Get our free extension to see links to code for papers anywhere online!

Chrome logo Add to Chrome

Firefox logo Add to Firefox


Using Noisy Extractions to Discover Causal Knowledge

Nov 16, 2017
Dhanya Sridhar, Jay Pujara, Lise Getoor


Share this with someone who'll enjoy it:


Knowledge bases (KB) constructed through information extraction from text play an important role in query answering and reasoning. In this work, we study a particular reasoning task, the problem of discovering causal relationships between entities, known as causal discovery. There are two contrasting types of approaches to discovering causal knowledge. One approach attempts to identify causal relationships from text using automatic extraction techniques, while the other approach infers causation from observational data. However, extractions alone are often insufficient to capture complex patterns and full observational data is expensive to obtain. We introduce a probabilistic method for fusing noisy extractions with observational data to discover causal knowledge. We propose a principled approach that uses the probabilistic soft logic (PSL) framework to encode well-studied constraints to recover long-range patterns and consistent predictions, while cheaply acquired extractions provide a proxy for unseen observations. We apply our method gene regulatory networks and show the promise of exploiting KB signals in causal discovery, suggesting a critical, new area of research.



   Access Paper Source



Share this with someone who'll enjoy it: