Bing Snapshot Adds More Entities, More Relationships

by Aaron Bradley on March 31, 2014

in Search Engines, Semantic Web

Bing Snapshot Adds More Entities, More Relationships

Bing today announced a number of additions to its Snapshot product.

Snapshot is Bing's Satori-powered version of Google's Knowledge Graph and, like the Knowledge Graph, information from this knowledge base typically appears in a panel to the right of the main search results, or in a carousel above them.

150 million more entities, with goodies

Bing reports that over the past few months they've added (or, in their words, "surfaced") "an additional 150 million entities encompassing web actives [sic], doctors, dentists, lawyers, and individual real estate properties."

This expansion in itself is fairly unremarkable, and even the real estate listings that now may be included directly in Snapshot results are hardly earth-shattering (the example is a Zillow listing, and marrying a street address to details about that address at a trusted URI is hardly rocket science).

The real estate listings are more of a segue to the way Bing is now framing the challenge to the Knowledge Graph that they're explicitly making with this latest round of Snapshot improvements.

And unlike other search providers, we let you do something with that information. We strive to provide more insights about entities to our users. For example, if you search for Winston Churchill, we show you comprehensive details about him such as birth and death dates, height and pivotal moments in his life and can connect you to his written work. But, we also let you play audio clips of his most famous speeches directly from the search results page. Similarly when you search for Princeton University, we provide online courses so you can take classes online.

These are indeed useful insights, and these changes appear a bit more substantive than many earlier Bing updates (especially those which have attempted to gain ground on Google by increasing the amount of information added to personalized search results that originates from a user's social networks).

I've long thought that associating educational organizations with their course listings was an opportunity for search engines, and here Bing appears to have beaten Google to the punch. A brief survey of Bing results indicates that at least the bulk of online course offerings are coming from Coursea: whether or not Bing is getting the data by parsing the Coursea website or through a partnership with them isn't known, but I'd agree with Bing that the inclusion of these results makes these college and university results more useful. You can judge for yourself.

university of british columbia - Bing Search Results, US IP, 31 March 2014

Bing search results for "university of british columbia" (US IP, 31 March 2014)
Click to enlarge

university of british columbia - Search Results, US IP, 31 March 2014 search results for "university of british columbia" (US IP, 31 March 2014)
Click to enlarge

There's some irony in this for search marketers who have been looking to Bing as an alternative to that evil, traffic-stealing, site-scraping search engine Google, and the propensity of its Knowledge Graph to bypass websites whenever possible to present answers directly in the search results: Bing appears headed down exactly the same pathway.

150 million more entities, with relationships

Bing is also highlighting its ability, via Snapshot, to better connect the relationships between entities.

While comprehensiveness is important, we are not only interested in the items themselves, but the relationships between them, helping you understand more about the world and take action more quickly. To help us better connect intent to actions, we have taken an open approach that takes advantage of rich data sources and services across the web, such as Wikipedia, LinkedIn, IMDB, and Netflix.

While it's great to see Bing making more of these connections between things, I don't think this represents a substantial difference from Google results for similar entities – and may even be a case of Bing playing catch up to a certain degree.

mission impossible ghost protocol - Bing and Google Search Results

Bing (left) and Google (right) search results for "mission impossible ghost protocol"
Click to enlarge

As one might expect when both Bing and Google are drawing on the same data sources, there's not a lot of difference between the two results except when it comes to which details each engine chooses to display.

On a personal note I can't help but take a moment to note with some satisfaction the similarity between Bing's message and, indeed, language to what I've previously said about semantic SEO – that is, that the true power of semantic web technologies for search engines isn't the simple identification of entities, but what you can do once you've identified them.

Semantic SEO is not about optimizing for strings, or for things, but for the connections between things.

Semantic SEO is optimizing for relationships.

The relationships between entities facilitated by the ability to uniquely and unambiguously identify them, and to provide unambiguous data about them.

Interestingly, while the Bing "Scroogled" campaign and the "Bing It On" challenge have been the highest profile efforts by Microsoft of late to take a bite out of Google's market share, from a product development point of view it is really in the semantic arena where Bing has been consistently presenting the biggest – and I think most credible – challenge to Google.

And search marketers familiar with Hummingbird and Google Now will see many parallels with these Google initiatives in the penultimate paragraph of the Bing announcement.

With advancements in Natural User Interface experiences such as gesture or voice, we believe that there’s a world just around the corner where Bing can be more helpful by proactively reminding you to pick up dinner on the way home, or helping you plan a vacation while keeping the weather, flight prices, and exchange rates all in mind.

Whether or not Bing is successful in gaining search engine market share as a result of these efforts, I think we're all benefiting from the pressure that Bing Snapshot is continually exerting on the Google Knowledge Graph team to step up their game. And I think Snapshot is one of the few areas where Bing, if only temporarily, can be reasonably credited with sometimes doing a better job than Google – as with these online course results.

Do elements like these online course listings appear in many Bing Snapshot results or in many Bing Snapshots? Microsoft has a history of unfortunate product names when it comes to search, and "Snapshot" is no exception.

The Bing Index and Knowledge Team apparently looks after the Knowledge Repository

I think when one neglects adding a "Snapshot" tag and instead makes reference to terms reminiscent of those employed by "other search providers" it might be time to consider a name change.

1 Luca Bove April 1, 2014 at 12:36 am

Only in USA? Or Worldwide?

2 Aaron Bradley May 8, 2014 at 4:48 pm

At time of writing it was only in the USA, but I’ve waited so long to reply that it now appears to be worlwide! Or at least I’m now seeing the same UBC result from Canada that I originally conducted from a US IP for the article.

Previous post:

Next post: