Korean Workplace Document Understanding Dataset

#information extraction #task and deadline identification #relationship extraction #Korean workplace communication #business Korean learning #workplace document understanding

5000
0.4GB
JSON
CATL
MOBIUSI INC

Updated:2026-10-19

AI Analysis & Value Prop

Designed for Korean for Business learning and document understanding, this resource contains JSON-structured samples of common workplace materials, including emails, meeting notices, schedules, internal process guides, leave policies, and handover documents. Annotations cover document categories and topics, tasks and assignees, deadlines, request intent, priority, execution conditions, and implied collaboration relationships, supporting models that identify key information and action requirements in workplace communication. Check the product page for the exact sample count, field format, and licensing terms.

Dataset Insights

Sample Examples

{}

Technical Specifications

Field	Type	Description
file_name	string	File name
priority	string	The overall handling priority inferred from the wording, deadlines, and business impact.
conditions	array	Conditions, prerequisites, and restrictions that affect task execution, leave requests, scheduling, or process applicability.
deadline_info	object	A structured representation of deadlines, time ranges, and recurring requirements mentioned in the document.
task_subjects	array	Tasks and action items identified in the document, including their responsible parties.
request_intent	string	The main action, confirmation, response, or information the author expects from the recipient.
source_document	object	The Korean workplace document to analyze, along with basic source information.
document_subject	string	A summary of the document's main topic or communication matter.
document_category	string	The workplace document category identified from its content.
collaboration_relations	array	Relationships among people, teams, or departments inferred from task assignment, approval, notification, handover, and feedback requirements.

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

Frequently Asked Questions

What types of Korean workplace documents are included?: Samples cover common workplace emails, meeting notices, schedules, internal process guides, leave policies, and handover documents, organized in JSON format.

Which understanding tasks can the annotations support?: Fields cover document category and topic, tasks and assignees, deadlines, request intent, priority, conditions, and implied collaboration relationships. These annotations can support document classification, information extraction, and semantic understanding.

How can I access the data?: Check the product page for available previews, download options, or delivery methods. The published sample count, file format, and delivery scope are authoritative.

What are the licensing terms?: The permitted uses and restrictions depend on the product's specific license. Review the terms on the product page and confirm that they cover your intended research, educational, or commercial use.

Can't find the data you need?

Post a request and let data providers reach out to you.

Get this Dataset

Verified for Enterprise Use

Cite this Work

@dataset{Mobiusi2026,
  title={Korean Workplace Document Understanding Dataset},
  author={MOBIUSI INC},
  year={2026},
  url={https://www.mobiusi.com/datasets/9024e5054eaf5321e6fcd74da99f7492},
  urldate={2026-10-16},
  keywords={Korean workplace document dataset,Korean document understanding,Korean for business learning,Korean email analysis,Korean information extraction,document semantic understanding,Korean task identification,workplace collaboration recognition},
  version={1.0}
}

Using this in research? Please cite us.

placeholder

Products

Scene

Domain

Modality

Task