遇见数据集

Medium-sized Multi-labeled Graph Datasets

收藏
Zenodo2025-12-20 更新2026-05-26 收录
官方服务:

资源简介:

This record contains five medium-sized multi-labeled graph datasets: dblp, enron, github, gowalla, and wikitalk. These datasets are originally from SNAP. These files follow the structure below: 1. Header Line (`t`): Starts with `t`, followed by two integers representing: Total number of vertices Total number of edges 2. Vertex Lines (`v`): Each line starts with `v`, followed by three integers per vertex: Vertex ID (0-based index) Label (arbitrary numeric value) Degree (number of edges connected to this vertex) 3. Edge Lines (`e`): Each line starts with `e`, followed by two vertex IDs (0-based) representing an undirected edge between them. Edges are unordered (e.g., `e 0 1` is the same as `e 1 0`). This format defines an undirected graph where vertices are labeled and degrees are explicitly listed for each node. For a graph $g$, the first line should be `t` $|V(g)|$ $|E(g)|$, followed by $|V(g)|$ vertex lines and $|E(g)|$ edge lines. Each vertex $v\in V(g)$ has a unique vertex ID from 0 to $|V(g)|-1$.

提供机构:
Zenodo
创建时间:
2025-12-19
二维码
社区交流群
二维码
科研交流群
商业服务