Data Mining: Text Mining, Visualization and Social Media

Strings are not Meanings : Part 5

I'm still chewing on Nöe's book (criticism of which has to revolve around his statements regarding 'the world [just] showing up [for us]'). In the meantime, I'm back to reading Turney's blog, and this post which covers a few of the points that I first read in Pinker's books. Turney (and those he quotes - Minsky, Quine) are certainly working with the model-the-world-then-reason approach to consciousness. In reading Nöe, I definitely get that feeling that he's defining a club that we're just not cool enough to join.

To better understand the book, and its position in the discussion, I'm looking for any reviews that summarize it and connect it with other work.

• Uncommon Descent
• American Scientist: "Out of Our Heads is a manifesto of hyperbolic claims resting on sketches of argument"
• Stark Reality

Science July 24th 2009

One of my blogosphere visualizations appears on the cover of this week's Science magazine. You can see more over in the gallery.

Quantifying and Visualizing Remoteness

Visuale points to a fascinating visualization published in New Scientist which measures the remoteness of locations on the planet.

[R]emoteness is calculated taking into account how long it takes to travel by land or water to the nearest place with a least a population of 50.000 inhabitants.

The research was carried out by the European Commission's Joint Research Centre.

The results are visualized in the following way.

The JRC doesn't yet appear to have any information on the project, but it has been picked up by Visual Complexity

Plotted onto a map, the results throw up surprises. First, less than 10 per cent of the world's land is more than 48 hours of ground-based travel from the nearest city. What's more, many areas considered remote and inaccessible are not as far from civilisation as you might think. In the Amazon, for example, extensive river networks and an increasing number of roads mean that only 20 per cent of the land is more than two days from a city - around the same proportion as Canada's Quebec province.

Rediscovering the Feed Reader

For some reason, while travelling, I've really noticed how I'd fallen into the bad habit of relying on Twitter and aggregaters like TechMeme to channel content for reading. While everyone is warm and fuzzy about WOC, recommendation systems,etc., frankly they scare the innovation right out of me. As I've mentioned in other posts, what gets things to the top of the pile is only partly what I vaguely call inherent value. A good part of what is going on is the force of social value, which, IMHO, can be the death of serrendipity.

Feed readers don't excatly cure this problem, but I've really enjoyed getting back to the long list of blogs that I've subscribed to and spending some time with the high quality, well thought-out opinions from the stable of blogs that I've spent years cultivating.

Infographics Tutorial

I'm a sucker for isometric infographics. Randy Krum at Cool Infographics points to this tutorial by Wendy Ding. What interests me is how organic the process is. Note also the intentional 'Where's Wally Effect' by design.

Hyperlocal's not Local

Max is searching for a definition of Hyperlocal. He asks if it isn't just the same as local. I think that the definition in wikipedia stands up to this question and answers it clearly: hyperlocal isn't just about targetting, it is also about the source of the content (it's not just beamed in). Once should be skeptical when witnessing the land rush caused by new buzzwords - but we don't need to be take these new terms too seriously, it is the meaning that we need to care about.

I suspect the real issue here is the challenge/opportunity that hyperlocal presents to advertising (Max's domain).

Break/Most Popular Blog Posts

I’ll be taking a short break so there will be no new posts for a time. The following posts are at the top of Feedburner’s list for the past 30 days.

1. The Long Tail of Text Mining
2. Bing haz Twitter
3. Twitter Trending Terms – Could Do Better
4. Measure, Don’t Guess – Growth in the Blogosphere
5. Steve Irwin, Michael Jackson

Microblogging and Astroturfing Hyperlocal Sites

Earlier this month, the Guardian reported a new venture into the hyperlocal space by the owners of the Daily Mail. The new sites launched by The Local People serve the following locations:

• Bedminster (36)
• Berkhamsted (39)
• Bideford (64)
• Bude (36)
• Chippenham (40)
• Christchurch (42)
• Clifton (46)
• Corsham (31)
• Dalston (54)
• Devonport (13)
• Dorchester (38)
• Exmouth (47)
• Falmouth (52)
• Finsbury Park (29)
• Hedge End (16)
• Keynsham (15)
• Kingswood (20)
• Locks Heath (22)
• Marlborough (10)
• Nailsea (23)
• Newton Abbot (38)
• Okehampton (25)
• Otley (29)
• Portishead (14)
• Quedgeley (12)
• Redland (30)
• Tavistock (15)
• Tiverton (55)
• Westbury On Trym (27)
• Weston Super Mare (24)
• Wimborne (7)
• Yate (24)

The sites have, among other features, a twitter like interface which allows the registered participants to microblog updates about the location. This is similar to the model that Neighborlight has adopted. Thus far, the Local People sites haven’t attracted large numbers of registrants. The numbers above indicate how many people have created a profile on the site.

I suspect that hyperlocal content requires a strong personality to help lead discussion about the community, and provide the dedication to both the community and the notion of hyperlocal required to maintain a high rate of quality relevant content. The blog format is perfect for this. Creating spaces online which provide a forum for hyperlocal content has the potential to miss out on this personality aspect, requiring a large number of people to contribute to make the site interesting.

In addition, locations select for residents, not for individuals who create content. Thus there are plenty of people willing to follow updates, share links about the site (e.g. via twitter) and chat in communities, but not necessarily to spontaneously generate the type of content required to get the community off the ground.

The question is, then: can large organizations (like the Daily Mail) succeed in creating compelling hyperlocal experiences?

Hyperlocal London

I’m starting to look at a few hyperlocal blogs in London town. Greenwich.co.uk, which is part of the Hyperlocal.co.uk network of hyperlocal blogs is a good place to start. In their ‘posts we like from other blogs’ section I see a link to a more straightforward blog called 853. While Greenwich has a lot more going on, and feels more like a business, the 853 blog is more my cup of tea: plain text and excellent pictures. Darryl, who runs 853, has developed a great list of other blog resources which appear in his blogroll. I’ve noticed that in some hyperlocal networks, there is a little reluctance to link out of network. Better to spread the love!

Of the blogs Darryl links to, Blackheath Bugle, Faded London and (the Guardian’s) Dave Hill’s London Blog look interesting. Faded London in particular shows that the hyperlocal scene is broad enough to support very specific locale oriented niches. The blog puts it this way:

London is a city constantly re-inventing itself, but every now and then there's a window on to the past. It might be a washed-out advert painted on a wall, a plaque, an architectural quirk or elaborate man-hole cover - all worth recording before they are finally removed or hidden.

TechCrunch’s Other Ethical Travesties

As TechCrunch celebrates the freedom of the press to pursue money and attention over basic civility, it makes you wonder what other stolen documents, business plans, private emails, etc. they have purchased or acquired from less well known companies. Perhaps they are doing this all the time but never quite had the pay off that Twittergate has given them.