The Google Panda process

Google Panda was built through an algorithm update that used artificial intelligence in a more sophisticated and scalable way than previously possible.

Human quality testers rated thousands of websites based on measures of quality, including design, trustworthiness, speed and whether or not they would return to the website.

Google’s new Panda machine-learning algorithm, made possible by and named after engineer Navneet Panda, was then used to look for similarities between websites people found to be high quality and low quality.

Many new ranking factors have been introduced to the Google algorithm as a result, while older ranking factors like PageRank have been downgraded in importance. Google Panda is updated from time to time and the algorithm is run by Google on regular basis.

Source. wikipedia