数据堂335,000组东南亚语种多轮对话文本数据

本数据集内容为印尼语,马来语,泰语,越南语等东南亚语种的多轮对话文本数据，共计335000组。涉及类目广泛，涵盖社交、情感、金融、法律、科技等72个类别。对话内容丰富，平均每组对话有8个轮次，本数据集有助于模型自然语言理解、自然语言生成、对话策略等能力的提升。

This dataset contains multi-turn dialogue text data in Southeast Asian languages including Indonesian, Malay, Thai, Vietnamese and others, with a total of 335,000 dialogue groups. It covers 72 diverse categories such as social interaction, emotion, finance, law, technology and more. The dialogues are rich in content, with an average of 8 conversational turns per group. This dataset contributes to enhancing the capabilities of models in natural language understanding, natural language generation, dialogue strategy and other relevant aspects.

搜集汇总

背景与挑战

背景概述

该数据集包含33.5万组东南亚语言（印尼语、马来语、泰语、越南语）的多轮对话文本，覆盖社交、情感、金融等72个类别，平均每轮对话8个回合。数据采用Jsonl格式，由DataTang版权所有并支持商业使用。