Automatic Knowledge Base Construction using Probabilistic Extraction, Deductive Reasoning, and Human Feedback

Daisy Zhe Wang, Yang Chen, Sean Goldberg, Christan Grant, Kun Li

Abstract: We envision an automatic knowledge base construction system consisting of three inter-related components. MADden is a knowledge extraction system applying statistical text analysis methods over database systems (DBMS) and massive parallel processing (MPP) frameworks; ProbKB performs probabilistic reasoning over the extracted knowledge to derive additional facts not existing in the original text corpus; CAMeL leverages human intelligence to reduce the uncertainty resulting from both the information extraction and probabilistic reasoning processes.