Update README.md
Browse files
README.md
CHANGED
|
@@ -1,3 +1,29 @@
|
|
| 1 |
-
|
| 2 |
-
|
| 3 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---# RideSearch Dataset
|
| 2 |
+
|
| 3 |
+
## Overview
|
| 4 |
+
RideSearch is a custom dataset created for a vehicle recommendation system.
|
| 5 |
+
It contains structured information about different car models, trims, years, and countries of origin, designed to support a search and recommendation application.
|
| 6 |
+
|
| 7 |
+
## Dataset Structure
|
| 8 |
+
- **Modality:** Tabular
|
| 9 |
+
- **Rows:** 1,250
|
| 10 |
+
- **Columns:**
|
| 11 |
+
- `name`: Car make and model
|
| 12 |
+
- `model`: Model name
|
| 13 |
+
- `trim`: Trim or edition
|
| 14 |
+
- `year`: Year of manufacture
|
| 15 |
+
- `origin_country`: Country where the car was manufactured
|
| 16 |
+
- Additional attributes for filtering and similarity search.
|
| 17 |
+
|
| 18 |
+
## Purpose
|
| 19 |
+
The dataset was built to power the RideSearch Hugging Face app, which allows users to find similar vehicles based on their selected input.
|
| 20 |
+
It is intended for educational purposes as part of a university AI application project.
|
| 21 |
+
|
| 22 |
+
## Usage
|
| 23 |
+
1. Input a car from the dataset into the RideSearch app.
|
| 24 |
+
2. The system returns the top 3 most similar vehicles based on attributes.
|
| 25 |
+
3. Designed for embedding-based similarity search.
|
| 26 |
+
|
| 27 |
+
## License
|
| 28 |
+
MITlicense: mit
|
| 29 |
+
---
|