RDRanking

Citation Author(s):
Ming
Xu
Submitted by:
Ming Xu
Last updated:
Sun, 04/27/2025 - 00:07
DOI:
10.21227/yzdw-8x75
License:
0
0 ratings - Please login to submit your rating.

Abstract 

The dataset is constructed using SUMO. It contains two road network datasets of different scales: a small-scale network (SR) and a larger regional network in Shenyang (SY). The dataset was constructed using the SUMO simulation platform, containing two road network datasets at different scales: a small-scale test network (SR) and a regional-level Shenyang network (SY). The SR network comprises 110 road segments, while the SY network contains 514 segments. Each segment includes features (e.g., speed limit, number of lanes, etc.) and importance labels. To obtain the ground-truth ranking of segment importance, we implemented the following procedure: First, a road segment was considered "failed" when its average travel speed dropped below 10% of the speed limit. We then artificially simulated failure conditions by sequentially reducing each segment's maximum speed limit to 10% of its original value. During specified time windows, we recorded the impact of each segment's failure on the overall network efficiency, using this Efficiency Degradation (ED) value as the importance score for the corresponding segment.. 

Instructions: 

# Dataset Description

 

## Overview

 

This dataset contains **two subsets**, named **RN** and **SY**.

 Each subset is organized as a separate folder and contains the following files:

 

- `feature.txt`

- `graph_entity.txt`

- `graph_entity_feature.txt`

- `graph_entity_feature_ld.txt`

- `link_forward_link.txt`

- `node_attribute.txt`

- `od.txt`

- `real.txt`

 

These datasets are designed to support research on heterogeneous graphs involving OD demands, paths, and road segments.

 They consist of node-level features, node attributes, link structures, and heterogeneous relational graphs.

 

All files are encoded in **UTF-8** format.

 

------

 

## Files and Detailed Descriptions

 

| Filename                      | Description                                                  |

| :---------------------------- | :----------------------------------------------------------- |

| `feature.txt`                 | Stores the features of each node (categorical or descriptive features rather than numerical feature vectors). |

| `graph_entity.txt`            | Represents the heterogeneous graph containing OD nodes, path nodes, and link nodes, and their interconnections. |

| `graph_entity_feature.txt`    | Represents an extended heterogeneous graph connecting OD/path/link nodes to their respective attribute nodes. |

| `graph_entity_feature_ld.txt` | A variant of `graph_entity_feature.txt`, adapted for experiments requiring latent dimension variations. |

| `link_forward_link.txt`       | Lists the composition of each path in terms of its ordered link segments. |

| `node_attribute.txt`          | Records the type of each node:  - `0`: OD, path, or link nodes  - `1`: Feature attribute nodes |

| `od.txt`                      | Contains the IDs of OD demand nodes.                         |

| `real.txt`                    | Contains the real-valued scores for each link (road segment), representing ground-truth evaluation metrics. |

 

## Node Types

 

Nodes in the dataset are categorized as follows:

 

- **OD Nodes**: Represent origin-destination points of traffic demand.

- **Path Nodes**: Represent specific paths composed of sequences of road segments (links).

- **Link Nodes**: Represent individual road segments.

- **Attribute Nodes**: Represent features or properties connected to the OD, path, and link nodes.

 

The node type is explicitly indicated in `node_attribute.txt`.

 

------

 

## Graph Structures

 

- **Heterogeneous Graph (Entity-Level)**:

   `graph_entity.txt` defines the direct relationships among OD, path, and link nodes.

   Each line typically contains a pair or triplet indicating a relationship (depending on format).

- **Extended Heterogeneous Graph (Entity + Attribute-Level)**:

   `graph_entity_feature.txt` adds attribute nodes to the entity graph, connecting OD/path/link nodes to their descriptive features.

- **Latent Dimension Variant**:

   `graph_entity_feature_ld.txt` provides a modified version for experiments that may require adjusted feature richness or compression.

- **Path Composition**:

   `link_forward_link.txt` specifies the ordered link sequence that constitutes each path.

- **Real Scores**:

    `real.txt` assigns real-valued scores to each link, which can be used as ground-truth supervision in downstream tasks such as link ranking or regression.

 

## Usage Notes

 

- Features are **descriptive** and are not directly numerical vectors.

- When constructing graphs, the node types should be referred to from `node_attribute.txt` to ensure correct relationship modeling.

- In `link_forward_link.txt`, the links are **ordered**, reflecting the actual sequence within the path.

- The real scores in `real.txt` can be used for supervised learning, evaluation, or ranking tasks.

Dataset Files

    Files have not been uploaded for this dataset

    Documentation

    AttachmentSize
    File readme(1).md3.69 KB