Google and the difference between information and knowledge

I am a regular reader of Frederic Filloux’s weekly commentary on media, The Monday Note. I cannot recommend it too highly for its trenchant originality.

Triggered by a new wave of complaint about Google in Europe, today’s note looks at Google’s interest in legacy news media. Why, Filloux asks, has Google maintained Google News for so long when it makes no money and when news sites are so relatively insignificant as sources in Google’s gigantic search business?

He thinks that the answer lies in Google’s planned move from being a search engine to being a knowledge engine: the ability to deliver more sophisticated and useful answers than most of us can dream automated search can now deliver. At the heart of that effort is something called Knowledge Graph. And the key to that is the boring-but-important issue of the structure of data. News media connect bits of information to make it knowledge people may want and need.

As Filloux points out, pure-play web news sites are often better at this than the ones built by established mainstream media – despite the fact that the legacy media often hold richer, bigger databases. New media’s data is easier to find because what is stored is better labelled and can be made sense more easily.

Well out of public view, many sizeable news organisations have been struggling quietly with this issue. It holds a clue to a very large change in the basic shape of information media. Printed and broadcast media once produced an ephemeral product which was of passing, temporary interest while curiosity about the contents peaked. People made regular appointments to consume the output. That output was archived, but consulting that store of knowledge was slow and effortful (unless you worked for the news organisation itself and the library of past news was in the building).

Now, all that accumulated information is stacked where anyone online can search it. In effect, news media are shifting from being a rolling, moving sequence of bulletins to becoming mountains or encyclopaedias of data which are added to, little by little, all the time. The whole idea of news has changed as what used to be scarce – timely (mostly) accurate information – is in glut.

Sensing this, news outfits such as major papers have been trying to make their archives searchable, to tag their data and to package it better (example of the New York Times topic pages here). The software and the resources to do this really well have often proved elusive. In short, news media are trying to find out if they can move into the broader knowledge business. Filloux suggests that this may be harder than it looks.

Sound a bit nerdy? Yes. But how data is structured is a pivot of how we acquire and use information in the future.

Update 6/11/12: Malcom Coles, the digital director of Trinity Mirror, shot back at Filloux’s argument defending the ability of legacy media companies to compete in search-engine optimisation. His irritated rant seemed to miss the point that this isn’t just about SEO tricks but about the long-range importance of how organisations structure the data they hold. (Have a look at Trinity Mirror’s 5-year share price here).




Tags: , , , ,

1 comment

  1. While journalists always joked that today’s newspaper was tomorrow’s chip-wrapper, they always forgot that it would also be next year’s, next decade’s, next century’s primary historical research material. The increasing availability of scanned newspapers going back three centuries or more on the web is already revolutionising historical research across all sorts of fields.