Cypher Query Generation Dataset

该数据集旨在训练和评估自然语言转换为Cypher查询的能力，其来源于一个开源的文本至Cypher数据集，该数据集包含了超过16种不同的图模式。此外，该数据集还包含了多种图模式以供评估。在规模上，训练数据是从文本至Cypher数据集中抽取的，测试集则来自“fincen”和“movies”数据库。该数据集的任务是生成自然语言至Cypher查询的转换。

This dataset is designed for training and evaluating the capability of converting natural language into Cypher queries. It is derived from an open-source text-to-Cypher dataset that contains over 16 distinct graph patterns. Additionally, this dataset also includes multiple graph patterns for evaluation purposes. In terms of scale, the training data is sampled from the aforementioned text-to-Cypher dataset, while the test set originates from the "fincen" and "movies" databases. The core task of this dataset is to generate transformations from natural language to Cypher queries.