A differential privacy workflow for inference of parameters in the rasch model

Teresa Anna Steiner; David Enslev Nyrnberg; Lars Kai Hansen

doi:10.1007/978-3-030-13463-1_9

A differential privacy workflow for inference of parameters in the rasch model

Teresa Anna Steiner, David Enslev Nyrnberg, Lars Kai Hansen^*

^*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceeding › Article in proceedings › Research › peer-review

Abstract

The Rasch model is used to estimate student performance and task difficulty in simple test scenarios. We design a workflow for enhancing student feedback by release of difficulty parameters in the Rasch model with privacy protection using differential privacy. We provide a first proof of differential privacy in Rasch models and derive the minimum noise level in objective perturbation to guarantee a given privacy budget. We test the workflow in simulations and in two real data sets.

Original language	English
Title of host publication	Proceedings of ECML PKDD 2018 Workshops - MIDAS 2018 and PAP 2018
Editors	Anna Monreale, Carlos Alzate
Publisher	Springer
Publication date	1 Jan 2019
Pages	113-124
ISBN (Print)	9783030134624
DOIs	https://doi.org/10.1007/978-3-030-13463-1_9
Publication status	Published - 1 Jan 2019
Event	European Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases 2018 - Dublin, Ireland Duration: 10 Sep 2018 → 14 Sep 2018 http://www.ecmlpkdd2018.org/

Conference

Conference	European Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases 2018
Country/Territory	Ireland
City	Dublin
Period	10/09/2018 → 14/09/2018
Internet address	http://www.ecmlpkdd2018.org/

Series	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	11054 LNAI
ISSN	0302-9743

Keywords

Differential privacy
Rasch model
Student feedback

Access to Document

10.1007/978-3-030-13463-1_9

SFX availability

Full text

Cite this

Steiner, T. A., Nyrnberg, D. E., & Hansen, L. K. (2019). A differential privacy workflow for inference of parameters in the rasch model. In A. Monreale, & C. Alzate (Eds.), Proceedings of ECML PKDD 2018 Workshops - MIDAS 2018 and PAP 2018 (pp. 113-124). Springer. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) Vol. 11054 LNAI https://doi.org/10.1007/978-3-030-13463-1_9

Steiner, Teresa Anna ; Nyrnberg, David Enslev ; Hansen, Lars Kai. / A differential privacy workflow for inference of parameters in the rasch model. Proceedings of ECML PKDD 2018 Workshops - MIDAS 2018 and PAP 2018. editor / Anna Monreale ; Carlos Alzate. Springer, 2019. pp. 113-124 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), Vol. 11054 LNAI).

@inproceedings{47f80231d7e24ca6b0e6095cf75309b6,

title = "A differential privacy workflow for inference of parameters in the rasch model",

abstract = "The Rasch model is used to estimate student performance and task difficulty in simple test scenarios. We design a workflow for enhancing student feedback by release of difficulty parameters in the Rasch model with privacy protection using differential privacy. We provide a first proof of differential privacy in Rasch models and derive the minimum noise level in objective perturbation to guarantee a given privacy budget. We test the workflow in simulations and in two real data sets.",

keywords = "Differential privacy, Rasch model, Student feedback",

author = "Steiner, {Teresa Anna} and Nyrnberg, {David Enslev} and Hansen, {Lars Kai}",

year = "2019",

month = jan,

day = "1",

doi = "10.1007/978-3-030-13463-1_9",

language = "English",

isbn = "9783030134624",

series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

publisher = "Springer",

pages = "113--124",

editor = "Anna Monreale and Carlos Alzate",

booktitle = "Proceedings of ECML PKDD 2018 Workshops - MIDAS 2018 and PAP 2018",

note = "European Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases 2018, ECML-PKDD 2018 ; Conference date: 10-09-2018 Through 14-09-2018",

url = "http://www.ecmlpkdd2018.org/",

}

Steiner, TA, Nyrnberg, DE & Hansen, LK 2019, A differential privacy workflow for inference of parameters in the rasch model. in A Monreale & C Alzate (eds), Proceedings of ECML PKDD 2018 Workshops - MIDAS 2018 and PAP 2018. Springer, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 11054 LNAI, pp. 113-124, European Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases 2018, Dublin, Ireland, 10/09/2018. https://doi.org/10.1007/978-3-030-13463-1_9

A differential privacy workflow for inference of parameters in the rasch model. / Steiner, Teresa Anna; Nyrnberg, David Enslev; Hansen, Lars Kai.

Proceedings of ECML PKDD 2018 Workshops - MIDAS 2018 and PAP 2018. ed. / Anna Monreale; Carlos Alzate. Springer, 2019. p. 113-124 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), Vol. 11054 LNAI).

Research output: Chapter in Book/Report/Conference proceeding › Article in proceedings › Research › peer-review

TY - GEN

T1 - A differential privacy workflow for inference of parameters in the rasch model

AU - Steiner, Teresa Anna

AU - Nyrnberg, David Enslev

AU - Hansen, Lars Kai

PY - 2019/1/1

Y1 - 2019/1/1

N2 - The Rasch model is used to estimate student performance and task difficulty in simple test scenarios. We design a workflow for enhancing student feedback by release of difficulty parameters in the Rasch model with privacy protection using differential privacy. We provide a first proof of differential privacy in Rasch models and derive the minimum noise level in objective perturbation to guarantee a given privacy budget. We test the workflow in simulations and in two real data sets.

AB - The Rasch model is used to estimate student performance and task difficulty in simple test scenarios. We design a workflow for enhancing student feedback by release of difficulty parameters in the Rasch model with privacy protection using differential privacy. We provide a first proof of differential privacy in Rasch models and derive the minimum noise level in objective perturbation to guarantee a given privacy budget. We test the workflow in simulations and in two real data sets.

KW - Differential privacy

KW - Rasch model

KW - Student feedback

U2 - 10.1007/978-3-030-13463-1_9

DO - 10.1007/978-3-030-13463-1_9

M3 - Article in proceedings

AN - SCOPUS:85063419666

SN - 9783030134624

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 113

EP - 124

BT - Proceedings of ECML PKDD 2018 Workshops - MIDAS 2018 and PAP 2018

A2 - Monreale, Anna

A2 - Alzate, Carlos

PB - Springer

T2 - European Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases 2018

Y2 - 10 September 2018 through 14 September 2018

ER -

Steiner TA, Nyrnberg DE, Hansen LK. A differential privacy workflow for inference of parameters in the rasch model. In Monreale A, Alzate C, editors, Proceedings of ECML PKDD 2018 Workshops - MIDAS 2018 and PAP 2018. Springer. 2019. p. 113-124. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), Vol. 11054 LNAI). https://doi.org/10.1007/978-3-030-13463-1_9

A differential privacy workflow for inference of parameters in the rasch model

Abstract

Conference

Keywords

Access to Document

SFX availability

Fingerprint

Cite this