UAS Multi-Temporal Disaster Change Reasoning QA Dataset
- Records
- 5750
- Size
- 0.4GB
- Format
- JSON
- Provider
MOBIUSI INC- Updated
- v1.02026-11-04
This dataset presents pre- and post-disaster UAS mapping observations as question-answer examples, with explicit rules for distinguishing newly observed changes, pre-existing damage, and cases where evidence is insufficient. Each example includes a question, answer, change classification, reasoning, and evidence summary, supporting reasoning SFT and disaster change attribution evaluation while helping reduce false attribution of existing conditions as new damage. The dataset size is not specified in the available information.
Overview
Dataset Insights
Preview
Sample Examples
Schema
Technical Specifications
| Field | Type | Description |
|---|---|---|
| file_name | string | File name |
| question | string | The question to answer by comparing the pre- and post-disaster survey records and applying the inference rules. |
| reasoning | string | A brief explanation of how the pre- and post-disaster records and stated rules support the classification. |
| answer_text | string | A concise natural-language answer grounded in the temporal survey evidence and inference rules. |
| inference_rules | string | Explicit rules for distinguishing newly observed changes, pre-existing damage, and cases where the evidence is insufficient. |
| evidence_summary | string | A concise summary of the observations supporting the conclusion, including relevant temporal comparisons or missing evidence. |
| pre_disaster_survey | string | Textualized UAS mapping observations recorded before the disaster, including any damage or conditions already present. |
| post_disaster_survey | string | Textualized UAS mapping observations recorded after the disaster. |
| change_classification | string | The inferred temporal status of the relevant condition. |
Trust
Compliance Statement
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
Frequently Asked Questions
- What does the dataset contain?
- Examples include pre- and post-disaster UAS mapping records, inference rules, questions and answers, as well as change classifications, reasoning, and evidence summaries.
- How does it distinguish new damage from pre-existing damage?
- Compare the pre- and post-disaster records over time and apply the supplied inference rules to classify a condition as a new change, pre-existing damage, or a case where the available evidence is insufficient.
- What are the dataset size and annotation quality?
- The available information does not specify the number of examples or quantitative quality metrics. The structured rules, classifications, and evidence summaries make it easier to verify whether conclusions are supported by temporal evidence.
- Which tasks is it suitable for, and how can I access it and check its license?
- It is suitable for reasoning SFT, disaster change attribution, change classification, and evidence-grounded question-answering evaluation. Refer to the dataset listing or license agreement for access details and permitted use.
Can't find the data you need?
Post a request and let data providers reach out to you.
Get this Dataset
High-quality data, ready for your next AI breakthrough.
Cite this Work
@dataset{Mobiusi2026,
title={UAS Multi-Temporal Disaster Change Reasoning QA Dataset},
author={MOBIUSI INC},
year={2026},
url={https://www.mobiusi.com/datasets/d07a8f5a2d437a17dcac4196c82c572f?dataset_scene_id=10},
urldate={2026-10-26},
keywords={UAS disaster dataset,pre and post-disaster change detection,disaster mapping QA,disaster damage assessment,multi-temporal reasoning data,UAS mapping data,disaster change attribution,reasoning SFT dataset},
version={1.0}
}Using this in research? Please cite us.