I was invited to this search-a-palooza, but I’m in NYC and then SD tending to some business and family issues. I won’t be covering it, but you can follow it here.
Today Is Google “Searchology” Day
I was invited to this search-a-palooza, but I'm in NYC and then SD tending to some business and family issues. I won't be covering it, but you can follow it here….
I hate websites which resize my main browser window. The link in this post goes to one such website. That link should have been a pop-up window link.
I was used to my browser window size and I can’t it back exactly as it was 🙁 Will take few days to get used to this size.
yikes, LOTS of feedback… and not the good kind. there’s so much echoing, it sounds like the podcast’s taking place atop the grand canyon. hopefully, this will be resolved soon for SoCal folks.
I caught the last half of this webcast. I have to say it is shocking to see Google finally catching on, finally starting to offer query clustering and explicit user feedback tools. For example, Marissa Mayer said something to the effect of users needing a way of browsing or navigating search results that goes beyond a single-dimensional ranked list. In particular, when she introduced the time/date-based clustering of results, I nearly fell over in my chair.
Why was I so shocked? Because this contradicts everything Google has been saying for years about users being lazy. Users, they have led us to believe, will not interact with query modification or query enhancement tools. Users will only type 1-2 words into the search box, and then when they don’t find what they are looking for, will then search again, and type a different set of words into the search box.
Thus if the user is looking for information on the Olympics in Munich, a typical interaction (at least according to the things I’ve heard Google say for years) is that a user will type “olympics”, see that their query has failed, and then type “olypmics 1972”.
Now, with this timeline view, Google lets the user type “olympics”, and then click on the 1970s time range. In other words, Google first does data mining/information extraction on time-features, clusters the top 100 or 500 or 1000(?) documents by those features, and then offers explicit query modification tools to let you choose which cluster is most relevant to you. This is a significant new direction for Google. This is Google starting to think like Vivisimo. Google no longer assumes that the user is too lazy to use these tools.
This is a huge positive step forward, and it is great of them to finally start doing something in this area. But I am frankly a bit floored, again because this contradicts years of Google statements to the contrary, years of saying that users are too lazy to use such tools.
I wonder what has finally changed? Google, or users?