---
task_categories:
- tabular-classification
---
## Towards Full Candidate Interaction: A Comprehensive Comparison Network for Better Route Recommendation

This is the dataset for our paper.<br>
The following table contains the feature dimensions and key features of our dataset.

|  Feature Type         | Interpretation  |  Shape  |    Some Key Features
|        :----          | :----         | :----:  |        :---          |
| Route Features        | Used to describe each route, including static features,<br> dynamic features, and trajectory statistical features              | N * 62 | The estimated time of arrival for the route<br>The total distance length of the route |
| Scene Features        | Represent the contextual information of route recommendation            |1 * 10 | Request time<br>User familiarity with the origin and destination |
| Comparison-Level Features  | Represents the attributes of the non-overlapping segments between two routes  | N * N * 27 | Comparison diff eta<br>Comparison diff distance |