Survey of the state of the art in natural language generation: Core tasks, applications and evaluation

A Gatt, E Krahmer - Journal of Artificial Intelligence Research, 2018‏ - jair.org
This paper surveys the current state of the art in Natural Language Generation (NLG),
defined as the task of generating text or speech from non-linguistic input. A survey of NLG is …

Measuring compositionality in representation learning

J Andreas - arxiv preprint arxiv:1902.07181, 2019‏ - arxiv.org
Many machine learning algorithms represent input data with vector embeddings or discrete
codes. When inputs exhibit compositional structure (eg objects built from parts or procedures …

Computational generation of referring expressions: A survey

E Krahmer, K Van Deemter - Computational Linguistics, 2012‏ - direct.mit.edu
This article offers a survey of computational research on referring expression generation
(REG). It introduces the REG problem and describes early work in this area, discussing what …

Reasoning about pragmatics with neural listeners and speakers

J Andreas, D Klein - arxiv preprint arxiv:1604.00562, 2016‏ - arxiv.org
We present a model for pragmatically describing scenes, in which contrastive behavior
results from a combination of inference-driven pragmatics and learned semantics. Like …

[ספר][B] Computational models of referring: a study in cognitive science

K Van Deemter - 2016‏ - books.google.com
An argument that computational models can shed light on referring, a fundamental and
much-studied aspect of communication. To communicate, speakers need to make it clear …

Factors causing overspecification in definite descriptions

R Koolen, A Gatt, M Goudbeek, E Krahmer - Journal of Pragmatics, 2011‏ - Elsevier
Speakers often overspecify their target descriptions and include more information than
necessary for unique identification of the target referent. In the current paper, we study the …

[PDF][PDF] The TUNA challenge 2008: Overview and evaluation results

A Gatt, A Belz, E Kow - 2008‏ - um.edu.mt
The TUNA Challenge was a set of three shared tasks at REG'08, all of which used data from
the TUNA Corpus. The three tasks covered attribute selection for referring expressions …

[PDF][PDF] The use of spatial relations in referring expression generation

J Viethen, R Dale - Proceedings of the Fifth International Natural …, 2008‏ - aclanthology.org
There is a prevailing assumption in the literature on referring expression generation that
relations are used in descriptions only 'as a last resort', typically on the basis that including …

The effect of scene variation on the redundant use of color in definite reference

R Koolen, M Goudbeek, E Krahmer - Cognitive Science, 2013‏ - Wiley Online Library
This study investigates to what extent the amount of variation in a visual scene causes
speakers to mention the attribute color in their definite target descriptions, focusing on …

Learning in the rational speech acts model

W Monroe, C Potts - arxiv preprint arxiv:1510.06807, 2015‏ - arxiv.org
The Rational Speech Acts (RSA) model treats language use as a recursive process in which
probabilistic speaker and listener agents reason about each other's intentions to enrich the …