Extracting Cultural Commonsense Knowledge at Scale

TL;DR

This paper introduces CANDLE, a scalable method for extracting high-quality cultural commonsense knowledge from web data, enhancing AI understanding of socio-cultural contexts for more human-centric applications.

Contribution

CANDLE is a novel end-to-end approach that automatically extracts and organizes cultural commonsense assertions across multiple domains and facets, outperforming prior datasets.

Findings

CANDLE produces higher quality CCSK than previous methods.

CANDLE improves GPT-3's understanding of cultural contexts.

The dataset covers diverse domains like geography, religion, and occupation.

Abstract

Structured knowledge is important for many AI applications. Commonsense knowledge, which is crucial for robust human-centric AI, is covered by a small number of structured knowledge projects. However, they lack knowledge about human traits and behaviors conditioned on socio-cultural contexts, which is crucial for situative AI. This paper presents CANDLE, an end-to-end methodology for extracting high-quality cultural commonsense knowledge (CCSK) at scale. CANDLE extracts CCSK assertions from a huge web corpus and organizes them into coherent clusters, for 3 domains of subjects (geography, religion, occupation) and several cultural facets (food, drinks, clothing, traditions, rituals, behaviors). CANDLE includes judicious techniques for classification-based filtering and scoring of interestingness. Experimental evaluations show the superiority of the CANDLE CCSK collection over prior works, and an extrinsic use case demonstrates the benefits of CCSK for the GPT-3 language model. Code and data can be accessed at https://candle.mpi-inf.mpg.de/.