故障溯源知识图谱补全的训练数据

该数据集主要包括柴油发动机故障诊断文本中的实体以及实体之间的关系，用于柴油发动机故障知识图谱相关模型的训练任务，即优化模型的权重参数。在该数据集中，实体类型主要有：发动机、组成系统、零部件、故障现象、故障原因；实体之间的关系定义为4类：组成、位于、原因、unknown，其中unknown指的是一个句子中的两个实体之间并不存在特定的关系；数据集中各个实体类别之间的关系。

This dataset mainly includes entities and their inter-entity relational links in diesel engine fault diagnosis texts, and is used for training models related to diesel engine fault knowledge graphs, specifically to optimize the weight parameters of such models. The main entity types in this dataset are as follows: engine, component system, parts and components, fault phenomenon, and fault cause. The relationships between entities are defined into four categories: compose, located_in, cause, and unknown. Here, 'unknown' denotes that there is no specific relational connection between two entities within a single sentence. The relationships between the various entity categories in the dataset.

搜集汇总

背景与挑战

背景概述

该数据集是用于故障溯源知识图谱补全的训练数据，主要针对柴油发动机故障诊断领域，包含实体（如发动机、零部件、故障现象等）和四种关系类型（组成、位于、原因、unknown），以jsonl和docx格式提供，旨在优化知识图谱模型的权重参数。数据来源于哈尔滨工业大学，属于国家重点研发计划项目，适用于软件工程和计算机科学技术领域的相关研究。