Google makes fundamental changes by introducing Bert

Google, the world’s most popular search engine,  is adopting artificial intelligence in its ranking algorithm. The change which Google explains as the most significant revision in at least five year, uses a new form of language analysis to understand users’ queries better and will affect its responses to one in `10 searches , according to Pandu Nayak, Vice-President of search.

Google had already began updating its software to make change for English language searches and will eventually apply it to other languages as well.

Till now Google’s algorithm has tried to single out the most important words in any search query, ignoring smaller or common words that seem less significant. This enables it to zero in on the main subject, but often results in it misunderstanding a precise request.

The new technique known as Bert, relies on language model built up from analysis of vast amounts of text online.  It no longer read a string of search words sequentially, but analyses all at the same time  – including smaller words that would have been ignored before.

Google gave the types of question it could handle  “How old was Taylor Swift  when Kanye jumped on the stage” –points to more  couples queries that have been beyond its reach before.

Google was founded in 1998 periodic updates to its ranking algorithm have sometimes resulted in big changes in the amount of traffic it directs to outside websites, angering companies whose business depend on the search engine. But Google predicted the latest revision would not have a large impact on online traffic. The update marks the first application of piece of research a natural language processing from last year that has drawn considerable attention in AI given the fluidity of language depending on the context and the person using it.

According to Google the Bert technique would return more useful responses to many queries. Google also said the Bert technique would return more useful responses to many queries and in some instances the new algorithm had produced worse results than before.  Jeff Dean, head of AI, said “ I don’t think we’re anywhere close to solving language, but this is a good step.”