In December last, Microsoft launched a new Artificial Intelligence (AI) powered Intelligent Search feature for its search engine Bing, that goes beyond traditional Q&A style of search and let you interact more naturally and provides multiple perspective to an answer.
Traditionally search results carries many blue links and sometimes one link and a snippet of text highlighted at the top. But, AI-powered intelligent answers goes beyond by showing musch faster two points of view for the same quesion, expalins bing.
“There are many questions where getting just one point of view is not sufficient, convenient or comprehensive,” wrote Bing’s Mir Rosenberg. This is where Multi-perspective answers, which is another format of the Intelligent Answers feature comes into the picture that uses artificial intelligence to provide two different and often opposing perspectives on a topic.
In addition to intelligent answers, also introduced was “intelligent image search” in which Bing suggests query refinements for broad searches.
Here are some examples provided by Rosenberg of Bing multi-perspective answers cards on bothe desktop and mobile (cards shown as vertically).
How Does Multi-perspective Answers Work?
Rosenberg explains that it uses several learning models to “prioritize reputable content from authoritative, high-quality websites that are relevant to the subject in question, have easily discoverable content and minimal to no distractions on the site.”
When a question is asked:
1. Passage candidates from web pages are selected from Bing’s Web Search and Question Answering engine.
2. Then, candidates are organized in clusters to determine similarity and sentiment using deep recurrent neural network (Deep RNN) models.
3. Finally, Bing ranks the most relevant passages from the web pages from each sentiment based cluster analysis and other features and serve in the multi-perspective answer cards.
The screen shot below summarizes above process:
The article also offers clues about the signals it looks to rank sites for intelligent snippets, including:
- Authoritative and high quality
- Relevant to the topic
- Content is easy to crawl and index
- Good user experience on the web page
These intelligent answer are currently limited to a few cards in the U.S. However, Bing plans to expand it to many more questions with increased coverage. Soon, these results will be coming to the United Kingdom as well.
Following Bing, Google is also exploring to add Sentiment Analysis to featured snippets, as it currently exploring to show multiple responses to a query. Recently Google’s Danny Sullivan has confirmed it by stating, “We’re exploring solutions to this challenge, including showing multiple responses.”
“There are often legitimate diverse perspectives offered by publishers, and we want to provide users visibility and access into those perspectives from multiple sources,” Matthew Gray, a software engineer at Google who leads the featured snippets team, explained.
In addition, Bing also recently extended their Speech API, a Microsoft Cognitive Service, to now offer six new TTS languages powered by latest AI technology to all developers, bringing the total number of available languages to 34.
Additionally, by the end of February, these new TTS languages will also become available through Microsoft Translator Speech API and Microsoft Translator apps, making these new languages text-to-speech output available for developers integrating the Translator speech API as well as end-users of the Microsoft Translator apps and Translator live feature.