Scene-Level TV metatags and metacontent. corp.lingospot.com

Powered by patented natural language processing, semantic search, image analysis and machine learning, our software analyzes multiple aspects of over 2,000 TV channels ...
See More to extract what's occurring inside each scene. We provide time coded scene envelopes containing topics and events metatags as well as rich metacontent about each item. We also recommend relevant, supplemental content and provide access to an extensive network of relationships and integrations with thousands of content sources from across the web.

Technology:

Lingospot's technology performs four types of analysis on the image, audio and closed caption streams of TV stations and networks: (1) natural language processing of the closed caption transcripts, to identify topics being discussed, (2) optical character recognition, to extract the words from text that is displayed on the screen, such as a location name in a movie, or a ticker tape in a news channel, (3) facial matching and image processing, to identify who is on the screen, such as the actor in a movie or the reporter in a newscast, and (4) audio analysis to identify the mood and other aspects of the program.

Our algorithms then combine the information collected via these four methods with other information and metadata we know about the show to more accurately identify the precise context. We then tag the video with topics and events that are relevant, whether mentioned on air or observed on screen. We make these scene-level metatags available via our API and, optionally, can include synched-to-broadcast metacontent related to the programing of each channel.