Thursday, February 28, 2008

Premal Shah's Kiva talk at PARC stresses the importance of Social Transparency

As mentioned previously, Premal Shah of Kiva.org recently spoke at the PARC's special speaker series on "Going Beyond Web2.0". During the talk, Premal gave many great stories of how Kiva got started and the growth of the system. It was a great talk, because it showed that Web2.0 can be used as a way to connect people directly and to help them to get out of poverty. The way Kiva.org does this is through microloans.

The key, according to Premal, is transparency. At the talk, he said, "The reason why people dig Kiva is because of the transparency." This coincides with our "social transparency" principle that went into the design of WikiDashboard---we make the data visible and easier to understand, and you decide on how to use the information. Premal used the same principle here in the design of Kiva. For example, he said that they expose on the website the risk rating of various microloan organizations, and it is up to the users to decide on how much risk they want to take with their loans. The idea here is that a "social investor" in microloans need all of the information and make the decision locally in a distributed way. If enough people votes with their $25 loans, then a higher-risk loan can still get funded. Indeed, "people funded $25 at a time can actually beat Citibank's $100M microfinance total fund."

Incidentally, Katie Payne, who is giving the PARC forum on how to measure social media effects, is just now talking about the importance of "transparency" in building credibility and trust.

In any case, here is the Premal's Kiva talk:

Wednesday, February 20, 2008

Can Technology Solve the Problem of Wikipidiots?

The ASC Research group at PARC has been working on a project called Wikidashboard (led by Bongwon Suh and Ed Chi) that is eventually aimed at helping users assess the credibility of articles they read in Wikipedia.

Those of us living in San Francisco were recently treated to the front-page headline "Wikipidiots" emblazoned on the freebie "SF Weekly" all over town. The Feb 13th article by Mary Spicuzza partly concerns the heightened "internet rage" that seems to afflict San Franciscans, but mostly focuses on "edit wars" on Wikipedia, and the difficulty of establishing the credentials, credibility, reputation, etc. of people hiding behind on-line personas.

Spicuzza reports on her attempts to track down and interview a Wikipedia user who seems to have gotten into quite a few Wiki-spats who goes by the handle "Griot". Who is this guy/gal? What makes them tick? Where do they live? Why do they write so much about the Greens? Much of the information Griot reveals about him/her-self is unverifiable, and purposely obfuscating. But Spicuzza does end up constructing a caricature of Griot--the online persona--based on digging through Wikipedia, the discussions, edit histories, and so on.

Most people aren't investigative reporters with the drive and time to do that kind of digging every time they read a possibly controversial wiki page. Wikidashboard is a tool/visualization that embeds itself into Wikipedia pages and offers a way of seeing the authors and their edit histories, and allowing drill-down on such information in a much more usable way than the standard Wikipedia interface.

As stated on the Wikidashboard page:

The idea is that if we provide social transparency and enable attribution of work to individual workers in Wikipedia, then this will eventually result in increased credibility and trust in the page content, and therefore higher levels of trust in Wikipedia.
You can check Wikidashboard out here.

Tuesday, February 12, 2008

Wikipedia users: We’d like to talk with you!

We're conducting ethnographic interviews on Wikipedia use, to help us create better tools for both readers and editors. Share your experiences and give us your opinions! The interview takes about an hour, can be done remotely or at PARC, and we can schedule at your convenience. We'll even give you an Amazon gift certificate as a token of our appreciation. Please contact dschiano@parc.com.
Thanks!

Wednesday, February 6, 2008

Kiva.org to speak at PARC's Web2.0 Speaker Series

Premal Shah, the President of Kiva.org, is speaking at PARC tomorrow, and I'm really excited. Kiva.org is a non-profit organization that provides microloans to individuals in the developing world. Over the last few months, as reported by the NYTimes Magazine, Kiva.org has had so much demand, that some visitors to the website was greeted with the message, "Thanks Kiva Lenders! You've funded EVERY business on the site!!". When was the last time you heard from a charity that it had enough money to do everything it wanted to do? That's the amazing popularity of Kiva.

After seeing a PBS Frontline special late at night in my hotel room while I was traveling out of town in Nov 2006, I immediately opened my web browser and joined other people in discovering the joy of being a microlender. The experience has been amazing. I've made 8 loans so far and tell everyone about Kiva whenever I can.



What's amazing about Kiva is that it uses Web2.0 application design principles to connect lenders to borrowers.

First, it builds a social network around a microloan, so you can see everyone who has also loaned to the same person. While I personally have not really experienced a lot of communication between lenders so far due to my busy schedule, the feel of the community is real. People build their lender profile pages, and some even appears to compete to see who can make more loans.

Second, it exploits the long tail of participation to reach people at all economic levels as potential lenders. If you are willing to part with $25 for a while, you can be a microlender too, and the risk is just a 0.14% default rate! It's hard to convince someone to part with large amounts of money, but it isn't hard to convince almost anyone to loan out $25 that have a good chance of being paid back (and you get to help someone in the mean time!) The concepts of microloans is pioneered by Bangladeshi economist Muhammad Yunus, who won a Nobel Peace Prize for his efforts.

Third, it builds and thrives on end-user participation. On the fansite, KivaFriends.org, user-generated content show screencasts of how to make a loan step-by-step, people post about their experiences, and organize fundraisers and sell calendars.

I'm really looking forward to the talk tomorrow, and will post the video of the talk here as soon as I can.

Saturday, February 2, 2008

Ross Mayfield's talk at PARC available as video stream

The PARC Forum Speaker Series called 'Going Beyond Web2.0' that ASC organized will continue to release video recordings of the talks at:
http://www.parc.com/events/forum/archive.php.
The first talk by Ross Mayfield was an entertaining talk about how Web2.0 is changing the way people think about enterprise software:
Social Software is made of people, and it is often about how the software needs to get of the way. Ross makes the point that much of Web2.0 is about bottom-up processes, and about the augmentation of the groups rather than automation of workflows. Indeed, he says that the average knowledge worker isn't to spend time to perform workflows, but actually dealing with exceptions in the workflow. More importantly, workers add their value by dealing with these exceptions.

View Ross' talk here.

Ross' talk plucked some patterns out of the current movement in Enterprise2.0 software. He said that one way to look at the current Web2.0 and Enterprise2.0 movement is to notice that these systems are all made of people, and it is important for the software to enable users to connect with each other and just get out of the way. The point here is to augment the people, not to automate them.

Indeed, one of the intriguing point he made is that "the average knowledge worker doesn't spend time to perform workflows, but actually to deal with exceptions." Indeed, much of new communication software and infrastructure enable a kind of emerging culture to occur. For example, when email was introduced to enterprises, it enabled a new kind of private small group communication and gradually developed its own culture---the appropriateness of topics and the amount of formality and pitfalls.

Another pattern he noted was that there is "abundant desire to share information", and that "Social goods are created when the means of production and/or distribution is broadly available". Ross mentioned the example of Craig's List, which was a community that was build bottom-up that eventually became a disruptive force in classified ad market. What's interesting about Craig's List is that it did this by sharing the control of what to publish with the end-user.

An interesting pattern here is that "to get the benefits of sharing control and being open, you have to move towards transparency". This obviously connects with our research on WikiDashboard, which enables a kind of social transparency in Wikis.

If you want to watch Ross' talk, I recently also uploaded Ross' talk to Google video:

Organizational learning and Knowledge Building

One of the things we have been doing at PARC in order to understanding this new field that we're calling Augmented Social Cognition is to go back to past research on understanding organizational memory and learning, and past systems that have attempted at collaborative knowledge building. Even though much of this research has mostly focused on group collaboration (typically on the order of a work or task group or perhaps a small organization), I believe the research points to directions for future Web2.0 research.

One of these papers we read in our ASC Reading Group recently is Mark Ackerman's work on AnswerGarden, which I consider to be pioneering work in this area. The research developed a tool that enables a database of frequently asked questions (FAQs) to grow "organically" over time. As new questions arise and are answered, they are inserted into an ontology that organizes past question and answer pair. A user interacts with the system by going through a branching network of diagnostic network of questions that might help them find the answers. If the question is not in the database, the question is automatically send to an expert, and the answer is both send back to the asker as well as inserted into this diagnostic network.

The research contained an outline of how to build such systems, which bears quite a resemblance to how people use Yahoo! Answers as well as ways people use Google/Wikipedia together, as well as the recent Google announced effort called Knol. Moreover, the research contained field studies of the system used in practice to answer questions about how to use the X11 Windowing System. This is all very cool research indeed, and I encourage more people to find out more about this work.

Ackerman, M. S. and Malone, T. W. 1990. Answer Garden: a tool for growing organizational memory. In Proceedings of the ACM SIGOIS and IEEE CS TC-OA Conference on office information Systems (Cambridge, Massachusetts, United States, April 25 - 27, 1990). F. H. Lochovsky and R. B. Allen, Eds. ACM, New York, NY, 31-39.
DOI= http://doi.acm.org/10.1145/91474.91485


Ackerman, M. S. 1998. Augmenting organizational memory: a field study of answer garden. ACM Trans. Inf. Syst. 16, 3 (Jul. 1998), 203-224. DOI= http://doi.acm.org/10.1145/290159.290160