At least, we would like to have a small community of students and professors to use this service. Then gathering a group around this site, that is large enough to matter, may prove challenging. But le’ts assume that we’re able to attract them here (we’re subscribed to massive UM campus email lists after all ) would we be able to retain them? make them spent time, battery life and clicks.. using the service so we can do some research?

We’d recall Mark Zuckerberg approach.  Zuckerberg, founder of Facebook, was invited in 2007 to the World Economic Forum International Media Council meeting in Davos, Switzerland where world leaders were discussing the future of Iraq, to share his view on how to leverage on creating a community that could help bring peace to the country.

Zuckerberg told a story about his sophomore year at Harvard, while starting his company, he failed to study at all for his Art course; he didn’t even go to class. So days before the final, already in panic, he pulled all the pictures he needed to analyze off the web and put them up on a page online with boxes underneath. He emailed the class and said he’d put up a study guide. Sure enough, in moments, the students filled in their essential knowledge on the art. Zuckerberg got an A. And the professor told him that the grades in the class improved 10 percent over previous years.

It’s a magnificent lesson in everybody winning with cooperation. (Complete story from Jeff Jarvis, available here)

But in essence, his advice goes like this: you can’t create communities, they already exist and they are doing whatever they want, if you are lucky they would let you help them organize themselves.

Maybe we can borrow from him on this one?  Instead of deciding ourself what kind of features the service should have and what would users do with it.. may we instead listen to their desires and observe their patterns so we can accommodate their needs?

Indeed, there may be surprises.. and is hard to predict all the interaction patters and ideas that may appear as a result. But it certainly would have an impact on campus life and more importantly it will be public! Google will index it all and links from this eventual explosion of content would provide more visibility to all the academic and social life in the university.  Student may discuss their satisfaction with this or that other course, which in turn leave professors with valuable feedback to tailor their lectures. It may even touch government policies in the long run. Who knows how far would it go? All in all, instead of working for a mere class project this work could be turned into a useful service to the community.

But back to our goals, as far as building this site is concern what we are basically proposing is to use the wisdom of the crowds, as recenty explained by James Surowiecki in his 2004 book, to aid in the process and gain necessary momentum. Then, indeed, have lots of fun in the process.

I just stumbled on this interesting video. It shows a spontaneous and unsuspected avalanche effect in the crowd, to start dancing during a Sasquath Music Festival, earlier this year.

The video is increasingly getting attention among complex system researchers, which find similitude with other avalanche processes in nature, e.g. physical and biological systems.

Notice that Guy#3 is clearly responsible for triggering an avalanche of the crowd, presumably after a descending threshold of inhibition in the public that is produced by him dancing freely.

We no longer have Ali Rafiei in the team for this project, however, i wonder is there is a way to potentially produce a parallel effect toward contributions from all of us in this attempt.

Wouldn’t it be fantastic?

It should be remembered that finding information online today is getting easier than ever.. thanks to search engines that index and sort global information, by crawling the web 24/7. There is lots of discussion/knowledge about how to optimize site’s content for good/relevant search rankings (i.e. see SEO). Location is just one more aspect that a ranking algorithm would consider to provide a search result. However we explore location information from a different perspective.

Despite the recent advances in mobile platforms to access the web, there is a remarkable trade-off to be considered: the lighter, smaller and more portable the device the more constrained the user interface is –all the way from Laptops to net-books to smart-phones.

One of the purposes of this project is to explore how location assisted browsing would impact user navigation experience.  Thus, in this context, we explore how the granularity of the location information provided would help ease the user interaction by automatically contextualizing browsing activities.

As previously discussed, in our experiments the location information is conveniently provided via the Wi-Fi campus infrastructure and a location broker provided by the DSI Research Group.

When a user accesses this website from within the campus,  its device’s network card is associated with an AP in the network. The location of this AP is known by the location broker service available at a given local url. The basic idea is then use that information to aid in browsing the site content.

The simple approach we propose is to have the user optionally point its browser to that known local url and let the broker redirect the petition to a specific link within this site. Such link would be a category or virtual repository associated with the location in question. The effective location granularity is to be determined. It could be the school of engineering, the bar or the “cantine” or maybe a building block.

As we will propose for many other issues in this project, the easier way to determine the appropriate granularity is to let the users them-self find out what is relevant and convenient. This requires to let the use patterns evolve so we can identify relevant location distinctions and thus group AP in relevant clusters, each of which will be associated with a corresponding section in the site.

A very similar approach is to be implemented to equally aid in choosing a posting category or section. It would be chosen automatically based on the url path used to reach the posting form, e.g. which section did the user came from to post? –which is ultimately linked to location.

More technical details of the location broker side implementation to be discussed later.

For convenience, I’m proposing to discuss the development process and ideas about the project inhere. I’ve created this Development category just for that purpose.

The original Mobile Computing project description is available here (pdf).

The project is based on four tasks, associated to phases, as described in the file above. We now focus on task 4: “Development and Deployment” of the project application. But we’ll opportunely backtrack a little and discuss our view on the project, namely: design issues and on-the-fly feature implementation as well as scenarios for real uses, most relevantly to the information access, publishing and annotation models from task 3.

The main design shift from previous proposals is to use WordPress to support publishing and content management for the LiveCampus project. It will further integrate with the Campus Wi-Fi based location broker service to provide location specific browsing, but that will require a separate discussion, to be addressed shortly.

WordPress offers enormous flexibility and provides elegant organization to both users and service administrators. Based on a large repository of plugins and themes, all contributed by a large community of bloggers and developers. This is particularly relevant as it permits to rapidly accommodate the site design and features to meet user needs, add further services and integrate more online tools.  All to be further discussed soonish.

A project description will be provided shortly. Keep tuned