As described in section 11.11.B of the BIFFUD Corporate Bylaws I (we) hereby invoke my (our) rights as a (a) sentient being(s) who has (have) not uploaded their mind(s) to the cloud for consideration of this project application by the BIFFUD Hive Mind. With this application I (we) submit my (our) interest in becoming a (a) Member(s) of BIFFUD and having this project supported and adored by all who can 🤔.
Project Name: Talking Point Tracker
Here is what was said:
blah blah blah blah blah blah blah
blah blah blah blah blah
Project Analogy: It's like the human brain for summarizing the day's talking points
The Bad Idea Factory, led by Dan Schultz, will assist the Duke Reporters’ Lab by developing Talking Point Tracker, which will include methods and technologies to integrate statements from the TV News Archive and other sources with fact checking work flows.
The Talking Point Tracker will extract the most frequent political statements aired on broadcast television in near real-time. The new tool will provide political fact-checkers with data to know which claims are catching on and being repeated to the public.
The tracker will incorporate elements that quantify soundbites, phrases, and key metadata.
Soundbites: Adapt the Duplitron tool, an open source audio fingerprinting tool, to track news coverage in near real-time by crunching through TV news broadcasts over a given time period and detecting which soundbites are being clipped the most. Develop ability to automatically distinguish commercials, improve soundbite detection of specific durations, and incorporating video fingerprinting into algorithms.
Phrase tracking: Explore ways to use natural language processing and analysis to enable the identification of common talking points and themed claims for a given time period through caption and transcript analysis. Continue to open our closed captioning data stores to others to help work toward jumping these technological hurdles while working on advancing these ourselves.
Key metadata. Enhance key metadata related to topic detection capabilities. Sentence clustering, entity extraction (essentially noun detection–finding persons, places, or things mentioned), all are areas to develop.
How is this project a bad idea?
It is trying to solve like ten impossibly hard problems at once!
If this project were a D&D Character, what alignment would it be and why?