Entitled "Automatic Discovery of Metadata," the filing was published by the U.S. Patent and Trademark Office early Thursday. It describes an automated system to alleviate the "burdensome task" of manually adding metadata.
According to Apple, examples of common metadata tags include the identity of users "generating or editing the data, the content of the data, the location or time at which the data was generated to edited, user-defined tags, or other attributes of the data." Other possible types of metadata include "orientation, time, temperature, or vibration/movement metadata," and data retrieved from the content of the file, such as "color, texture, face detection, image recognition output, and voice recognition metadata."
Apple notes that some forms of metadata are often automatically associated with files when they are generated, while others can be added "at a later time." However, the limitations of a device, such as a lack of GPS functionality, can limit which fields it could automatically populate.
The invention would overcome those limitations by "discovering other data sharing characteristics" of stored data and applying them to a file. In one embodiment, the date and time of a captured image on a device could be checked against events inputted into a calendar app to automatically add event information metadata to the photo.
"In some embodiments, the processing module can identify entries in one or more social networks that relate to the data. For example, the processing module can identify social network streams associated with the user accessing the data, or with friends of the user accessing the data and extract metadata from the social network streams," the filing read.
Specific examples of possible metadata include people who may have been at the event and the location of the event.
Apple details an example where the system searches a social networking stream to identify a user in a video. Once identified, the processing module could then "retrieve a social networking handle associated with the identified person from an address book" and review the relevant social streams for that user. The module could then automatically tag people present in the video and add any relevant metadata.
In the filing, Apple further expanded the invention to include additional searches from "remote sources as a whole," such as "some or all of the Internet." One embodiment would entail searching the Web with a color profile from an image in order to identify the location where it was taken. For example, large amounts of orange banners could mean a photo was snapped at Golden Gate Park, while clothing worn by people at an event could help a processing module deduce where and when the event took place.
The filing also suggests a process that would start from the user's data and search a social network of the Internet for "some or all of the data that corresponds to the user's data." Though Apple stops short of mentioning specific use cases and services, the invention could presumably help a device search through a user's friends on a social networking service like Facebook to identify faces in a photo. The filing does mention an example of analyzing audio data from a video to "extract voice prints" that could be compared against a library of voice prints in a device's address book.
Apple filed for the application on February 27, 2012. Michael Ingrassia, Jeffery Lee and Dave Rempel are listed as the inventors.
The patent application could be taken as evidence of Apple's increased interest in social integration. Late last month, Apple CEO Tim Cook promised that Apple would add more social aspects to its products.
"Apple doesn't have to own a social network," he said. "But does Apple have to be social? Yes."
Cook also admitted at the time that Ping, a social music discovery service the company launched in 2010, had not worked out as well as originally planned. He added that the company was looking into whether to "kill" the service. Last week, a report claimed that Apple would ditch Ping this fall with the next iteration of iTunes and iOS 6.