SlideShare a Scribd company logo
1 of 15
Input Privacy Preserving
Techniques: UNECE
project experiences
Massimo De Cubellis, Mauro Bruno, Fabrizio De Fausti,
Monica Scannapieco
Tecnologo | ISTAT
30.11//2021
o Input Privacy definition
o Input privacy vs Output privacy
o Motivations
o Input privacy preserving techniques (HE, SMPC, TEE, FL, etc.)
o UNECE IPP project overview
o Use cases / mini-pilots
o Conclusions
Outline
2
INPUT PRIVACY PRESERVING TECHNIQUES: UNECE PROJECT EXPERIENCES | M. DE CUBELLIS, M. BRUNO, F. DE FAUSTI, M. SCANNAPIECO
Input privacy: a definition
INPUT PRIVACY PRESERVING TECHNIQUES: UNECE PROJECT EXPERIENCES | M. DE CUBELLIS, M. BRUNO, F. DE FAUSTI, M. SCANNAPIECO
3
Input privacy means that the Computing Party cannot access or derive any input value provided by Input
Parties, nor access intermediate values or statistical results available at Result parties during processing of
the data (unless the value has been specifically selected for disclosure). [UN Handbook on Privacy-
Preserving Computation Techniques, 2019]
Source Data
Statistical
Analysis
Statistical
products
Computing parties Output parties
Input parties
Input privacy
o Input privacy techniques are based on data
«transformations» that preserve source data
privacy
o Examples of input privacy techniques: Secure
Multi-Party Computation (SMC), Homomorphic
Encryption (HE), Trusted Execution
Environment (TEE), Federated Learning (FL)
Output privacy
o Output privacy aims at reducing the risk of
privacy breaches in the phases of
disseminating or exchanging statistical
products
o Examples of output privacy techniques:
Differential Privacy, Statistical Disclosure
Control
4
INPUT PRIVACY PRESERVING TECHNIQUES: UNECE PROJECT EXPERIENCES | M. DE CUBELLIS, M. BRUNO, F. DE FAUSTI, M. SCANNAPIECO
Input Privacy vs Output Privacy
Output parties
Source Data
Statistical
Analysis
Statistical
products
Computing parties Output parties
Input parties
Statistical
products
Motivations: why NSIs need to apply IPP techniques?
INPUT PRIVACY PRESERVING TECHNIQUES: UNECE PROJECT EXPERIENCES | M. DE CUBELLIS, M. BRUNO, F. DE FAUSTI, M. SCANNAPIECO
5
Official Statistics is moving towards new production scenarios involving the use of data owned by external
parties [i.e.: Private parties like Mobile Networks Operators (MNOs), public parties like Central Banks or
others NSIs]
New needs (i.e. giving NSIs access to new sources of Big Data – enabling Big data collaboration across
multiple NSIs) have led NSIs to take into account input privacy, (i.e. protect the privacy of data acquired by
parties external to the NSIs)
Private/Public sector
NSI Central
Authority
NSO-1 NSO-2 NSO-3
The main input privacy preserving techniques are:
o Secure Multy-party computation (SMPC)
o Federated Learning (FL)
o Homomorfic Encryption (HE)
o Trusted Execution Environment (TEE)
Input Privacy Techniques: overview (1 of 3)
INPUT PRIVACY PRESERVING TECHNIQUES: UNECE PROJECT EXPERIENCES | M. DE CUBELLIS, M. BRUNO, F. DE FAUSTI, M. SCANNAPIECO
6
o Secure Multy-party computation (SMPC)
o Secure Multi Party Computation (SMPC) is a paradigm based on communication between the parties
o Data are split, and each party sends a few portions of her data
o Other parties can not reconstruct the initial data, but can make some computation on portions of data received
by other parties
o Then once each party has finished, everything can be aggregated and the result of the output is known to each
party
o Federated Learning (FL)
o Machine learning system based on distributing the algorithm to where the data, is instead of gathering the data
where the algorithm is (decentralized/distributed computation)
Input Privacy Techniques: overview (2 of 3)
INPUT PRIVACY PRESERVING TECHNIQUES: UNECE PROJECT EXPERIENCES | M. DE CUBELLIS, M. BRUNO, F. DE FAUSTI, M. SCANNAPIECO
7
o Homomorfic Encryption (HE)
o Homomorphic cryptography is a type of cryptography based on techniques that allow the manipulation of
encrypted data.
For example, having two numbers X and Y (encrypted with the same homomorphic algorithm starting from two
numbers A and B) it is possible to calculate the "encrypted" sum of A and B by directly adding the two encrypted
numbers X and Y without the need to decrypt them
o Trusted Execution Environment (TEE)
Input Privacy Techniques: overview (3 of 3)
INPUT PRIVACY PRESERVING TECHNIQUES: UNECE PROJECT EXPERIENCES | M. DE CUBELLIS, M. BRUNO, F. DE FAUSTI, M. SCANNAPIECO
8
o unlike other IPP techniques, TEE represents a hardware
solution
o the enclave technology allows programs to be executed in
isolation of other programs.
o all data inbound and outbound is encrypted, and
computation in clear only happens within the enclave.
o the enclave code and integrity can then be checked
externally.
Example: Privacy-preserving genotype imputation
in a trusted execution environment
Context:
Project name: Input Privacy Preservation Project
Period: July 2020 – December 2021 (probably will be extended)
Organizations involved: Eurostat, GSO-Vietnam, INEGI-Mexico , Istat-Italy, ONS-UK, Statistics Canada,
Statistics Netherlands, UNECE
Project Goals:
o investigate statistical use cases that require protection on the input side
o assess applicability of selected classes of techniques for main scenarios
o identify opportunities for sharing across statistical community
o create community across statistical organizations and external partners (academia, private sector)
UNECE Input Privacy Preserving: Project overview
INPUT PRIVACY PRESERVING TECHNIQUES: UNECE PROJECT EXPERIENCES | M. DE CUBELLIS, M. BRUNO, F. DE FAUSTI, M. SCANNAPIECO
9
WP1 – Document the use cases
o Existing use cases documentation (original version / generalyzed version)
o Use cases «description» through a logical framework to specify Input Privacy for OS Use Cases
WP 2 – Elaboration for the use cases
o Track 1: Private Set Intersection
o PSI with analytics (ISTAT)
o PSI + Analytics using HE (CBS)
o PSI - Measures the coverage of a data source from a third party in privacy preserving way,
(StatCan)
o Track 2: Private Machine Learning
o Track 3: Organize external consultation
(to suppot NSIs in the design and development of IPP services)
UNECE Input Privacy Preserving: Project structure
INPUT PRIVACY PRESERVING TECHNIQUES: UNECE PROJECT EXPERIENCES | M. DE CUBELLIS, M. BRUNO, F. DE FAUSTI, M. SCANNAPIECO
10
Simulate a runtime environment with several NSIs gathering sensors data with the aim to perform a private
machine learning training
o Architecture: distributed architecture with a central coordinator
o Data: sensors open data by accelorometer
o Task: A machine learning classification task: to predict human activities starting from accelerometer
data
o IPP Technology: Federated Learning
o Framework: Flower Federated Learning
UNECE IPP Use cases: Track 2 – Private Machine Learning (1/2)
INPUT PRIVACY PRESERVING TECHNIQUES: UNECE PROJECT EXPERIENCES | M. DE CUBELLIS, M. BRUNO, F. DE FAUSTI, M. SCANNAPIECO
11
An example of Federated Learning approach
UNECE IPP Use cases: Track 2 – Private Machine Learning (2/2)
INPUT PRIVACY PRESERVING TECHNIQUES: UNECE PROJECT EXPERIENCES | M. DE CUBELLIS, M. BRUNO, F. DE FAUSTI, M. SCANNAPIECO
12
Central
Authority
NSI-1 NSI-2 NSI-3 NSI-4
A Central Authority (server) wants to start a FL training with 4 clients (NSIs)
2 - Central Authority sends the model and parameter
1 - The server sets the FL strategy (i.e. federated average) and connects to
clients. Client data stay on his side
3 - each client train his model and sends it back to the server
4 - the server aggregates the models following the selected strategy
Strategy
o Other configurations are possible
o In each configuration:
o the data are not shared with other NSIs
o different strategies are possible
o more iterations are possible
Parameters
Aggregation
Results obtained:
o Sharing of previous experiences on the subject of IPP
o Development of new mini-pilots / use-cases applying PSI and Federated Learning techniques)
o Know-how acquisition on how to apply IPP techniques within the official statistics
o Ongoing collaboration with UN Privacy Enhancing Technology Lab (UN PET-Lab)
Further development:
o Extension of the UNEEC IPP project in which further mini-pilots and use cases can be carried out
o Design a platform to provide Input Privacy Preserving services
o Apply the lessons learned in the official statistics production within the NSIs
Conclusions
INPUT PRIVACY PRESERVING TECHNIQUES: UNECE PROJECT EXPERIENCES | M. DE CUBELLIS, M. BRUNO, F. DE FAUSTI, M. SCANNAPIECO
13
o UN Handbook on Privacy-Preserving Computation Techniques, 2019
o https://towardsdatascience.com/homomorphic-encryption-intro-part-1-overview-and-use-cases-
a601adcff06c
o https://www.sciencedirect.com/science/article/abs/pii/S2405471221002891
References
INPUT PRIVACY PRESERVING TECHNIQUES: UNECE PROJECT EXPERIENCES | M. DE CUBELLIS, M. BRUNO, F. DE FAUSTI, M. SCANNAPIECO
14
Thanks for your attention !
UNECE Input Privacy Preserving
15

More Related Content

What's hot

Data-intensive decision making in the era of big data and artificial intellig...
Data-intensive decision making in the era of big data and artificial intellig...Data-intensive decision making in the era of big data and artificial intellig...
Data-intensive decision making in the era of big data and artificial intellig...samossummit
 
Verso le trusted smart statistics - prospettive di sviluppo e risultati del e...
Verso le trusted smart statistics - prospettive di sviluppo e risultati del e...Verso le trusted smart statistics - prospettive di sviluppo e risultati del e...
Verso le trusted smart statistics - prospettive di sviluppo e risultati del e...Istituto nazionale di statistica
 
DISCOVERY DAY 2017: MAKE IT HAPPEN!
DISCOVERY DAY 2017: MAKE IT HAPPEN!DISCOVERY DAY 2017: MAKE IT HAPPEN!
DISCOVERY DAY 2017: MAKE IT HAPPEN!FAO
 
Approaches and use of ai in the public sector by the european member states a...
Approaches and use of ai in the public sector by the european member states a...Approaches and use of ai in the public sector by the european member states a...
Approaches and use of ai in the public sector by the european member states a...PanagiotisKeramidis
 
Open Forum Summit June 2010
Open Forum Summit June 2010Open Forum Summit June 2010
Open Forum Summit June 2010Jerry Fishenden
 
Horizon research project opportunities final
Horizon research project opportunities finalHorizon research project opportunities final
Horizon research project opportunities finalPanagiotisKeramidis
 
Governing algorithms – perils and powers of ai in the public sector1(1)
Governing algorithms – perils and powers of ai in the public sector1(1)Governing algorithms – perils and powers of ai in the public sector1(1)
Governing algorithms – perils and powers of ai in the public sector1(1)PanagiotisKeramidis
 
6. Capturing technological transformations in the ECS CEDEFOP
6. Capturing technological transformations in the ECS CEDEFOP6. Capturing technological transformations in the ECS CEDEFOP
6. Capturing technological transformations in the ECS CEDEFOPBEYOND4.0
 
Presentation by the OECD - Session 3: Towards a new generation of indicators ...
Presentation by the OECD - Session 3: Towards a new generation of indicators ...Presentation by the OECD - Session 3: Towards a new generation of indicators ...
Presentation by the OECD - Session 3: Towards a new generation of indicators ...Marie-Claude Gohier
 
Digital ready policymaking and the digital screening process(1)
Digital ready policymaking and the digital screening process(1)Digital ready policymaking and the digital screening process(1)
Digital ready policymaking and the digital screening process(1)PanagiotisKeramidis
 
Verso le trusted smart statistics - prospettive di sviluppo e risultati del e...
Verso le trusted smart statistics - prospettive di sviluppo e risultati del e...Verso le trusted smart statistics - prospettive di sviluppo e risultati del e...
Verso le trusted smart statistics - prospettive di sviluppo e risultati del e...Istituto nazionale di statistica
 
Digital Governance Master Programme. The Gov3.0 Master’s Curriculum
Digital Governance Master Programme. The Gov3.0 Master’s CurriculumDigital Governance Master Programme. The Gov3.0 Master’s Curriculum
Digital Governance Master Programme. The Gov3.0 Master’s Curriculumsamossummit
 
ICT observatories for better governance
ICT observatories for better governanceICT observatories for better governance
ICT observatories for better governancezalisova
 
1. BEYOND4.0 Intro
1. BEYOND4.0 Intro1. BEYOND4.0 Intro
1. BEYOND4.0 IntroBEYOND4.0
 
The evolution of the interoperability concept a european quest for simplicity...
The evolution of the interoperability concept a european quest for simplicity...The evolution of the interoperability concept a european quest for simplicity...
The evolution of the interoperability concept a european quest for simplicity...PanagiotisKeramidis
 
Using administrative data to measure public procurement of R&D: Opportunities...
Using administrative data to measure public procurement of R&D: Opportunities...Using administrative data to measure public procurement of R&D: Opportunities...
Using administrative data to measure public procurement of R&D: Opportunities...STIEAS
 
From eGov 2.0 to eGov 3.0: The Research Agenda
From eGov 2.0 to eGov 3.0: The Research AgendaFrom eGov 2.0 to eGov 3.0: The Research Agenda
From eGov 2.0 to eGov 3.0: The Research Agendasamossummit
 
Digital Transformation of Public Administration
Digital Transformation of Public AdministrationDigital Transformation of Public Administration
Digital Transformation of Public Administrationsamossummit
 
Artificial intelligence (ai) multidisciplinary perspectives on emerging chall...
Artificial intelligence (ai) multidisciplinary perspectives on emerging chall...Artificial intelligence (ai) multidisciplinary perspectives on emerging chall...
Artificial intelligence (ai) multidisciplinary perspectives on emerging chall...PanagiotisKeramidis
 
BPPT Experience - TIKoMeter
BPPT Experience - TIKoMeterBPPT Experience - TIKoMeter
BPPT Experience - TIKoMeterTatang Taufik
 

What's hot (20)

Data-intensive decision making in the era of big data and artificial intellig...
Data-intensive decision making in the era of big data and artificial intellig...Data-intensive decision making in the era of big data and artificial intellig...
Data-intensive decision making in the era of big data and artificial intellig...
 
Verso le trusted smart statistics - prospettive di sviluppo e risultati del e...
Verso le trusted smart statistics - prospettive di sviluppo e risultati del e...Verso le trusted smart statistics - prospettive di sviluppo e risultati del e...
Verso le trusted smart statistics - prospettive di sviluppo e risultati del e...
 
DISCOVERY DAY 2017: MAKE IT HAPPEN!
DISCOVERY DAY 2017: MAKE IT HAPPEN!DISCOVERY DAY 2017: MAKE IT HAPPEN!
DISCOVERY DAY 2017: MAKE IT HAPPEN!
 
Approaches and use of ai in the public sector by the european member states a...
Approaches and use of ai in the public sector by the european member states a...Approaches and use of ai in the public sector by the european member states a...
Approaches and use of ai in the public sector by the european member states a...
 
Open Forum Summit June 2010
Open Forum Summit June 2010Open Forum Summit June 2010
Open Forum Summit June 2010
 
Horizon research project opportunities final
Horizon research project opportunities finalHorizon research project opportunities final
Horizon research project opportunities final
 
Governing algorithms – perils and powers of ai in the public sector1(1)
Governing algorithms – perils and powers of ai in the public sector1(1)Governing algorithms – perils and powers of ai in the public sector1(1)
Governing algorithms – perils and powers of ai in the public sector1(1)
 
6. Capturing technological transformations in the ECS CEDEFOP
6. Capturing technological transformations in the ECS CEDEFOP6. Capturing technological transformations in the ECS CEDEFOP
6. Capturing technological transformations in the ECS CEDEFOP
 
Presentation by the OECD - Session 3: Towards a new generation of indicators ...
Presentation by the OECD - Session 3: Towards a new generation of indicators ...Presentation by the OECD - Session 3: Towards a new generation of indicators ...
Presentation by the OECD - Session 3: Towards a new generation of indicators ...
 
Digital ready policymaking and the digital screening process(1)
Digital ready policymaking and the digital screening process(1)Digital ready policymaking and the digital screening process(1)
Digital ready policymaking and the digital screening process(1)
 
Verso le trusted smart statistics - prospettive di sviluppo e risultati del e...
Verso le trusted smart statistics - prospettive di sviluppo e risultati del e...Verso le trusted smart statistics - prospettive di sviluppo e risultati del e...
Verso le trusted smart statistics - prospettive di sviluppo e risultati del e...
 
Digital Governance Master Programme. The Gov3.0 Master’s Curriculum
Digital Governance Master Programme. The Gov3.0 Master’s CurriculumDigital Governance Master Programme. The Gov3.0 Master’s Curriculum
Digital Governance Master Programme. The Gov3.0 Master’s Curriculum
 
ICT observatories for better governance
ICT observatories for better governanceICT observatories for better governance
ICT observatories for better governance
 
1. BEYOND4.0 Intro
1. BEYOND4.0 Intro1. BEYOND4.0 Intro
1. BEYOND4.0 Intro
 
The evolution of the interoperability concept a european quest for simplicity...
The evolution of the interoperability concept a european quest for simplicity...The evolution of the interoperability concept a european quest for simplicity...
The evolution of the interoperability concept a european quest for simplicity...
 
Using administrative data to measure public procurement of R&D: Opportunities...
Using administrative data to measure public procurement of R&D: Opportunities...Using administrative data to measure public procurement of R&D: Opportunities...
Using administrative data to measure public procurement of R&D: Opportunities...
 
From eGov 2.0 to eGov 3.0: The Research Agenda
From eGov 2.0 to eGov 3.0: The Research AgendaFrom eGov 2.0 to eGov 3.0: The Research Agenda
From eGov 2.0 to eGov 3.0: The Research Agenda
 
Digital Transformation of Public Administration
Digital Transformation of Public AdministrationDigital Transformation of Public Administration
Digital Transformation of Public Administration
 
Artificial intelligence (ai) multidisciplinary perspectives on emerging chall...
Artificial intelligence (ai) multidisciplinary perspectives on emerging chall...Artificial intelligence (ai) multidisciplinary perspectives on emerging chall...
Artificial intelligence (ai) multidisciplinary perspectives on emerging chall...
 
BPPT Experience - TIKoMeter
BPPT Experience - TIKoMeterBPPT Experience - TIKoMeter
BPPT Experience - TIKoMeter
 

Similar to 14a Conferenza Nazionale di Statistica

FIBRE (legacy) testbed Future Perspectives
FIBRE (legacy) testbed Future PerspectivesFIBRE (legacy) testbed Future Perspectives
FIBRE (legacy) testbed Future PerspectivesFIBRE Testbed
 
Representing and Managing Informed User Consent with Knowledge Graphs
Representing and Managing Informed User Consent with Knowledge GraphsRepresenting and Managing Informed User Consent with Knowledge Graphs
Representing and Managing Informed User Consent with Knowledge GraphsAnelia Kurteva
 
Selected Cloud Security Patterns For Improving End User Security and Privacy ...
Selected Cloud Security Patterns For Improving End User Security and Privacy ...Selected Cloud Security Patterns For Improving End User Security and Privacy ...
Selected Cloud Security Patterns For Improving End User Security and Privacy ...PRISMACLOUD Project
 
Summer school bz_fp7research_20100708
Summer school bz_fp7research_20100708Summer school bz_fp7research_20100708
Summer school bz_fp7research_20100708Sandro D'Elia
 
Analysis of Homomorphic Technique and Secure Hash Technique for Multimedia Co...
Analysis of Homomorphic Technique and Secure Hash Technique for Multimedia Co...Analysis of Homomorphic Technique and Secure Hash Technique for Multimedia Co...
Analysis of Homomorphic Technique and Secure Hash Technique for Multimedia Co...IJERA Editor
 
Model-Driven Security with Modularity and Reusability for Engineering Secure ...
Model-Driven Security with Modularity and Reusability for Engineering Secure ...Model-Driven Security with Modularity and Reusability for Engineering Secure ...
Model-Driven Security with Modularity and Reusability for Engineering Secure ...Phu H. Nguyen
 
Towards a distributed framework to analyze multimodal data.pdf
Towards a distributed framework to analyze multimodal data.pdfTowards a distributed framework to analyze multimodal data.pdf
Towards a distributed framework to analyze multimodal data.pdfCarlosRodrigues517978
 
ITU-T Study Group 17 Introduction
ITU-T Study Group 17 IntroductionITU-T Study Group 17 Introduction
ITU-T Study Group 17 IntroductionITU
 
Pedro medeiros citi-cloudviews
Pedro medeiros citi-cloudviewsPedro medeiros citi-cloudviews
Pedro medeiros citi-cloudviewsEuroCloud
 
Pedro medeiros citi-cloudviews
Pedro medeiros citi-cloudviewsPedro medeiros citi-cloudviews
Pedro medeiros citi-cloudviewsEuroCloud
 
CTO-CybersecurityForum-2010-Andrea Gloriso
CTO-CybersecurityForum-2010-Andrea GlorisoCTO-CybersecurityForum-2010-Andrea Gloriso
CTO-CybersecurityForum-2010-Andrea Glorisosegughana
 
Research Group 'Multimedia Communication' Presentation (March 2015)
Research Group 'Multimedia Communication' Presentation (March 2015)Research Group 'Multimedia Communication' Presentation (March 2015)
Research Group 'Multimedia Communication' Presentation (March 2015)hellwagner
 
H2020 project WITDOM overview
H2020 project WITDOM overviewH2020 project WITDOM overview
H2020 project WITDOM overviewElsa Prieto
 
EUBrasilCloudFORUM - Concertation Meeting
EUBrasilCloudFORUM - Concertation MeetingEUBrasilCloudFORUM - Concertation Meeting
EUBrasilCloudFORUM - Concertation MeetingEUBrasilCloudFORUM .
 
SC7 Workshop 1: BigDataEurope Approach to Security
SC7 Workshop 1: BigDataEurope Approach to SecuritySC7 Workshop 1: BigDataEurope Approach to Security
SC7 Workshop 1: BigDataEurope Approach to SecurityBigData_Europe
 
ZONeSEC: critical infrastructure protection in real practice
ZONeSEC: critical infrastructure protection in real practice ZONeSEC: critical infrastructure protection in real practice
ZONeSEC: critical infrastructure protection in real practice José Ramón Martínez Salio
 

Similar to 14a Conferenza Nazionale di Statistica (20)

FIBRE (legacy) testbed Future Perspectives
FIBRE (legacy) testbed Future PerspectivesFIBRE (legacy) testbed Future Perspectives
FIBRE (legacy) testbed Future Perspectives
 
FIBRE testbed
FIBRE testbed FIBRE testbed
FIBRE testbed
 
NIS.docx
NIS.docxNIS.docx
NIS.docx
 
Representing and Managing Informed User Consent with Knowledge Graphs
Representing and Managing Informed User Consent with Knowledge GraphsRepresenting and Managing Informed User Consent with Knowledge Graphs
Representing and Managing Informed User Consent with Knowledge Graphs
 
Selected Cloud Security Patterns For Improving End User Security and Privacy ...
Selected Cloud Security Patterns For Improving End User Security and Privacy ...Selected Cloud Security Patterns For Improving End User Security and Privacy ...
Selected Cloud Security Patterns For Improving End User Security and Privacy ...
 
Summer school bz_fp7research_20100708
Summer school bz_fp7research_20100708Summer school bz_fp7research_20100708
Summer school bz_fp7research_20100708
 
Analysis of Homomorphic Technique and Secure Hash Technique for Multimedia Co...
Analysis of Homomorphic Technique and Secure Hash Technique for Multimedia Co...Analysis of Homomorphic Technique and Secure Hash Technique for Multimedia Co...
Analysis of Homomorphic Technique and Secure Hash Technique for Multimedia Co...
 
Model-Driven Security with Modularity and Reusability for Engineering Secure ...
Model-Driven Security with Modularity and Reusability for Engineering Secure ...Model-Driven Security with Modularity and Reusability for Engineering Secure ...
Model-Driven Security with Modularity and Reusability for Engineering Secure ...
 
Towards a distributed framework to analyze multimodal data.pdf
Towards a distributed framework to analyze multimodal data.pdfTowards a distributed framework to analyze multimodal data.pdf
Towards a distributed framework to analyze multimodal data.pdf
 
ITU-T Study Group 17 Introduction
ITU-T Study Group 17 IntroductionITU-T Study Group 17 Introduction
ITU-T Study Group 17 Introduction
 
Pedro medeiros citi-cloudviews
Pedro medeiros citi-cloudviewsPedro medeiros citi-cloudviews
Pedro medeiros citi-cloudviews
 
Pedro medeiros citi-cloudviews
Pedro medeiros citi-cloudviewsPedro medeiros citi-cloudviews
Pedro medeiros citi-cloudviews
 
CTO-CybersecurityForum-2010-Andrea Gloriso
CTO-CybersecurityForum-2010-Andrea GlorisoCTO-CybersecurityForum-2010-Andrea Gloriso
CTO-CybersecurityForum-2010-Andrea Gloriso
 
Research Group 'Multimedia Communication' Presentation (March 2015)
Research Group 'Multimedia Communication' Presentation (March 2015)Research Group 'Multimedia Communication' Presentation (March 2015)
Research Group 'Multimedia Communication' Presentation (March 2015)
 
H2020 project WITDOM overview
H2020 project WITDOM overviewH2020 project WITDOM overview
H2020 project WITDOM overview
 
Smit Flyer
Smit FlyerSmit Flyer
Smit Flyer
 
EUBrasilCloudFORUM - Concertation Meeting
EUBrasilCloudFORUM - Concertation MeetingEUBrasilCloudFORUM - Concertation Meeting
EUBrasilCloudFORUM - Concertation Meeting
 
SC7 Workshop 1: BigDataEurope Approach to Security
SC7 Workshop 1: BigDataEurope Approach to SecuritySC7 Workshop 1: BigDataEurope Approach to Security
SC7 Workshop 1: BigDataEurope Approach to Security
 
Intrusion detection systems for internet of thing based big data: a review
Intrusion detection systems for internet of thing based big data:  a reviewIntrusion detection systems for internet of thing based big data:  a review
Intrusion detection systems for internet of thing based big data: a review
 
ZONeSEC: critical infrastructure protection in real practice
ZONeSEC: critical infrastructure protection in real practice ZONeSEC: critical infrastructure protection in real practice
ZONeSEC: critical infrastructure protection in real practice
 

More from Istituto nazionale di statistica

More from Istituto nazionale di statistica (20)

Censimenti Permanenti Istituzioni non profit
Censimenti Permanenti Istituzioni non profitCensimenti Permanenti Istituzioni non profit
Censimenti Permanenti Istituzioni non profit
 
Censimenti Permanenti Istituzioni non profit
Censimenti Permanenti Istituzioni non profitCensimenti Permanenti Istituzioni non profit
Censimenti Permanenti Istituzioni non profit
 
Censimenti Permanenti Istituzioni non profit
Censimenti Permanenti Istituzioni non profitCensimenti Permanenti Istituzioni non profit
Censimenti Permanenti Istituzioni non profit
 
Censimenti Permanenti Istituzioni non profit
Censimenti Permanenti Istituzioni non profitCensimenti Permanenti Istituzioni non profit
Censimenti Permanenti Istituzioni non profit
 
Censimenti Permanenti Istituzioni non profit
Censimenti Permanenti Istituzioni non profitCensimenti Permanenti Istituzioni non profit
Censimenti Permanenti Istituzioni non profit
 
Censimenti Permanenti Istituzioni non profit
Censimenti Permanenti Istituzioni non profitCensimenti Permanenti Istituzioni non profit
Censimenti Permanenti Istituzioni non profit
 
Censimento Permanente Istituzioni Pubbliche
Censimento Permanente Istituzioni PubblicheCensimento Permanente Istituzioni Pubbliche
Censimento Permanente Istituzioni Pubbliche
 
Censimento Permanente Istituzioni Pubbliche
Censimento Permanente Istituzioni PubblicheCensimento Permanente Istituzioni Pubbliche
Censimento Permanente Istituzioni Pubbliche
 
Censimento Permanente Istituzioni Pubbliche
Censimento Permanente Istituzioni PubblicheCensimento Permanente Istituzioni Pubbliche
Censimento Permanente Istituzioni Pubbliche
 
Censimento Permanente Istituzioni Pubbliche
Censimento Permanente Istituzioni PubblicheCensimento Permanente Istituzioni Pubbliche
Censimento Permanente Istituzioni Pubbliche
 
14a Conferenza Nazionale di Statisticacnstatistica14
14a Conferenza Nazionale di Statisticacnstatistica1414a Conferenza Nazionale di Statisticacnstatistica14
14a Conferenza Nazionale di Statisticacnstatistica14
 
14a Conferenza Nazionale di Statistica
14a Conferenza Nazionale di Statistica14a Conferenza Nazionale di Statistica
14a Conferenza Nazionale di Statistica
 
14a Conferenza Nazionale di Statistica
14a Conferenza Nazionale di Statistica14a Conferenza Nazionale di Statistica
14a Conferenza Nazionale di Statistica
 
14a Conferenza Nazionale di Statistica
14a Conferenza Nazionale di Statistica14a Conferenza Nazionale di Statistica
14a Conferenza Nazionale di Statistica
 
14a Conferenza Nazionale di Statistica
14a Conferenza Nazionale di Statistica14a Conferenza Nazionale di Statistica
14a Conferenza Nazionale di Statistica
 
14a Conferenza Nazionale di Statistica
14a Conferenza Nazionale di Statistica14a Conferenza Nazionale di Statistica
14a Conferenza Nazionale di Statistica
 
14a Conferenza Nazionale di Statistica
14a Conferenza Nazionale di Statistica14a Conferenza Nazionale di Statistica
14a Conferenza Nazionale di Statistica
 
14a Conferenza Nazionale di Statistica
14a Conferenza Nazionale di Statistica14a Conferenza Nazionale di Statistica
14a Conferenza Nazionale di Statistica
 
14a Conferenza Nazionale di Statistica
14a Conferenza Nazionale di Statistica14a Conferenza Nazionale di Statistica
14a Conferenza Nazionale di Statistica
 
14a Conferenza Nazionale di Statistica
14a Conferenza Nazionale di Statistica14a Conferenza Nazionale di Statistica
14a Conferenza Nazionale di Statistica
 

Recently uploaded

How to Make a Pirate ship Primary Education.pptx
How to Make a Pirate ship Primary Education.pptxHow to Make a Pirate ship Primary Education.pptx
How to Make a Pirate ship Primary Education.pptxmanuelaromero2013
 
Science lesson Moon for 4th quarter lesson
Science lesson Moon for 4th quarter lessonScience lesson Moon for 4th quarter lesson
Science lesson Moon for 4th quarter lessonJericReyAuditor
 
Presiding Officer Training module 2024 lok sabha elections
Presiding Officer Training module 2024 lok sabha electionsPresiding Officer Training module 2024 lok sabha elections
Presiding Officer Training module 2024 lok sabha electionsanshu789521
 
EPANDING THE CONTENT OF AN OUTLINE using notes.pptx
EPANDING THE CONTENT OF AN OUTLINE using notes.pptxEPANDING THE CONTENT OF AN OUTLINE using notes.pptx
EPANDING THE CONTENT OF AN OUTLINE using notes.pptxRaymartEstabillo3
 
A Critique of the Proposed National Education Policy Reform
A Critique of the Proposed National Education Policy ReformA Critique of the Proposed National Education Policy Reform
A Critique of the Proposed National Education Policy ReformChameera Dedduwage
 
CARE OF CHILD IN INCUBATOR..........pptx
CARE OF CHILD IN INCUBATOR..........pptxCARE OF CHILD IN INCUBATOR..........pptx
CARE OF CHILD IN INCUBATOR..........pptxGaneshChakor2
 
ENGLISH5 QUARTER4 MODULE1 WEEK1-3 How Visual and Multimedia Elements.pptx
ENGLISH5 QUARTER4 MODULE1 WEEK1-3 How Visual and Multimedia Elements.pptxENGLISH5 QUARTER4 MODULE1 WEEK1-3 How Visual and Multimedia Elements.pptx
ENGLISH5 QUARTER4 MODULE1 WEEK1-3 How Visual and Multimedia Elements.pptxAnaBeatriceAblay2
 
POINT- BIOCHEMISTRY SEM 2 ENZYMES UNIT 5.pptx
POINT- BIOCHEMISTRY SEM 2 ENZYMES UNIT 5.pptxPOINT- BIOCHEMISTRY SEM 2 ENZYMES UNIT 5.pptx
POINT- BIOCHEMISTRY SEM 2 ENZYMES UNIT 5.pptxSayali Powar
 
Employee wellbeing at the workplace.pptx
Employee wellbeing at the workplace.pptxEmployee wellbeing at the workplace.pptx
Employee wellbeing at the workplace.pptxNirmalaLoungPoorunde1
 
Sanyam Choudhary Chemistry practical.pdf
Sanyam Choudhary Chemistry practical.pdfSanyam Choudhary Chemistry practical.pdf
Sanyam Choudhary Chemistry practical.pdfsanyamsingh5019
 
Final demo Grade 9 for demo Plan dessert.pptx
Final demo Grade 9 for demo Plan dessert.pptxFinal demo Grade 9 for demo Plan dessert.pptx
Final demo Grade 9 for demo Plan dessert.pptxAvyJaneVismanos
 
Introduction to ArtificiaI Intelligence in Higher Education
Introduction to ArtificiaI Intelligence in Higher EducationIntroduction to ArtificiaI Intelligence in Higher Education
Introduction to ArtificiaI Intelligence in Higher Educationpboyjonauth
 
Science 7 - LAND and SEA BREEZE and its Characteristics
Science 7 - LAND and SEA BREEZE and its CharacteristicsScience 7 - LAND and SEA BREEZE and its Characteristics
Science 7 - LAND and SEA BREEZE and its CharacteristicsKarinaGenton
 
BASLIQ CURRENT LOOKBOOK LOOKBOOK(1) (1).pdf
BASLIQ CURRENT LOOKBOOK  LOOKBOOK(1) (1).pdfBASLIQ CURRENT LOOKBOOK  LOOKBOOK(1) (1).pdf
BASLIQ CURRENT LOOKBOOK LOOKBOOK(1) (1).pdfSoniaTolstoy
 
Blooming Together_ Growing a Community Garden Worksheet.docx
Blooming Together_ Growing a Community Garden Worksheet.docxBlooming Together_ Growing a Community Garden Worksheet.docx
Blooming Together_ Growing a Community Garden Worksheet.docxUnboundStockton
 
How to Configure Email Server in Odoo 17
How to Configure Email Server in Odoo 17How to Configure Email Server in Odoo 17
How to Configure Email Server in Odoo 17Celine George
 
Call Girls in Dwarka Mor Delhi Contact Us 9654467111
Call Girls in Dwarka Mor Delhi Contact Us 9654467111Call Girls in Dwarka Mor Delhi Contact Us 9654467111
Call Girls in Dwarka Mor Delhi Contact Us 9654467111Sapana Sha
 
internship ppt on smartinternz platform as salesforce developer
internship ppt on smartinternz platform as salesforce developerinternship ppt on smartinternz platform as salesforce developer
internship ppt on smartinternz platform as salesforce developerunnathinaik
 

Recently uploaded (20)

How to Make a Pirate ship Primary Education.pptx
How to Make a Pirate ship Primary Education.pptxHow to Make a Pirate ship Primary Education.pptx
How to Make a Pirate ship Primary Education.pptx
 
Science lesson Moon for 4th quarter lesson
Science lesson Moon for 4th quarter lessonScience lesson Moon for 4th quarter lesson
Science lesson Moon for 4th quarter lesson
 
Presiding Officer Training module 2024 lok sabha elections
Presiding Officer Training module 2024 lok sabha electionsPresiding Officer Training module 2024 lok sabha elections
Presiding Officer Training module 2024 lok sabha elections
 
EPANDING THE CONTENT OF AN OUTLINE using notes.pptx
EPANDING THE CONTENT OF AN OUTLINE using notes.pptxEPANDING THE CONTENT OF AN OUTLINE using notes.pptx
EPANDING THE CONTENT OF AN OUTLINE using notes.pptx
 
A Critique of the Proposed National Education Policy Reform
A Critique of the Proposed National Education Policy ReformA Critique of the Proposed National Education Policy Reform
A Critique of the Proposed National Education Policy Reform
 
CARE OF CHILD IN INCUBATOR..........pptx
CARE OF CHILD IN INCUBATOR..........pptxCARE OF CHILD IN INCUBATOR..........pptx
CARE OF CHILD IN INCUBATOR..........pptx
 
ENGLISH5 QUARTER4 MODULE1 WEEK1-3 How Visual and Multimedia Elements.pptx
ENGLISH5 QUARTER4 MODULE1 WEEK1-3 How Visual and Multimedia Elements.pptxENGLISH5 QUARTER4 MODULE1 WEEK1-3 How Visual and Multimedia Elements.pptx
ENGLISH5 QUARTER4 MODULE1 WEEK1-3 How Visual and Multimedia Elements.pptx
 
POINT- BIOCHEMISTRY SEM 2 ENZYMES UNIT 5.pptx
POINT- BIOCHEMISTRY SEM 2 ENZYMES UNIT 5.pptxPOINT- BIOCHEMISTRY SEM 2 ENZYMES UNIT 5.pptx
POINT- BIOCHEMISTRY SEM 2 ENZYMES UNIT 5.pptx
 
Employee wellbeing at the workplace.pptx
Employee wellbeing at the workplace.pptxEmployee wellbeing at the workplace.pptx
Employee wellbeing at the workplace.pptx
 
TataKelola dan KamSiber Kecerdasan Buatan v022.pdf
TataKelola dan KamSiber Kecerdasan Buatan v022.pdfTataKelola dan KamSiber Kecerdasan Buatan v022.pdf
TataKelola dan KamSiber Kecerdasan Buatan v022.pdf
 
Sanyam Choudhary Chemistry practical.pdf
Sanyam Choudhary Chemistry practical.pdfSanyam Choudhary Chemistry practical.pdf
Sanyam Choudhary Chemistry practical.pdf
 
Final demo Grade 9 for demo Plan dessert.pptx
Final demo Grade 9 for demo Plan dessert.pptxFinal demo Grade 9 for demo Plan dessert.pptx
Final demo Grade 9 for demo Plan dessert.pptx
 
Staff of Color (SOC) Retention Efforts DDSD
Staff of Color (SOC) Retention Efforts DDSDStaff of Color (SOC) Retention Efforts DDSD
Staff of Color (SOC) Retention Efforts DDSD
 
Introduction to ArtificiaI Intelligence in Higher Education
Introduction to ArtificiaI Intelligence in Higher EducationIntroduction to ArtificiaI Intelligence in Higher Education
Introduction to ArtificiaI Intelligence in Higher Education
 
Science 7 - LAND and SEA BREEZE and its Characteristics
Science 7 - LAND and SEA BREEZE and its CharacteristicsScience 7 - LAND and SEA BREEZE and its Characteristics
Science 7 - LAND and SEA BREEZE and its Characteristics
 
BASLIQ CURRENT LOOKBOOK LOOKBOOK(1) (1).pdf
BASLIQ CURRENT LOOKBOOK  LOOKBOOK(1) (1).pdfBASLIQ CURRENT LOOKBOOK  LOOKBOOK(1) (1).pdf
BASLIQ CURRENT LOOKBOOK LOOKBOOK(1) (1).pdf
 
Blooming Together_ Growing a Community Garden Worksheet.docx
Blooming Together_ Growing a Community Garden Worksheet.docxBlooming Together_ Growing a Community Garden Worksheet.docx
Blooming Together_ Growing a Community Garden Worksheet.docx
 
How to Configure Email Server in Odoo 17
How to Configure Email Server in Odoo 17How to Configure Email Server in Odoo 17
How to Configure Email Server in Odoo 17
 
Call Girls in Dwarka Mor Delhi Contact Us 9654467111
Call Girls in Dwarka Mor Delhi Contact Us 9654467111Call Girls in Dwarka Mor Delhi Contact Us 9654467111
Call Girls in Dwarka Mor Delhi Contact Us 9654467111
 
internship ppt on smartinternz platform as salesforce developer
internship ppt on smartinternz platform as salesforce developerinternship ppt on smartinternz platform as salesforce developer
internship ppt on smartinternz platform as salesforce developer
 

14a Conferenza Nazionale di Statistica

  • 1. Input Privacy Preserving Techniques: UNECE project experiences Massimo De Cubellis, Mauro Bruno, Fabrizio De Fausti, Monica Scannapieco Tecnologo | ISTAT 30.11//2021
  • 2. o Input Privacy definition o Input privacy vs Output privacy o Motivations o Input privacy preserving techniques (HE, SMPC, TEE, FL, etc.) o UNECE IPP project overview o Use cases / mini-pilots o Conclusions Outline 2 INPUT PRIVACY PRESERVING TECHNIQUES: UNECE PROJECT EXPERIENCES | M. DE CUBELLIS, M. BRUNO, F. DE FAUSTI, M. SCANNAPIECO
  • 3. Input privacy: a definition INPUT PRIVACY PRESERVING TECHNIQUES: UNECE PROJECT EXPERIENCES | M. DE CUBELLIS, M. BRUNO, F. DE FAUSTI, M. SCANNAPIECO 3 Input privacy means that the Computing Party cannot access or derive any input value provided by Input Parties, nor access intermediate values or statistical results available at Result parties during processing of the data (unless the value has been specifically selected for disclosure). [UN Handbook on Privacy- Preserving Computation Techniques, 2019] Source Data Statistical Analysis Statistical products Computing parties Output parties Input parties
  • 4. Input privacy o Input privacy techniques are based on data «transformations» that preserve source data privacy o Examples of input privacy techniques: Secure Multi-Party Computation (SMC), Homomorphic Encryption (HE), Trusted Execution Environment (TEE), Federated Learning (FL) Output privacy o Output privacy aims at reducing the risk of privacy breaches in the phases of disseminating or exchanging statistical products o Examples of output privacy techniques: Differential Privacy, Statistical Disclosure Control 4 INPUT PRIVACY PRESERVING TECHNIQUES: UNECE PROJECT EXPERIENCES | M. DE CUBELLIS, M. BRUNO, F. DE FAUSTI, M. SCANNAPIECO Input Privacy vs Output Privacy Output parties Source Data Statistical Analysis Statistical products Computing parties Output parties Input parties Statistical products
  • 5. Motivations: why NSIs need to apply IPP techniques? INPUT PRIVACY PRESERVING TECHNIQUES: UNECE PROJECT EXPERIENCES | M. DE CUBELLIS, M. BRUNO, F. DE FAUSTI, M. SCANNAPIECO 5 Official Statistics is moving towards new production scenarios involving the use of data owned by external parties [i.e.: Private parties like Mobile Networks Operators (MNOs), public parties like Central Banks or others NSIs] New needs (i.e. giving NSIs access to new sources of Big Data – enabling Big data collaboration across multiple NSIs) have led NSIs to take into account input privacy, (i.e. protect the privacy of data acquired by parties external to the NSIs) Private/Public sector NSI Central Authority NSO-1 NSO-2 NSO-3
  • 6. The main input privacy preserving techniques are: o Secure Multy-party computation (SMPC) o Federated Learning (FL) o Homomorfic Encryption (HE) o Trusted Execution Environment (TEE) Input Privacy Techniques: overview (1 of 3) INPUT PRIVACY PRESERVING TECHNIQUES: UNECE PROJECT EXPERIENCES | M. DE CUBELLIS, M. BRUNO, F. DE FAUSTI, M. SCANNAPIECO 6
  • 7. o Secure Multy-party computation (SMPC) o Secure Multi Party Computation (SMPC) is a paradigm based on communication between the parties o Data are split, and each party sends a few portions of her data o Other parties can not reconstruct the initial data, but can make some computation on portions of data received by other parties o Then once each party has finished, everything can be aggregated and the result of the output is known to each party o Federated Learning (FL) o Machine learning system based on distributing the algorithm to where the data, is instead of gathering the data where the algorithm is (decentralized/distributed computation) Input Privacy Techniques: overview (2 of 3) INPUT PRIVACY PRESERVING TECHNIQUES: UNECE PROJECT EXPERIENCES | M. DE CUBELLIS, M. BRUNO, F. DE FAUSTI, M. SCANNAPIECO 7
  • 8. o Homomorfic Encryption (HE) o Homomorphic cryptography is a type of cryptography based on techniques that allow the manipulation of encrypted data. For example, having two numbers X and Y (encrypted with the same homomorphic algorithm starting from two numbers A and B) it is possible to calculate the "encrypted" sum of A and B by directly adding the two encrypted numbers X and Y without the need to decrypt them o Trusted Execution Environment (TEE) Input Privacy Techniques: overview (3 of 3) INPUT PRIVACY PRESERVING TECHNIQUES: UNECE PROJECT EXPERIENCES | M. DE CUBELLIS, M. BRUNO, F. DE FAUSTI, M. SCANNAPIECO 8 o unlike other IPP techniques, TEE represents a hardware solution o the enclave technology allows programs to be executed in isolation of other programs. o all data inbound and outbound is encrypted, and computation in clear only happens within the enclave. o the enclave code and integrity can then be checked externally. Example: Privacy-preserving genotype imputation in a trusted execution environment
  • 9. Context: Project name: Input Privacy Preservation Project Period: July 2020 – December 2021 (probably will be extended) Organizations involved: Eurostat, GSO-Vietnam, INEGI-Mexico , Istat-Italy, ONS-UK, Statistics Canada, Statistics Netherlands, UNECE Project Goals: o investigate statistical use cases that require protection on the input side o assess applicability of selected classes of techniques for main scenarios o identify opportunities for sharing across statistical community o create community across statistical organizations and external partners (academia, private sector) UNECE Input Privacy Preserving: Project overview INPUT PRIVACY PRESERVING TECHNIQUES: UNECE PROJECT EXPERIENCES | M. DE CUBELLIS, M. BRUNO, F. DE FAUSTI, M. SCANNAPIECO 9
  • 10. WP1 – Document the use cases o Existing use cases documentation (original version / generalyzed version) o Use cases «description» through a logical framework to specify Input Privacy for OS Use Cases WP 2 – Elaboration for the use cases o Track 1: Private Set Intersection o PSI with analytics (ISTAT) o PSI + Analytics using HE (CBS) o PSI - Measures the coverage of a data source from a third party in privacy preserving way, (StatCan) o Track 2: Private Machine Learning o Track 3: Organize external consultation (to suppot NSIs in the design and development of IPP services) UNECE Input Privacy Preserving: Project structure INPUT PRIVACY PRESERVING TECHNIQUES: UNECE PROJECT EXPERIENCES | M. DE CUBELLIS, M. BRUNO, F. DE FAUSTI, M. SCANNAPIECO 10
  • 11. Simulate a runtime environment with several NSIs gathering sensors data with the aim to perform a private machine learning training o Architecture: distributed architecture with a central coordinator o Data: sensors open data by accelorometer o Task: A machine learning classification task: to predict human activities starting from accelerometer data o IPP Technology: Federated Learning o Framework: Flower Federated Learning UNECE IPP Use cases: Track 2 – Private Machine Learning (1/2) INPUT PRIVACY PRESERVING TECHNIQUES: UNECE PROJECT EXPERIENCES | M. DE CUBELLIS, M. BRUNO, F. DE FAUSTI, M. SCANNAPIECO 11
  • 12. An example of Federated Learning approach UNECE IPP Use cases: Track 2 – Private Machine Learning (2/2) INPUT PRIVACY PRESERVING TECHNIQUES: UNECE PROJECT EXPERIENCES | M. DE CUBELLIS, M. BRUNO, F. DE FAUSTI, M. SCANNAPIECO 12 Central Authority NSI-1 NSI-2 NSI-3 NSI-4 A Central Authority (server) wants to start a FL training with 4 clients (NSIs) 2 - Central Authority sends the model and parameter 1 - The server sets the FL strategy (i.e. federated average) and connects to clients. Client data stay on his side 3 - each client train his model and sends it back to the server 4 - the server aggregates the models following the selected strategy Strategy o Other configurations are possible o In each configuration: o the data are not shared with other NSIs o different strategies are possible o more iterations are possible Parameters Aggregation
  • 13. Results obtained: o Sharing of previous experiences on the subject of IPP o Development of new mini-pilots / use-cases applying PSI and Federated Learning techniques) o Know-how acquisition on how to apply IPP techniques within the official statistics o Ongoing collaboration with UN Privacy Enhancing Technology Lab (UN PET-Lab) Further development: o Extension of the UNEEC IPP project in which further mini-pilots and use cases can be carried out o Design a platform to provide Input Privacy Preserving services o Apply the lessons learned in the official statistics production within the NSIs Conclusions INPUT PRIVACY PRESERVING TECHNIQUES: UNECE PROJECT EXPERIENCES | M. DE CUBELLIS, M. BRUNO, F. DE FAUSTI, M. SCANNAPIECO 13
  • 14. o UN Handbook on Privacy-Preserving Computation Techniques, 2019 o https://towardsdatascience.com/homomorphic-encryption-intro-part-1-overview-and-use-cases- a601adcff06c o https://www.sciencedirect.com/science/article/abs/pii/S2405471221002891 References INPUT PRIVACY PRESERVING TECHNIQUES: UNECE PROJECT EXPERIENCES | M. DE CUBELLIS, M. BRUNO, F. DE FAUSTI, M. SCANNAPIECO 14
  • 15. Thanks for your attention ! UNECE Input Privacy Preserving 15