Alexandria: Unsupervised High-Precision Knowledge Base Construction using a Probabilistic Program

Creating a knowledge basethat is accurate, up-to-date and complete remains a significant challenge despite substantial efforts in automated knowledge baseconstruction. In this paper, we present Alexandria -- a system for unsupervised, high-precision knowledge baseconstruction. Alexandria uses a probabilistic program to define a process of converting knowledge basefacts into unstructured text. Using probabilistic inference, we can invert this program and so retrieve facts, schemas and entities from web text. The use of a probabilistic program allows uncertainty in the text to be propagated through to the retrieved facts, which increases accuracy and helps merge facts from multiple sources. Because Alexandria does not require labelled training data, knowledge basescan be constructed with the minimum of manual input. We demonstrate this by constructing a high precision (typically 97\%+) knowledge basefor people from a single seed fact.