Yandex has released a new search ranking algorithm to better understand user intent. The algorithm has been named Korolyov and, according to Yandex, is a huge improvement on the company's preceding algorithm, Palekh. Korolyov is neural network-based and matches the meaning of a search query to the entire contents of a page, as opposed to just the headline as had been the case in the past. Additionally, the algorithm is being applied to 200,000 of the top relevant pages per search, compared to just the 150 pages that Palekh was applied to. Users are seeing benefits in the way it handles long-tail queries, and how the search engine interprets the intent behind the search also allows it to understand searches with missing information. Using artificial intelligence, the algorithm is able to improve its function further as more people use it.Â
