Empath: Understanding Topic Signals in Large-Scale Text

TL;DR

Empath is a scalable tool that generates and validates lexical categories from seed words using neural embeddings trained on a large corpus, enabling nuanced topic analysis in text.

Contribution

It introduces a method to create and validate new lexical categories on demand from small seed sets using deep learning and crowd validation.

Findings

Empath's categories correlate highly with LIWC categories (r=0.906).

It can generate relevant new categories from minimal seed words.

Empath analyzes text across 200 pre-validated categories.

Abstract

Human language is colored by a broad range of topics, but existing text analysis tools only focus on a small number of them. We present Empath, a tool that can generate and validate new lexical categories on demand from a small set of seed terms (like "bleed" and "punch" to generate the category violence). Empath draws connotations between words and phrases by deep learning a neural embedding across more than 1.8 billion words of modern fiction. Given a small set of seed words that characterize a category, Empath uses its neural embedding to discover new related terms, then validates the category with a crowd-powered filter. Empath also analyzes text across 200 built-in, pre-validated categories we have generated from common topics in our web dataset, like neglect, government, and social media. We show that Empath's data-driven, human validated categories are highly correlated (r=0.906) with similar categories in LIWC.