Author: Rebuffel, Clément; Soulier, Laure; Scoutheeten, Geoffrey; Gallinari, Patrick
Title: A Hierarchical Model for Data-to-Text Generation Cord-id: h5q6rx5a Document date: 2020_3_17
ID: h5q6rx5a
Snippet: Transcribing structured data into natural language descriptions has emerged as a challenging task, referred to as “data-to-textâ€. These structures generally regroup multiple elements, as well as their attributes. Most attempts rely on translation encoder-decoder methods which linearize elements into a sequence. This however loses most of the structure contained in the data. In this work, we propose to overpass this limitation with a hierarchical model that encodes the data-structure at the e
Document: Transcribing structured data into natural language descriptions has emerged as a challenging task, referred to as “data-to-textâ€. These structures generally regroup multiple elements, as well as their attributes. Most attempts rely on translation encoder-decoder methods which linearize elements into a sequence. This however loses most of the structure contained in the data. In this work, we propose to overpass this limitation with a hierarchical model that encodes the data-structure at the element-level and the structure level. Evaluations on RotoWire show the effectiveness of our model w.r.t. qualitative and quantitative metrics.
Search related documents:
Co phrase search for related documents, hyperlinks ordered by date