After Google released the “Panda” update there has been some discussion about advertisements on the Internet. Many think the Panda update will look at how much advertising a site has and where the advertising is located to help determine the quality of a site. The update is meant to reduce content farms that offer little content on their site that is mostly drawn from other websites without permission and the goal is to get people to click on the ads.
Microsoft has filed a patent application that explores the differences between images in advertising and the images that aren’t. Microsoft points out that there are ways for the search engine to tell the difference between ads in images and those that aren’t. Many believe that Google will be doing the same thing and making it easier to determine if an image is in an ad. Obviously this is important because Google would not want ads showing up in their image results.
The four things that search engines would look at are text features, link features, visual layout features, and content features.
Text features provide an indication that the picture is an advertisement due to words like “pop up ad” or “advertisement.”
The search engine will look at what links are associated with the image. Does the link go to the same website or another page on the same site as the image? If it goes to a different website or server it may be an indication of an ad.
The layout refers to how the image is positioned on the page. There are many content features that can help determine if the image is part of an ad, including image format, whether it is a photograph or graphic, the size, number of colors in the image, gray area of the image, and whether it has high contrast. Ads usually use graphics, have a bigger size to get noticed, use fewer colors, have more gray area, and have a sharp contrast.