Monday, May 18, 2009

WolframAlpha Vs Google

There is sudden buzz about WolframAlpha being a "Google killer" and I was quite surprised how that can happen when Google has been around for so long and it has vast amount of data already indexed.


Then, it turns out that WolframAlpha is a computational engine which targets at delivering the numbers in a meaningful format. 

I searched for "distance from New York to New Jersey" on WolframAlpha and it gave me the exact distance in miles (52.61 miles) and it also gave me results in meters, kilometers, nautical miles (??) and centimeters (??????). On top of it, it gave me direct travel time taken by flight, sound, light in a fibre and light in vacuum. Now most of the data that it returned are something that I do not care about. When I search the same on Google, I get links to distance calculators. 

After going through the examples, I could gather that, WolframAlpha in its current form is a great resource for research analysts who play with data. It is not that Google cannot provide this data, but WolframAlpha makes this data understandable and much more accessible 

Search for "India literacy" on Google and you will get 61% as the first result (without even digging the link) and this was present before WolframAlpha could do it with it computation engine based on Mathematica.

Is WolframAlpha a semantic search? I don't agree entirely. The structured queries as per their examples are fine, but here is why I am saying it is not entirely semantic.

I did see some weird responses as well. I searched for "voting age US" and it did not give me any result, whereas "voting US" gave me the result as "18 years of age; universal".

I searched for "US viting" and WolframAlpha interpreted it as "US Vitina" and showed me the distance from the US to Vitina in Bosnia and Herzegovina. Google prompted me and asked "Did you mean: US voting".

WolframAlpha has taken online computing to a different level, but not yet usable by common internet user who is used to Google. They are working on a lot of stuff right now and may soon come up with something far more robust, but will it challenge Google at any point? No I don't think so, just because the purpose of WolframAlpha is completely different from what Google is designed for. 

Again, Google Squared is in the news now and is being termed as "WolframAlpha killer'.

Note to Wolfram: ... I am not looking for distance from New York to New Jersey IN CENTIMETERS OR NAUTICAL MILES, and I am definitely not interested in how soon light will reach NJ from NY in vacuum. 


Wednesday, May 13, 2009

Advent Of Real-Time Search - Checking Scoopler

Scoopler Social Media Search
Real-time web is the next big thing and there are quite a few real-time search services that are creating buzz around. Tweetmeme, OneRiot, Twitscoop, Scoopler etc. Most of these like Tweetmeme and Twitscoop focus around Twitter updates, but OneRiot and Scoopler takes it further to include other services like Digg. Scoopler, which I would really call a "social media search engine" covers Flickr, Youtube, Delicious and other services.

Scoopler keeps it simple and structures the information very effectively. It shows tweets in real-time along with the popular videos, links and images from Youtube, Flickr, Digg etc. Along with it, it also stores your last three searches thereby making it easier to switch between your results.

While your at discovering information through the search, Scoopler makes it easy by showing the entire headline and instead of going to the URL, you can get a "peek" of the content in the target URL and "share" through a host of other services.

Scoopler Peek

More than real-time web I am inclined towards onDemand web as written in the post by @rickmans. With that in mind, Scoopler comes up with too many tweets if you search for a hot topic (Eg: Adam Lambert while American Idol is being aired) and keeps pushing information onto your screen. There is no easy way, time or provision to make sense of the flowing information especially when it is a hot keyword. A "pause" button is critical for Scoopler (or for any real-time service), like in a Tweetmeme or a oneRiot.

Its not eye catchy from a look and feel perspective and I think a lot can be done to improve it.

But, I like Scoopler more than the others right now available.

... and before I finish, Scoopler uses a less known Cloudant, a distributed database service firm (a new type of database) funded by Y Combinator.