This is an excerpt from the Mashable! post Wolfram Apha Approaches 100 Million Queries by Ben Parr:
“…I recently had the rare pleasure of chatting with the founder of Wolfram Research and the man behind the Wolfram Alpha engine and its algorithms. For 45 minutes we discussed the concept of Wolfram Alpha, the difficulties of parsing language, how social media helps filter information, and even the ability to analyze multimedia. Dr. Wolfram also revealed that Wolfram Alpha is set to exceed 100 million queries. Below is an analysis of the interview. If you’re interested in the full audio, we’ve provided it at the end, embedded via Yahoo Media Player…
It’s clear that Dr. Wolfram is fascinated with how users interact with the Wolfram Alpha system. Beyond discussing some ways users have used Wolfram Alpha in unanticipated ways (6 mins, 30 secs), Dr. Wolfram helped define why Wolfram Alpha is not a search engine:
‘At a practical level, it means that it doesn’t make a lot of sense to cache things because you keep on getting new and different things you’ve never seen before; it kind of reveals the fact that [Wolfram Alpha] really has to compute stuff.’ (8 mins, 40 secs)
Because Wolfram Alpha is designed to parse natural speech and natural language (full sentences and questions), Dr. Wolfram and his team have had to address and understand the difficulties of creating a web tool based on natural language. This becomes difficult when people are trained to use online input fields and search boxes in a specific way:
‘People see an input field and they think search engine and they think two or three or four words…the main point is that they’re saying let’s search for a concept, not let’s try and get a specific answer for a question that we actually have.’ (16 mins, 20 secs)
In the end though, it seems that people are quickly grasping better and more effective ways to interact with Wolfram Apha – in fact, quicker than Dr. Wolfram expected. People are finding diverse and unique ways to find the answers they seek via Wolfram Alpha…
There are several important lessons and takeaways that we can take from the Dr. Wolfram interview:
- First, that human language is deceptively complex and that analyzing it has been a challenge. In fact, Dr. Wolfram wondered if it was even possible many times during the process.
- Second, that there is a necessity to provide people practical access to this type of computational information. This may be part of why Dr. Wolfram and Wolfram Research built this engine, and even why they decided to live stream the launch.
- Finally, that social media is a vital tool for filtering information and opinions. It’s useful to the Wolfram team already for understanding what users want, and may be useful later on as actual data points for the Wolfram Alpha algorithms.
Wolfram Alpha is simply a project we’re all going to have to watch (and use) in the coming months. It’s not a search engine nor is it a social sharing tool like Twitter, but we still haven’t seen this much excitement for a new web product in a long while.
Because we just love multimedia, we have embedded the interview in its entirety here. You’re also welcome to download it, stick it on your iPod, and listen to it while you’re on the road, on the train, or just relaxing. Dr. Wolfram’s insights and analysis are rare gems that you should not miss.”
Wolfram Alpha Mashable Interview (43 minutes, 48 seconds)