Dynamic tagging of media for service sessions

Inventors

Jayapalan, Vijay • Yarbrough, Gregory • Smith, Eric J.

Assignees

United Services Automobile Association USAA

Abstract

Techniques are described for real time monitoring and tagging of media content generated during a service session between an individual and a service representative (SR). Media content received during a service session may be dynamically tagged with one or more metadata tags during the service session, the tag(s) indicating portion(s) of the media content that potentially include sensitive information. Tagging may be based on a SR's gestures, clicks, keystrokes, mouse-overs, and/or other actions performed in a SR user interface (UI) during the service session. The media content may be redacted based on the tags to remove and/or obfuscate potentially sensitive information.

Core Innovation

A method and a system access tagged media content including a stream of media content. The stream includes a first tag added at a first location that corresponds to a first time in which a first action is detected and a second tag added at a second location that corresponds to a second time in which a second action is detected. The first tag and the second tag define at least one portion of the tagged media content.

The stream is analyzed to determine information that needs to be redacted. The tagged media content is compared with the stream of media content to determine whether the at least one portion defined by the first tag and the second tag indicates the presence of the information that needs to be redacted. Responsive to the comparing, at least one of the first tag or the second tag is adjusted such that a portion of the tagged media content having the adjusted at least one of the first tag or the second tag indicates the presence of the information that needs to be redacted.

The portion of the tagged media content having the adjusted tag is redacted to generate redacted media content. In the documented embodiments, the stream of media content includes audio content and corresponding audio tags, or includes video content and tags that correspond to video frames. The document also states that the media content can include text transcription of audio and/or video corresponding to a communication session.

Claims Coverage

The document includes three independent claim types: a computer-implemented method, a system, and non-transitory computer-readable media. Across these independents, the claims share the same core flow of accessing tagged media, analyzing to determine information that needs to be redacted, comparing to determine whether tagged portions indicate that information, adjusting tags, and redacting to generate redacted media content. The independent claims each recite three inventive actions: tag-based portion definition, tag adjustment based on presence determination, and redaction to generate redacted media content.

Across the independent claims, the inventive subject matter is directed to redaction of tagged portions in a stream by defining portions using first and second tags tied to detected actions, determining what information needs to be redacted, comparing to detect whether the tagged portions indicate that information, adjusting the tags responsive to that determination, and redacting the adjusted tagged portions to generate redacted media content.

Stated Advantages

Documented Applications