Doukan/trendyol-turkish-product-reviews

⸻

license: cc-by-nc-4.0 language: • tr tags: • turkish • nlp • sentiment-analysis • ecommerce • text-classification • reviews pretty_name: Trendyol Turkish Product Reviews

⸻

582K+ Turkish E-Commerce Reviews from Trendyol

A large-scale Turkish product reviews dataset collected from Trendyol, designed for Natural Language Processing (NLP), sentiment analysis, and consumer opinion mining. The dataset captures real-world, user-generated Turkish text, including informal language, spelling variations, emojis, and very short comments.

This dataset aims to help researchers and practitioners working on low-resource Turkish NLP tasks by providing a rich and realistic corpus of e-commerce reviews.

⸻

📦 Dataset Overview • Total reviews: 582,601 • Total products: 500 • Categories: 5 • Language: Turkish • Source: Publicly available product review pages on Trendyol • License: CC BY-NC 4.0 (Non-commercial use only)

⸻

🗂️ Product Categories • Aksesuar • Giyim • Kozmetik & Kişisel Bakım • Kitap • Elektronik

Each category includes reviews from 100 distinct products, ensuring balanced product coverage while preserving natural differences in review volume.

⸻

📊 Data Characteristics • Short-text heavy: ~37% of reviews contain 5 words or fewer • Noisy real-world text: spelling errors, elongations, emojis, informal expressions • Rating-aligned: each review includes a numerical rating score • Multi-purpose: suitable for both classification and exploratory analysis

This makes the dataset especially useful for: • Short-text sentiment analysis • Turkish text normalization • Robust NLP model evaluation • Review usefulness and consistency studies

⸻

🧾 Data Fields

Each entry represents a single user review and contains the following columns: • category – Product category • producturl – Source product page URL • user – User identifier (as collected) • date – Review date (string format) • point – Rating score • comment – Review text (Turkish) • seller – Seller name • imagecount – Number of images attached to the review • seller_rate – Seller rating (if available)

⸻

🎯 Intended Use

This dataset is intended for research and educational purposes, including but not limited to: • Sentiment analysis and opinion mining • Text classification and topic modeling • Category-aware NLP modeling • Benchmarking Turkish language