Industry Event

* This page is in draft mode, there will be updates. Thanks.

As a new addition to the standard program of previous editions of the conference, the Industry Day section (separate from the industry research track) consists of talks given by visionaries from industry and from the VC (Venture Capital) community located (mostly) in Silicon Valley. The talks are meant to give the conference audience a sense of what it takes to move an idea from its initial write-up into a trend-changing success that improves the lives of millions of people around the world. Speakers and Panelists for Industry Day are by invitation only.

List of Speakers
Organizers

Abstracts

List of Speakers

Andrei Z. Broder (Yahoo! Research, USA)

Steven Gary Blank (Stanford University, USA)

Christopher J. C. Burges (Microsoft Research, USA)

Doug Cutting (Yahoo!, USA)

Omar Hamoui (AdMob, USA)

Ronny Kohavi (Microsoft, USA)

Peter Norvig (Google, USA)

Daniel E. Rose (A9.com, USA)

Peter Rip (Crosslink Capital, USA)

Organizers

Marius Pasca (Google, USA)

James G. Shanahan (Independent Consultant, USA)

Abstracts

Andrei Z. Broder

Affiliation: Fellow and VP of Computational Advertising, Yahoo! Research

Bio: Andrei Broder is a Fellow and Vice President for Computational Advertising in Yahoo! Research. Previously he was an IBM Distinguished Engineer and the CTO of the Institute for Search and Text Analysis in IBM Research. From 1999 until 2002 he was Vice President for Research and Chief Scientist at the AltaVista Company. Previously he was a senior member of the research staff at Compaq's Systems Research Center in Palo Alto. He was graduated Summa cum Laude from Technion, the Israeli Institute of Technology, and obtained his M.Sc. and Ph.D. in Computer Science at Stanford University under Don Knuth. His current research interests are centered around computational advertising, web search, context-driven information supply, and randomized algorithms. Broder is co-winner of the Best Paper award at WWW6 (for his work on duplicate elimination of web pages) and at WWW9 (for his work on mapping the web). He has authored more than eighty papers and was awarded over twenty patents. He is an ACM Fellow, an IEEE fellow, and past chair of the IEEE Technical Committee on Mathematical Foundations of Computing.

Presentation title: TBD

Presentation abstract: TBD

Steven Gary Blank

Affiliation: Stanford University, Graduate School of Engineering and University of California Berkeley, Haas Business School

Bio: Over the last 30 years, Steve has been part of, or co-founded eight Silicon Valley startups. These have run the gamut from semiconductors, video games, personal computers, and supercomputers. (MIPS, Zilog, Rocket Science, SuperMac, Convergent Technologies, Ardent, ESL) Steve's last company was E.piphany, an enterprise software company. Steve is on the board of CafePress.com an on-line marketplace, and IMVU a 3D IM social network. Steve also serves on the California Coastal Commission and is on the board of Audubon National & is Chairman of Audubon California and Peninsula Open Space Trust (POST.) Steve currently teaches entrepreneurship at U.C. Berkeley Haas Business School, the joint Berkeley/Columbia Executive MBA program, and at the Stanford University Graduate School of Engineering. Steve teaches a methodology of starting and managing marketing, sales and business development in high technology startups. Course text at http://www.cafepress.com/kandsranch

Presentation title: The Secret History of Silicon Valley

Presentation abstract: The origins of Entrepreneurship in Silicon Valley has many threads; world class universities, regional climate and culture, capital markets, etc.

This talk tells the story of how World War II set the stage for the growth of defense industries in Silicon Valley during the Cold War. It describes the seminal role of Frederick Terman and Stanford in working with the CIA and the National Security Agency during the Cold War to invent the entrepreneurial culture in the valley.

Christopher J. C. Burges

Affiliation: Principal Researcher, Microsoft Research

Bio: Chris Burges started out with a PhD in theoretical physics, and has been conducting a random walk in career space ever since. He was at Bell Labs for 14 years, working on network performance modeling, network routing algorithms, and machine learning: AT&T used an algorithm he designed to route their CCS7 signaling network. He also worked on OCR and neural nets for handwriting and machine print recognition for check and ZIP code reading; he contributed to work that is still used to process millions of checks daily in bank back offices. Chris was then fortunate to have an opportunity to work with Vladimir Vapnik on an ARPA contract to develop Support Vector Machines. After joining Microsoft Research in 2000, Chris has worked on a variety of projects, principally audio fingerprinting, which currently ships as part of Windows Media Player, and also on ranking for information retrieval, which currently ships in Live Search.

Presentation title: The Mountains or The Street Lamp: Search, Research, and Research Again

Presentation abstract: How do the fruits of research find their way from the halls of large industrial research labs to the outside world? Should real world impact even be a criterion that an academic researcher should care about? How can companies maximize the effectiveness of their research efforts, from both the company's and the researcher's perspectives? I will start by describing some examples gathered from over two decades of experience in industrial research labs, and I will distinguish between goal-driven and exploratory research. My observations are informal and are offered in the spirit of personal reminiscences of an involved player, rather than through any claim of formal expertise in the management of large research projects.

One way to mitigate risk in research is to try to maintain a balanced portfolio of projects, and to openly expect a significant failure rate from the higher risk projects. To make these ideas concrete I will describe two current projects, the first briefly and the second in more depth. The first is a very high risk effort, whose chances of success are extremely small, but whose consequences if successful would be far reaching: that of factoring large semiprimes. The second is a lower risk project, in which I have applied learning methods to the problem of ranking web search results. The latter project has led to new information retrieval techniques that are currently used in a major search engine.

Doug Cutting

Affiliation: Yahoo!

Bio: TBD

Presentation title: TBD

Presentation abstract: TBD

Omar Hamoui

Affiliation: Founder and CEO, AdMob

Bio: Omar is an entrepreneur with deep roots in the mobile industry. After earning a degree in Computer Science at UCLA, Omar founded and ran several companies in mobile software and services. At the time, he was frustrated by the lack of available options when it came to promoting and monetizing mobile products. Omar realized that mobile service and content firms needed better ways to advertise and make money on the mobile Web. While earning his MBA at the Wharton School, Omar began to think more seriously about the untapped potential of mobile advertising, and created AdMob to address this growing opportunity. Today, Omar has taken a leave from Wharton to focus on AdMob, and is now running one of the fastest growing and most innovative companies in the mobile world.

Presentation title: The Data Driven Challenges and Opportunities of Mobile Advertising

Presentation abstract: For most online businesses, data is becoming an increasingly important strategic asset. For AdMob, data is at the center of almost everything that we do. AdMob operates an ad network wherein millions of consumers, thousands of mobile content and service providers, and hundreds of advertisers participate in the distribution of commercial messages. AdMob's role in this network is to match ads to content and consumers billions of times a month in a way that is relevant to consumers, that produces a high return on investment for advertisers with varying business models, and that maximizes revenue for providers of relatively disparate content and services.

In addition to these already complicated constraints and the high volume of traffic in our network, we operate our network in the mobile ecosystem. This ecosystem consists of many hundreds of service providers in all parts of the world. The resulting confederation of networks and servers is somewhat less homogenous and standardized than we are now accustomed to with the Internet. For AdMob, this means that we're trying to solve a set of constrained optimization problems, using enormous volumes of data, thousands of time a second, in the presence of noise, where success or failure is metered in dollars.

I will devote the majority of this talk to a discussion of how AdMob handles some of its data problems. I will talk about a few of our successes and best practices that have emerged from them. Perhaps of more importance, I will talk about a few problems related to statistical machine learning, convex optimization, and information visualization that continue to vex us.

Ronny Kohavi

Affiliation: General Manager, Experimentation Platform, Microsoft

Bio: Ronny Kohavi is the General Manager for Microsoft's Experimentation Platform, a team whose mission is to build a platform that will accelerate software innovation through trustworthy experimentation. Controlled experiments, also called A/B tests, allow evaluating ideas through randomized assignment of users to a Control group or different Treatment groups. The methodology is practically the only scientific method we know to establish causal relationships between ideas and metrics of interest.

Prior to joining Microsoft in 2005, Ronny was the director of data mining and personalization at Amazon.com, where he was responsible for personalization, automation, search engine marketing (SEM), consumer behavior / data mining, site experimentation, and automated e-mail. His teams introduced several features estimated to be worth several hundred million dollars in incremental revenue. Prior to Amazon, Ronny was the Vice President of Business Intelligence at Blue Martini Software, where he led the engineering group responsible for the data collection, analysis, visualization, reporting, and campaign management modules in Blue Martini's applications. Prior to joining Blue Martini, Kohavi managed the MineSet product, Silicon Graphics' award-winning product for data mining and visualization. MineSet was based in part on MLC++, a machine learning library developed at Stanford University.

Ronny received a Ph.D. in Machine Learning from Stanford University and a BA from the Technion, Israel. He was the General Chair for KDD 2004, he co-chaired KDD 99's industrial track with Jim Gray, and he co-chaired the KDD Cup 2000. He was an invited speaker at Emetrics 2007, the SF ACM Data Mining SIG in 2006, Emetrics 2004, KDD 2001's industrial track, the National Academy of Engineering in 2000.

More information about Ronny is available at www.kohavi.com.

Presentation title: Practical Guide to Controlled Experiments on the Web: Listen to Your Customers not to the HiPPO

Presentation abstract: The web provides an unprecedented opportunity to evaluate ideas quickly using controlled experiments, also called randomized experiments or A/B tests (and their generalizations). Controlled experiments embody the best scientific design for establishing a causal relationship between changes and their influence on user-observable behavior. We provide a practical guide to conducting online experiments, where end-users can help guide the development of features. Our experience indicates that significant learning and return-on-investment (ROI) are seen when development teams listen to their customers, not to the Highest Paid Person's Opinion (HiPPO). We provide multiple real-world examples of controlled experiments with surprising results. We review the important ingredients of running controlled experiments, and discuss their limitations (both technical and organizational). We share key lessons and pitfalls that will help practitioners in running trustworthy controlled experiments.

The talk is based on joint work with Randy Henne and Dan Sommerfield: http://exp-platform.com/hippo.aspx

Peter Norvig

Affiliation: Director of Research, Google

Bio: Peter Norvig is a Fellow of the American Association for Artificial Intelligence and the Association for Computing Machinery. At Google Inc he was Director of Search Quality, responsible for the core web search algorithms from 2002-2005, and has been Director of Research from 2005 on.

Previously he was the head of the Computational Sciences Division at NASA Ames Research Center, making him NASA's senior computer scientist. He received the NASA Exceptional Achievement Award in 2001. He has served as an assistant professor at the University of Southern California and a research faculty member at the University of California at Berkeley Computer Science Department, from which he received a Ph.D. in 1986 and the distinguished alumni award in 2006. He has over fifty publications in Computer Science, concentrating on Artificial Intelligence, Natural Language Processing and Software Engineering, including the books Artificial Intelligence: A Modern Approach (the leading textbook in the field), Paradigms of AI Programming: Case Studies in Common Lisp, Verbmobil: A Translation System for Face-to-Face Dialog, and Intelligent Help Systems for UNIX. He is also the author of the Gettysburg Powerpoint Presentation and the world's longest palindromic sentence.

Presentation title: TBD

Presentation abstract: TBD

Daniel E. Rose

Affiliation: Director, Search Relevance, A9.com, a subsidiary of Amazon.com

Bio: Daniel E. Rose is the Director of Search Relevance at A9.com, which provides the product search engine used on Amazon.com and other retail sites. Previously, he was Director of Search Presentation, Assistance, and Interaction at Yahoo! Inc., where he oversaw several projects for improving user interaction with web search. Dan has been working in the area of search for over fifteen years. At Apple's Advanced Technology Group, he spent several years leading the Information Access Research team, which explored document search, clustering, summarization, and collaborative filtering. His group also developed the core search technology later used in Apple's Sherlock and Spotlight products. Dan has also held key positions at web search pioneer AltaVista and at Xigo, a startup where he architected a search and alert service for realtime financial news. He holds a Ph.D. in Cognitive Science and Computer Science from UC San Diego and a B.A. in Philosophy from Harvard University.

Presentation title: Crowdsourcing for Relevance Evaluation

Presentation abstract: Relevance evaluation is an essential part of the development and maintenance of information retrieval systems. Yet traditional evaluation approaches have several limitations; in particular, conducting new editorial evaluations of a search system can be very expensive. I'll describe a new approach to evaluation based on the crowdsourcing paradigm, in which many online users, drawn from a large community each performs a small evaluation task.

Peter Rip

Affiliation: General Partner, Crosslink Capital

Bio: Peter brings over 25 years of experience as a successful entrepreneur, venture investor, and institutional investor. He has focused exclusively on early stage technology companies.

Before becoming a venture investor, Peter co-founded Silicon Compiler Systems, a major IC design automation software company acquired by Mentor Graphics in 1991. He began venture investing in 1992. He has specialized in investing in early stage companies, several of which have been acquired by AOL, CA, HP, Microsoft and others. Prior to joining Crosslink, Peter was a Managing Director at Leapfrog Ventures and the Managing Director of Knight Ridder Ventures.

Presentation title: NA (Panelist)

Presentation abstract: NA (Panelist)