Published November 25, 2025 | Version v1
Dataset Open

Standards for transparent AI in Human Resource Management (TRANKI) – Research data

  • 1. ROR icon HTW Berlin - University of Applied Sciences

Description

Disclaimer

The following data-set was collected in February 2025 for the research Project “TRANKI – Standards for transparent AI”, funded by the Hans-Böckler-Foundation (Grant No. 2022-797-2).

The goal was to assess the effects of AI literacy on the interpretation of AI-enhanced user interfaces of HR software, and how explainable AI (XAI) elements affect interpretation results. The project’s research questions were:

  • What approaches at the user interface level help make AI systems in human resources transparent for users?
  • What differences in transparency requirements and perceptions exist between employees, managers, and HR experts?

All items are included with their German labels and English translations. The translation process was automated, using LLM model mistral-medium-2508, and proof-read by the authors. Please feel free to translate further as needed.

Folder structure and file tree


📂 Tranki-data/

├── 📄 tranki_data.csv    # CSV Data set
├── 📄 varnames.csv       # Variable names German
├── 📄 varlabels.csv      # Variable labels German
├── 📄 varnames_engl.csv  # Variable names English
├── 📄 varlabels_engl.csv # Variable labels English

├── 📄 LICENSE.md         # License
├── 📄 README.pdf         # Further details and instructions
└── 📄 CHANGELOG.md       # Version history and changes

For detailed information and guidelines regarding the research dataset, refer to the accompanying README.pdf.

Relevant publication

cite with our publication: 

Kalff, Y., & Simbeck, K. (2025). Explained, yet misunderstood: How AI Literacy shapes HR Managers’ interpretation of User Interfaces in Recruiting Recommender Systems. In M. Kaya, T. Bogers, G. Bied, C. Johnson, & J.-J. Decorte (Hrsg.), Proceedings of the 5th Workshop on Recommender Systems for Human Resources (RecSys in HR 2025) (Bd. 4046). Gehalten auf der RecSys in HR 2025, Prague: CEUR. https://ceur-ws.org/Vol-4046/RecSysHR2025-paper_3.pdf. Zugegriffen: 28. September 2025

 

 

 

Files

README.pdf

Files (1.5 MB)

Name Size Download all
md5:2b9a2e984ee77cf091cbb0f424b5875b
56 Bytes Preview Download
md5:9a788123ae6c2a2d9b2b4cf673e950e7
18.7 kB Preview Download
md5:cea2a5d595c42079fc588d1eb46a7e09
1.1 MB Preview Download
md5:c13504c848a609bb8a6c2139a4988eb4
222.7 kB Preview Download
md5:b8621024987243879df840de73bb9d94
34.2 kB Preview Download
md5:5e9d6f73a5f2a0692b537497afa8acb0
30.3 kB Preview Download
md5:148f96df8ab08e2bcfbe35fed97c35c4
51.6 kB Preview Download
md5:f7c72613778d6d14faa1e88bafa153ec
49.1 kB Preview Download

Additional details

Related works

Is supplement to
Conference paper: arXiv:2509.06475 (arXiv)
Conference proceeding: urn:nbn:de:0074-4046-X (URN)
Conference paper: https://ceur-ws.org/Vol-4046/RecSysHR2025-paper_3.pdf (URL)

Funding

Hans Böckler Foundation
2022-797-2

Dates

Other
2024-12
Pretested
Collected
2025-02
Available
2025-11

Software

Programming language
R
Development Status
Active