morizon/TCGA_Reports_ja_structured

这是一个基于TCGA病理报告的数据集，包含英语原文、英语原文的日语翻译以及从日语翻译中提取的结构化信息。数据集的目的是用于自然语言处理研究，特别是机器翻译和信息抽取任务。每个样本都包括一个唯一标识符、英语原文、日语翻译和结构化的JSON信息，其中结构化信息包含器官、采集方法、诊断、分化度、病期、肿瘤径、浸润范围、断端和其他所见等详细信息。

This dataset is based on TCGA pathology reports, including English original text, Japanese translation of the English original, and structured information extracted from the Japanese translation. The dataset is intended for natural language processing research, particularly machine translation and information extraction tasks. Each sample includes a unique identifier, English original, Japanese translation, and structured JSON information, which includes details such as organ, collection method, diagnosis, differentiation, stage, tumor diameter, infiltration range, tumor status at resection margins, and other findings.