According to media reports Tuesday,at the International World Wide Web Conference,Goolge has disclosed a cutting-edge image ranking technology “VisualRank” for ranking similar images.

VisualRank is positioned as PageRank for images. Google’s PageRank helps determine a site’s value, based on content and scaled from 0-10. The higher the PageRank, the higher the site appears in organic search listings for related keywords.

Image search today largely relies on analyzing the text near the image, the image’s file name and the words in ALT text associated with them.

Despite decades of effort, image analysis remains a largely unsolved problem in computer science, researchers said. For example, while progress has been made in automatic face detection in images, finding other objects such as mountains or tea pots, which are instantly recognizable to humans, has lagged.

The method in Google’s paper changes that. A group of images retrieved for a query using traditional search methods is then further analyzed. Image recognition software finds which images in the group seem most similar to each other. It then estimates “visual hyperlinks” between them to produce a final ranking.

“We wanted to incorporate all of the stuff that is happening in computer vision and put it in a Web framework,” said Shumeet Baluja, a senior staff researcher at Google.

To develop VisualRank, Google researchers focused on the 2,000 most popular product queries on Google, including iPod, Xbox and Zune. They later determined the top 10 images from its ranking system, gleaned in part from Google Image Search results.

Spread the word: related/bookmark it/readit

At the 2008 O’Reilly Web 2.0 Conference here, Google spam maven Matt Cutts’ session on “What Google knows about “ identified a new threat that is keyword spam on websites and scamp blog comments that harass online communities.

Keyword spam is the use of words, frequently having nothing to do with the site content where they are placed, put into a web page in order for the page creators to get traffic directed to them from search engines. These pages are then used to drive advertising clicks from uncomplicated users or for spreading viruses. Typically these sites contain hundreds of misspelled words to attract users that quickly typed entries in search engines.

This keyword text spamming does not have to be visible, Cutts said. Font and background web page colors can be matched so that they are invisible to browsers, but picked up on by computers and search engines that read the implicit in code. But these tactics can be used for “good” under the inference of search engine optimization. Known as search-engine optimization (SEO), this is technically not spam – it’s the effort to rise above the fold in search results.

Google’s “PageRank” – a key to what makes their search engine so effective – employs an recursive method of trust and reputation to help prevent this type of spam. While this is done though the company’s monitoring cross-promotional links e-Bay and Amazon use a manual user feedback mechanism to let people know that their community members can be trusted for traffic activity. Large sites can increase traffic by adjusting internal links and URL names. Small sites can get more traffic by fan and community cross- linking.

But with the vast number of bloggers today, a second type of spam is much more widespread: “Comment Spam”

Cutts offered these tips for society developers to eliminate spam in their platforms:

• Be less of a target with hosted solutions
• Build reputation and trust into the service
• Make the spammers send money, time or effort - in other words, scotch them

You can prevent spam attacks by using a CAPTCHA, or obscure series of words which most users find very difficult to read, Cutts said. A better idea, such as a simple math problem like what’s 3+5 is a preferred solution, he said.

Many platforms like Google’s hosted service Blogger can be made to require a valid email address or Google login for users to comment. With the vast number of GMail, Yahoo, or Hotmail users out there, hosting your blog on one of these platforms will make commenting easier and more prevailing. As these platforms are hosted services, they are less likely to have rogue code attacks on the server. As spam is getting increasingly dangerous with script attacks, if you run your own blog make sure that your operating system, blog software and database is frequently updated.

Google has created a resource for webmasters at its Web site that will notify you if your site is being spammed or taken over by villain commenter’s and provide education on how to prevent this in the future. It also is a home base for statistics and shows you how users are coming onto your site from their engine.

Spread the word: related/bookmark it/readit

The act of purposely clicking on ads (either banner ads or paid text links) on pay-per-click programs with no interest of purchasing the product. If ads are based on click-through (pay-per-click), the Web site publishing the ads and clicking the ads countless times can make a dishonest profit. This can be done manually, by automated tools, robots, or other deceptive software.

Google’s Process of Detecting Invalid Clicks:

Google is dedicated to a number of resources to protect account against invalid activity. Google examine each click on an AdWords ad by looking IP address, the time of the click, duplicate clicks and various other click patterns to isolate and filter out invalid clicks. This detection and filtering occurs over a number of levels including Real-time systems filter out activity fitting a profile of invalid behavior (such as excessively repetitive clicks) and Clicks and impressions from known sources of invalid activity are automatically discarded.

Various unique and innovative methods are applied at each stage of the filtering process, to maximizing proactive detection of invalid activity. Google engineers are also constantly improving monitoring technology, enhancing filters, and examining a growing set of signals.

Google also has a team that uses specialized tools and techniques to examine invalid clicks. When system detects potentially invalid clicks, a member of this team examines the affected account to pick the important data about the source of the potentially invalid clicks.

Google team make invalid activity very difficult and unrewarding for unethical users, thereby decreasing their chance of success. If Google found that invalid clicks have been charged in the past two months, it credit advertisers’ accounts.

To view credits for invalid clicks:

• Sign in to your AdWords account at https://adwords.google.com
• Select the My Account tab
• Click Billing Summary
• Click Advertising costs and adjustments for a particular month. Any invalid click credit you’ve received will be labeled Adjustment - Click Quality.

Spread the word: related/bookmark it/readit