Last Friday the much talked about Wolfram Alpha finally went live to the public. The news of an impressive search machine that utilizes Natural Language Processing has captivated the world of search for quite some time now. Being a SOSG myself, I naturally couldn’t wait to get my keyboard humming with queries! Here’s what I found:
My first query was for my company’s primary brand name. This was when I realized that Wolfram Alpha isn’t a search engine by the mainstream definition. Instead of posting a list of web sites Wolfram’s software analyzed the term itself and gave a unique interpretation including the term’s etymology. This would be impressive to say the least, however, my brand name is a completely made up word! That’s extremely impressive!
Wolfram Alpha’s biggest potential is its amazing computation engine. This type of technology is ideal for quantitative analysis and in-depth factual research. If you take a look at the many examples that they provide on their site, you will see that the areas that it excels in the most, are rich with calculations.
So, after it was clear that Wolfram could crunch the numbers I wanted to see how it could handle topics that have little to no quantitative value. Therefore, my next query was my favorite TV show: Doctor Who (yes, I am a complete geek, I have accepted this, you should too). Wolfram returned a one page analysis on the 1996 Doctor Who movie. Sounds good right? Wrong! See, every Doctor Who fan will tell you that the ’96 movie was a very brief footnote in the Doctor Who universe. Along with the movie there have been 31 television seasons with a total of 11 different actors playing the lead role. None of that was included in Wolfram’s results.
So I know what you are thinking: Get a life, Joe! It’s just geeky TV. It may be just TV with this query, but what this tells me is that Wolfram might encounter scalability problems in the future if they can’t acquire a growing repository of data. Search engines like Google have no problem expanding their index as the web grows because Google’s index is basically an ever expanding copy of the relevant areas of the web. However, Wolfram Alpha’s index seems to be made up of data taken from different databases and sources. It doesn’t appear that their methods for data collection are organic, making scalability a challenge. However, if Tim Berners-Lee’s vision of linked data, one day becomes a reality, then Wolfram Alpha may not see a scalability challenge at all.
In the end, Wolfram Alpha is a lot like my 5-year-old niece. My niece is amazing, just to name a few things: she plays the violin, can already read and do division, and is an amazing ballerina (yes, I am a proud uncle). She has loads of potential. If given the right environment she will grow up to be an amazing human being that will accomplish big things. In this regard, if Wolfram Alpha is given the tools and capabilities to grow and expand it will leave a lasting impact on the way that we search and understand information.