Bing can now answer some yes or no questions - which is incredibly cool. Bing announced that it deployed large neural network models that allows it to answer yes or no questions, expand its intelligent answers globally and better understand complex and ambiguous concepts. This new search feature is live in the United States and will be expanded to more markets soon, Bing said.
Here is a yes or no question in Bing for a search on [can dogs eat chocolate]:
Bing said it synthesize across sources to generate an unambiguous answer to this question.
Bing also expanded its intelligent answers globally to support more regions. For example, "rapa rossa benefici" is "red turnip benefits" in Italian and Bing is using the exact same QnA model it operates in English-speaking markets, which has never seen any labeled data in Italian! It's called a zero-shot approach and it gives Bing the ability to expand this globally. The following example is a query in Italian that translates to “red turnip benefits”. The intelligent answer returned by Bing is generated from a top search result, using the very same universal model that we operate in English-speaking markets.
Bing is also better at understanding query intent to improve search relevance. Specifically Bing created a NLR-based model that is fine-tuned to rate potential search results for a given query, using the same scale as our human judges. This model is able to understand complex and ambiguous concepts much better than its predecessor. The following query is “brewery germany from year 1080”. It turns out there is no known German brewery founded that exact year. However, Bing can assume the user was looking for a very old (millennium-old!) brewery in Germany, even though they may have misremembered or mistyped the year.
Frederic Dubut from Bing wrote on Twitter "Very proud to share some of what we've been working on recently: an improved Bing search experience thanks to Microsoft Turing and large language models!"
Forum discussion at Twitter.
