Please share with me so that I can beat you

[Ed note: The Connectedness staff is behind in blogging, due to an overload of Web Science synthesis. We apologize for any inconvenience this may cause.]

Martin Nowak is one of the world's pre-eminent gurus of evolutionary dynamics (which I called "die and adapt" a few months ago). Yesterday at the Harvard Kennedy School Complexity Series, Nowak brilliantly outlined five reasons why I might willingly die so that you can live, and and what adaptations occur in our population as a result:

1. Kin selection: you are genetically similar enough to me (e.g., child, sibling) that I could rationally decide to sacrifice my own body for the sake of helping your/my DNA
2. Direct reciprocity: you and I interact repeatedly, and I expect that over time you will repay the favors I give you now
3. Indirect reciprocity: I may never see you again, but I care about my reputation (and know about yours), and I gain enough in reputation to make it worth doing you a favor.

The above three mechanisms of cooperations were rounded out by two more: graph selection and group selection, which allow one to model (1) a world where people don't randomly mix with the entire world but instead have specific neighborhoods of interaction, and (2) a world where tribes develop collective strategies. This last one (group selection) is still somewhat controversial, and Nowak himself has flip-flopped over the course of his career with respect to it. Currently he believes it is an important dimension of a complete model of the evolution of cooperation.

The talk and audience were theoretical-minded, but the person sitting on my right shared a very pragmatic reason for attending:

"The successful push-pull of collaboration and competition ... [in] both open source programming and wikis and is certain to find its way into many enterprises as collaborative design becomes commonplace."

More about that in his paper, "In Praise of Tweaking," where he makes clear one specific enterprise that is deliberately investing in theoretical evolutionary dynamics in order to redesign its own process of production. Thank you, Ned Gulley.

This work is licensed under a Creative Commons Attribution-ShareAlike 3.0 License and is copyrighted (c) 2007 by Connective Associates LLC except where otherwise noted.

Economic externalities and network science

Thanks to Six Degrees: The Science of a Connected Age, which I use as the textbook for my course on web science, I have learned about economic externalities. I don't always include the topic of economic externalities in the class, but last fall I did--with intriguing results. It was all part of the major course redesign that preoccupied me more or less 24/7 from Columbus Day until Kwanza.

Here are the four types of economic externalities according to Duncan Watts, with explanations and examples by yours truly:

1. Information externalities: Knowing how others have acted under similar circumstances saves me the effort of evaluating all the options "objectively." Example: I am hungry. McDonalds has sold 30 billion Big Macs. They must be OK.
2. Coercive externalities: Anticipating the impact of my decision on others influences my choice. Example: Everyone is drinking at this party. What will they think of me if I don't drink?
3. Market externalities: As a particular option is chosen by more and more people, that option becomes more and more valuable to all those who have chosen it. Example: In 1980 very few people had email and so email was of very limited use. In 2007 many people have email and that popularity makes email exponentially more useful.
4. Coordination externalities: I will sacrifice my short-term selfish interests for long-term gains that depend on favors from others, to the extent that (1) I care about the future, and (2) I believe my actions affect the decisions of others. Example: When my friend lends me $10, I will pay him back the next time I see him. I lose $10 when I pay him back but gain more than that in the long run.

Watts discusses the above concepts almost exclusively in terms of their relationship to "tipping point" network dynamics. He also emphasizes the top of the list at the expense of the bottom. Having swallowed Watts' taxonomy of economic externalities into the same pipeline that is slowly digesting Tim Berners-Lee's framework for web science, I now have (1) a chronic case of indigestion, and (2) the following mapping of economic externalities onto basic concepts of network science:

1. Information externalities --> cardinality and centrality
2. Coercive externalities --> connectivity and clustering
   
3. Market externalities --> structural equivalence
   
4. Coordination externalities --> symmetry and asymmetry
   

As I prepare for spring '08, the above appears to be the backbone of my web science curriculum. If it doesn't make sense, dear reader, please be patient with me and stay tuned!

This work is licensed under a Creative Commons Attribution-ShareAlike 3.0 License and is copyrighted (c) 2007 by Connective Associates LLC except where otherwise noted.

Performance reviews

Now that grading is done, I get to see my fall student evaluations. I skimmed a handful earlier today and spotted a few "five-star" marks and a couple stinging criticisms. Time to pause now. The forms are resting outside in my car, so that tomorrow I can bring them inside and absorb them with an appropriately cool frame of mind.

The very same day my students wrote their anonymous evaluations, they handed in their final project reports. Each student had the entire semester to build any kind of website she wanted (e.g., Japanese, business, entertainment, sports, travel, cars). Each final project report told the story behind the website, concluding with the most valued lessons the student learned over the course of the project. One of my favorite endings was this:

I had, before this class, become obsessed with learning code "hardcore", but I see that I'll pretty much always be able to find useful code snippets in resources online.... For coding a whole website, it's not so much about a limitless knowledge of code, but about persistence. With trial and error, focus, solid consideration, and just a little work, you can usually make what you envision a reality (or something that works just as well as what you'd envisioned).

Cool!

The flip side of this lesson came out in the anonymous evaluations. (Written the same day.) One student said it simply: "Everything I learned I had to teach myself." Another said, "Sometimes it seemed like there were students who knew more than the instructor." In the context of my consulting work, those two quotes could very easily be compliments, but my students wrote them in response to the question, "What were the most significant weaknesses of this course?"

My impulse in these moments of paradox is WWLTD -- "What Would Lao Tsu Do?" I am not sure that Lao Tsu's paycheck depended on anonymous student evaluations, though.

This work is licensed under a Creative Commons Attribution-ShareAlike 3.0 License and is copyrighted (c) 2007 by Connective Associates LLC except where otherwise noted.

Delete all your links, except to me

As part of the course re-design I mentioned last time, I have been digesting this excellent summary of Google PageRank that my students found for me a few weeks ago. I have boiled it down to this. Another way I teach my students PageRank is by entering all their semester web-programming projects into a PageRank contest, where the PageRank is determined by student recommendations entered into our class wiki.

The contest has one more week to go. You can see the front-runners vying for votes (or not) in the wiki excerpt below. They are listed in order of PageRank:

Site Link / Description	Incoming Links
http://people.bu.edu/gatewood/index.html Quit editing this wiki, get out and do something! My site is a December calendar of events, from concerts to festivals, that would interest a college age student in Boston.	choy alliew rgm ajhlee jihong88 cohens11 xzhang08 dbrien lbdunn
http://people.bu.edu/afacini/gaijin/ A resource site for learning how to read and write basic Japanese. Learn to write both hiragana and katakana, the basic alphabet scripts of Japanese.	jasonhu gaep13 wcmundel icc arthim edflem yasmin89 kevshea slee0903 dan013 abg cnm lbdunn
http://www.tormor.org The homepage of me and a friend of mine and our upcoming vegetable oil fueled cross-continental adventure.	jasonhu alliew tjpirog billba edflem yuliyab cohens11 koates andrak lbdunn
http://people.bu.edu/xzhang08/psych/index.html Enjoy looking at optical illusions? This website has lots of optical illusions and explains how each works!	gaep13 gillianb arthim yasmin89 timwalsh ajhlee mgro27 slee0903 dbrien jfsr lbdunn
http://people.bu.edu/anishk/Kattukaran-Webproject/Index.html (Firefox recommended) A website for a startup company that i am working on. The new venture will deal with educational services targetted at students from developing countries and aims to improve the standards of education in many of these countries.	aecook gaep13 icc haysher edflem dechat eeun zucchig bross55 lbdunn eskizzle
http://people.bu.edu/ajhlee/cs103/index.html This page is about how hip hop is dead	jihong88 cohens11 gatewood eeun jfsr edflem lbdunn

My favorite feature of PageRank is this strategy suggested by the PageRank article: "Outbound links are a drain on a site's total PageRank... But there are 'abnormal' ways of linking to other sites that don't result in leaks. PageRank is leaked when Google recognizes a link to another site. The answer is to use links that Google doesn't recognize or count."

This is true enough. To dramatize the point, I have provided my class with an Excel spreadsheet that takes the above wiki table as a copy-and-paste and instantaneously provides the resulting PageRank scores and relative class rankings. Each student can edit the resulting spreadsheet and see exactly how much his score goes up when he deletes all his outbound links and stops recommending other students' websites.

Hopefully, by the time I do the final PageRank calculation, all my students will have taken this lesson to heart and deleted all their links. You should too. Boost your visibility on the Internet by deleting all outbound links from all your blogs and other sites--except links to me. Thanks!

This work is licensed under a Creative Commons Attribution-ShareAlike 3.0 License and is copyrighted (c) 2007 by Connective Associates LLC except where otherwise noted.

Experiencing technical difficulty--please stand by

Major course redesign happening in front of a live student audience has Connectedness staff preoccupied. Regular blogging will resume shortly.

This work is licensed under a Creative Commons Attribution-ShareAlike 3.0 License and is copyrighted (c) 2007 by Connective Associates LLC except where otherwise noted.

Use Visone to make your first network map

I often get asked about network mapping software. I switch back and forth between several programs depending on what I want to do. For most, I usually recommend NetDraw. However, for those wishing to make very simple maps, another excellent choice is Visone.

You can easily run Visone with the "webstart" option available here. Once you have it running, the picture below hints at how easy it is to (1) give yourself a starting set of nodes by creating a "random" graph with no edges, and (2) use "edit mode" to make your nodes look as you wish and add edges between them. Then you can use "analysis mode" to drag nodes around and try different automatic layouts, etc.I use Visone to make all the illustrations in my Introduction to Network Mathematics, but I don't use it for consulting because it is licensed only for non-commercial use.

This work is licensed under a Creative Commons Attribution-ShareAlike 2.5 License and is copyrighted (c) 2007 by Connective Associates except where otherwise noted.

Information advantage

I attended a talk recently about computational sociology and data mining. The speaker began with a claim that technology is never policy-agnostic but almost always advocates for some policy or other.

Half-way through the talk, someone referred to the impressive array of technology employed by the speaker's research and asked what policy that technology was advocating. The speaker deftly avoided the question by raising policy questions without answering any of them. He was policy-agnostic, you might say.

In such situations (and many others), it is a safe bet that the policy being advocated by the technologist is "I deserve your respect, money, and/or votes."

The best case I have seen for this argument was put forth by Robert Thomas in his book, What Machines Can't Do, which I originally mentioned here with respect to user-driven innovation.

I agree with Thomas, and certainly hope that my blog wins me your respect, money, and/or votes. Let the world know how much you admire my wisdom and power:

This work is licensed under a Creative Commons Attribution-ShareAlike 2.5 License and is copyrighted (c) 2007 by Connective Associates except where otherwise noted.