"Jack Dorsey from Twitter, Mark Zuckerberg from Facebook, all those Google nerds: They’re monumentally screwed, because they have no idea how to tame the monsters they have created. The way I see it, these guys — and they are mostly guys — were arbitrarily chosen. They started with some good ideas, some luck, great timing, got lots of people to believe in their rosy vision, and they won the unicorn lottery. Little did they know or care what problems they were creating."
Note that one of the themes of the professional search workshop at SIGIR was transparent, explicit interaction. So there is at least an acknowledgment that for certain users, this sort of thing is a given. I'm just not convinced that more than 3% of users wouldn't also engage, if the system were built in a way so as to allow them to get immediate, transparent feedback on what the system is doing for them, as a result of each explicit interaction. Thoughts? Comments?
So I continue to remain a skeptic of the belief that web users won't interact, in an explicit manner, with search systems. I don't know that it has ever really been tried. Not just in the sense of throwing up a button or two that the user probably doesn't understand. But in the sense of really, transparently walking the user through what the system is doing for him or her, when the user interacts with it.
I've never believed that story, because I've always though that the experiments were never done in a way that allowed the users to really understand what the system was doing for them, and never really gave the users immediate, transparent feedback on their explicit interactions. It was thus confirmed to me by this engineer that such transparency and immediate feedback was not present in the system.
At SIGIR last week I met one of the engineers who worked on explicit relevance feedback at Altavista in the late 90s. This is the apocryphal story about only 3% of users using it, which influenced multiple generations of web search companies in believing that users would never explicitly engage/interact with such features.
The "unofficial" Information Retrieval Mastodon Instance.
Goal: Make idf.social a viable and valuable social space for anyone working in Information Retrieval and related scientific research.
Everyone welcome but expect some level of geekiness on the instance and federated timelines.