Friday, March 29, 2013

Online Question Asking

 I do information retrieval research, which means that I try to figure out how people use search engines, and use what I learn from that to build better search algorithms. My goal is to take the short, one or two word query that you enter into a search box and use it to identify a few relevant results from among the billions of documents that make up the web. Something happened a few years ago, however, to change the way that I think about search.

I was in Hawaii with a friend, and as we drove through the town of Wailuku we crossed a street named Keanu Street. This led to a big discussion about whether the actor Keanu Reeves was born in Hawaii or not. I don’t know how these kind of discussions got resolved before the Internet – there were probably big arguments and friendships lost. But, fortunately for Sheila and me, computers are perfect for this. Sheila pulled out her smart phone, and I thought, “Great, she’ll search for the answer and our conflict will be resolved.” But instead she surprised me by texting a handful of her friends to ask where Keanu was born. Within a few minutes someone wrote her back with the answer that, “No, he was not born in Hawaii, but his father is Hawaiian.”

Prior to this incident, I had been thinking about the process of looking for information as entering a query into a search box. But, as Sheila’s text illustrates, most of the time you want to know something you actually ask somebody a question. Over the past decade we have built amazing tools to help people search better, but we have not yet built new tools that change the way that we ask questions. However the Internet and social networks offer real opportunities for us to do this.

Thus Meredith Ringel Morris and I set out to try to understand and support social online question asking. Technology changes question asking in that it enables us to broadcast our questions to many people. Sheila, for example, asked a number of her friends where Keanu was born, and this is probably why she was able to get a response quickly. We likewise observed that many people use social networking sites, like Facebook, Twitter, or LinkedIn, to mine their networks for information, and as a result focused our efforts on social network question asking. Over the next couple of weeks I will share a little about what we learned about why people ask questions rather than search, how to ask questions to get the best possible responses, and how technology can better support social network question asking.

This information is also summarized in a TEDx talk.


  1. I think that was because searching is LIKELY to get a bunch of results than unique exact information. And some people still think human resource is more reliable than net. But if I were in place of her, I would rather do a brief search or ask via Google Now.

  2. I hate when people ask "their network" rather than just searching themselves. (As a librarian, I would then go to IMDB or google to get the answer and be thinking to myself, why didn't she just google it herself?) Maybe because I'm an introvert so it doesn't occur me to use people that way, unless I am asking a specific person who I know is an expert at the topic in question.

    Found your blog via the SearchReSearch blog btw.

  3. "... most of the time you want to know something you actually ask somebody a question." Thank heavens for librarians (and thank heavens for job security for librarians)!

    I think most people who would go to the trouble of coming to the library to ask such questions have either had bad experiences with online searching, or are complete strangers to technology. As a librarian, I couldn't be happier when people ask me such fairly straight-forward research question; it gives me an opportunity to explore search engine choice, search phrase construction, choice of results, and credibility with someone who may have had poor experiences in the past.

  4. Thanks for the comments. :) I just added a post about what we have learned about why people might ask v. search, since that seems to be a topic of particular interest.