Taobao Serendipity

Survey-based recommendation feedback that captures surprise, relevance and satisfaction beyond click accuracy.

What the resource contains

  • 11,383 respondent records from a 2017–2018 study
  • Anonymized user, item and category IDs with click/purchase indicators
  • Perceived novelty, serendipity, satisfaction and personality questionnaire responses

Potential uses

  • Evaluate recommendation diversity
  • Study satisfaction versus clicks
  • Research perceived novelty

Access & formats

CSV. Repository files. Follow the project’s instructions on GitHub; some resources require external files, registration or approval.

License & reuse

The dataset permits research with attribution; commercial or revenue-bearing use needs prior researcher permission. Review its redistribution wording directly.

Important limitations

Survey and behavioral records concern people. Use aggregate analysis and do not attempt identification or sensitive profiling.

Source & attribution

greenblue96 — original GitHub project ↗

Documentation reviewed 2026-09-24. Review scope and licensing notes. Suggest a correction.

Related resources

Sessions & recommendations

OTTO Sessions

otto-de

An ecommerce interaction benchmark for predicting what a shopper will click, add to cart, or order next.

Open license stated

JSONL · External dependencies