HarambeeNet Advisory Board Meeting September 24

Time: 9:00am-4:00pm on September 24, 2007

Place: North Building room 314 on Duke University’s West Campus

• Board Members
  • Eytan Adar, Ph.D. Student in Computer Science at the University of Washington and creator of GUESS
  • Noshir Contractor, Jane S. & William J. White Professor of Behavioral Sciences Professor of Industrial Engineering & Management Science, McCormick School of Engineering; Professor of Communication Studies, School of Communication; and Professor of Management & Organizations, Kellogg School of Management at Northwestern University
  • Jonathon Cummings, Associate Professor of Management at Fuqua School of Business at Duke University
  • Jennifer Golbeck, Assistant Professor of Information Studies at University of Maryland
  • Balachander Krishnamurthy, Researcher at AT&T Labs
  • Deepak Kumar, Professor of Computer Science at Bryn Mawr College
  • Ellen Spertus, Associate Professor of Computer Science at Mills College
  • Fred Stutzman, Ph.D. Student at the School of Information and Library Science at University of North Carolina
• Principal Investigators
  • Jeff Forbes, Assistant Professor of the Practice of Computer Science at Duke University
  • Susan Rodger, Associate Professor of the Practice of Computer Science at Duke University
• Undergraduate assistants
  • Kandis Wood
  • Samantha Jones
  • Dametrious Peyton

As the final part of your duties from this workshop, you should submit a report on your proposed module.

In summer 2008, we will have a 1.5 day workshop for educators interested in the science of networks. Advisory board members would attend this workshops and meet for additional day for an advisory board meeting.

Proposed date:

• 2nd week of June, 2008

Insert your proposed module here.

I do stuff (http://www.cond.org)

Think Networks!

http://www.netvis.org/tutorial.php

tutorial for how to analyze and visualize social networks (social science research methods)

I study social networks on the web http://trust.mindswap.org/

I don’t teach a course.

• Recommender System
• Shortest Path

My course http://www.ibiblio.org/fred/inls_490/index.php/Main_Page

• At the beginning stages, but we have a lot of questions:
  • Interest in Computer Science over time graph → Why does this pattern occur?
    • The gap expanded between men and women’s interest over time; computer science was the only field to have that happen → Why are fewer & fewer women attracted to computer science?
    • Why was computer science so popular in 1984? –the intro of the PC
    • Second peak in interest: due to the internet
    • Recent decline in interest: how do we explain it?
      • Outsourcing
      • Maybe there is a perception of what occurs in the field that does not appeal to college kids
      • The way computer science is taught…
      • Computers are more ubiquitous now; they are embedded in other departments/jobs
      • People cannot define Computer Science; think that it is only programming
      • People see their major as a stepping stone to an ultimate goal (i.e. biology → becoming a doctor)
      • Computer Science has a bad reputation; students believe employees look for Engineering, i.e.
      • There is a marketing/image issue
      • BUT, other fields such as Chemical Engineering has not had the same decline interest & it does not have the same gender gap AND it does not have an “attractive” image either
    • How can we stop the decline in interest in Computer Science?
    • The COHFE schools: lots of big private schools—Interest in Computer Science going down in those schools as well
      • Why is Computer Science below Math & Physics? –People have taken math & physics before, but most students are unfamiliar with computer science
    • NSF CPATH Program: Wants to revitalize computer science education on a national scale
      • CPATH Community Building
      • Evaluation, Adoption, Extension Program
• Our project
  • Build community
  • A diverse Advisory Board (different stages; different backgrounds)
  • Faculty Learning Communities: Bring faculty together to meet regularly& discuss a topic. People do not do this often with education (usually do it with research)
    • 8 Duke faculty will meet this year
    • 8 faculty from Duke & 6 faculty from other NC schools - will meet next academic year
    • Will look at modules, then narrow in on projects to work on
  • Goal: To create curricular modules. It’s hard to adopt a course. But, we will create problem-based modules. Specific problems.
  • Also proposing to build a new course: Quantitative. Will not be about programming. Aimed at non-majors. The idea would be that lots of students take the course. Appeal to the students who major in social science.
• People who aren’t interested in computer science—never take a computer science course—will not major in it
• A future for Computer Science?
  • The science of networks
    • Is there a science of networks?
• What’s new about networks?
  • Graph theory is old—1736
  • But, the recent study of networks is different
• The hypothesis: most networks will have similar structures
• Social Networks
  • i.e. The Oracle of Bacon – we have the actual data → How do we visualize this?
• Web-based Social Networks
  • MySpace – 150 million
  • Adult friend finder – 23 million
  • Facebook – 24 million
  • Bebo – 30 million
  • Orchid

1. Lunch and Introductions: Everyone introduced themselves, their affiliation, and explained the aspect of networks that interested them.

• Jeff Forbes, Computer Science, Duke, is interested in leveraging student interest in music to look at social networks, have students use collaborative filtering algorithms, and nearest neighbor analysis (example: DukeScrobbler).
• Eytan Adar, University of Washington., is interested in large scale analysis of web data using social networks, visualization, homophily, and algorithms to see how homophily spreads.
• Fred Stutzman, School of Information and Library Science at UNC-Chapel Hill is teaching a course on online social networks and is interested in individual relationships with information (ie via search engines), using social networks and thinking critically about search engines connecting social behaviors.
• Andrea Novicki, Center for Instructional Technology, Duke, brings CIT’s experience in conducting faculty groups to facilitate the FLC.
• Robert Duvall, Computer Science, Duke, is interested in having relevant and realistic data sets to visualize and explore, to increase student interest in computer science.
• Richard Lucic, Computer Science, Duke, is exploring virtual worlds (Second Life, Croquet) to enhance and study social networks in the real world.
• Susan Rodger, Computer Science, Duke, uses Facebook examples in her class to look at who is at the center of a graph or a path.
• Jonathon Cummings. Fuqua School of Business, Duke is interested in paths, short paths, and how information flows through an organization. He would like to introduce ideas of centrality, betweeness and network properties.
• Ellen Spertus, Mills College is interested in comparing algorithms and results and collaborative filtering.
• David Banks, Statistics, Duke, has three interests: 1. Using feedback in social networks similar to that used in gene regulation, 2. Identifying holes in Wikipedia networks and 3. Dimensionality of agents.
• James Moody, Sociology, Duke, is interested in network methods for sociologists, the dynamic evolution of networks and how it affects carrying goods (like ideas).
• Joshua Socolar, Physics, Duke, is interested in dynamic processes in networks and exploring algorithms for networks. He is interested in using networks that underlie common phenomenon in his teaching.
• Deepak Kumar, Computer Science, Bryn Mawr College, would like to make computer science more attractive to a wider population. He described his work in Serendip which arose from a network of individuals to provide resources for education and for exploring personal interests. He is interested in emergent properties.
• Balachander Krishnamurthy, AT&T Labs, focuses on internet measurements, and metrics for social networks.
• Jennifer Golbeck, Information Studies, University of Maryland, is interested in trust in web-based social networks, and how to generate and identify small world networks
• Noshir Contractor, numerous departments and schools, Northwestern University, is interested in injecting theories of social science into creating communities of self interest, social exchange, and homophily. He is interested in developing a tool in computer science to analyze networks based on different social motivations.
• Owen Astrachan, Computer Science, Duke, is the facilitator of the FLC; he’s responsible for making sure that the participants are on track, and the point person for resources for the FLC. He is interested in problems that could be used in a variety of courses, and in how might results change when people realize they are being studied, which ties into student awareness of privacy issues.

1. Goals/ Modules for FLC and for Advisory board.

• What is a module? Modules can be used to infiltrate computer science into other courses, because interest and enrollment in computer science has decreased. The modules should be at least reasonably self contained, not just slides. They should also be interactive, and available online.
• Discussion ensued about whether the webpage for the module has to be interactive, or if interactivity can arise from activities within the module.
• Owen stated that the web page for the module does not have to be interactive, but that all of the materials you would need to fit the module into the course should be included.
• Discussion continued about designing a course rather than modules, and the course would dictate the modules. Jeff thought the modules should be designed first, to drop into courses, and a course can be built from the modules.
• Noshir Contractor asked if there were modules available already, that could be mined. Jeff said that initially, the FLC will survey courses and other materials available.
• Should Duke (or anyone) own the content of the modules? Consensus was that the content would be freely available.
• Deepak Kumar mentioned a model for distributing modules based on Serendip, which provides web modules not created with any particular course in mind. The problems addressed by the modules are designed to be interesting to anyone, not just within a course.
• Jeff provided two handouts, a list of science of networks courses, and portions of Network Science, from the National Academies Press.
• Discussion continued about the audience for modules, and what should be included. Jeff said the audience is not grad students or seniors, but you can expect some mathematical sophistication. Can build on the previous user experience with networks, but students don’t conceptualize networks, to them they are just communicating with friends.
• Noshir Contractor pointed out that the issue of prerequisites is important, because the modules should be included in gateway courses, therefore the prerequisites should be minimal.
• Owen pointed out that the most popular majors at Duke are economics and biology, therefore the students have calculus. Joshua and others discussed the problems with relying on the student’s ability to use calculus, even if they have had classes.
• Minimally, the modules can be aimed at students who are not math challenged.
• Robert pointed out the modules can be targeted at several levels – incorporating extra support for targeting weak backgrounds. Jim pointed out that people across different disciplines will use the materials in different ways.
• What about assessments, should they be embedded in the modules? David thought that might be thinking a bit far ahead. Richard asked if the assessments were to determine if the module attracted students to computer science. Others asked if this was the correct goal, and Owen did not believe this was possible or why the grant was funded.
• Joshua pointed out that network science is a growing field in itself that spans disciplines in interesting ways. The goal of modules is to help students realize that networks underlie many social and physical phenomena, and they may draw people into thinking about networks.
• Jennifer Golbeck related her experience teaching a nonmajors course on networks, which was popular and successful in that students were excited about computer science; however, she was not permitted to teach the course again.
• What is the goal of the faculty learning community and of the modules?
  • Jeff explained that it is a community building project, with the first outcome being around networks. Joshua suggested one possible outcome could be to convince the computer science department to create more connections and courses to attract students.
• Jeff gave an overview of the FLC – the first 3rd of the year the participants will survey what’s out there and identify a problem, then they will develop a module, and finally evaluate and test their module. The problems will have to selected.
• What are the canonical resources? Books mentioned:
  • Social Network Analysis: Methods and Applications (Structural Analysis in the Social Sciences) by Stanley Wasserman and Katherine Faust
  • Nexus: Small Worlds and the Groundbreaking Theory of Networks by Mark Buchanan
  • The Structure and Dynamics of Networks: (Princeton Studies in Complexity) by Mark Newman, Albert-Laszlo Barabasi, and Duncan J. Watts
  • Small Worlds: The Dynamics of Networks between Order and Randomness (Princeton Studies in Complexity) by Duncan J. Watts
• Discussion continued about how much of a time commitment from students modules should be, and the conclusion was that it was open.
• Jeff introduced the Harambenet, showed the materials on it. He asked everyone to sign up for an account and login, using the netid as the user name.
• One of the goals for the FLC participants is to use the wiki, and before the next FLC meeting, to post their goals under their names on the wiki here.
• Jeff asked for a recommendation of a conference for the Faculty Learning Community. Unanimous recommendation is Sunbelt (INSNA). It’s a very interdisciplinary conference, particularly with participants from business schools and public health. They accept most submissions. Papers are currently 20 minutes. The schedule of meetings is here. The abstract submission deadline for this January’s meeting is October 5, 2007. The FLC could present a panel discussion on teaching social networks at the meeting in early 2009, in southern California.