Cross-Register Projection for Headline Part of Speech Tagging

TL;DR

This paper introduces a method for improving part of speech tagging on news headlines by projecting tags from corresponding sentences in news bodies, resulting in significant accuracy gains and aiding information extraction.

Contribution

It proposes a joint training approach on both headlines and long-form text, and introduces POSH, a new annotated corpus for headline POS tagging.

Findings

23% relative error reduction per token

19% error reduction per headline

Improved headline POS tags enhance information extraction

Abstract

Part of speech (POS) tagging is a familiar NLP task. State of the art taggers routinely achieve token-level accuracies of over 97% on news body text, evidence that the problem is well understood. However, the register of English news headlines, "headlinese", is very different from the register of long-form text, causing POS tagging models to underperform on headlines. In this work, we automatically annotate news headlines with POS tags by projecting predicted tags from corresponding sentences in news bodies. We train a multi-domain POS tagger on both long-form and headline text and show that joint training on both registers improves over training on just one or naively concatenating training sets. We evaluate on a newly-annotated corpus of over 5,248 English news headlines from the Google sentence compression corpus, and show that our model yields a 23% relative error reduction per token and 19% per headline. In addition, we demonstrate that better headline POS tags can improve the performance of a syntax-based open information extraction system. We make POSH, the POS-tagged Headline corpus, available to encourage research in improved NLP models for news headlines.