Thought this may be interesting to both my social media and user interface students. Even if you are not submitting, read the call for participation just to become informed about trends and research ideas.
Call for Participation: 1st International Workshop on User Modeling from Social Media
In conjunction with IUI 2012, Lisbon – Portugal
Massive amounts of data are being generated on social media sites, such as Twitter and Facebook. People from all
walks of life share data about social events, express opinions, discuss their interests, publicize businesses,
recommend products, and, explicitly or implicitly, reveal personal information.
This workshop will focus on the use of social media data for creating models of individual users from the content that
they publish. Deeper understanding of user behavior and associated attributes can benefit a wide range of intelligent
applications, such as social recommender systems and expert finders, as well as provide the foundation in support
of novel user interfaces (e.g., actively engaging the crowd in mixed-initiative question-answering systems). These
applications and interfaces may offer significant benefits to users across a wide variety of domains, such as retail,
government, healthcare and education. User modeling from public social media data may also reveal information that users would prefer to keep private. Such concerns are particularly important because individuals do not have complete control over the information they share about themselves. For example, friends of a user may inadvertently divulge private information about that user in their own posts. In this workshop we will also discuss possible mechanisms that
users might employ to monitor what information has been revealed about themselves on social media and obfuscate
any sensitive information that has been accidentally revealed.
In this workshop, we will discuss related topics:
• What aspects of an individual can be modeled from their public social media postings?
• What aspects cannot be modeled?
• What aspects should not be modeled?
• How accurate are the models that can be extracted?
• What are the best techniques for creating models?
• How might the creation of such models be thwarted?
(e.g. to preserve privacy while still allowing participation on a social network)
We hope to bring together researchers and practitioners from diverse areas, such as user modeling, intelligent user
interaction, social media analysis, natural language processing, data mining, machine learning, privacy and
security, to discuss these issues and share results.
TOPICS OF INTERESTS
Topics of interest may include but not limited to:
• Domain-specific user modeling using public social
media, including twitter, facebook, myspace, social
Q&A sites, and Amazon.com reviews for
• Domain-independent user modeling using public
social media, such as twitter, facebook, myspace, and
foursquare, to derive a wide variety of user traits
• Enterprise-focused user modeling using social media
data on public social networks and communications
(e.g., emails and blogs) within an enterprise:
- Employees’ social and collaboration patterns in a
- Work-related personality traits such as innovativeness, versatility, adaptiveness,
leadership quality, and level of expertise.
• Task-specific user modeling for
- Information recommendation
- Expert finding
- Social Q&A
We plan to propose a special issue of ACM Transactions on
Interactive Information Systems (TIIS) on this topic after
We invite submissions in two categories:
• Position papers (2 pages)
• Short papers (4 pages)
We will hold a full-day workshop program on the first day
of IUI 2012. The program begins with a madness session
during which participants introduce themselves and their
work in 5 minutes. Participants will be encouraged to make
slides for the madness session, but this will not be required.
Each paper session will conclude with a discussion led by a
pre-chosen workshop participant. These discussions will tie
together common themes of the presentations and hopefully
lead to insightful discussions about further research
directions. The program will end with a panel discussion
where panelists will discuss the current state of the art,
focus areas, and opportunities for future research.
• Paper Submissions – January 6, 2012
• Author Notification – January 20, 2012
• Camera-Ready version – January 27, 2012
• Workshop – February 14, 2012
CONTACT & WEBSITE
Jalal Mahmud, IBM Research – Almaden
Jeffrey Nichols, IBM Research – Almaden
Michelle Zhou, IBM Research – Almaden