Print 13 comment(s) - last by mindless1.. on Apr 23 at 1:10 AM

A peek at Similar Images, a new search engine from Google
Google continues to improve how users scrounge the web for information, searching smarter, not harder

When Google came along, it redefined the search experience to the point where its name has become a colloquial verb ("to Google"), meaning to search.  Even though it has since diversified greatly, from applications to hardware, it remains committed to continuing to lead the search world.

While competitors like Microsoft hone their sharpest new products, Google is equally hard at work and has just released two fruits of its recent labor: new searches for images and news.

The first is a site called Similar Images, which eschews actual image recognition like Apple's recent iPhoto (which can recognize faces) in favor of a user refined search, driven by metadata.  What does this mean?  If searching for a picture related to Apple Inc., a user might type in Apple and see pictures of apples (fruit), Apple computers, Apple corp. headquarters, Apple Corp. (the Beatles record label), etc.  By clicking on one of these results and searching for similar images, the search can be refined. 

The approach arguably works as good or better than graphical image recognition, while being far less computationally intensive.  The downside is that someone has to type in the metadata, which a home user is unlikely to do (so Apple's software still makes sense), but which an internet author is likely to do (as metadata already drives page searches).

The second new service is called News Timeline.  This search engine provides refined time searches for news only (crude time-specific searches are already in Google's base search engine, but often return inaccurate results for when the page was published).  The result is that you can specify a time period of a specific day, week, month, year, or decade, and get news results tagged with specific keywords from that time period.

This search in particular should be a godsend for tired college students looking to get their Sociology or History reports done.  To bring the Apple theme full circle, the new search was actually designed by Andy Hertzfeld, a former member of the original Apple Macintosh design team, lured away by Google.  He pointed out at a special news conference that the search not only encompasses web articles, but scanned books, print newspapers, and magazines

Some magazines such as Popular Science have given Google special permission to show old articles; for others it will take you to the publication's web page and provide you with information on what issue to look in.

Both of the new search engines are part of Google Labs, Google's research and development efforts.  Google Labs recently got a shiny new webpage.  Describes
R.J. Pittman, director of product management at Google, "We actually gutted it and rebuilt it from the ground up."

Comments     Threshold

This article is over a month old, voting and posting comments is disabled

Simply incredible...
By quiksilvr on 4/21/2009 9:40:50 PM , Rating: 2
When I first saw this:

I thought to myself, "What more could they possibly do at this point?"

And once again I'm amazed. Google really will take over the world.

RE: Simply incredible...
By Chiisuchianu on 4/21/2009 11:17:49 PM , Rating: 1
Let's hope so.

RE: Simply incredible...
By mindless1 on 4/22/2009 1:00:25 AM , Rating: 2
... or just amass a large pile of money being first to implement what others will eventually, kinda like MS did.

Let's hope Google doesn't get much bigger, competition is good for the consumer. I prefer and use Goggle far more often than any other search engine but features must at some point be competition to meet consumer demand, it is in our interests to have this state of industry.

RE: Simply incredible...
By nafhan on 4/22/2009 8:41:34 AM , Rating: 4
When Google starts to push others out of the marketplace largely through litigation and/or giving money to the right people (i.e. bribery), that's when I'll have a problem with Google.
Right now, they are growing by innovating. As long as they keep doing that, more power (and growth) to them!

RE: Simply incredible...
By mindless1 on 4/23/2009 1:10:14 AM , Rating: 2
Regardless of which way others are forced out of the market, the effect can still be bad.

Of course they are growing some by innovation, some by sheer word of mouth popularity, but some by squeezing other companies dry with too few advertisers, which prevents competitive innovation from competitors and reduces Google's need to innovate.

Right now, Google is lacking enough competition to even feel the need to take any of their services out of beta stage or give users enough assurance the service will be around long term or remain static enough to rely on for business purposes.

Google is offering great services but at the same time it is more like a huge science experiment. The growth was good but at some point we have to look beyond size. Take GM for example.

RE: Simply incredible...
By mmcdonalataocdotgov on 4/22/2009 11:23:53 AM , Rating: 2
Competition? I am not spending any dollars on Google.

Ad revenue is what drives Google, and as long as they can get advertisers, they will succeed. That is where the competition is for search engines. I don't think they care about the user beyond what makes us use their site rather than others.

RE: Simply incredible...
By KCjoker on 4/22/2009 6:47:34 PM , Rating: 2
No one spends money on browsers either but MS gets in trouble with IE all the time.

RE: Simply incredible...
By mindless1 on 4/23/2009 1:06:12 AM , Rating: 2
Competition for features users desire is what causes rise and fall of userbase size, which effects their ad revenue.

It's all tied together, regardless of whether the money comes out of your pocket or not.

RE: Simply incredible...
By 67STANG on 4/22/2009 2:58:02 AM , Rating: 2
What's so brilliant about this? It's just image tagging (done for free by google users by the way), and it's not too hard to do.... they are just applying a second layer of data and associating it to their indexed data... I have been doing this in .Net for at least 2 years now. Microsoft can do this too-- if they had anyone with ideas working for them.

RE: Simply incredible...
By corduroygt on 4/22/2009 11:25:56 AM , Rating: 2
if it's done by the users, what's stopping anyone from mislabeling pictures? How is it moderated?

RE: Simply incredible...
By 67STANG on 4/22/2009 12:41:07 PM , Rating: 2
Because Google has each picture tagged by at least 5-10 different individuals. Non-uniform results are dropped.

For example if you had an image of a Steve Jobs and 9 out of 10 people tagged it as "CEO of Apple", "Steve Jobs", etc. but one person tagged him as "Snake-oil Salesman", "Crook", etc. Google drops the minority tagging. A fully automated process, and again-- not difficult.

What's in it for me?
By jskirwin on 4/22/2009 9:38:50 AM , Rating: 2
The downside is that someone has to type in the metadata, which a home user is unlikely to do (so Apple's software still makes sense), but which an internet author is likely to do (as metadata already drives page searches).

No, metadata already boosts bandwidth usage. If I make it easier to find my online images, I'll have to buy more bandwidth. It won't translate into increased traffic which could eventually (magically?) become cash. That's why many bloggers block others from linking to their images.

Turn on your filters again
By rdeegvainl on 4/22/2009 12:19:03 PM , Rating: 2
So now when people search teabagging rednecks, they are in for a real surprise.

“And I don't know why [Apple is] acting like it’s superior. I don't even get it. What are they trying to say?” -- Bill Gates on the Mac ads

Copyright 2016 DailyTech LLC. - RSS Feed | Advertise | About Us | Ethics | FAQ | Terms, Conditions & Privacy Information | Kristopher Kubicki