I’ve spent three days this last week at ALT 2007, the annual conference for learning technologists working primarily in higher education. With 196 papers, posters, presentations and workshops to attend, there were plenty of signs of ingenuity and creativity in the sector. The event was well planned and organised by the ALT team, though I was sorry they had not managed to attract more participants from further and adult education.
There were compelling keynote speeches from Michelle Selinger, Cisco’s Global Education Strategist and from Dylan Wiliam, Deputy Director of the Institute of Education on Assessment, but the one that had caught my eye before arrival was the concluding one, “Learning in an Open World” by Peter Norvig, an acknowledged expert in Artificial Intelligence and Director of Research at Google, where he oversees a range of their cutting edge research initiatives. [Update16/9/07: There's some comment on the details of these keynotes by ALT's CEO Seb Schmoller in his fortnightly mailing - together with a link to an upcoming interview with Peter Norvig.]
I was interested to hear Peter’s views on current developments in search strategy and algorithm generation, as there’s been quiet beating on the tom-toms on this theme over these last three months. Google claims that it ‘stands alone in its focus on developing the “perfect search engine”‘, but in a New York Times article, ‘The Human Touch that may loosen Google’s Grip’ (June 24th), Randall Stross reported a host of small start ups were kicking at Google’s heels.
He highlighted in detail the case of Mahalo, which is developing a search technique with human editing for improved and more focused results. This was picked up by Matt Cutts, head of Google’s Webspam team who indicated that changes were in the air at Google and was himself reflecting comments earlier in the week from Marissa Meyer, Google’s Vice President (Search Products and User Experience) [Guardian blogs].
In the event the point wasn’t covered in Peter’s speech, so I asked him at the conclusion if he thought user generated content could help to improve and refine their search algorithms, the key instruments of this $160 billion company, now the largest media company in the world. His answer, interesting for its rationale and conclusion is below.
“I think the problem that there is too much information is that there really is that much information. It’s not just a problem of overloading people with the presentations …… No one really cares that there’s a million resolved and they’re not going to go to the end and look at that.
“So the results are up there and really the question is how much time do they have to investigate this area and how much time are they willing to put in; and can we find the good stuff for them in that allocated amount of time.
“And given that, I think there is room for user generated commentary to help that process. To an extent Google has always been driven by user generated content – so there are users who happen to be web masters who publish material, there are other users who are web masters who link between the materials; and that provides sort of codes, that we go on, to judge appropriateness.
“We also go by looking at our users’ address activity to judge what is important and so on. So it’s done by algorithms but all these algorithms have, after inputs, actions that offset the punts of the user.
“Now I think we can use additions to those sources that we’re already using. We can use some more explicit ones – people loading ‘Yes – This is a good site for this topic, or for this keyword, or for this area'; and getting more people than just people who traditionally have had access, network access to try to open up that form to democratise it more, so that other people can put in their voices as well. So that’s an area we are certainly looking at”.
The Economist’s recent article, ‘Inside the Googleplex’ (1 September) chose to use “The Simpsons” TV programme portrayal of Marge Simpson typing her own name into a Google search engine as a sign of the company entering mainstream US culture, but her amazement to find 629,000 results doesn’t quite hit the right note now. Companies like Amazon, epinions and Trip Advisor have been using UGC with marked success for a good few years, so it will be fascinating to see how quickly Google follows the trend and moves in this new direction.