Application of Large Language Model in Connecting Non-Player Characters Intelligent Conversations

Peiyuncheng Ding

2024

Abstract

The large language model is a kind of Advanced natural language processing model. By inserting the large language model, non-player characters can understand the question and flow the personal character of non-player characters from players. This kind of intelligent dialogue not only can improve the game experience of players but also can improve the interaction and fun of the games. This paper will link the large language model to the non-player characters to communicate with non-player characters. By using Unity, Ollama and ChatGPT3.5 and build API communication interface and some optimization. This paper has solved some problems of Performance optimization of large language models and let the reply be closer to the character's personality and also improve the interactivity and the experience of players. Finally, this paper has linked the large language model with non-player characters in the game successfully and chats with it fluently and continually. In the future, there must be a more efficient and perfect solution to link the large language model to the non-player characters of games to provide a better experience for players.

Paper Citation

in Harvard Style

Ding P. (2024). Application of Large Language Model in Connecting Non-Player Characters Intelligent Conversations. In Proceedings of the 1st International Conference on Modern Logistics and Supply Chain Management - Volume 1: MLSCM; ISBN 978-989-758-738-2, SciTePress, pages 135-138. DOI: 10.5220/0013235000004558

in Bibtex Style

@conference{mlscm24,
author={Peiyuncheng Ding},
title={Application of Large Language Model in Connecting Non-Player Characters Intelligent Conversations},
booktitle={Proceedings of the 1st International Conference on Modern Logistics and Supply Chain Management - Volume 1: MLSCM},
year={2024},
pages={135-138},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0013235000004558},
isbn={978-989-758-738-2},
}

in EndNote Style

TY - CONF

JO - Proceedings of the 1st International Conference on Modern Logistics and Supply Chain Management - Volume 1: MLSCM
TI - Application of Large Language Model in Connecting Non-Player Characters Intelligent Conversations
SN - 978-989-758-738-2
AU - Ding P.
PY - 2024
SP - 135
EP - 138
DO - 10.5220/0013235000004558
PB - SciTePress