Pokérator - Unveil your inner Pokémon

Dominique Geissler, Elisa Nguyen, Daphne Theodorakopoulos, Lorenzo Gatti

Research output: Chapter in Book/Report/Conference proceedingConference contributionAcademicpeer-review

214 Downloads (Pure)

Abstract

The Pokérator is a generator of Pokémon names and descriptions, based on user input. The names are generated by blending words based on syllables or characters according to a bigram language model. An accompanying description is generated by filling a template with ConceptNet answers. This sentence is then used as a prompt for text generation with the GPT-2 language model which was finetuned on Pokédex entries. The evaluation of the generated Pokémon names shows that the names are not realistic, but appreciated and creative.
Original languageEnglish
Title of host publicationProceedings of the 11th International Conference on Computational Creativity (ICCC’20)
Pages500-503
Number of pages4
ISBN (Electronic)978-989-54160-2-8
Publication statusPublished - 2020
Event11th International Conference on Computational Creativity, ICCC 2020 - Virtual Conference
Duration: 7 Sept 202011 Sept 2020
Conference number: 11

Conference

Conference11th International Conference on Computational Creativity, ICCC 2020
Abbreviated titleICCC 2020
CityVirtual Conference
Period7/09/2011/09/20

Fingerprint

Dive into the research topics of 'Pokérator - Unveil your inner Pokémon'. Together they form a unique fingerprint.

Cite this