来自今天的爱可可AI前沿推介

[CL] Ontologically Faithful Generation of Non-Player Character Dialogues

N Weir, R Thomas, R D'Amore, K Hill, B V Durme...
[Johns Hopkins University & Microsoft Semantic Machines & Microsoft Gaming]

非玩家人物本体忠实对话生成

要点:

  1. 提出KNUDGE(知识受限用户-NPC对话生成)任务，根据自然语言段落中提供的游戏知识和任务相关知识规范生成对话树；
  2. 构建KNUDGE数据集，由Obsidian Entertainment的《The Outer Worlds》游戏数据提供的45个副任务对话树组成，每个树最多可达到100个角色发言节点，包含复杂的分支和循环；
  3. 一系列基于神经语言模型的知识约束对话编写基线，表明约束信息可以合理地反映在真实生成的树中。

摘要：
本文提出一种基于流行电子游戏环境的语言生成任务。KNUDGE(知识受限用户-NPC对话生成)涉及生成以自然语言段落中捕获的本体为条件的对话树，提供任务和实体规范。KNUDGE是由直接从Obsidian Entertainment的《The Outer Worlds》的游戏数据中提取的支线任务对话构建的，带来了现实世界的复杂性：(1) 对话是分枝树，而不是线性话语链；(2) 话语必须忠实于游戏背景——角色人物、背景故事和实体关系；以及 (3) 对话必须向人类玩家准确揭示与任务相关的新细节。本文报告了监督和上下文学习技术的结果，发现未来在创建逼真的游戏质量对话方面有很大的空间。

We introduce a language generation task grounded in a popular video game environment. KNUDGE (KNowledge Constrained User-NPC Dialogue GEneration) involves generating dialogue trees conditioned on an ontology captured in natural language passages providing quest and entity specifications. KNUDGE is constructed from side quest dialogues drawn directly from game data of Obsidian Entertainment's The Outer Worlds, leading to real-world complexities in generation: (1) dialogues are branching trees as opposed to linear chains of utterances; (2) utterances must remain faithful to the game lore--character personas, backstories, and entity relationships; and (3) a dialogue must accurately reveal new quest-related details to the human player. We report results for supervised and in-context learning techniques, finding there is significant room for future work on creating realistic game-quality dialogues.

论文链接：https://arxiv.org/abs/2212.10618
图片
图片
图片
图片

内容中包含的图片若涉及版权问题，请及时与我们联系删除