Wolfram Alpha
Sunday, May 17, 2009
There is a lot of hype on Wolfram Alpha that launched on 15th - a brand new search engine, from the company famous for its Mathematica software.
But this search engine is math based (no surprise) - which calculate 'answers' to your questions - or rather tries to figure out answers to the questions you are querying.
They even posted their infrastructure details - having the 44th largest supercomputer, mirrored at 5 different locations and can handle 175+ million queries a day.
I tried a little more than a couple of queries, most of which returned Wolfram|Alpha isn't sure what to do with your input.
I tried queries based on my hometown, Trivandrum, Kerala which I keep track of on the net from time to time. Given "Where is Trivandrum", it showed the map of India with Trivandrum marked. But it doesn't seem to know anything about kerala - one of the top 50 tourist locations in the world (National Geographic). If this isn't in their current database of 10+ trillion of pieces of data, its going to be hard for me to find information using it.
Then I tried input queries on stuff it could 'definitely' search on - maths. 10000! gave the first bunch of 35660 digits but you can easily extend the number of digits. But even expressions like "5+132!-59^5*sin(32)/5log(12)15" showed results with neat details. Over 20 years in the making of Mathematica, Wolfram would probably have the best number search engine. But "Knowledge" based ? It doesn't seem to have a lot of knowledge as of now.
After a while it just got tired of searching.
Everyone is curious about who or what will be the next Google. There is a big surety among many, even by Tim Berners Lee, that semantic web search will overtake statistical based web search (GYM). Wolfram Alpha is going to take a while to catch up, but results from statistical data is still more important to users with the current available data on the net. Right now, Wolfram Alpha is a really cool Internet Calculator.
But this search engine is math based (no surprise) - which calculate 'answers' to your questions - or rather tries to figure out answers to the questions you are querying.
They even posted their infrastructure details - having the 44th largest supercomputer, mirrored at 5 different locations and can handle 175+ million queries a day.
I tried a little more than a couple of queries, most of which returned Wolfram|Alpha isn't sure what to do with your input.
I tried queries based on my hometown, Trivandrum, Kerala which I keep track of on the net from time to time. Given "Where is Trivandrum", it showed the map of India with Trivandrum marked. But it doesn't seem to know anything about kerala - one of the top 50 tourist locations in the world (National Geographic). If this isn't in their current database of 10+ trillion of pieces of data, its going to be hard for me to find information using it.
Then I tried input queries on stuff it could 'definitely' search on - maths. 10000! gave the first bunch of 35660 digits but you can easily extend the number of digits. But even expressions like "5+132!-59^5*sin(32)/5log(12)15" showed results with neat details. Over 20 years in the making of Mathematica, Wolfram would probably have the best number search engine. But "Knowledge" based ? It doesn't seem to have a lot of knowledge as of now.
After a while it just got tired of searching.
Everyone is curious about who or what will be the next Google. There is a big surety among many, even by Tim Berners Lee, that semantic web search will overtake statistical based web search (GYM). Wolfram Alpha is going to take a while to catch up, but results from statistical data is still more important to users with the current available data on the net. Right now, Wolfram Alpha is a really cool Internet Calculator.
1 comments:
absolutely correct.. Semantic web is the next best thing... :). Its gonna be web 3.0 or maybe web 4.0.
Post a Comment