VisualRank
:by Google
VisualRank
Some Google researchers at a web conference in Beijing announced they are currently working on some kind of PageRank specifically aimed at images indexed by Google. They have named it VisualRank. At first the technology was only applied to a smaller test set of images, because applying it to all images would make it too computing-intensive. According to the New York Times yesterday, visual rank is an algorithm “for blending image-recognition software methods with techniques for weighting and ranking images that look most similar,†and in Google’s internal scoring tests it achieved far higher quality results.
If I understand the gist of the research paper (Download the PDF format) right, then it looks like the core of Google’s VisualRank algorithm now consists of not only looking at textual cues in regards to images, but also image content itself. After identifying the most authoritative set of picture candidates for a given keyword, Google then tries to improve the ranking of images found to be sharing the most visual characteristics with the group at large, by creating a similarity network. Center node images or those images containing large resolution versions would then determined to be the most relevant. In 1000 sample queries – taken from the top Google Product search queries – 762 VisualRank results were tested to be more relevant than Google’s old approach, with 202 equal quality results and only 70 results that were worse.
