product-thai-114k / README.md
Porameht's picture
Update README.md
526bcb2 verified
metadata
dataset_info:
  features:
    - name: title
      dtype: string
    - name: name
      dtype: string
    - name: product_url
      dtype: string
    - name: price
      dtype: float32
    - name: original_price
      dtype: float32
    - name: unit
      dtype: string
    - name: overall_rating
      dtype: float32
    - name: rating_count
      dtype: int32
    - name: image
      dtype: image
  splits:
    - name: train
      num_bytes: 1762857315.375
      num_examples: 113677
  download_size: 1118819788
  dataset_size: 1762857315.375
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*

Product Thai Dataset

Dataset Description

The Product Thai Dataset is a comprehensive collection of Thai product information, including images and ratings, designed for e-commerce and product analysis tasks specific to the Thai market. This dataset contains 113,677 examples, providing a rich source of information for researchers and developers working on product recommendation systems, image-based product search, and e-commerce analytics in the context of Thai products.

Dataset Summary

  • Number of examples: 113,677
  • Dataset size: 1.76 GB
  • Download size: 1.12 GB

Features

The dataset includes the following features for each product:

  1. title (string): The title of the product in Thai.
  2. name (string): The name of the product (may differ from title) in Thai.
  3. product_url (string): URL link to the product page.
  4. price (float32): Current price of the product in Thai Baht (฿).
  5. original_price (float32): Original price of the product in Thai Baht (฿) (before any discounts).
  6. unit (string): Unit of measurement or sale for the product in Thai.
  7. overall_rating (float32): Average rating of the product.
  8. rating_count (int32): Number of ratings received by the product.
  9. image (image): Product image.

Usage

This dataset can be used for various tasks, including but not limited to:

  • Thai product image classification
  • Price prediction for Thai market
  • Rating prediction for Thai products
  • Product recommendation systems for Thai e-commerce
  • Visual search and similarity for Thai products
  • E-commerce trend analysis in Thailand

To load the dataset using the Hugging Face datasets library:

from datasets import load_dataset

dataset = load_dataset("Porameht/product-thai-114k")

Data Splits

The dataset is provided as a single train split.

Dataset Creation

Source

The data was collected from various Thai e-commerce platforms and product listings. The specific sources and collection methodology are not provided to maintain privacy and data integrity.

Preprocessing

The dataset has undergone cleaning and preprocessing to ensure consistency and usability:

  • Missing values have been handled appropriately for each feature type.
  • Image data has been standardized and preprocessed for machine learning applications.
  • Numerical values (price, rating, etc.) have been normalized and checked for consistency.
  • Thai text has been preserved and standardized where necessary.

Considerations for Use

When using this dataset, please consider the following:

  • The dataset may contain biases inherent in Thai e-commerce data, such as popularity bias or category imbalance.
  • Product information and pricing data may become outdated over time.
  • Images and product details should be used in compliance with copyright and fair use guidelines.
  • The dataset is specific to the Thai market and may not be generalizable to other regions without adaptation.

Citation

If you use this dataset in your research or applications, please cite it as follows:

@dataset{product_thai_dataset,
  author = {Porameht},
  title = {Product Thai Dataset},
  year = {2023},
  publisher = {Hugging Face},
  url = {https://huggingface.co/datasets/Porameht/product-thai-114k}
}

Feedback and Contributions

For any questions, feedback, or contributions to improve this dataset, please open an issue on the dataset's GitHub repository or contact the dataset maintainer through Hugging Face.