W3C Workshop on Web & Machine Learning

[POSTPONED] Berlin, Germany

Ryzhi

Call For Participation

What is the purpose of this workshop?

The primary goal of the workshop is to bring together providers of machine learning toolkits and framework providers with Web platform practitioners to enrich the Open Web Platform with better foundations for machine learning.

The secondary goals of the workshop are as follows:

  • Understand how machine learning fits into the Web technology stack,
  • Understand how browser-based machine learning fits into the machine learning ecosystem,
  • Explore the impact of machine learning technologies on Web browsers and Web applications,
  • Evaluate the opportunities for standardization around machine learning APIs and formats.

We won't just be listening to presentations, but we will be actively participating in breakout sessions and working discussions covering topics identified as relevant to the participants.

What topics will be covered?

  • Is there a need for browser-implemented dedicated machine learning APIs? How high or low level should such browser APIs be? What are the prospects of dedicated machine learning processors (NPU, TPU, etc) deployment on the market?
  • How should machine learning APIs and frameworks integrate with data sources provided by Web browsers (sensor APIs, audio and video input from devices and WebRTC streams)?
  • How can machine learning be layered on top of existing dedicated high-performance computing APIs (WebGL, WebGPU, WebAssembly, …)?
  • How should machine learning primitives relate to domain-specific ML APIs (e.g. speech recognition, shape detection)?
  • What are the portability challenges and interoperability considerations for model exchange formats in on-device in-browser use?
  • How can client- and cloud-based machine learning interactions help adapt ML applications to improve user experience based on user environment (network, available computing power, available battery) while preserving their privacy? What client-side APIs would be required to enable such usages?
  • How much (if any) of training for machine learning can usefully be pushed to browsers, and if so, what considerations should apply in using that data? What aspects of AI/ML bias may apply, and how can those risks be mitigated?
  • What benefits and risks machine learning usage in browsers create from privacy, security and accessibility perspectives? How can the risks be mitigated?
  • How can machine learning primitives help browsers, authoring tools and evaluation tools improve accessibility of Web pages and Web applications?
  • What are the key considerations when choosing between ML application deployment on client-side vs. server-side? What are the gaps that prevent ML applications from moving to the client?

Suggestions for further workshop topics? Submit a pull request on GitHub or email Dominique Hazael-Massieux <dom@w3.org>.

References

How can I attend?

Attendance is free for all invited participants and is open to the public, whether or not W3C members.

If you wish to express interest in attending, please fill out the registration form. We want to fill the room with people with insights on the intersections of machine learning and Web technologies, either from practice experience in deploying solutions in that space, or from their involvement in relevant standardization activities.

Because the venue can accommodate at most 120 attendees, you must receive an acceptance email in order to attend. Also, be sure to keep an eye on these important dates.

As an alternative to the registration form, you are encouraged to submit a topic in the form of a position statement.

Our aim is to get a diversity of attendees from a variety of industries and communities, including:

  • ML-based application and service providers,
  • Experts on the intersection of machine learning a privacy, security, accessibility,
  • Machine learning frameworks and platforms,
  • Cloud computing providers,
  • Machine-learning hardware and chip providers,
  • Browser vendors,
  • SDOs involved in related standard setting.

This workshop, as other W3C meetings, operates under its Code of Ethics and Professional Conduct.

How can I suggest a presentation?

This is a workshop, not a conference, and any presentations will be short, with topics suggested by submissions and decided by the chairs and program committee. Our goal is to actively discuss topics, not to watch presentations.

In order to best facilitate informed discussion, we encourage attendees to read the accepted topics prior to attending the workshop.

If you wish to present on a topic, you can send us a position statement to the Program Committee at <group-machine-learning-pc@w3.org> by the deadline (see important dates). Our program committee will review the input provided, and select the most relevant topics and perspectives.

A good position statement should be a few paragraphs long and should include:

  • Your background on machine learning and Web technologies,
  • Which topic you would like to lead discussion on,
  • Links to related supporting resources.
  • Any other topics you think the workshop should cover in order to be effective.
  • Position statements must be in English, preferably in HTML or plain-text format; images should be included inline in HTML using base64-encoded data URIs. You may include multiple topics, but we ask that each person submit only a single coherent position statement. The input provided at registration time (e.g., bio, goals, interests) will be published and linked to from this workshop page.

What is W3C?

W3C is a voluntary standards consortium that convenes companies and communities to help structure productive discussions around existing and emerging technologies, and offers a Royalty-Free patent framework for Web Recommendations. We focus primarily on client-side (browser) technologies, and also have a mature history of vocabulary (or “ontology”) development. W3C develops work based on the priorities of our members and our community.

Location

The W3C Workshop on Web & Machine Learning is located at the Microsoft Atrium in Berlin, Germany.

Venue

Microsoft Atrium

Unter den Linden 17

Berlin, 10117

Germany

Program

Program Committee

Chairs

  • Kelly Davis (Mozilla)
  • Anssi Kostiainen (Intel)

Committee

  • Göran Eriksson (Ericsson)
  • Dominique Hazaël-Massieux (W3C)
  • Ningxin Hu (Intel)
  • Dean Jackson (Apple)
  • Sangwhan Moon
  • Roy Ran (W3C)
  • Georg Rehm (DFKI)
  • Amy Siu (Beuth University of Applied Sciences, Berlin)
  • Nikhil Thorat (Google)

Host

Microsoft

Becoming a Sponsor

W3C Workshops, meetups, and other events bring you into direct contact with leading Web technology experts: representatives from industry, research, government, and the developer community.

Whether your interests are focused on a particular topic being discussed by a Working Group, or you wish to reach a diverse international audience setting, your sponsorship will help your organization engage W3C's participants in its strategic direction.

Sponsorships offset a portion of our meeting costs, so W3C welcomes multiple sponsors for each event. All proposals for sponsorship are subject to W3C approval.

If you're interested in being a sponsor of the W3C Web & Machine Learning Workshop, please contact J. Alan Bird, Global Business Development Leader, at <abird@w3.org> or +1 617 253 7823.

For details on the available sponsorship opportunities for this workshop, see our Sponsorship Packages.