Complexity and Social Networks Blog

« August 2007 | Main | October 2007 »

30 September 2007

"Marshall Van Alstyne on "Diffusion, Network Structure & Information Advantage"

Tomorrow (Monday Oct 1) Marshall Van Alstyne will be lead off the latest year of The Cambridge Colloquium on Complexity and Social Networks of the Program on Networked Governance. Full announcement below.

"Diffusion, Network Structure & Information Advantage"

Marshall Van Alstyne
Boston University
School of Management

Monday, October 1, 2007
12:00 – 1:30 p.m.
Fainsod Room, 3rd Floor Littauer Building

Marshall Van Alstyne, Professor Van Alstyne works in the area of Information Economics. His research interests include the economics of networks, valuing information, equity and growth effects of information sharing, and integration effects of access to technology. The underlying theme is information: how to value it; how does it affect productivity, product design and competitive advantage; how does it alter property rights; what happens when it is shared, and why access alone may not lead to everyone having it independent of preferences. (for more information click here).

Abstract: The talk will be a summary of 2 papers (one on drivers of information diffusion, the other on how social networks affect access to novel information and the effect on productivity). Both papers are currently available below. In these papers we examine relationships between social network structure, information diversity, and individual performance. Specifically, we investigate which network structures influence access to novel information, and whether these relationships explain performance in information intensive work. We trace the word level diffusion using a ten month panel of email communication. Then we build and validate an analytical model of information diversity, develop hypotheses linking size and diversity to the distribution of novel information among information workers. We test our theory using statistical evidence linking message content to project revenue among employees at a medium sized executive recruiting firm.
Our results indicate that: (1) The total amount of novel information flowing to actors increases in their network size and network diversity. (2) The marginal increase in information diversity decreases in actors' network size. (3) Network diversity contributes to performance even when controlling for the positive performance effects of access to novel information. This suggests additional benefits to network diversity beyond those conferred through information advantage. (4) Traditional demographic and human capital variables have surprising effects on access to diverse information, highlighting the importance of network structure for information advantage. The methods and tools developed are replicable and can be readily applied to other settings in which email is widely used and available, opening a new frontier for the analysis of social networks and information content.

Aral, Sinan, Brynjolfsson, Erik and Van Alstyne, Marshall W., "Productivity Effects of Information Diffusion in Networks"
Aral, Sinan and Van Alstyne, Marshall W., "Network Structure & Information Advantage"

Posted by David Lazer at 1:09 PM

27 September 2007

More on computational social science...

I presented some of my work on "computational social science" at the Applied Stats workshop yesterday. One of the questions that came up was what are the best tools to deal with unusual and massive data sets. Clearly, part of the answer is that there is nothing truly "off the shelf" that you have to write a lot of code from scratch. But the other part is that there are some flexible platforms/tools that are vastly better than others, and I would be interested in comments on what you think is useful for datasets, say, with millions of observations, or pulling text and link structure off of the web, etc.

Posted by David Lazer at 10:25 AM

Overview of U.S. and Canadian 311 city and county service center

I have created a mash-up of the U.S. and Canadian 311 projects (last update: 9/26/07) which I would like to share with you. There are currently around 70 service centers (311) in the U.S.. Most of the 311 projects have been realized on the municipal level and in most of the U.S. biggest cities. While 11 countys have decided to offer 311, not all of them are multi-jurisdictional, that is information and services from the municipalities within a county are not integrated. Furthermore, 311 services can have various levels of sophistication and may either be operated by the police department or by newly generated 311 service units/departments.

The first city to test 311 was Baltimore in 1996, however, it was Chicago which used 311 in 1999 in a much broader way for public service provision, city management and accountability. The City of Chicago's 311 still is the first place to visit and learn from for many elected officials and public managers. Today, New York City is the biggest 311 implementation in the U.S. (size of the service center/ population served) and probably the most well known implementation due to the global media coverage it received. With a pop of approx. 5400 Alaska's City of Bethel is the smallest place to use the number.

Except for the City of Somverille none of the cities in Massachusetts have implemented 311. Given the close proximity of cities in the greater Boston metro area its really hard to understand from a citizen's perspective why there couldn't be a single 311 solution for the whole area. After all, there would be around 3.6 million people less to serve than in NYC and there should be many information redundancies.

View Larger Map

Blue = Municipal 311 (Realized)
Red = County 311 (Realized)
Yellow = Planning or implementation stage
Green = 311 in Canada

If you know about new 311 projects please email me.

Posted by Alexander Schellong at 12:05 AM

20 September 2007

More on networks in political science and sunbelt...

FYI, we have spoken to the Sunbelt organizers, who have reserved two + panel slots for political science related presentations. So, if you do register to present at Sunbelt, and wish to be part of a political science oriented panel, when you select a panel, write in:

Networks in Political Science (NIPS I)

or, if you are interested in policy networks, in particular:

Networks in Political Science (NIPS II): Policy Networks

Posted by David Lazer at 12:53 PM

19 September 2007

Finding those mutual friends

There has been a spate of stories (also this story from globe) recently about the use of cell phones to track the locations of your friends. There has also been some talk of linking data from the social networking sites to cell phones, so, for example, you could walk into a room and instantly find someone who you had never met, but who was a friend of a friend (through facebook or some other database that was stored in the phone).

These are clever ideas, but let me throw out another idea: a program for phones that facilitates those “you grew up in Long Island? Did you know John Smith, by any chance?” moments. (A small digression—this exact question (except for the name) was posed to me once. And, surprisingly, I actually did know the person in that case.)

The basic idea is quite simple—if you are talking to someone, and you both have the program, have the phones link via Bluetooth and:

1) find overlapping phone numbers and report them back. A more extensive version of this would also match any incoming or outcoming calls the phones have ever made.

2) collect and match structured data about the owners of the phones—where and when you went to school, where you’ve lived different years, etc., and report back matches. This is all of that information that people initially exchange when they meet. While no substitute, this could be faster and more thorough way to those “Do you know” questions. That is, you would both instantly find out if you both happened to live in Ann Arbor for the same two years in the 1990s.

This would not be a hard program to write, but there is the classic chicken-egg problem of how to get enough people to sign on to make it work. Not my problem—but if you write such a program, let me know.

Posted by David Lazer at 4:34 PM

Networks in political science

I have been meaning to write about the incipient rise of network ideas within political science since the political science (APSA) meetings over Labor Day weekend. There has always been a (very) thin thread of network ideas within the discipline. The most robust has been on political behavior, most notably the work by Robert Huckfeldt, John Sprague, and collaborators, and Samuel Patterson wrote about networks in legislatures. And there has always been a networky component of the federalism literature, e.g., in Jack Walker’s work on diffusion among states. And of course, the recent work from Robert Putnam on social capital has a relational foundation. Recent years have definitely witnessed an uptick in network-related publications in political science, but nowhere is this trend more evident than at APSA. As a point of comparison, back in the mid 90s, there would be at most a handful of network-related papers (some years, just me, I believe). This year, I am guessing there were 50-60. This included a thematic panel to lead off the conference, with Michael Heaney, John Scholz, Scott McClurg, John Padgett, Christopher Ansell, James Fowler, Sarah Reckhow and myself.

I will note that those of us on the thematic panel met after to talk about what we could do to catalyze the network agenda within political science. Approaches like this face a particular challenge within political science because of the fairly rigid division of the discipline into subfields across which there is relatively little communication. Thus, cross-subfield methods and theories face a critical mass problem. It’s not insurmountable, but it does require some self conscious efforts to get the ball rolling.

Anyhow, while we have floated the possibility of doing a conference on networks in political science (stay tuned on that), one immediate step I would recommend is getting a solid group of political scientists to attend Sunbelt. I have now spoken to a number of folks, and I think we are pretty much assured of getting 10-15 political scientists there (instead of < 5), and I would encourage any interested political scientists to attend. Note that the deadline to submit abstracts is coming up fast—October 5.

And in the mean time, I would be interested in hearing suggestions as to how to catalyze things (either through e-mails to me, or preferably as a comment to this posting).

Posted by David Lazer at 2:36 PM

17 September 2007

Governance and Information Technology: From Electronic Government to Information Government

It is my pleasure to announce that Governance and Information Technology: From Electronic Government to Information Government, edited by Viktor Mayer-Schönberger and myself, is now available. I am also happy to say that we were able to negotiate a 20% discount from MIT for readers of this blog-- from an already low $15 to $12. Just follow this link.

I have also asked our contributors to step in as guest bloggers, so the next three months we will have a weekly posting on "Information Government"-- starting next week with a posting by Viktor and myself.

A brief description of the volume:

Developments in information and communication technology and networked computing over the past two decades have given rise to the notion of electronic government, most commonly used to refer to the delivery of public services over the Internet. This volume argues for a shift from the narrow focus of "electronic government" on technology and transactions to the broader perspective of information government--the information flows within the public sector, between the public sector and citizens, and among citizens--as a way to understand the changing nature of governing and governance in an information society.

Contributors discuss the interplay between recent technological developments and evolving information flows, and the implications of different information flows for efficiency, political mobilization, and democratic accountability. The chapters are accompanied by short case studies from around the world, which cover such topics as electronic government efforts in Singapore and Switzerland, the U.S. Environmental Protection Agency’s effort to solicit input on planned regulations over the Internet, and online activism "cyberprotesting" globalization.

Contributors:
Robert D. Behn, Maria Christina Binz-Scharf, Herbert Burkert, Lorenzo Cantoni, Cary Coglianese, Martin J. Eppler, Jane E. Fountain, Monique Girard, Åke Grönlund, Matthew Hindman, Edwin Lau, David Lazer, Viktor Mayer-Schönberger, Ines Mergel, Gopal Raman, David Stark, Sandor Vegh, and Darrell M. West.

And some of the generous things that have been said about the book:

"So you thought information technology in the form of 'e-government' would save taxpayer dollars, improve government performance, increase transparency and accountability, and promote democratic participation--and all in a hurry too? Some first-rate scholars of the subject show how the several truths about these matters are much more complicated, and the reasons for them sometimes paradoxical."
--Eugene Bardach, Department of Public Policy, University of California, Berkeley

"Through a rich set of essays by leading thinkers, this book advances the next generation of ideas about information technology and government, moving the literature beyond the original, transactional conception of 'electronic government.' The authors bring up to date a thesis extending back to Federalist thought in the U.S., which is that flows of information are central to the exercise of power and indeed form one of the foundations of government."
--Bruce Bimber, Department of Political Science, University of California, Santa Barbara

"The editors of Governance and Information Technology have assembled a strong juxtaposition of general overviews and concrete case studies to critically examine ways in which information and communication technologies are reconfiguring access to information both within government and between governments and citizens. This book not only challenges the idea that new technologies are democratizing access, but also presents alternatives conceptions, such as the development of an 'information class,' that will shape debate and research on the political implications of e-government"
--William H. Dutton, Director Oxford Internet Institute, University of Oxford

"The e-governance revolution has transformed the way that government commonly delivers basic services. But has it transformed democracy? This is a first-class study of the complex processes of information flows between citizens and government. Drawing upon well-known experts and a diverse range of cases, the study provides provocative and important insights into processes of political communications, the uses and limits of information technologies, and the transformation of modern governments."
--Pippa Norris, Director, Democratic Governance Group, Bureau for Development Policy United Nations Development Programme

Posted by David Lazer at 9:34 PM

Offline networking course for the Facebook generation

While the baby boomers are slowly taking to online networking (see my earlier post here), some youngsters should probably do a little less of it.
Apparently too much use of Facebook does not have a positive effect on "real-life" (or traditional) networking. An article by Michael Schulman in this week's edition of the New Yorker talks about an NYU freshman seminar entitled "Facebook in the Flesh". The aim of the seminar, part of a series of seminars during freshman orientation, was to re-introduce the Facebook generation to face-to-face meetings. Participants were given a few questions ("What drew you to NYU?") they had to ask each other in pairs. This excerpt from the article nicely summarizes the outcome:

"[The facilitator] blew a whistle. 'Thoughts? Feelings? Reactions?' he said. 'Was it hard?' 'Harder than Facebook,' one girl said."

Some food for thought.

Posted by Maria Binz-Scharf at 1:00 AM

13 September 2007

2 Post-Doc positions

The Institute for Prospective Technological Studies (IPTS) is one of the seven scientific institutes of the European Commission's Joint Research Centre (JRC). IPTS is located in Seville, Spain.

This short entry is to inform you that IPTS published 2 Post-Doc positions in the context of the following projects:

- eInclusion Strategic Support 2020
- New innovation models in eApplications (web2.0 and innovation in public services)

The deadline is September 16th.

Posted by Alexander Schellong at 11:35 AM

"Older people are sticky" - social sites for baby boomers

An article in today's New York Times discusses the emergence and popularity of social networking sites aimed at the 55+ population. Very interesting, I thought, in particular the hypothesis that while these sites might take longer than myspace or facebook to reach high levels of usage, baby boomers are likely to "hang around". This resonates well with investors...read the full article below.

The Graying of the Web
By MATT RICHTEL

SAN FRANCISCO, Sept. 11 — Older people are sticky.

That is the latest view from Silicon Valley. Technology investors and entrepreneurs, long obsessed with connecting to teenagers and 20-somethings, are starting a host of new social networking sites aimed at baby boomers and graying computer users.

The sites have names like Eons, Rezoom, Multiply, Maya’s Mom, Boomj, and Boomertown. They look like Facebook — with wrinkles.

And they are seeking to capitalize on what investors say may be a profitable characteristic of older Internet users: they are less likely than youngsters to flit from one trendy site to the next.

“Teens are tire kickers — they hang around, cost you money and then leave,” said Paul Kedrosky, a venture capitalist and author of the blog “Infectious Greed.” Where Friendster was once the hot spot, Facebook and MySpace now draw the crowds of young people online.

“The older demographic has a bunch of interesting characteristics,” Mr. Kedrosky added, “not the least of which is that they hang around.”

This prospective and relative stickiness is helping drive a wave of new investment into boomer and older-oriented social networking sites that offer like-minded (and like-aged) individuals discussion and dating forums, photo-sharing, news and commentary, and chatter about diet, fitness and health care.

Last week, VantagePoint Ventures, an early investor in MySpace, announced that it had led a $16.5 million round of financing for Multiply, a social networking site aimed at people who are settled.

In August, Shasta Ventures led a $4.8 million financing round for TeeBeeDee, a site coming out of its test stage this month. The name is short for “To Be Determined” (as in: just because you’re not trolling for a mate on MySpace doesn’t mean your life is over.)

Also in August, Johnson & Johnson spent $10 million to $20 million to acquire Maya’s Mom, a social networking site for parents, according to a person briefed on the deal. The site has been in existence about a year.

Social networking has so far focused mainly on businesspeople and young people because they are tech-savvy and are treasured by Madison Avenue.

But there are 78 million boomers — roughly three times the number of teenagers — and most of them are Internet users who learned computer skills in the workplace. Indeed, the number of Internet users who are older than 55 is roughly the same as those who are aged 18 to 34, according to Nielsen/NetRatings, a market research firm.

TeeBeeDee’s founder is Robin Wolaner, who in 1987 created Parenting magazine. That year, at least seven magazines focused on being a parent were started, and Ms. Wolaner said she was seeing the same sudden recognition of a need for Internet publishers to respond to the demands of older Americans.

She came up with the idea for the site, she said, “when I was sitting around with friends and we said, ‘We’re not going to hang out at the AARP site. What is there for us?’ ” (Plus, she said, she wanted to find a community where she could discuss her interest in getting an eye lift).

“There’s a recognition that this generation now uses the Internet just like younger people,” she said. “The one thing this generation hasn’t done yet is network online.”

The question is whether they’ll want to network in large enough numbers to justify the tens of millions of dollars going into the space. Indeed, the interest from entrepreneurs and venture capitalists has led to a mini-boom in sites that cater to baby boomers, creating what they say is both critical mass and a likely falling out.

Some of the older users of the sites say the experience feels more comfortable to them than when they tried MySpace, Facebook or Friendster.

“I’ve discussed my divorce, my medical issues, and when do I dare go dating again,” said Martha Starks, 52, a retired optician in Tucson, who spends an hour or two each evening on a site called Eons. “I sure wouldn’t discuss that stuff with a 20-year-old.”

She says she talks about lighter things, too, like movies and music, with an audience that gets what she is saying.

“They don’t even know who Aretha is — she’s the queen of soul!” she said.

Meg Dunn, 38, who is raising three children in Fort Collins, Colo., said she had tried MySpace and Facebook but had found that the short attention span of users didn’t suit her either. She now uses Multiply, where she shares family photos with her relatives, and gets into discussions on substantive topics, like health issues and illnesses affecting elderly people.

“I feel like I’m putting down roots, building relationships,” she said. “My feeling on MySpace is that people give you a poke, and then they’re gone and you never see them again.”

Peter Pezaris, president and chief executive of Multiply.com Inc., based in Boca Raton, Fla., said he believed that older customers were stickier than younger ones, but said the evidence so far was anecdotal. He said 96 percent of the company’s active users returned each month, a statistic that he said impressed the venture capitalists who considered investing in the site.

David Carlick, a managing director with VantagePoint, which led the latest investment round in Multiply, said he believed that social networking sites in general had a bright business future as advertisers start to gravitate to them. He also said he believed that targeted sites, like those focused on an age demographic, could be particularly effective.

He said he had some concern that sites focusing on younger users could be vulnerable to the whims and caprice of fashion.

“That was on our minds when Murdoch came in with an offer,” he said of the decision to sell MySpace to the publishing tycoon Rupert Murdoch for around $550 million.

But venture capitalists and entrepreneurs have been slow to embrace the interests of older Internet users, said Susan Ayers Walker, a freelance technology journalist for AARP and founder of SmartSilvers Alliance, which offers consultant services to businesses looking to connect with older consumers.

She said that Silicon Valley investors have seen themselves as eternally youthful, and identified with ever-new gadgets. But they are starting to accept their age — and to invest in it.

“They’ve all got high blood pressure,” she said. “They’re starting to understand their age group — they’re living it.”

Peter Ziebelman, a partner at Palo Alto Venture Partners, joked that the interest in sites aimed at aging Americans represented the end of a state of denial for venture capitalists.

“Perhaps there aren’t many V.C.’s who want to be in the newspaper saying they’re backing the 5o-and-over population,” he said. “They’d rather say they’re attending the next keg party.”

(Mr. Ziebelman is an investor in www.agis.com, which is not a social networking site, but focuses on delivering information and services to people who need help with elder care.)

Ms. Ayers said that the investors are learning that social networks aimed at older users are a big draw for investors, consumer products and services companies. “Not only do we have a lot more money, we pay a lot more attention to advertisers,” she said.

The advertisers on Eons include Humana health care insurance, Fidelity Investments and the pharmacy chain CVS. Lee Goss, president and chief operating officer of Eons Inc., which received backing from the venture capital firms Sequoia Capital and General Catalyst, said that the sites aimed at an older audience may not grow as quickly as MySpace, but could have longevity.

“Our audience, while it is harder to attract, is more durable and sticky over time,” he said.

Posted by Maria Binz-Scharf at 1:00 AM

12 September 2007

Campaign contributions and the network of hedgers

I’ve blogged in the past about the fund raising networks underlying the campaign. Well, one of my students last term, Katie Selenski, did a nice paper examining the pattern of campaign contributions to Giuliani, Obama, and Clinton. She examined contribution patterns as two mode (contributor x candidate) data. In particular, who maxed out in contributing to at least two of those three candidates? 344 individuals fell into that category in the first quarter of this year (clearly people with more money to spare than I do). To give a flavor, below is a list of California Obama-Clinton hedgers (88 in all).

While there are some big name entertainers on the list (see Streisand, Spielberg, Manilow, etc), there is a remarkable number of top executives (big circles)—although we don’t know how this compares to nonhedgers. One suspects that for top executives there is a particular instrumental value to these contributions, which is why we would see so many.

Below is the picture of hedgers nationally.

Again, there are a large number of top executives. And, unsurprisingly, there are more Clinton-Obama hedgers than Giuliani-Obama or Giuliani-Clinton. But there are still a remarkable number of folks in the latter two categories. Some interesting patterns: there are more attorneys who do a Clinton-Giulianai hedge than Obama-Giuliani, and more people from finance who go the Obama-Giuliani path. Presumably this reflects the underlying social structure of campaign contributions—e.g., one imagines that many individuals are recruited to contribute by associates, and may have associates who belong to different candidate camps.

In any case, these are nice illustrations of the potential power of these data. I have seen a fair amount of news reporting based on these data—would be interested in postings on academic or professional (i.e., campaign fundraising) uses of these data.

Posted by David Lazer at 10:47 AM