Social media has allowed people to share all manner of content with friends, family, and others with similar interests. One popular form of sharing is livestreamed video, where one can share a live video with others over the Internet. The sharing can be used to share media presentations to teach a class, to engage in mobile journalism (mo jo), to provide public service announcements (PSAa), or simply to have fun. A media presentation such as a livestreamed video can be played to an individual. Mental state data and facial data from the individual can be captured and compared to time-based mental state event signatures. The time-based signatures indicate the onset, intensity, duration, and decay of a given mental state. Mental states are inferred from the collected mental state data and the facial data. The facial data is analyzed for eyebrow furls and other facial expressions. The individual’s mental state data is correlated with the mental state data from others who have also viewed the media presentation. The correlating identifies and uses maximally dissimilar responses which rank the presentation relative to other presentations, and recommends additional media presentations to the individual.
Read the full patent here.
Title: Video Recommendation Using Affect
Patent Number: US 9,503,786 B2
Application Number: 14/821,896
Filing Date: Aug. 10, 2015
Issue Date: Nov. 22, 2016
Inventors: el Kaliouby, Rana; Abdelrahman Mahmoud; and Panu James Turcot
Related Application Data: Continuation-in-part of application No. 13/406,068, filed on Feb. 27, 2012, now Pat. No. 9,106,958, and a continuation-in-part of application No. 14/796,419, filed on Jul. 10, 2015, which is a continuation-in-part