Potato I have

The latest Economist Technology Quarterly (free access for the moment) has a brief status on video search. Searching the image itself is very difficult, but there is a lot of activity in searching associated text: the closed captions, the audio portion (via speech recognition), and human or machine produced video metadata.

IBM has a test version of a more sophisticated analysis/retrieval engine called Marvel; after manually tagging 1-5% of a set of video content, the system learns to recognize the rest — this is a huge reduction of the expensive, error-prone human tagging effort. IBM’s web has good detail on the project. From those pages I learned about MPEG-7, which is a video metadata (rather than encoding) standard.

The ETQ acknowledges that business models are lacking, but singles out Critical Mention as one that is working; they offer web-based search and alert (”reputation management”) services for TV news. I see in the news last week that they now have a licensing agreement with AP (Associated Press) Digital.

Since it looks like metadata is playing a big role in video search, it should be interesting to see which producers/consumers are doing the tagging, and where the analysis and search logic resides. The big search engines are an obvious home, fancy services like CM are another, but why shouldn’t I also have something to find stuff on my computer or LAN?

I found a fine article about Miyazaki by Margaret Talbot, one of the few people to manage to interview him. Along the way, I discovered the New America Foundation, which looks like a place where I could seriously spend a lot of time reading; check out the list of the best articles of 2005. Its Board of Directors is chaired by James Fallows. And from its mission:

Now, more than ever, our nation needs a robust public debate, one that does justice to the complex challenges and opportunities of this unfolding era. […] The purpose of the New America Foundation is to bring exceptionally promising new voices and new ideas to the fore of our nation’s public discourse. Relying on a venture capital approach, the Foundation invests in outstanding individuals and policy ideas that transcend the conventional political spectrum.

The Talbot article was originally published in the New Yorker, but they seem to have very perishable links.

France was late to the internet, partly due to an already established online services network called Minitel. But lately it seems they’ve become quite a leader, at least in high-bandwidth offerings. My DSL provider’s plan is 30€/month and includes a connection of up to 20Mbit (in practice I get about 10Mbit), a VoIP solution that has free unlimited calls in France and is about 1.80€/hour to California, and >200 TV channels (>80 free, including the new digital-over-airwaves channels). The satellite content providers are now delivering over DSL too. Stability and service are not rock-solid — and they’ve been downright nightmarish for some — but I’ve had a pretty good experience overall. Connection-only, high-bandwidth (20Mbit) offerings are about 15€/month, from multiple providers.

My provider has been aggressive with new services, and I especially like this free one; it’s based on the free and open-source (GNU GPL) VLC media player project, and means I can stream video from any computer on my LAN (Mac and PCs for me, but most OSes are supported) to my TV without any additional hardware. No computer in the living room. DVDs, mpgs, etc all presented as playlists which can be selected with the TV remote. Well, any multimedia can be streamed, but video is the interesting one for me.

I’m probably mentioning Wikipedia too often, but my previous post about the semantic web offers an interesting comparison. The Forbes article and the Wikipedia entry are both approximately 1200 words, and the former is perhaps a better non-technical introduction, but I’m vastly more satisfied with the latter.

Why? The Wikipedia entry:

• has had an awful lot of editing, by many people
• has an ongoing discussion to improve the entry; anyone who can read the entry can participate
• being hypertext, directly links all of its numerous internal and external references
• also links its ‘Knowledge representation’ category, which has another 37 now 44 topically-related entries

The one is a feature article, the other an encyclopedia entry, they’re not the same thing, but it seems to me a good illustration of the different authoring processes and results.

This Forbes article describes the semantic web in simple terms, starting by mentioning how rudimentary our search technology is:

We’ve become a society of information managers, navigating huge amounts of data with ease and expertly tracking down obscure facts and figures.

But as far as we’ve come, all we’ve really done is become good at finding needles in haystacks.

Semantic web means smarter search, but that is just the start:

[S]emantic technology also holds great promise for all kinds of businesses. […] With data growth rates averaging between 20% and 30% annually, many businesses are drowning under the weight of their own files […] “We want to get the human out of the loop for obvious reasons–they cost money, and they make errors.”

So machines can handle a lot of the work in a rapid-growth area.

Google’s comparison shopper Froogle is given as an example of the first step to the semantic web: agreed-upon “tags” help us all find a price, for example, regardless of what language or what terms the web site uses. But it doesn’t do the harder things: relating terms and building context.

(via wired)

I’ve noticed for a while that Firefox browser share at sites I visit is much higher than the 5-10% we’ve been hearing; this InformationWeek article tries to find numbers for the blogging community, and they are considerably higher — in some cases Firefox is exceeding IE:

At Boing Boing, among the most popular blogs, the most recent statistics for the month of March indicate that 35.9% of visitors are using Firefox, compared with 34.5% using Internet Explorer.
[…]
Kottke.org, another popular blog, reported on Feb. 27 that 41% of visitors sported Mozilla-based browsers (one of which is Firefox), while 31% of visitors arrived with Internet Explorer.

I find the productivity boost, just to pick one thing, so compelling with Firefox that I never understood the 90/10 numbers. Now I’m hoping that blogger browsing is somehow a leading indicator for browsing in general, and that Firefox won’t remain a weak minority for the duration.

Ben Hammersley has an article in the Guardian that claims Yahoo has eclipsed Google. For him, Google’s real differentiators until now have been their lab and prototypes, and their API. He says Yahoo is better on the first, because they are more transparent:

Research.yahoo.com, launched last month, is the same idea as labs.google.com - a showcase for new and interesting projects - but it’s better. Unlike Google, Yahoo publishes its papers, names its researchers and says what it is up to.

He also claims Yahoo’s new API is better too, because it “has more features, it’s more complete, it’s technically more elegant, and it’s easier to use than Google’s alternative.”

Finally, he compares a number of product moves, e.g. Picasa vs Flickr, but it was this one that especially caught my attention:

Yahoo has quietly launched search.yahoo.com/cc, a search engine engineered to find and index Creative Commons material. To do this, Yahoo must be indexing the web for data called RDF - a highly advanced, potentially powerful technology that Google has said it isn’t going to touch.

Yahoo even wins the googlefight.

The US State Department provides a synthesis of foreign media reaction:

Each business day, the U.S. Department of State’s Office of Research produces an Issue Focus of foreign media commentary on a major foreign policy issue or related event. These reports provide a global round-up of editorials and op-ed commentary from major newspapers, magazines and broadcast media around the world.

(via neat new stuff)

The best find in a slashdot thread on comparing search engines is Phil Bradley’s task-oriented engine list; go bookmark it now.

From the same thread was a graphical comparison tool of google and yahoo search results for your query string.

A friend just asked in jabber what I am up to, and I said “I’m reading a history of wikipedia.” The project started in 2001, and already

the nonprofit venture is the largest encyclopedia on the planet. Wikipedia offers 500,000 articles in English - compared with Britannica’s 80,000 and Encarta’s 4,500 - fashioned by more than 16,000 contributors. Tack on the editions in 75 other languages, including Esperanto and Kurdish, and the total Wikipedia article count tops 1.3 million.

(via robotwisdom; seeing Jorn Barger filtering links again after a year and a half away is the best thing that’s happened to me today)

(more…)