Project

Liquid Movies

Groups

Video has evolved from an esoteric production to the default means of communication both online and as broadcast. We accept a bad or lengthy video where we would object to poor writing. And because moving images impact us viscerally as well as intellectually, we are susceptible to bias, misinformation and falsity in video by dint of its presentation rather than its argument.  We propose a two-year research program that builds on seven years of work recording and parsing broadcast news to develop machine-learning based analysis and navigational aids to understanding the media with which we live, learn, and communicate.  The engine that drives this is a suite of recording and analysis programs by which we parse and add metadata to video to reveal elements of style and content that is interfaced with viewing and presentation tools to facilitate analysis.  We intend to apply this processing to both news and informational video to better understand both portrayed and subliminal cues.  We also plan to develop applications to improve how normal media consumers can peruse, search, learn from the material, and build their own analytical engines.

Video has evolved from an esoteric production to the default means of communication both online and as broadcast. We accept a bad or lengthy video where we would object to poor writing. And because moving images impact us viscerally as well as intellectually, we are susceptible to bias, misinformation and falsity in video by dint of its presentation rather than its argument.  We propose a two-year research program that builds on seven years of work recording and parsing broadcast news to develop machine-learning based analysis and navigational aids to understanding the media with which we live, learn, and communicate.  The engine that drives this is a suite of recording and analysis programs by which we parse and add metadata to video to reveal elements of style and content that is interfaced with viewing and presentation tools to facilitate analysis.  We intend to apply this processing to both news and informational video to better understand both portrayed and subliminal cues.  We also plan to develop applications to improve how normal media consumers can peruse, search, learn from the material, and build their own analytical engines.