CausalFusion: Causal Knowledge Extraction and Fusion from Multiple Modalities
Abstract
Combining information from multiple sources into a single usable representation, termed information fusion, is crucial for effective support of decision making both by humans and machines. Given the increasing number, types and modalities (text, images, geographic, video, social) of data sources it is challenging to produce representations that do not overwhelm users. This project investigates a unified theoretical framework for the extraction of causal knowledge, which helps to reduce spurious information. Specifically, we explore principles for determining actual cause between static entities in different modalities.
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- Jan 21, 2022
- Source ID
- FA86552217155XX0
Entities
People
- Paul Groth
Organizations
- Air Force Office of Scientific Research
- United States Air Force
- University of Amsterdam