BAAGZ (www.baagz.com), from the French Web and enterprise search company Exalead (http://corporate.exalead.com), is a new search and collaboration system, still in beta, that applies Web 2.0 features to social networking. While the idea of using search topics and vetted results as nodes around which dynamic topical networks could form is interesting, the current early beta implementation gives just a hint of what may be possible.
Once added and indexed, the contents of the baagz are easy to search—searches can be limited to one’s personal baagz or to all the baagz on the system. In this way, members can find valuable links and experts in a field. It’s a live network of shared interests, using advanced evaluation technology developed for Exalead’s Semantic Search to notify members of related content or experts on their shared topics. These algorithms allow the contents of the baag to describe it in ways that the members could never express. There is tight integration with Exalead’s Web, Image, Video, and Wikipedia searches, so it’s extremely easy to add items from these results to the appropriate baag.
By connecting members through a multitude of complex shared interests, rather than schools or workplaces or even user-created tags, BAAGZ can bring together search and networking. As the system grows, it will become richer by the nature of the "network effect" (coined by Robert Metcalf, the inventor of Ethernet), which means that the benefits grow along with the number of other members.
The first beta of BAAGZ runs only within the browser Firefox version 2, although compatibility with the browsers Internet Explorer 6+ and Safari is coming soon. It’s not a public beta; entry is by request, so the system can work out any problems and scale at a reasonable rate. There’s a link to leave your email address at www.baagz.com/beta.html.
As innovative new services like BAAGZ start up, there is one major stumbling block—the people most likely to join have probably already joined at least one other social networking service. They have located their classmates, colleagues and friends, and filled out their profiles, and perhaps even joined some of the groups there. But the world of a single social network is considered a "walled garden," like AOL before its members could get to the World Wide Web. If people have a friend or wish to follow a blog or photo postings on another service, they must create an account on that service and remember to check it (and remember the password). The other problem is that if members wish to leave a social networking service, because of conflict or expense, they must re-create their accounts on a new service and convince their friends to move. This means that new services have a hard time getting enough members to enable the network effect, because so many people have accounts on other services.
OpenID, the brainchild of LiveJournal founder Brad Fitzpatrick, offers a way around this problem. Instead of an identity as a member of a service, it would be a single user name and password, which could be used on any social networking system or Web site requiring login—an Internet-wide secure identity. This works somewhat like the Internet’s DNS (Domain Name Service), which matches names to IP addresses, making it possible to change Web site hosting without breaking every link into the site. Because OpenID is decentralized and open source, it’s not a single password database, vulnerable to hacking, corporate misuse, or legal discovery. Each person will make a choice of OpenID server, and should take security and policies and can move from one server to another by choice.
Services that support OpenID now include LiveJournal (and its open source clones GreatestJournal, InsaneJournal, and DeadJournal), Vox, Microsoft, Symantec, AOL, Digg, WordPress, TypeKey, and more every day. Services accepting OpenID use it as an alternate form of login; instead of looking up the user name and password in their own databases, they send that data to the OpenID server indicated by the login and wait for an answer. Once the OpenID account is confirmed, that account is treated like a native account, including group memberships and any access to private content granted by other users. Orange (France Telecom) just announced that it is enabling OpenID for its users based on their handsets and other hardware.
OpenID Hosting services, such as myOpenID.com and VeriSign’s PIP (http://pip.verisignlabs.com), offer free secure OpenID accounts (others are at http://openid.net/wiki/index.php/OpenIDServers). Many of the OpenID services, such as LiveJournal, Vox, AOL, Technorati, and WordPress include OpenID services as part of their normal accounts—members of LiveJournal can use Technorati with their LiveJournal account and vice versa. For those even more concerned about security, there are easy instructions for adding an OpenID server to any Web site that can run PHP (www.intertwingly.net/blog/2007/01/03/OpenID-for-non-SuperUsers).
This leads into a proposed Bill of Rights for Users of the Social Web (http://opensocialweb.org/2007/09/05/bill-of-rights). This asserts that users should retain rights to ownership of their personal data, control over who can see this data, and freedom to share this data. As I read it, this does not preclude advertising, but it does mean that the minority of users who view the content out of the service might not see the ads. It’s more permission-based marketing than blasting ads at everyone. Services should allow users to share their data, friends, and friends’ data, link from outside the service, and look up friends from other services.
With support for OpenID and these rights for users, new services such as BAAGZ could get off the ground, attracting users who will try them out if the psychological cost of entry is low and the services concentrate on providing innovative features and special services. A service will only work if there are enough participants to provide both synergy and community.