Seekysense/IT-Running-Shoes-Descriptions

Italian Running Shoes Dataset (Multimodal)

This dataset contains a collection of running shoe products specifically curated from the Italian market. It is designed for tasks such as Computer Vision (Image Classification), Natural Language Processing (NLP) in Italian, and E-commerce recommendation systems.

📂 Dataset Structure

The dataset consists of a central metadata file and a folder containing product images.

• images/: Directory containing .jpg images of the shoes.
• shoes_metadata.csv: A CSV file containing product details in Italian.

📊 Data Dictionary

The shoes_metadata.csv file uses a comma-separated format with the following columns:

Column	Description	Example
brand	The manufacturer of the shoe.	Adidas
nome	The product name and colorway.	Runfalcon 5 Nero
descrizione	A detailed description in Italian, often including discounts.	Adidas Scarpa da Running Donna...
prezzo	The price in Euro (€).	60,00
immagine	The filename of the corresponding image in the /images folder.	ADIDIE8826_VAR_01_3480.jpg

📝 Sample Data

"brand","nome","descrizione","prezzo","immagine"
"Adidas","Runfalcon 5 Nero","Adidas Scarpa da Running Donna Runfalcon 5 Nero","60,00","ADIDIE8826_VAR_01_3480.jpg"
"Adidas","Adistar 3 Blu Fucsia","Adidas Scarpa da Running Donna Adistar 3 Blu Fucsia Sconto 10%","126,00","ADIDJI1230_VAR_01_38b8.jpg"

🚀 Use Cases

1. Image Classification: Training models to recognize shoe brands or specific models.
2. Price Prediction: Analyzing the correlation between product descriptions and their retail price.
3. Italian NLP: Fine-tuning LLMs or NER (Named Entity Recognition) models to extract features (gender, color, discount) from Italian commercial text.

🛠 How to Use with Python

import pandas as pd

# Load the metadata
df = pd.read_csv('shoes_metadata.csv')

# Display the first few rows
print(df.head())