BIOSNAP_DDI_and_genes_data

<strong># NamE Datasets</strong><br> Jeffrey Seathrún Sardina<br> <br> <strong>## Information and Use</strong><br> Datasets for the paper "NamE: Named Graph Embeddings for Context Modelling and Expert Knowledge Integration in Knowledge Graph Embeddings". Datasets are given with all variant forms, and with the original train-test-valid spits used in the paper. Train, test, and valid files are marked as such in the file prefix (e.g. "*-train.csv.gz").<br> <br> Datasets beginning in "triples-" contain triples in a comma-separated format; i.e.<br> ```<br> subject,predicate,object<br> subject,predicate,object<br> subject,predicate,object<br> ```<br> <br> Datasets beginning in "quads-" contain quads in a comma-separated format; i.e.<br> ```<br> subject,predicate,object,namedGraph<br> subject,predicate,object,namedGraph<br> subject,predicate,object,namedGraph<br> ```<br> <br> All datasets are gzipped; on most systems (Linux / MAC) you can uncompress these with "gunzip (file)" if needed.<br> <br> <strong>## Citation</strong><br> If you use these datasets or NamE in your work, please cite "NamE: Named Graph Embeddings for Context Modelling and Expert Knowledge Integration in Knowledge Graph Embeddings"<br> ```<br> BIBTEX citation pending<br> ```<br> <br> Included below are citations to the data sources from which these datasets were created. Note that the triples form of FB15K-237, as it appears in this repository, is unmodified from its original version.<br> <br> <strong>**BIOSNAP dataset citation:**</strong><br> ```<br> @misc{biosnap,<br> author = {Marinka Zitnik, Rok Sosi\v{c}, Sagar Maheshwari, and Jure Leskovec},<br> title = {{BioSNAP Datasets}: {Stanford} Biomedical Network Dataset Collection},<br> howpublished = {\url{http://snap.stanford.edu/biodata}},<br> month = aug,<br> year = 2018<br> }<br> ```

# NamE 数据集
作者：Jeffrey Seathrún Sardina

## **数据集信息与使用说明**
本数据集对应论文《NamE：面向知识图谱嵌入（Knowledge Graph Embeddings）中上下文建模与专家知识融合的命名图嵌入（Named Graph Embeddings）》。所有数据集均提供完整变体形式，并附带论文中使用的原始训练-测试-验证划分方案。训练集、测试集与验证集的文件前缀会标注对应用途（例如"*-train.csv.gz"）。

前缀为"triples-"的数据集采用逗号分隔格式存储三元组，格式如下：

subject,predicate,object
subject,predicate,object
subject,predicate,object

即：主语、谓语、宾语。

前缀为"quads-"的数据集采用逗号分隔格式存储四元组，格式如下：

subject,predicate,object,namedGraph
subject,predicate,object,namedGraph
subject,predicate,object,namedGraph

即：主语、谓语、宾语、命名图（namedGraph）。

所有数据集均经gzip压缩；在多数Linux或macOS系统中，可通过`gunzip <文件名>`命令按需解压。

## **引用说明**
若您在研究工作中使用本数据集或NamE方法，请引用论文《NamE：面向知识图谱嵌入中上下文建模与专家知识融合的命名图嵌入》。

BIBTEX citation pending

以下为本数据集所依赖的原始数据源引用。需注意，本仓库中FB15K-237的三元组形式与原始版本完全一致，未作任何修改。

**BIOSNAP数据集引用：**

@misc{biosnap,
author = {Marinka Zitnik, Rok Sosiv{c}, Sagar Maheshwari, and Jure Leskovec},
title = {{BioSNAP Datasets}: {Stanford} Biomedical Network Dataset Collection},
howpublished = {url{http://snap.stanford.edu/biodata}},
month = aug,
year = 2018
}