Tag Archives: The Internet


EDITED 10/19/11: Google has since shut down the Aardvark site. Damn these corporate bigwigs who think they can stamp out the little innovators and I wish would just hire me already!

Cool new thing you should check out. Ever had a question but didn’t know how to get an answer? Sure, you can hop online and see if you can turn your question into a search for which Google will return some results, but Google is an engine, a program, not a human. For example, Colleen and I are new to North Liberty and it’s time for our six month teeth cleaning. Since we don’t know anybody in North Liberty that we can ask for recommendations, we pop on the web to see what we can find. But how do you use Google for this kind of query? You can’t exactly search “friendly dentist, north liberty” and hope to get any sort of legitimate return that you can use. So much for the most popular search engine in the world.

You should check out Aardvark. This is the new generation of searching. In the last five years we’ve seen social media and networking blow up like nothing anyone expected. It only makes sense that search and discovery would find a way to incorporate the social aspect into itself. So here’s the deal with Aardvark. There’s a text box in which you type your question in natural human language, anything like “Which is the friendliest dentist office in North Liberty?” or “What’s the most reliable brand of tires?” or “What’s the best bbq sauce that pairs well with Sam Adams beer?” Aardvark takes your query and finds another human user that can provide an answer. So rather than just getting a list of links that might or might not be what you’re looking for, you get a response in human language from another human who has a better idea of what you’re looking for.

It works like this, and this might be what turns some new users off initially. You do have to sign up to even ask your question. But it’s nothing more than an email and a zip code. It also asks you what subjects about which you’d consider yourself able to answer questions. So you sign up, and Aardvark then takes your query and sends it on to its ever-growing set of users, hopefully one of which will answer your question. Depending on the type of question and the key words that are found within the query, your query is sent to users who have specified that they have some knowledge about the key word subjects. Real life example, I recently queried “What is the most trustworthy auto repair shop in North Liberty, IA that can repair a broken exhaust pipe?” The main key words picked out of this query were car maintenance and North Liberty and the question is sent to other users who have specified that they have some knowledge about car repair, and even more optimal, users who are located close to North Liberty. My query was received by Josiah M. in Iowa City, IA and he replied back through Aardvark “I can highly recommend All Season’s Auto”. Boom. Now I’ve got at least an idea of where to start looking for auto repair from someone locally.

This is such an awesome idea, even though there are some obvious cons. The pros are pretty great though. Firstly, the design and user interface is just perfect. The main page has great colors and hardly anything except just the one text box to enter in the query. It’s very Google-esque in its design without immediately recalling Google. The other huge pro is that returns are going to be far more user specific than you’d get from Google. Having your query sent to other human users who can understand your natural language question lets them answer it in a way that you, another human user, will be able to understand. It’s awesome.

Plus, by becoming a user of Aardvark, you also become an answerer. When you sign up, you pick a list of subjects in which you’ve got a little expertise, and if another user asks a question that involves one of your subjects, you might get an email or instant message (or whatever specific type of communication you’ve listed with the site) from Aardvark that says something like, “Hi (your name), want to try answering this question? (then it lists the user’s question)” This is a great system because it allows the answerer many options. The answerer can go ahead and answer the question with an easy reply email, they can simply choose to ignore it and the query will get sent on to another user, they can flag the question as inappropriate, or they can refer the query to another user they have in their network that might be able to answer it. Example: I get a question from a user about the best sushi around North Liberty. Unfortunately, I don’t like sushi so I know I can’t give a good answer. Luckily, I have a friend in my Aardvark network that does live in the area and loves sushi, so I refer the question on to them to see if they can answer it. Also, as you get answers to your questions, you can add the users whose responses you like the best to your network to continue to query those users most often. That’s what makes this interface so great; it is so user-centric and really boosts the social networking aspect of search and discovery.

More about the communication that this interface allows between users. With a site like Google, you query some terms and get some returns. That’s the end of it. You can tweak your query a bit to get different returns but it’s still querying a generic machine search engine. With Aardvark, the whole search/return system is like a giant conversation between users. I query something, another human user responds, and if I want, I can respond back to continue on a conversation between us. This allows a huge amount communication, conversation, and concept searching among users.

Here is a great example of this. I submitted a query, “What are some good blues albums that are stylistically and sonically like the album ‘Bluesbreakers With Eric Clapton’ by John Mayall?” I got four responses:

1. Vera S. (27, female, from Dnepropetrovsk, Ukraine) said: “you can check out ‘Similar Artists’ section at http://www.last.fm/music/John%2BMayall%2B%2526%2BThe%2BBluesbreakers

2. Jim B. (59, male, from Bowling Green, KY) said: “Hi, Try Stevie Ray Vaughan, The Paul Butterfield Blues Band, Derek and the Dominoes, Otis Rush, Cream, Johnny Winter, and one of my all time favorites Albert King. Jim. Also, try Peter Green Jim”

3. Tolgahan K. (22, male, from Rockville, MD) said: “John Mayall has plenty of albums with great guitarists, Clapton is just one of them. “A Hard Road” with Peter Green, and “Blues Breakers” with Walter Trout are very similar to the one with Clapton, just pure electric blues. Also check out Clapton’s “From The Cradle”, one of the best blues albums that is modern, yet pays respects to the traditional stuff.”

4. Michael E. (male, Chickawauga, GA) said: “Try Jonny Lang, Buddy Guy, Jimmie Vaughan, Taj Mahal, Little Ed and the Imperials, Steve Vai and the rattlers, Robert Cray, Dr. John. Just to name a few Mayhal was a St. Louis style Blues. Try a little Chicago style with South Side Johnny(has a brass section) Hope this helps.”

This is a pretty good sampling of responses I’d hope to receive from most queries. The first one is pretty dumb; if I had wanted to use Google for this query that’s what I would’ve found. Not helpful. But the other three are great; people who have actually listened to the album and know what other albums sound like it suggest those and artists with similar styles to John Mayall. The last guy even got specific with different genres of blues style. Awesome. And I can respond back to any of these that I want to, basically starting an online dialogue.

Clearly this engine is not without its set of disadvantages. The fact that the entire system is based upon communication with other human beings rather than just pulling up a list of links means that timing is an issue. For every query I’ve tried so far, responses have taken upwards of an hour to come back. Not optimal if you’re looking for the best Mexican in town and you’re hungry right now. But for less urgent queries, this is such a unique way to find answers to questions.

Another big con for me is that Aardvark was recently acquired by Google. I was a little disappointed to hear this, even though it will mean the system getting more publicity and more users, thereby improving the quality of its output. Google is such a giant monopolistic company, and while the products that they offer are absolutely great, it’s always nice to see an upstart with an idea that actually has something to offer that Google doesn’t. But this isn’t a dealbreaker for me, and I’ll still continue to use it when I need.

Here’s the link again: give it a try. It’s not a perfect replacement for Google; for most queries, you’ll probably want to stick to Google. But for those times when you need an opinion from someone, have some time to spare before you get an answer, and want more returns than the top ten corporate sponsors from Google, try Aardvark. And then let me know what you think. It’s a pretty cool thing.


Leave a comment

Filed under Librarianship

Jon AKA web designer.

How sweet is this? I have created my own web page.

Take a peek.

For my Computing Foundations class, we get to use the library science program’s online web server to upload things like websites. Our first assignment was to learn a bit about HTML by reading a textbook and writing a website in HTML. It’s pretty cool, programming in HTML isn’t really “designing” a website, but rather taking the template for a website and telling it what to do. So this is website is very basic but I’m still pretty happy with it. Obviously it could use some work and I will hopefully be updating it as I come up with new stuff; any suggestions?

1 Comment

Filed under Random