Headline Generation as a Sequence Prediction with Conditional Random Fields

Carlos A. Colmenares, Marina Litvak, Amin Mantrach, Fabrizio Silvestri, Horacio Rodríguez

Research output: Chapter in Book/Report/Conference proceedingChapterpeer-review

4 Scopus citations

Abstract

Automatic headline generation, or automatic one-line summarization, is a sub-task of document summarization with many reported applications in several domains. In this chapter we present a sequence-prediction technique for learning how editors title their news stories. The introduced technique models the problem as a discrete optimization task in a feature-rich space, where the global optimum can be found in polynomial time by means of dynamic programming. We train and test our model with an extensive corpus of financial news, and compare it against a number of baselines by using standard metrics from the document summarization domain, as well as some new metrics proposed in this work. The obtained results are very appealing and substantiate the soundness of the approach.

Original languageEnglish
Title of host publicationMultilingual Text Analysis
Subtitle of host publicationChallenges, Models, and Approaches
PublisherWorld Scientific Publishing Co.
Pages201-243
Number of pages43
ISBN (Electronic)9789813274884
ISBN (Print)9789813274877
DOIs
StatePublished - 1 Jan 2019
Externally publishedYes

ASJC Scopus subject areas

  • General Computer Science

Fingerprint

Dive into the research topics of 'Headline Generation as a Sequence Prediction with Conditional Random Fields'. Together they form a unique fingerprint.

Cite this