MOBIUSI INCThis dataset connects questions from educational question banks and their standard answers with webpage text, candidate passages, and precise supporting evidence spans. Each structured record can include the question and answer, a source webpage, retrieved passages, and the relationship between evidence spans and the standard answer, supporting the development and evaluation of evidence retrieval and answer attribution in education. It helps models locate relevant information in web content and makes the grounding of educational answers more traceable.
Overview
Preview
Schema
| Field | Type | Description |
|---|---|---|
| file_name | string | File name |
| webpage_input | object | The original webpage record used as a source for locating evidence that supports the answer. |
| evidence_spans | array | Precise text spans within retrieved passages, including their relationship to the standard answer. |
| retrieved_passages | array | Passages selected from the source webpage as potentially relevant to the question and answer. |
| question_answer_input | object | The original question and its standard answer, provided as the input for evidence retrieval. |
Trust
Authorization Type
Proprietary - Commercial AI Training License (No Redistribution)
Commercial Use
Requires exclusive subscription or authorization contract (monthly or per-invocation charging)
Privacy and Anonymization
No PII, no real company names, simulated scenarios follow industry standards
Compliance System
Compliant with China's Data Security Law / EU GDPR / supports enterprise data access logs
FAQ
Post a request and let data providers reach out to you.
High-quality data, ready for your next AI breakthrough.
@dataset{Mobiusi2026,
title={Educational Question Evidence Retrieval Dataset},
author={MOBIUSI INC},
year={2026},
url={https://www.mobiusi.com/datasets/fb3cb0d7e72c2abf6f7d01f62a798378?dataset_scene_id=7},
urldate={2026-10-22},
keywords={educational question dataset,evidence retrieval data,question answer grounding,educational QA dataset,web text retrieval,evidence span annotation,answer attribution,learning resource linking},
version={1.0}
}Using this in research? Please cite us.