Apple May Evolutionize Map Searches With Crowd Sourcing


Apple Patents

A new patent application shows that Apple may be planning on rewriting the rules for ranking map search results through the use of anonymous crowd sourcing techniques. The patent application, titled, “Relevancy Ranking for Map-Related Search,” was filed in March of 2010, and it describes a method of ranking search results for, say, a restaurant according to how many customers that restaurant gets.


The Method

In the patent application, which was first noticed by AppleInsider, Apple makes the case directly for why its method is deserving of a patent by calling all of the existing methods for ranking search results flawed.

“Each [of the existing methods for ranking search results] has certain limitations associated with it,” the company wrote. “Search results ordered by proximity do not account for quality of the search result relative to the query. Search results ordered by average-user-ranking are based upon opinions of relatively few people whom take the time to review the location. Search results that are ordered based on advertising dollars also do not take into account quality or desirability and sometimes broaden the criteria for relevance beyond a desirable measure.”

Apple argues that it can rank search results better by collecting anonymous user location data, and then collecting that data into what it called a “location popularity index.”

“The location popularity index can be used to rank search results by making various assumptions based on the data,” the company wrote. “For example, it can be assumed that a person who visited a restaurant for over an hour ate at the restaurant. Based on these and other assumptions, search results can be ranked according to the locations having the most visitors. Accordingly, in a search for restaurants, the location that had the most number of visitors can be ranked the highest.”

Apple also said it could then use this data to recommend other established based correlating where these anonymous users go. If two people eat at the same restaurant, they must have overlapping tastes in restaurants. Add in more users to the mix and Apple can make sophisticated recommendations based on those overlapping tastes.

While the application doesn’t say so, this part of the company’s method sounds similar to the way online music service Pandora makes recommendations for new music.

Apple Patent Application Figure

A figure from Apple’s patent application showing the flow chart for its method


As noted above, this application was filed in March of 2010, long before concerns over the way that Apple, Google, and other mobile platform providers were collecting user location data. Even still, the application plainly spells out that the method involved is based on collecting anonymous data that can not be attached to the actual user.

“Importantly,” Apple wrote, “no user information is sent from the handheld communication device and the system cannot associate the received data with the user.”

The Big Picture

Such a method would give Apple a way to differentiate its search results from Google Maps and other competing services, and just in case it needs to be spelled out, if the company is granted the patent, it would prohibit other companies from using that method.

Apple has been working on developing its own mapping resources for several years. Two of the company’s extraordinarily limited acquisitions were map-related companies Placebase (purchased in 2009) and Poly9 (purchased in 2010). Apple has also posted job listings for map-related positions, and even the company’s purchase of artificial intelligence firm (or “mobile assistant: firm) Siri (purchased in 2010) could involve location searches in one way or another.

As with all of the many patents and patent applications The Mac Observer covers, that Apple filed for the patent doesn’t necessarily mean that the company plans or even can bring the related technology to market. This one, however, makes a compelling argument for how Apple might differentiate its own service from Google’s.

Some images made with help from iStockPhoto.