5 Mar

Playing with algorithms 1: Proposal

I will explore what happens with algorithms when I search as I usually do, on the Google open web (influenced by location etc data, but not me specifically).

I will then carry out the same search logged in on my personal Google account. I rarely use this account, but I do occasionally fail to log out for a few days after I’ve logged in. I would therefore expect a longitudinal series of random snapshots of my search behaviour.

Finally I will carry out the same search using my work account. I am usually logged into this account at work, and most of my search and activity should be stored there. As it is a work account, I expect all my activity to be passively logged by my employer anyway. I also expect to be doing the same searches over and over again to carry out routine professional tasks, and so a ┬ápreference ‘bubble’ or ‘me loop’ is actually useful.

I have predictive search turned off on my laptop, but on on my mobile. I may get different answers depending on the research.

2 thoughts on “Playing with algorithms 1: Proposal

  1. Katherine – I thought this sounded like a really good way of approaching this – did you get anywhere with it?

    • Yes, I tried various things–things that would be influenced by geolocation, things that would be influenced by typical searches. All my searches came up the same. I suspect that years of private browsing, carefully not logging my search history, means I don’t have enough of a trace.
      I don’t use online recommendations much, except for Amazon, and they are not very useful–or rather, they reflect what other people who are in my field, or follow the book bloggers I do, buy. And I am more likely to get that information from the source.

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>