The Intelligence Advanced Research Projects Activity (IARPA) will hold a proposers’ day on Jan. 19th to provide stakeholders information on technical requirements and objectives of a program that intends to develop novel artificial intelligence systems to help protect author privacy and attribute authorship through identification and use of explainable linguistic fingerprints.
The Human Interpretable Attribution of Text using Underlying Structure (HIATUS) program seeks to advance the development of “novel techniques to generate representations that capture author-level linguistic variation and will use these representations to build human-interpretable algorithms to perform authorship attribution and ensure author privacy,” IARPA said Friday.
IARPA said technical approaches selected through the HIATUS program will be scalable across languages, genres and topic domains. Interested stakeholders have until Jan. 14th to register for the event.
