On Wednesday, 2:30pm at CHI 2010 (in Regency 5), Anna Wu and I will be presenting a paper that touches on this topic. Anna and I spent last summer figuring out which user behaviors on Beehive indicate that two IBMers have a strong or weak relationship, and then further determining which behaviors reflect a professional versus a personal closeness. As you might expect, many behaviors on the site indicate general closeness, but a few are reflective of personal closeness only. Come to our talk to find out the details! (or, read the paper.)

A Wu, JM DiMicco, DR Millen. (2010) “Detecting Professional versus Personal Closeness Using an Enterprise Social Network Site.” Proceedings of CHI 2010, April 2010.

In social networks, people can increase their defenses against identification by adopting tight privacy controls on information in personal profiles. Yet an individual’s actions, researchers say, are rarely enough to protect privacy in the interconnected world of the Internet.

You may not disclose personal information, but your online friends and colleagues may do it for you, referring to your school or employer, gender, location and interests. Patterns of social communication, researchers say, are revealing.

Something we’ve been thinking about lately are the people you actually forgot you friended. You may have had a brief interaction with them which was meaningful in the moment, so led to immediate friending, following or exchange of contact information. But as time and space increased, you forget the specifics of the interaction and when you look at the person’s name it doesn’t trigger a memory for you. Have you looked at your friend lists lately and can you remember who everyone is? What are you revealing about yourself to those people? Do you mind sharing your life with non-familiar strangers? On the flip side, are you keeping up with the lives of any total strangers just because they are on one of your friend lists? (If you can’t relate at all to these issues, maybe you have a better social memory than I do? Or you are more selective in your friending from the start?)

I find this to be a particular problem for me on LinkedIn because I meet people at professional conferences and then never see them again. Another aspect of this is if the stranger is an infrequent posters (or limits the visibility of their posts), then there is nothing to remind me that I’m following them. They are silent wallflowers that may or may not be observing my posts.

Let’s say you go through the effort to pick out these strangers on your friend list. Then there is the whole issue of not wanting to offend these people by de-friending them. What if this total stranger remembers your meeting and hopes to reconnect again? What if they have been following your updates and when you defriend them are painfully aware you have removed them? Or what if they have no idea who you are either, and is just too polite to de-friend you? Awkward!!

What I find really interesting about this argument is that there are no fundamental truths about how online communication must happen. Why is email private? Because it has been since the 70’s? Why are status message public? Because that is twitter’s default setting? Status messages can be private (e.g. protected tweets) and emails could be public (maybe that’s what blogs are). These private/public distinctions are pretty arbitrary in my mind. Sure, I assume that the email I send today is private because that is how my email software behaves today, but I can imagine email environments that don’t operate this way.

Google also has enough market share and influence that the way they design their applications can fundamentally change how we think about email and status messages. If Buzz takes off, we may look back a few years from now and have trouble remembering that there used to be a clear distinction between emails and status messages. It is all just communication with our network after all.

• 5 Easy Steps to Stay Safe (and Private!) on Facebook Impression management guide for Facebook hits the NYTimes. “When the President of the United States warns schoolchildren to watch what they say and do on Facebook, you know that we’ve got a problem.” Uh, yes! Glad this is now considered a mainstream problem.
  
  
• Twitter Serves Up Ideas From Its Followers Consumers = Producers! Of the next feature set! This is also not news for a Twitter user, but I like it.
  
  
• And since I still have a few minutes, apparently, here’s some self-promotion: I was interviewed by the Financial Times about Beehive (now called “Social Blue”) and the podcast is available. On the Financial Times’ podcast page, it is the Oct 20th podcast.

The scientists looked at how many students in the longitudinal study named a given student as a friend, which it termed “in-degree” affinity; how many students a given student named as friends (out-degree affinity); what the odds were of a given student’s friends knowing each other (transitivity); and how central or peripheral to a network a given student might be (centrality).

The researchers found that in-degree, transitivity and centrality are “significantly heritable.” This means that your genetic background may help determine not only how many people count you as a friend, but also how many of your friends are friends among themselves. This sheds light on the kind of social network you inhabit, and whether your presence is central to it, or not.

There are related studies that find that levels of innovation, obesity, smoking and depression can be linked to who you are friends with. If we break apart the causal link between genetics and innovation/obesity/smoking/depression research outcomes, we might find that it isn’t your genes, but rather your genetically pre-determined set of friends that are influencing the course of your life. Better break the genetic determinism by stepping up that Facebook friending!

I maintain that because Twitter’s earliest adopters were adults, not teenagers or college kids, it is in a much better position to find a profitable business model compared to other social network services. Not only are Twitter’s core users in the business of web 2.0, they are using Twitter to support their work. It is only a matter of time before Twitter figures out how to make money off of that.

• Ross Mayfield says SocialText will be a better sell in this economy than IBM and Microsoft’s solutions because SocialText is cheaper. I don’t know the prices, but this sounds pretty logical.
• Facebook and Twitter don’t plan to become profitable any time soon. I think this is a good sign that the fun isn’t over yet.
• LinkedIn has released some “all work, no play” applications. Will be very interesting to see where this takes LinkedIn.

As you watch the debates, even though CNN has told you that these uncommitted Ohioans are not a statistically representative sample, you are influenced by their knob turning. Just like you can be influenced by your friends’ laughter and comments in a face-to-face group setting, as you watch those lines moving up and down, you are convinced that McCain is angry and Obama is presidential. There is lots of experimental evidence of this from previous elections (read the OTM transcript).

My PhD thesis focused on this exact issue: real-time visualizations of social behavior and how this information changes the dynamics and your perceptions of an interaction. For my project, Second Messenger, I focused on small, face-to-face groups receiving feedback about their group dynamics, but the phenomenon is universal and extends beyond social visualization. Behavioral economists refer to it as framing.

Besides being interesting to think about, it is important to consider the influence of the squiggly lines. First, CNN should be aware of its power to influence viewers. Second, viewers should be aware that their attitudes can be manipulated by this relatively simple and unrepresentative data point. They discussed this a bit on OTM:

BOB GARFIELD (interviewer): We have established that the squiggly lines do not represent any statistically significant data. And … we’ve established that people are influenced by the squiggly lines. Does that not mean that it is journalistically irresponsible to put them on the TV? I mean, is CNN doing the right thing in adding this feature to its coverage?

SAM SCHECHNER: We’re social animals. We process information in the context of the people around us. We do watch our debates with friends or at a bar or in some sort of group environment, frequently, in part because we want to hear the reactions of other people. That’s in part why … you and I can’t stop looking at that CNN line on the bottom of the screen.

I think it’s hard to argue, necessarily, that it’s a bad thing. In fact, the people who do it – the professors at SMU – see it as a way of shifting some of the balance of influence from pundits, who have been shown to be able to have a pretty strong effect with their pre- and post-debate spin and how they set expectations and frame the debates to actual voters.

I’m not sure I agree that uncommitted Ohio voters are the antidote to polarizing pundits, but something to think about.

More information on the report:

WSJ, 10/10/08: “When Your Political Opinion Isn’t Yours Alone” Broadcasts of Political Debates That Include Live Audience Feedback Can Influence What You’re Thinking

More on my research:

Joan Morris DiMicco, Kate J Hollenbach, Anna Pandolfo, Walter Bender. (2007) “The Impact of Increased Awareness while Face-to-Face.” Special Issue on Awareness Systems Design, Human-Computer Interaction. Volume 22 (2007), Number 1.

If you create the coolest widget, it can’t have impact unless people see it, use it, experience it. I guess that is basically Marketing 1.0. But something computer scientists aren’t always so savvy to.

Ok, time to go review CHI papers for publication. It would be so much more fun to be reviewing YouTube videos…

The focus of the center will be on a style of research we are calling “Venture Research.”

Venture Research can be distinguished from other types of technology research by two major features. First, a venture project involves a large-scale deployment of a research project “in the wild,” for example, on the web. Typically the reason for doing a deployment to a large population is to test hypotheses about interactions between thousands of people and to allow for appropriation of the technology for unanticipated purposes.

Second, at a practical level, recruiting thousands of participants for a project means that the venture project’s application must provide a tangible value to its users. As a result, they are typically long-lasting projects and users often expect that the services will be available consistently and indefinitely. This is where the practical funding issues arise. While you can build a Web 2.0 app relatively quickly, maintaining a community of thousands, over extended periods of time takes funding that is separate and on top of base funding that supported the initial research idea’s formulation. In this way, the projects require funding more akin to venture capital — high risk, but high reward.

The reason IBM is interested in doing this style of research is that we see this scale of user interaction more the norm than the exception. As IBM is in the business of building solutions for companies, the new Center’s intention is to lead the field towards understanding how companies will communicate, interact, collaborate, and exchange at these scales of interaction.

There’s one more free and easy way to improve the design of your site — using HTML programming code. Fortunately, you don’t need to have programming skills to use HTML. All you need to know is that a block of HTML — essentially, a bunch of gobbledygook words and symbols — can add extra features to your site.

I guess they didn’t have room for the 3 sentences it would take to explain what HTML is, but dismissing it as “gobbledygook” seems a little extreme. The section in the article about how to get your website found by search engines was actually kind of useful. No gobbledygook required.

The CNET article points out that “Facebook’s insistence on being a ‘neutral platform provider’ in the situation” is a little odd, “because, to state the matter bluntly, it isn’t.”

“[Facebook] has a history of tightly policing activity on its developer platform, banning “Secret Crush” over a spyware claim, locking down applications believed to be spamming users, and occasionally raising developer ire with some of its more stringent regulations. The site even temporarily blocked Top Friends, a creation of widget powerhouse Slide, when a security hole was discovered…

“Facebook wanted to keep Scrabulous around. In claiming a ‘neutral’ stance, the company was actually taking the activist route.”

So it makes me wonder, is it possible for a community site or social network site to remain neutral about its content? When you build a site, and especially when you are also studying the behavior of the users of the site, you do not want to be the police, monitoring and deleting content deemed inappropriate. But, as this situation on Facebook reveals, by doing nothing, you can’t claim neutrality because you are implicitly endorsing the content. It could be argued that it is only because Facebook removed some content that they can’t claim neutrality. But what if they did nothing about illegal content such as child pornography? Do they need to comply with laws? Copyright laws? If yes, then that means they have to be the police, a role no web2.0 community developer wants. It is a tough situation.