|
|
 |
Home > About SI > News > Article
News of SI
Mark Ackerman's SocialWorlds research group: Melding social science and technology
(Apr 2008) Though he did his graduate work in computer science and information technologies, Mark Ackerman's undergraduate degree was in "history in the social sciences," and both the technology and the social science influences are clear in the work of his research group.
An Associate Professor at SI and in electrical engineering and computer science, Ackerman heads up the "SocialWorlds" group. This team of faculty and doctoral students conducts research in the areas of collaborative technologies, pervasive computing, privacy, organizational memory, and online communities.
Ackerman describes the group's research as "combining social science and technology in a way that's very SI'ish but very different from other groups."
Their projects range across this spectrum, from ethnographic studies of how organizations build memory to the development of network simulation software.
Heirloom vegetables and heirloom computers
One SocialWorlds group researcher, doctoral student Jina Huh, is doing a study of users of the HP200LX, a palmtop computer that was discontinued almost a decade ago. She is working to understand how and why groups of users keep this piece of obsolete technology humming.
Huh studies online forums where devotees of such "abandonware" congregate and exchange information. "It's a whole world where people keep things alive," Ackerman says, drawing connections with other classes of relics kept alive only through the passion of devoted followers, such as heirloom vegetables or classic cars.
The analogy may be dramatic, but Ackerman is trying to draw a parallel between the kinds of communities that form around such items and the work people do collectively to maintain them. He sees Huh's project as an example of the new and different research that can be done when the technology becomes so pervasive that the issue of computerization itself fades into background.
"What happens when it stops being computerization and just becomes the fabric of our everyday lives," he asks, "when it's not special anymore?"
One of Ackerman's fascinations is the way such changes in infrastructure bring about changes in human behavior. As communications technology morphs, he points out, so does the nature of the communications conducted. For example, the near ubiquity of cameras today means more photos. "More photos," he says, "means change in ways we remember things."
The SocialWorlds group's work on memory in the digital age includes doctoral student Xiaomu Zhou's ethnographic study of the role played by "memory artifacts" in domestic environments -- heirlooms, souvenirs, photos, collections -- and their digital analogs. With people spending ever more time in the digital world, Zhou is interested in how the issues associated with memory, and the objects that trigger memories, will change.
Pervasive vs. privacy
The down side of the growth of pervasive computing is a commensurate shrinkage of personal privacy. Ackerman's own work in privacy strives to reconcile the notion of pervasive, ubiquitous computing technology with the human need to control what we consider to be personal data.
"Some have called privacy the killer threat in pervasive environments," writes Ackerman. "Finding a solution to this real or perceived threat is of considerable importance."
SI doctoral students Ben Congleton and Kevin Nam worked with Ackerman on issues of privacy in pervasive computing environments. Together they are developing a "privacy policy simulator," a software tool for simulating the effects of privacy policies in a given environment.
This tool allows people to see the implications of their choices when setting their privacy preferences within a system. As users adjust their settings, the tool applies these against past data to show the user, in effect, what the results of these policies would have been if they had been in force in the past. Such simulations can help users make more informed privacy decisions.
Jun Zhang, Mark Ackerman, and Lada Adamic developed the CommunityNet Simulator to study how various network characteristics will affect the structure of the network in an online help-seeking community. They've used it to test the effectiveness of automated methods of identifying the top experts in such a community.
|
Getting help online: Whom do you believe?
Anyone who has spent time scouring online forums seeking answers to life's persistent questions knows that all experts are not created equal. What if there were an automated means of separating the knowledgeable from the merely imaginative in these support forums? Ackerman and colleagues are working on it.
Ackerman has done recent work, much of it with Assistant Professor Lada Adamic and then-Ph.D.-candidate (now Dr.) Jun Zhang, analyzing and improving online help communities and forums, or "expertise networks." The researchers have worked to develop automated methods that will identify the most valuable and reliable experts in these networks. Their methods have scored nearly as well as human raters in identifying the best sources of information in a given online help community. This work has also resulted in a number of productive, targeted approaches to searching social networks.
The SocialWorlds team's work encompasses several other projects focused on automating expertise finding or knowledge retrieval. The Arkose system, an effort led by doctoral student Kevin Nam, helps human editors make sense of and distill knowledge from online discussion forums. Arkose brings together a number of visualization and information retrieval tools with an authoring tool and a navigator to help editors traverse the information space. It grew from an earlier system called iDiag, which was built to record and distill content from online community brainstorming sessions.
As Ackerman himself might say, "very SI'ish."
[Recent grants in support of the SocialWorlds group's work have come from the National Science Foundation, the National Institutes of Health, the U.S. Army Research Institute, and Intel Corporation.]
Visit the School of Information News Archive
Home > About SI > News > Article
|
|
|
|
|
|
|
Associate Professor Mark Ackerman is co-editor of a 2007 volume on computer-supported cooperative work (CSCW) titled Resources, Co-Evolution, and Artifacts: Theory in CSCW. The book looks at how software resources get created, adopted, modified, and abandoned.
Ackerman's SocialWorlds group has worked with some interesting data sets in doing its analysis:
- a set of more than half a million e-mail messages from the senior management of the late Enron Corporation (graphed above)
- 8.4 million answers to more than a million questions from the Yahoo! Answers Web site
- 50,000 threads (comprising more than 300,000 messages) in the Java Developer Forum
A "bow tie" analysis of the Java Developer Forum shows that more than half of all users only ask questions (the "In" users) while only about one out of eight users both asks and answers questions (the "Core" users).
|
|
|
|
|
|
