Sunday, December 19, 2010

Emotional stress changes search

Jetlag, bad Internet connectivity, behind on my to do list -- I haven't been myself recently, but I'm working to get back the balance. However, there's one thing over which I have no control: the deteriorating political situation in the Cote d'Ivoire. The worries turn over and over in my mind and I find myself scouring the Internet for information about the current situation. In a moment of clarity, I had some insight that what I am doing has larger ramifications in terms of search:

Emotional stress changes search. I am not myself and I my search-self also appears to be different.

There seem to be two main dimensions to this difference: first, the reason why I am searching is different. I want information, but not because I can use it in any particular way, but rather because knowing has the potential to give me back a sense of control and, second, my search strategies are completely different. Effectively, it's like I've forgotten most of what I know about how to find things on the Internet.

In particular, I found myself going back to the New York Times over and over, like I expected them to publish more than one article a day on the subject. I gradually expanded, just typing "Ivory Coast" into Google, and then clicking the news option.

I don't know how many searches I fired off before realizing that I was only reading English results, and could expect to find more, perhaps, new information from the French press. That is ridiculous! I spent two years of my life working a a project called MultiMatch dedicated to multilingual, multimodal information access. If there is something I should do automatically, it is turn to cross-lingual or multi-lingual search to give me more results and more variety. I should remember that our current search engines don't give us an easy way to tell them which languages we would like to receive results in.

I started searching "Ivory Coast", "Cote d'Ivoire", "Elfenbeinkueste". Sorry, Google, I can't type the umlaut easily with this keyboard -- argh and you seem to not interpret my work around, let me try again "Elfenbeinkuste". Each of these steps that I would usually do without thinking seems particulary painful.

And then I remember social media. Twitter. This is ridiculous! My current project, PetaMedia, has a significant social media component.

I know full well that if a country's government clamps down on the press, then you back off to who's tweeting. I add "twitter" to my Google query and find an article discussing the emergence of a twitter hash tag: #CIV2010. But is there information here coming directly from the Cote d'Ivoire? Is this really (still) the most authoritative and informational hash tag?

Hmm. If I could compare what else is out there. Can I search Twitter directly for other hash tags that are related to this one so that I can compare? Apparently not. How can Twitter not let me search for hash tags? Usually, I would be able to address this question in a sensible fashion -- but now, I just leave it as a search dead-end and zap off to look at individual tweets.

I end up just kind of reading throuh individual tweets and finally find one that contains a recent release from Reuters UPDATE 1-U.N. peacekeepers won't leave Ivory Coast - Ban. Long exhale. It looks like at the moment the world is still with us and won't look away. It's a thin thread, but it's enough to let me break out of the cycle. I didn't know that this was what I was looking for, but when I find it, I also stop searching. At least for today.

Search under stress also happened to me in 2009 in Corfu after the CLEF workshop. There was a storm and the airport was closed down and my flight was canceled. I need to get to Athens to hook up to another ticket I had to Rome...but everyone wanted to get out and planes were booked full. It was one in the morning and I was hooked up to some vague wisps of wireless at a near empty airport. I don't remember the details but about the only query I could come up with was "Is there a boat out of here tomorrow?"

Search under stress is probably correlated with topic: violence, health, shelter, mobility, citizens rights. These are critical topics for our well being and happiness. Search under emotional stress happens when search matters most. It seems to make sense to tackle the challenge of search under stress as different from search in other states of mind. But first -- let me get some sleep.