines and cognitive stylists for fashion
  • Balancing art and science in fashion recommendation algorithms
  • Personal styling with humans and machines: recommendations with humans in the loop
  • Assembling outfit recommendations: interactions and serendipity
  • Algorithmic clothing: design by data
  • Predicting fashionability scores
  • Social networking for fashion
  • Fashion retail analytics
  • Interactive textiles
  • Digital wardrobe
  • Mining style rules
  • Assessing fashion personality (from social media platforms)
  • Virtual trial rooms
  • Plagiarism detection in fashion
  • Fashion and wearable computing

We also invite submissions in other retail domains where design, trends, styling, recommendations are important (for example, jewelry, furniture etc.).


Program Committee (PC) Members

  • Vikas C. Raykar, IBM Research
  • Brad Klingenberg, Director of Styling Algorithms, Stitch Fix, San Francisco
  • Heng Xu, Associate Professor, Information Sciences and Technology, Pennsylvania State University
  • Raghavendra Singh, IBM Research
  • Amrita Saha, IBM Research
  • Priyanka Agrawal, IBM Research
  • Qiang Ma, Research Scientist, Advertising Data Science, Yahoo! Inc.
  • Deep Ganguli, Stitch Fix, San Francisco
  • Jay Wang, Stitch Fix, San Francisco
  • Chris Moody, Stitch Fix, San Francisco

Submission Guidelines

We solicit submission of papers of papers of 4 to 10 pages representing reports of original research, preliminary research results, case studies, proposals for new work and position papers. We also seek poster submissions based on recently published work (please indicate the conference published).

All papers will be peer reviewed, single blind (i.e. author names and affiliations should be listed). If accepted, at least one of the authors must attend the workshop to present the work. The submitted papers must be written in English and formatted in the double column standard according to the ACM Proceedings Template, Tighter Alternate style. The papers should be in PDF format and submitted via the EasyChair submission site. The workshop website will archive the published papers.

For more information or any clarifications please email

  • Paper Submission Deadline: June 3, 2016 (Submission Closed)
  • Acceptance Notifications: June 20, 2016
  • Camera-Ready Submission Date: July 1, 2016
  • Workshop date: August 14, 2016

All deadlines are at 11:59 PM Pacific Standard Time.