Surfarama

Via Waxy, Similicio.us helps the discovery of related content by checking del.icio.us tags and seeing what other content (represented by a URL) has been similarly tagged. Apparently it also uses the EasyUtil engine, although it isn’t clear to me how that fits into this…I guess it uses del.icio.us tags as a proxy for similarity, and then uses EasyUtil to refine this somehow. I need to grok EasyUtil properly to answer this question.

In any event, I found it interesting because it seeks to achieve the same goal as Tagalyzer but uses a different method…

While Similicio.us checks del.icio.us to see what tags have been used to categorize an article, and then returns other articles similarly tagged, Tagalyzer uses Yahoo’s content analysis API to identify keywords, and then various APIs to find related content for those keywords (including the del.icio.us API).

The question this begs is the difference between keywords and tags. Are they two different things? ‘Keywords’ are words from the article which the Yahoo! content analysis considers to be important from a search perspective. Tags on the other hand may or may not be words from the article itself, but they do represent the article content or themes as perceived by the individual. I’d expect considerable overlap, but there is a subtle difference and it most certainly has an impact on relevance and usefulness.

Thinking about it in these terms Tagalyzer may have been poorly named, conflating ‘keywords’ and ‘tags’. But then again who really cares. I’d just be keen for some feedback about Tagalyzer. Useful or a load of bollocks?

By way of a demo, here’s the tagalyzer results for the Jesse James Garrett article which Similicio.us also uses as a demo.

Classic, Ted’s been farked. Love it.

Cool man, we’ve just upgraded Delivr to use the Flickr interestingness API for all recent images. We’re still filtering for Creative Commons of course, but the interestingness thing returns a way higher class of photo…these are super cool, and make great digital postcards.

The image search still covers all CC licensed images (so there’s great breadth), and you can still access your own images…although I’ve discovered a quirk which I think stems from the point that Flickr was rolled into Yahoo! and they started using alphanumeric strings to indentify user pages. Luckily this doesn’t break Delivr, it just means I need to add some instruction to make it clear, which I will get to eventually.

So, there you go. Good stuff…as you were.

Nice review of publi.sh…it’s cool that someone gets the utility of it. They also got a nice screen grab of the posting interface.

Vast is pretty killer aswell. Havn’t yet had time to fully grok this, but they’ve apparently got an API which allows for unlimited commercial use so it will be interesting to see what that leads to…tempted to play I am.

Gabbly is way cool, and ought to be bought by Google (yes, that’s a prediction), or maybe Yahoo!

I’ve often thought about something like this, but always imagined users would simply link to the site and the referring url would be used to drop the visitor into the chat…but how these guys have implemented is way more practical. Relying on referring urls is problematic and being able to link to gabbly.com/http://surfarama.com means we can easily incorporate gabbly into blog templates and the like to create individual chat rooms for each post…love it.

Boing Boing should do this, but even if they don’t readers can easily do it themselves. I also imagined it would be super handy for conference attendees…just drop the URL for the conference page into gabbly and voila! an instant back channel. The possibilities are endless.

UPDATE: Yep, Gabbly is gonna be huge. The more I play with it the more I like it. Each of my posts now has a Gabbly chat link…so all three of my readers (Mum, that includes you and me!) can get together and chat about anything I write.

Of course I don’t ever expect that to actually happen, I just love that it could. Like I said though this will be killer for popular sites like Boing Boing, or Digg, or Flickr even. Or for communities surrounding conferences, or events with a web presence. Let’s just hope they can scale (which is where Google or Yahoo! could come in handy).

I used to be in the petrochemicals business and we used telemetry alot to manage supply and demand. In that context it was all about visibility into our customers business so that we could be proactive, and it worked well, and customers loved it.

So when I first came across The Trixie Update I was pretty impressed. It is a cool effort to track data related to the sleep, eat, poop patterns of a new born baby. That, and I was impressed that new parents would have the discipline to keep this up. The data is facinating, it’s baby telemetry.

So anyway, I was interested to note (via Matt Haughey) that the guy behind The Trixie Update, has released the The Trixie Tracker: Baby Tracking Software as a service.

It might sound kinda odd, but I am willing to bet that there’s enough nerdy parent types who’d be into this to make it succesful. And imagine if it did turn out to be popular, what an amazing data set they’d have, research as a service. IF we have another baby I’d signup for sure.