Neural Wikipedian: Generating Textual Summaries from Knowledge Base Triples

Most people need textual or visual interfaces in order to make sense of Semantic Web data. In this paper, we investigate the problem of generating natural language summaries for Semantic Web data using neural networks. Our end-to-end trainable architecture encodes the information from a set of tripl...

Full description

Saved in:
Bibliographic Details
Published inWeb semantics Vol. 52-53; pp. 1 - 15
Main Authors Vougiouklis, Pavlos, Elsahar, Hady, Kaffee, Lucie-Aimée, Gravier, Christophe, Laforest, Frédérique, Hare, Jonathon, Simperl, Elena
Format Journal Article
LanguageEnglish
Published Elsevier B.V 01.10.2018
Elsevier
Subjects
Computer Science
Natural language generation
Neural networks
Semantic web triples
Natural language generation
Neural networks
Semantic web triples
Online AccessGet full text

Cover

More Information
Summary:Most people need textual or visual interfaces in order to make sense of Semantic Web data. In this paper, we investigate the problem of generating natural language summaries for Semantic Web data using neural networks. Our end-to-end trainable architecture encodes the information from a set of triples into a vector of fixed dimensionality and generates a textual summary by conditioning the output on the encoded vector. We explore a set of different approaches that enable our models to verbalise entities from the input set of triples in the generated text. Our systems are trained and evaluated on two corpora of loosely aligned Wikipedia snippets with triples from DBpedia and Wikidata, with promising results.
ISSN:1570-8268
1873-7749
DOI:10.1016/j.websem.2018.07.002