Toward Unsupervised Text Content Manipulation

  • 2019-02-08 17:08:56
  • Wentao Wang, Zhiting Hu, Zichao Yang, Haoran Shi, Frank Xu, Eric Xing
  • 0

Abstract

Controlled generation of text is of high practical use. Recent efforts havemade impressive progress in generating or editing sentences with given textualattributes (e.g., sentiment). This work studies a new practical setting of textcontent manipulation. Given a structured record, such as `(PLAYER: Lebron,POINTS: 20, ASSISTS: 10)', and a reference sentence, such as `Kobe easilydropped 30 points', we aim to generate a sentence that accurately describes thefull content in the record, with the same writing style (e.g., wording,transitions) of the reference. The problem is unsupervised due to lack ofparallel data in practice, and is challenging to minimally yet effectivelymanipulate the text (by rewriting/adding/deleting text portions) to ensurefidelity to the structured content. We derive a dataset from a basketball gamereport corpus as our testbed, and develop a neural method with unsupervisedcompeting objectives and explicit content coverage constraints. Automatic andhuman evaluations show superiority of our approach over competitive methodsincluding a strong rule-based baseline and prior approaches designed for styletransfer.

 

Introduction (beta)

None

 

Conclusion (beta)

None