Posted by: creativerebel
Through our Powerset division, Bing has access to semantically indexed Wikipedia content to help with “reference” searches when people are looking for information about a given entity (person, place or thing) such as Albert Einstein or TechCrunch. According to our research, reference search sessions are typically the most time- and query-intensive, averaging more than 9 minutes in length and more than 6 queries per session before someone finds the information they need.
Bing is here to help with what we call our Reference vertical. The easiest way to find the Reference vertical within Bing is to look for the Wikipedia result in Bing’s algorithmic listings. Each one of those results has an “Enhanced View” link (see below) that will take you to the that entity’s Wikipedia page, within the Bing Reference vertical.
Some queries will trigger a “Reference” Quick Tab in the Explore Pane. Click “Reference” and you will land on the best Wikipedia page for that query. Other queries that trigger the Reference vertical include Facebook, Sikhism, and squirrel monkey, to name a few.
Sometimes, a web query will trigger an instant answer that includes the answer to your question along with more information about the topic. From here, you can click on the image, or “More” to be taken to the Reference vertical within Bing.
This vertical provides a richer version of Wikipedia, with a medley of semantically enhanced features to help you scan, read and find exactly what you’re looking for. Plus this enriched version of Wikipedia can be browsed without leaving Bing, and with the rest of the web one click away.
You can use the Article Outline on the right to quickly scan the article. Click on a header to go to the corresponding part of the page. Click “pin” to keep the outline stationary as you scroll through the article and click “top” to navigate back to the top of the page.
Another task people often encounter when in research mode is looking for images. A great feature of the reference vertical is the ability to easily scan images from within the Wikipedia article itself or images from Bing Image Search on that same topic using the Images tab at the top of the article.
Searching within the Reference vertical will bring back semantically highlighted Wikipedia results, based on Powerset technology. Instead of just searching keywords, you can ask full questions. So, you don’t have to ask “earthquakes san Francisco,” just type in the information you want: “When did earthquakes hit San Francisco.” Notice a couple of advantages in the below screenshot. We’re often able to highlight the exact answer to your question, in this case in the form of a date, because we understand the sentence. Also, note that we understand that “strike” and “hit” mean the same thing, so we’ll return a result that you’d otherwise miss. We try to match the meaning of your query to the meaning of sentence in the reference vertical, instead of just matching the words. Finally, we also respect the grammar of your question. For example, the questions “who did Texaco acquire” and “who acquired Texaco,” are almost the same in a traditional search engine, since they contain the same words. We understand that these are grammatically (or “syntactically,” as we like to say in linguistics) very different.
The team at Powerset is proud of all the semantic contributions we’ve made in Bing and we’re excited to expand the features for exploring each article, and to attract more and more Bing users who are looking for reference material. Be sure to check out the Powerset Blog (part of the Bing community site) for updates and follow Powerset on Twitter for the latest and greatest from the Powerset team of Bing.
-Mark Johnson, Senior Program Manager, Powerset division of Bing