README.md

metadata

license: apache-2.0
language:
  - id
  - en
tags:
  - e-commerce,
  - product-reviews

Tokopedia Product Reviews 2019

Dataset Description

This dataset contains 40,607 product reviews from Tokopedia, one of Indonesia's largest e-commerce platforms, scraped in 2019. The dataset provides valuable insights into customer sentiment and shopping behavior in the Indonesian e-commerce market.

Dataset Summary

• Language: Indonesian (Bahasa Indonesia)
• Task: Sentiment Analysis, Product Review Analysis, E-commerce Research
• Size: 40,607 reviews
• Categories: 5 product categories
• Unique Products: 3,647
• Collection Period: 2019

Dataset Structure

Data Fields

• text (string): The review text written by customers
• rating (int): Rating given by the reviewer (typically 1-5 scale)
• category (string): Product category, one of:
  • pertukangan (tools/hardware)
  • fashion (fashion)
  • elektronik (electronics)
  • handphone (mobile phones)
  • olahraga (sports)
• product_name (string): Name of the product
• product_id (string): Unique identifier for the product
• sold (int): Number of items sold
• shop_id (string): Unique identifier for the shop/seller
• product_url (string): URL link to the product page

Data Splits

The dataset comes as a single split containing all 40,607 reviews.

Dataset Statistics

Category	Count
Total Reviews	40,607
Unique Products	3,647
Product Categories	5
Language	Indonesian

Usage

Loading the Dataset

from datasets import load_dataset

# Load the dataset
dataset = load_dataset("farhamu/tokopedia-product-reviews-2019")

# Access the data
reviews = dataset['train']
print(f"Number of reviews: {len(reviews)}")
print(f"First review: {reviews[0]}")

Potential Use Cases

• Sentiment Analysis: Classify customer sentiment based on review text and ratings
• Product Recommendation: Analyze product preferences across different categories
• Market Research: Understand Indonesian e-commerce customer behavior
• Natural Language Processing: Train Indonesian language models for e-commerce domain
• Category Classification: Predict product categories from review text
• Rating Prediction: Predict customer ratings from review text

Data Collection

The data was collected through web scraping of Tokopedia product pages in 2019. The scraping process captured genuine customer reviews across five major product categories, providing a representative sample of customer feedback on the platform.

Ethical Considerations

• This dataset contains public reviews that were posted on Tokopedia's platform
• Personal information has been anonymized (shop_id and product_id are anonymized identifiers)
• The data reflects genuine customer opinions and experiences
• Users should be mindful of potential biases in the data (e.g., selection bias, temporal bias from 2019)

Limitations

• Temporal Limitation: Data is from 2019 and may not reflect current market trends
• Platform Specific: Limited to Tokopedia platform, may not generalize to other Indonesian e-commerce platforms
• Category Limitation: Only covers 5 product categories
• Language: Primarily in Indonesian, limiting applicability to other languages

Citation

If you use this dataset in your research, please cite:

@dataset{tokopedia-product-reviews-2019,
  title={Tokopedia Product Reviews 2019},
  author={[M. Farhan]},
  year={2025},
  url={https://huggingface.co/datasets/farhamu/tokopedia-product-reviews-2019},
  note={Indonesian e-commerce product reviews dataset from Tokopedia platform}
}

Contact

For questions or issues regarding this dataset, please open an issue in the dataset repository or contact [email protected].

Acknowledgments

• Thanks to Tokopedia for providing a platform that enables customer reviews
• This dataset was originally shared on Kaggle but is now maintained on Hugging Face for better accessibility and integration with ML workflows