Work

Extracting Commonsense Properties from Embeddings with Limited Human Guidance

Year: 2018

Type: conference-paper

Abstract: Intelligent systems require common sense, but automatically extracting this knowledge from text can be difficult.We propose and assess methods for extracting one type of commonsense knowledge, object-property comparisons, from pretrained embeddings.In experiments, we show that our approach exceeds the accuracy of previous work but requires substantially less hand-annotated knowledge.Further, we show that an active learning approach that synthesizes common-sense queries can boost accuracy.

Source: Proceedings of the conference - Association for Computational Linguistics. Meeting

DOI: 10.18653/v1/p18-2102

Authors: Yiben Yang, Larry Birnbaum, Ji-Ping Wang, Doug Downey

Institution: Northwestern University

Language: English

Topic: Topic Modeling

Subfield: Artificial Intelligence

Field: Computer Science

Domain: Physical Sciences

Open Access status: hybrid

Funder: National Science Foundation

Award: CAREER: Web Information Extraction: Integration and Scaling