3 Matching Annotations
  1. Nov 2017
    1. having students do a basic Google image search for terms like “doctor” “teacher” “baby”

      It may sound obvious but it actually works. Just did it with each of these three words (on DuckDuckGo) and the results, though unsurprising, bring home the point. Tried switching on the Canadian filter, to check if their might be a difference, and it mostly reorders the results, for some reason. Also tried “student” and “musician” which provide an interesting contrast. Doing this exercise in class, would probably start by asking learners to write down what they expect to get. (Might even do it in my applied anthro class, tomorrow.)

    2. In this particular case, Google worked as a kind of amplifier of distortion.
  2. May 2015
    1. That is, the human annotators are likely to assign different relevance labels to a document, depending on the quality of the last document they had judged for the same query. In addi- tion to manually assigned labels, we further show that the implicit relevance labels inferred from click logs can also be affected by an- choring bias. Our experiments over the query logs of a commercial search engine suggested that searchers’ interaction with a document can be highly affected by the documents visited immediately be- forehand.