Decision-Theoretic Salience Interactions in Language Generation
Proceedings of the Thirteenth International Joint Conference on Artificial Intelligence. Natural Language. Pages 1246-1251.
Abstract
The salience of an entity is interpreted in natural language generation (NLG) as a measure of how well it stands out from other entities and biases the preference of the generator in selecting words and phrases. We delineate two categories, viz ., canonical salience and instantial salience, characterizing the former as a builtin preference in the conceptual- and linguistic knowledge of the speaker, and the latter as the prominence of specific objects in the context of NLG. This paper proposes that the interaction between these categories be formalized in terms of decision-theoretic combinations of probabilities and utilities, and construes phrase selection and basic level/entry level effects accordingly.
@inproceedings{ijcai1993p172,
title = {Decision-Theoretic Salience Interactions in Language Generation},
author = {Pattabhiraman, T. and Cercone, Nick},
booktitle = {Proceedings of the Thirteenth International Joint Conference on
Artificial Intelligence, {IJCAI-93}},
publisher = {International Joint Conferences on Artificial Intelligence},
editor = {Ruzena Bajcsy},
pages = {1246--1251},
year = {1993},
note = {Natural Language},
url = {https://ijcai.org/proceedings/1993/172},
}
TY - CPAPER TI - Decision-Theoretic Salience Interactions in Language Generation AU - Pattabhiraman, T. AU - Cercone, Nick T2 - Proceedings of the Thirteenth International Joint Conference on Artificial Intelligence, IJCAI-93 PB - International Joint Conferences on Artificial Intelligence PY - 1993 SP - 1246 EP - 1251 UR - https://ijcai.org/proceedings/1993/172 ER -Copied
Machine-readable versions of this paper, for tools and AI agents:
