Google Goes It Again

Google Inc.
Image via Wikipedia

More interesting news on the search front as Google goodies continue to trickle forth, this time in the form of what’s-coming-down-the-road-for-the-search-giant tips. I read about this on ReadWriteWeb today: Google is focusing attention on new ways to broaden and narrow search, all at the same time. When I say broaden, I am talking about enabling the translation of search queries into different languages so that all of the world’s websites and databases may be tapped. More on the translator gadget can be found here. When I say narrow, I mean tightening  search results to make them more personalized, individualized and social from your subjective perspective. What this really means is tapping into your friend’s content to focus results on sources you trust the most. More on the launching of Google Social Search here. Finally, Google is striving to make your search experience more intuitive: like a benevolent super-computer talking in soothing Hal-like tones, Google aims to be there when you open your browser, presenting you with information you didn’t even yet realize you were interested in. The beginnings of this trick are at play in the recently-announced Google mobile search product, but apparently are expected to extend much further. Telepathy, anyone? Whether it scares or excites, there is no question that Google is on a roll.

Reblog this post [with Zemanta]

Alpha’s Not Beta (Anymore)

Have you been hankering to try Wolfram Alpha’s new computational, knowledge-based search engine? Are you wondering if it is the next best thing in search? Well, I am playing with it right now, in another tab, and I have my own impressions.

For those who are still unsure of what Wolfram Alpha is, you can get the long, broad and sweeping answer at their “about” page. The short answer is that it is a search engine that “computes” answers for you by accessing a knowledge base treated or overlayed with algorithms / coding that Alpha can easily “read”. Or, if you ask Alpha this question “wolfram alpha”, you get this:

wolfram alpha

If you click on the “use as an internet domain link” on the above page, you get this:

Wolfram Alpha domain

I am fascinated with this new form of search interaction. But it is not the quick and dirty search service that Google offers and any comparison should end with the fact that they are both described as “search engines.” First, it takes some time to load the search page, much longer than it takes to load a Google page. Then it takes much longer to get an answer to your search inquiry, but that is expected. Alpha is not merely responding to your “query” with a list of pages ranked by hits. It is actually computing your answer by dipping into its information resources and providing a calculated response. The plus side is that you can be assured that your answer is most likely correct (unlike Google where the only assurance is that your answer has been hit by tens of thousands of other searchers).

The downside is that you have to consider whether Alpha is the best source for the particular information you are looking for and then you have to be very careful how you phrase your query. Take a look at Alpha’s search tips box:

Wolfram Alpha Help Tips

Alpha is great for quantitative questions, but not so good for qualitative questions. And even for quantitative questions, you have to be careful how you pose them. Talking to Alpha is much like talking to Spock – you will be treated to literal logic that is responsive to your request but may not be what you were looking for.

For example, my first search query was “distance from earth to sun.” Alpha got hung up on the words “distance” and “sun.” It defaulted on distance to astronomical units (AU) – how was Alpha to know that I meant miles? It did default to our Sun, but not before noting that it could not be sure that our Sun is what I meant. Fair enough.

I then asked it for “miles from earth to sun.” No answer. I then entered “miles from earth to sol.” No answer. I then entered “earth to sol” and, voila, I got astronomical units, kilometers, meters and miles. I also got a “corresponding quantity” of 4.5 micropascals, with an equation for how to get this amount.

When I finally got my answer, I got everything I was looking for and more. No question that my answer was accurate and scientifically solid. But getting there is more than half the battle and it helps to throw any notions about how to search in the more traditional formats right out the window.

Wolfram Alpha now has its place in my search arsenal – I added it to my Firefox search bar tool. But I will not be turning to Alpha for answers on every query. No doubt Alpha will evolve and will become more powerful as more information is made accessible to it. Give it a try yourself – I would love to hear about your impressions and results!

Reblog this post [with Zemanta]

Alpha's Not Beta (Anymore)

Have you been hankering to try Wolfram Alpha’s new computational, knowledge-based search engine? Are you wondering if it is the next best thing in search? Well, I am playing with it right now, in another tab, and I have my own impressions.

For those who are still unsure of what Wolfram Alpha is, you can get the long, broad and sweeping answer at their “about” page. The short answer is that it is a search engine that “computes” answers for you by accessing a knowledge base treated or overlayed with algorithms / coding that Alpha can easily “read”. Or, if you ask Alpha this question “wolfram alpha”, you get this:

wolfram alpha

If you click on the “use as an internet domain link” on the above page, you get this:

Wolfram Alpha domain

I am fascinated with this new form of search interaction. But it is not the quick and dirty search service that Google offers and any comparison should end with the fact that they are both described as “search engines.” First, it takes some time to load the search page, much longer than it takes to load a Google page. Then it takes much longer to get an answer to your search inquiry, but that is expected. Alpha is not merely responding to your “query” with a list of pages ranked by hits. It is actually computing your answer by dipping into its information resources and providing a calculated response. The plus side is that you can be assured that your answer is most likely correct (unlike Google where the only assurance is that your answer has been hit by tens of thousands of other searchers).

The downside is that you have to consider whether Alpha is the best source for the particular information you are looking for and then you have to be very careful how you phrase your query. Take a look at Alpha’s search tips box:

Wolfram Alpha Help Tips

Alpha is great for quantitative questions, but not so good for qualitative questions. And even for quantitative questions, you have to be careful how you pose them. Talking to Alpha is much like talking to Spock – you will be treated to literal logic that is responsive to your request but may not be what you were looking for.

For example, my first search query was “distance from earth to sun.” Alpha got hung up on the words “distance” and “sun.” It defaulted on distance to astronomical units (AU) – how was Alpha to know that I meant miles? It did default to our Sun, but not before noting that it could not be sure that our Sun is what I meant. Fair enough.

I then asked it for “miles from earth to sun.” No answer. I then entered “miles from earth to sol.” No answer. I then entered “earth to sol” and, voila, I got astronomical units, kilometers, meters and miles. I also got a “corresponding quantity” of 4.5 micropascals, with an equation for how to get this amount.

When I finally got my answer, I got everything I was looking for and more. No question that my answer was accurate and scientifically solid. But getting there is more than half the battle and it helps to throw any notions about how to search in the more traditional formats right out the window.

Wolfram Alpha now has its place in my search arsenal – I added it to my Firefox search bar tool. But I will not be turning to Alpha for answers on every query. No doubt Alpha will evolve and will become more powerful as more information is made accessible to it. Give it a try yourself – I would love to hear about your impressions and results!

Reblog this post [with Zemanta]