基于深度学习的自适应游戏剧情生成系统研究

Research on adaptive game scenario generation system based on deep learning

为了使游戏玩家从拥有固定文本的文字游戏中体验新鲜感,提出采用深度学习Seq2Seq模型,根据已有文本产生新的符合剧情的文本,增加游戏的趣味性和交互性。本文构建深度学习Seq2Seq模型,并编写Python代码实现了文本生成模型的仿真,对生成词与原文本之间的评价标准进行了对比分析,以自创的游戏验证了该方法的有效性。该论文的研究成果,是对自适应文本生成系统研究的有益尝试和补充,具有一定的应用价值。

In order to make game players experience novelty from text games with fixed-text,a deep learning Seq2Seq model is put forward,which can generate new scenario text based on existing text to increase the interesting and interactive nature of the game.The paper constructs a deep learning Seq2Seq model and writes Python code to simulate the text generation model.Besides,the paper also compares and analyzes the evaluation criteria between the generated words and the original text.The effectiveness of the method is verified by the self-created game.The research result is a beneficial trial and supplement to the research of adaptive text generation system,which has certain application value.