Programs
| 8:30-8:45 | Welcome and Introduction |
| 8:45-9:30 | Invited talk: "User Modeling on the World Wide Web" - Presentation by Andrew Tomkins |
| 9:30-10:00 | "Inferring Clickthrough Rates on Ads from Click Behavior on Search Results" by Sreenivas Gollapudi, Rina Panigrahy and Moises Goldszmidt |
| 10:00-10:30 | Break |
| 10:30-11:00 | Invited talk: "Social Media Analytics: Tracking the Flow of Information in Networks" - Presentation by Jure Leskovec |
| 11:00-11:30 | "Learning to Predict Web Collaborations" by Lilyana Mihalkova, Walaa Eldin Moustafa and Lise Getoor |
| 11:30-12:00 | Industrial Talk: "Targeting for Computation Market Research" by Frank Smadja |
| 12:00-13:30 | Lunch |
| 13:30-14:00 | Invited talk: "Leveraging Temporal Variability of Users and Content" by Eytan Adar |
| 14:00-14:30 | Position paper: "Towards a science of user engagement" by Mounia Lalmas, Simon Attfield, Gabriella Kazai and Benjamin Piwowarski |
| 14:30-15:00 | "Different Users and Intents: An Eye-tracking Analysis of Web Search" by Cristina González-Caro and Mari-Carmen Marcos |
| 15:00-15:30 | Break |
| 15:30-16:00 | "Web Queries: the Tip of the Iceberg of the User's Intent" by Cristina González-Caro, Liliana Caldero-Benavides and Ricardo Baeza-Yates |
| 16:00-16:30 | Invited talk: "The Moving Target of Mobile User Modeling" - Presentation by Irwin King |
| 16:30-17:00 | Panel and Discussion with the Audience |
Invited Speakers
Jure Leskovec
Social Media Analytics: Tracking the Flow of Information in Networks
Abstract: The information we experience online comes to us continuously over time, assembled from many small pieces, and conveyed through our social networks. This merging of information, network structure, and flow over time requires new ways of reasoning about the large-scale behavior of information networks. I will discuss approaches to tracking information as it travels and mutates in online networks. We show how to capture temporal patterns in the news over a daily time-scale -- in particular, the succession of story lines that evolve and compete for attention. We then develop models to quantify the influence of individual media sites and predict the popularity of online content. Since information spreads in many different ways we also introduce a method for tracing paths of diffusion through networks and inferring the networks over which information propagate. This leads to the last part where we efficiently detect information outbreaks and tackle the question of what news sites to follow to avoid missing important information?
Bio: Jure Leskovec is an Assistant Professor of Computer Science at Stanford University. His research focuses on mining and modeling large social and information networks, their evolution, and diffusion of information and influence over them. Problems he investigates are motivated by large scale data, the Web and on-line media. He received three best paper awards, a ACM KDD dissertation award, won the ACM KDD Cup in 2003 and topped the Battle of the Sensor Networks competition. Jure also holds three patents.
Eytan Adar
Leveraging Temporal Variability of Users and Content
Abstract: The World Wide Web is generally treated as a single snapshot in time. The consequence of this view of the Web as only existing in the "now" is that it throws away valuable data and imposes limits on users and Web services alike. By taking into account the temporal variability of both content and use I will explore some of the value that can be derived from treating the Web as a dynamic process. Specifically, I'll explore how certain cyclical processes such as revisitation and change can give us valuable clues about what dynamic information is actually important. Understanding this has implications to browser technologies, crawlers, and search engines when it is important to identify not all changing data, but rather interesting changing data. I'll also describe the broader value of nonreactive measures of behavior (also called unobtrusive measures or information side-effects) where large-scale patterns of use can be more broadly applied to both understand behavior and develop new technologies.
Bio: Eytan Adar is an Assistant Professor in the School of Information & Computer Science and Engineering at the University of Michigan. He completed his doctoral work in the Computer Science and Engineering Department at the University of Washington. He works in the area of temporal-informatics, studying how large populations interact with the dynamic Web and how those interactions can be enhanced. His interests are in understanding the dynamics of user behavior and data on the Web through text and log analysis, visualization, and the creation of new tools. Before entering graduate school, Eytan was a researcher at HP Labs and Xerox PARC for a number of years (spinning out a company called Outride somewhere in there). He received his Master of Engineering and Bachelor of Science degrees from the Massachusetts Institute of Technology. His website is at http://www.cond.org
Andrew Tomkins
User Modeling on the World Wide Web
Abstract: Online web pages present material in three different levels of personalization: generic, data, and model. The generic level performs no personalization; the data level employs simple queries based on the user's id; and the model level employs more sophisticated models of the user. I will discuss the prevalence of these modes, summarize the techniques that are common for model-based personalization and give some thoughts about future trends and the research problems they suggest.
Irwin King
The Moving Target of Mobile User Modeling
Abstract: The mobile industry has seen an explosive growth in recent years. Ways to integrate mobile user behavior patterns, interactions, intentions, etc. from various data sources are challenging research problems that may have significant impact to human-computer interface, network performance, business decisions, etc. In this talk, I plan to provide some background on these issues and demonstrate our initial work on commercial intention discovery and multi-task feature selection for better mobile user modeling.
Jure Leskovec
Social Media Analytics: Tracking the Flow of Information in Networks
Abstract: The information we experience online comes to us continuously over time, assembled from many small pieces, and conveyed through our social networks. This merging of information, network structure, and flow over time requires new ways of reasoning about the large-scale behavior of information networks. I will discuss approaches to tracking information as it travels and mutates in online networks. We show how to capture temporal patterns in the news over a daily time-scale -- in particular, the succession of story lines that evolve and compete for attention. We then develop models to quantify the influence of individual media sites and predict the popularity of online content. Since information spreads in many different ways we also introduce a method for tracing paths of diffusion through networks and inferring the networks over which information propagate. This leads to the last part where we efficiently detect information outbreaks and tackle the question of what news sites to follow to avoid missing important information?
Bio: Jure Leskovec is an Assistant Professor of Computer Science at Stanford University. His research focuses on mining and modeling large social and information networks, their evolution, and diffusion of information and influence over them. Problems he investigates are motivated by large scale data, the Web and on-line media. He received three best paper awards, a ACM KDD dissertation award, won the ACM KDD Cup in 2003 and topped the Battle of the Sensor Networks competition. Jure also holds three patents.
Eytan Adar
Leveraging Temporal Variability of Users and Content
Abstract: The World Wide Web is generally treated as a single snapshot in time. The consequence of this view of the Web as only existing in the "now" is that it throws away valuable data and imposes limits on users and Web services alike. By taking into account the temporal variability of both content and use I will explore some of the value that can be derived from treating the Web as a dynamic process. Specifically, I'll explore how certain cyclical processes such as revisitation and change can give us valuable clues about what dynamic information is actually important. Understanding this has implications to browser technologies, crawlers, and search engines when it is important to identify not all changing data, but rather interesting changing data. I'll also describe the broader value of nonreactive measures of behavior (also called unobtrusive measures or information side-effects) where large-scale patterns of use can be more broadly applied to both understand behavior and develop new technologies.
Bio: Eytan Adar is an Assistant Professor in the School of Information & Computer Science and Engineering at the University of Michigan. He completed his doctoral work in the Computer Science and Engineering Department at the University of Washington. He works in the area of temporal-informatics, studying how large populations interact with the dynamic Web and how those interactions can be enhanced. His interests are in understanding the dynamics of user behavior and data on the Web through text and log analysis, visualization, and the creation of new tools. Before entering graduate school, Eytan was a researcher at HP Labs and Xerox PARC for a number of years (spinning out a company called Outride somewhere in there). He received his Master of Engineering and Bachelor of Science degrees from the Massachusetts Institute of Technology. His website is at http://www.cond.org
Andrew Tomkins
User Modeling on the World Wide Web
Abstract: Online web pages present material in three different levels of personalization: generic, data, and model. The generic level performs no personalization; the data level employs simple queries based on the user's id; and the model level employs more sophisticated models of the user. I will discuss the prevalence of these modes, summarize the techniques that are common for model-based personalization and give some thoughts about future trends and the research problems they suggest.
Irwin King
The Moving Target of Mobile User Modeling
Abstract: The mobile industry has seen an explosive growth in recent years. Ways to integrate mobile user behavior patterns, interactions, intentions, etc. from various data sources are challenging research problems that may have significant impact to human-computer interface, network performance, business decisions, etc. In this talk, I plan to provide some background on these issues and demonstrate our initial work on commercial intention discovery and multi-task feature selection for better mobile user modeling.