Email Plugs Into Social Networking
An anonymous reader writes "Microsoft Research recently released SNARF, the Social Network and Relationship Finder. It works in the Outlook email client to prioritize and sort emails based on the relationship to the sender and other characteristics of incoming email messages. Trusted Reviews wonders if 2006 is the year of ordering information and reports on ClearContext, which does similar prioritization of emails as well as some email driven task management."
This was Google's plan with Gmail and Orkut. However, Orkut never seems to have really gotten off the ground in the way they'd hoped.
The theory of relativity doesn't work right in Arkansas.
(Insert emo Thundercat joke here)
My text based communications have moved more to SMS and IM than e-mail, especially in the last 6 months or so. I've even seen many of my non-geek friends and family moving to SMS and IM over e-mail, there is definitely a much high signal-to-noise ratio over the spam cluttered e-mail Inboxes that many people have.
I moved from my own server (which we ran for almost 9 years) to gmail recently, and couldn't be happier -- I wouldn't doubt that my tiny company is saving thousands per year of maintenance and upgrades, and having our own domain name isn't a big deal anymore. It also offers transportability if one of my employees moves on or if we bring someone on for a contract gig.
The downside to e-mail is still the signal-to-noise ratio. Spam filters are helping, but lately gmail has been losing the battle (but hey, my gmail address is publicly listed on slashdot and other forums, so I can't complain). I also have to wade through what is important right now and what isn't, and marking people with a star hasn't helped much.
I don't know if I trust Microsoft to design and build the necessary algorithms and heuristics to sort e-mails in an effective way. This is the same company that has one of the worst letter writing analyzers in word, and we all remember Clippy, who probably still exists. Sure, Microsoft has an intense amount of data they can sort from Hotmail and MSN accounts, but I'm not sure if it will be enough. E-mails, in my opinion, are incredibly unfriendly for PCs to analyze -- it's like the game Go. Humans can wade through e-mails in microseconds, but AI programs have never shown me to be intelligent enough to get mistakes to number close to zero. Microsoft's other problem is I wonder how many people still use Outlook for the desktop? Most of my Exchange customers -- nearly all of them -- use Outlook Web Access. I doubt you can install a SNARF MSI somewhere in the chain to support OWA, right?
Google might have a step up against Microsoft (especially now with AOL and gmail), but even their server AI isn't ready for primetime.
From what I can tell, though, the person who makes the best e-mail sorting AI will definitely come out on top and they could also save e-mail as the prime communication method. I prefer SMS and IM for the instantaneous communication, high signal-to-noise ratio and ability to truly limit who contacts me. Maybe the solution is some odd combo of IM, SMS and e-mail?
I been using SNARF for a few weeks and have gotten some value out of it.
If you get an extreme amount of email like I do, its a great way to get up to speed on things. You can prioritize them based on who is CC'd or see a nice graphical thread view that makes it easier to figure out what is going in.
Its definately not something that's fully baked yet, but SNARF is a very interesting tool with alot of potential.
Conformity is the jailer of freedom and enemy of growth. -JFK
First of all, it isn't even an Outlook plugin, it runs as a separate app that starts when Windows does [just what Windows needs another of].
Secondly, it is completely empty of useful features, has almost no real ability to customize based on user preferences, and the interface is bad even by Microsoft's standards [for lack of a better term.]
If this is Microsoft's idea of innovation, I can see why they usually just find it easier to buy other people's technologies and then "extend" them.