Data Mining: Text Mining, Visualization and Social Media

Riff vs RSS/Tags/Microformats - putting it together

Jeff Jarvis agrees with Crunchnotes:

Do we need Riffs when everyone seems very happy writing reviews directly on their blogs?

He then continues:

The service I’ll pay attention to is the one that lets me find the riffs and reviews (and recipes and whatever else) that people put on their own blogs. That can be a search engine or an aggregator or both that gets people to swarm around tags so they know their stuff will be found. It works inside Flickr and Del.icio.us. It can work outside, in the distributed web.

In a related post, he talks about Google Base:

I wish I were hearing more noise from the microformats guys to act as competitors — or at least as pressure on Google for openness and standards.

Imagine if you could go to a page that lets you put in your resume or house ad or job ad and it spits out tagged XML you could put on the web anywhere to be found by anyone.

Or imagine putting tags on restaurant reviews you post on your blog so anyone could aggregate or search for, say, all the cuisine=mexican restaurants in location=chicago. Well, you don’t really have to imagine that. If you aggreed on the tags, you could start doing that today via Del.icio.us and Technorati.

And imagine if you could go to Google or other services — e.g., Indeed and SimplyHired for jobs or Baristanet for three Jersey towns — and see the tags they use so you can swarm around those tags and find and be found. That’s the openness we need. If Google spearheads that with a truly open API that can be adapted by the community, then great. That is our distributed marketplace. But if not, then Google is only trying to recreate the centralized marketplaces of old — otherwise known as newspapers. That worked for newspapers when they had monopolies. They don’t anymore. Does Google think it has a monopoly?

There are two big things going on here. Firstly, the association of people (and therefore opinions) with content. This is the fundamental contribution of the combination of Blogging (or to be more specific, blogging software) and the right distribution mechanism (RSS). I say that this is the right distribution mechanism because it encourages applications (e.g. search, of course) that work thanks to the right encoding of the data: The association of people with content is something that Web 1.0 missed.

The fundamental motivation behind the original web content model was the logical structure of documents and the relationships between the (content within) documents. The functional aspect of the documents (which parts were content, which were navigation, which were adverts) was not captured, nor were notions of creation (including the content creator, the time, the subject matter, etc.)

Secondly, and this is what we are all about in this post, there is the notion of associating meaning with content. Meaning refers to the meaning of the object content (i.e. the meaning of the words authored) and the meaning of the navigational and other meta-document elements (e.g. the use of the rel tag).

Thirdly, there is associating intent with content - something which structured blogging and Riff are designed to do - they capture the notion of document genre aligned with certain authorial intentions - generally reviewing things.

Systems like Riff make tie all these things together explicitly. The data is encoded in the application, but that encoding pretty much is the application (at least on the user facing side). What Jeff is arguing for is a distribution of these representations that allow for applications like Riff to be layerd on top. Riff is an important step at getting at the functionality. Making the representation more open and fluid seems like the right thing to do, but how long do we have to wait to see powerful applications built on top. I mean, can you stand another maps mashup that puts a pin here and a pin there?

Finally, I'm going to requote Jeff:

If you aggreed on the tags, you could start doing that today via Del.icio.us and Technorati.

Agreeing on tags - organization around tags in the way Jeff suggests, is more akin to ontological and taxonomical semantic information (which I like) and not what the folksonomy taggers are about.

 

Where is this going?

If you have ever played (or watched) the game Half-Life 2, you will be familiar with state of the art urban modeling, physical models in an interactive context etc. Now have a look at some of the images from Google Earth, which includes data describing the shape of buildings, mix in some of Microsoft's Virtual Earth, which has high resolution data of buildings taken from a 45 degree angle and imagine the possibilities (taking a good does of geocoded data, GPS, RFID and so on).

You search for a hotel, a map tells you it is in striking distance of your business meeting or conference, so you flip your rich client in to walk-around mode and walk around the neighbourhood. It turns out there are just too many broken windows, and what is more, the pool of the hotel is constantly in the shadow of  surrounding buildings. So you explore - just walking the streets, guided by your local search which tells you which way to head for other hotels. Finally, you find one you like. You walk up to the door and interact with an online receptionist to book your stay.

Geocoding leading the way to rich clients?

The upcoming battle between Microsoft (Virtual Earth) and Google (Keyhole/Google Earth) is already receiving a lot of attention. What is interesting to me, beside the raw eye candy and fun of the whole thing, is the idea that we are moving to a rich client interaction with web/internet data. Browsers know a lot about a very small space of object data - document encodings. Rich clients will know more about different types of encoding, the way they interact and appropriate ways for users to interact with them. Imagine coming across a table of data in a document - in an html page - and being able to view the data in different ways, sort the data, pull in related data and so on. Adding scripting to clients, as a bolt on to the object level document layer, is not the way to do this horizontally.

Message Boards - remember them?

The hype around blogs has had an interesting effect on the perception of message boards. This post from Common Craft takes an amusing look at the rise of blogs in comparison with message boards. The takeaway (regardless of whether or not you agree with conclusion) is that community  is key. Blogs have an open but loosely coupled community, message boards have a closed but explicit community. The point of the post is that different communities enable different visibility and the potential for different focus vs breadth.

I'm not sure I agree with much of the post. For example, message boards can be very personal as well as highly focused - the same goes for blogs. One of the aspects of both of these publishing mechanisms that is often ignored is that the mechanism itself - the platform - is just a syntactic structure. The needs of the participants or authors will find expression somehow through that structure.