Using genetic algorithms to create meaningful poetic text

Ruli Manurung, Graeme D Ritchie, Henry Thompson

Research output: Contribution to journalArticle

38 Citations (Scopus)
8 Downloads (Pure)

Abstract

This article presents a series of experiments in automatically generating poetic texts. We confined our attention to the generation of texts which are syntactically well-formed, meet certain pre-specified patterns of metre and broadly convey some given meaning. Such aspects can be formally defined, thus avoiding the complications of imagery and interpretation that are central to assessing more free forms of verse. Our implemented system, McGONAGALL, applies the genetic algorithm to construct such texts. It uses a sophisticated linguistic formalism to represent its genomic information, from which can be computed the phenotypic information of both semantic representations and patterns of stress. The conducted experiments broadly indicated that relatively meaningful text could be produced if the constraints on metre were relaxed, and precise metric text was possible with loose semantic constraints, but it was difficult to produce text which was both semantically coherent and of high quality metrically.
Original languageEnglish
Pages (from-to)43–64
Number of pages22
JournalJournal of Experimental & Theoretical Artificial Intelligence
Volume24
Issue number1
Early online date20 Jan 2012
DOIs
Publication statusPublished - 20 Jan 2012

Fingerprint

Genetic algorithms
Semantics
Genetic Algorithm
Linguistics
Experiments
Complications
Experiment
Genomics
Text
Metric
Series

Keywords

  • natural language generation
  • genetic algorithms
  • poetry
  • creative language

Cite this

Using genetic algorithms to create meaningful poetic text. / Manurung, Ruli; Ritchie, Graeme D; Thompson, Henry.

In: Journal of Experimental & Theoretical Artificial Intelligence, Vol. 24, No. 1, 20.01.2012, p. 43–64.

Research output: Contribution to journalArticle

@article{d6d148ea6fda408bb29cbac3ea13ed08,
title = "Using genetic algorithms to create meaningful poetic text",
abstract = "This article presents a series of experiments in automatically generating poetic texts. We confined our attention to the generation of texts which are syntactically well-formed, meet certain pre-specified patterns of metre and broadly convey some given meaning. Such aspects can be formally defined, thus avoiding the complications of imagery and interpretation that are central to assessing more free forms of verse. Our implemented system, McGONAGALL, applies the genetic algorithm to construct such texts. It uses a sophisticated linguistic formalism to represent its genomic information, from which can be computed the phenotypic information of both semantic representations and patterns of stress. The conducted experiments broadly indicated that relatively meaningful text could be produced if the constraints on metre were relaxed, and precise metric text was possible with loose semantic constraints, but it was difficult to produce text which was both semantically coherent and of high quality metrically.",
keywords = "natural language generation, genetic algorithms, poetry, creative language",
author = "Ruli Manurung and Ritchie, {Graeme D} and Henry Thompson",
note = "Work carried out when all authors were at the University of Edinburgh.",
year = "2012",
month = "1",
day = "20",
doi = "10.1080/0952813X.2010.539029",
language = "English",
volume = "24",
pages = "43–64",
journal = "Journal of Experimental & Theoretical Artificial Intelligence",
issn = "1362-3079",
publisher = "Taylor and Francis Ltd.",
number = "1",

}

TY - JOUR

T1 - Using genetic algorithms to create meaningful poetic text

AU - Manurung, Ruli

AU - Ritchie, Graeme D

AU - Thompson, Henry

N1 - Work carried out when all authors were at the University of Edinburgh.

PY - 2012/1/20

Y1 - 2012/1/20

N2 - This article presents a series of experiments in automatically generating poetic texts. We confined our attention to the generation of texts which are syntactically well-formed, meet certain pre-specified patterns of metre and broadly convey some given meaning. Such aspects can be formally defined, thus avoiding the complications of imagery and interpretation that are central to assessing more free forms of verse. Our implemented system, McGONAGALL, applies the genetic algorithm to construct such texts. It uses a sophisticated linguistic formalism to represent its genomic information, from which can be computed the phenotypic information of both semantic representations and patterns of stress. The conducted experiments broadly indicated that relatively meaningful text could be produced if the constraints on metre were relaxed, and precise metric text was possible with loose semantic constraints, but it was difficult to produce text which was both semantically coherent and of high quality metrically.

AB - This article presents a series of experiments in automatically generating poetic texts. We confined our attention to the generation of texts which are syntactically well-formed, meet certain pre-specified patterns of metre and broadly convey some given meaning. Such aspects can be formally defined, thus avoiding the complications of imagery and interpretation that are central to assessing more free forms of verse. Our implemented system, McGONAGALL, applies the genetic algorithm to construct such texts. It uses a sophisticated linguistic formalism to represent its genomic information, from which can be computed the phenotypic information of both semantic representations and patterns of stress. The conducted experiments broadly indicated that relatively meaningful text could be produced if the constraints on metre were relaxed, and precise metric text was possible with loose semantic constraints, but it was difficult to produce text which was both semantically coherent and of high quality metrically.

KW - natural language generation

KW - genetic algorithms

KW - poetry

KW - creative language

U2 - 10.1080/0952813X.2010.539029

DO - 10.1080/0952813X.2010.539029

M3 - Article

VL - 24

SP - 43

EP - 64

JO - Journal of Experimental & Theoretical Artificial Intelligence

JF - Journal of Experimental & Theoretical Artificial Intelligence

SN - 1362-3079

IS - 1

ER -