Bing ‘intelligent search’ capabilities continue to expand, include facts from multiple sources
Bing also shares more about the platform it built to power intelligent search.Bing has announced several upgrades to its AI-powered intelligent search capabilities. Bing first launched intelligent search in December, bringing artificial intelligence to deliver richer search answers, and enhance image and conversational search.
Now, Bing is launching more intelligent search features with:
Aggregated facts across multiple sources.
Hover-over definitions for uncommon words.
Multiple answers for how-to questions.
Object detection for all common top fashion categories.
Using Intel’s FPGA chips allows Bing to quickly read and analyze billions of documents across the entire web and provide the best answer to your question in less than a fraction of a second.
When searching for facts, Bing may show information from multiple sources, all in one featured snippet. Here is a screen shot:
Also, if Bing recognizes a word in a featured snippet that isn’t common knowledge, it will show its definition when the user hovers over it with the mouse cursor:
Bing said in the next few weeks it will start showing multiple answers for how-to questions. “We found that having multiple answers is especially helpful in situations where users struggle to write a specific enough query, such as when they have DIY questions but may not know the right words to ask,” the team wrote in the announcement. Bing has not shared a screen shot of how this will look.
In December, Bing’s intelligent image search launched with clickable hotspots on objects within images, the focus was on a few fashion items like shirts and handbags. Now, Bing said it has expanded object detection to cover all common top fashion categories.
Bing also disclosed that these features, which require huge amounts of computational power, are powered by a platform that uses Intel chips. Bing said, “we built [intelligent search] on a deep learning acceleration platform, called Project Brainwave, which runs deep neural networks on Intel® Arria® and Stratix® Field Programmable Gate Arrays (FPGAs) on the order of milliseconds.”