Analyses and studies
Find out about recent studies in the field.
Study on privacy-enhancing technologies
31.03.2023
Among other things, privacy-enhancing technologies (PETs) help to ensure data protection, for example, at the service level, and avoid unnecessary data processing (minimal processing) or potential leaks. The implementation of PETs in Estonia has begun, albeit at a slow pace. It would, therefore, be beneficial to have a broader understanding of the potential applications and risks of PETs, considering where and which purpose they can be used for.
The findings of the study are applicable not only in the public sector but are also valuable for fostering the development of expertise in this field within the private sector. The study helps to provide clarity for private enterprises interested in offering solutions based on PETs. The study thus has a broader impact, including on data processing between the private and public sectors.
You can find the PET study and supplementary materials below (in Estonian):
Algorithmic Bias Risk Management Materials (ENG)
The Algorithmic Bias Risk Management Toolkit focuses on managing the risks of bias in algorithmic and AI systems. The toolkit consists of three parts: a guideline, which explains the nature and background of AI systems and algorithmic bias and explores ways to identify and mitigate bias; a methodology, which provides detailed instructions for setting up and carrying out the risk management process; and a checklist, which facilitates the documentation of information required for risk management. The toolkit primarily focuses on systems used within organisations.
The Algorithmic Bias Risk Management Guidelines help users of the methodology understand key concepts related to bias as well as the risks that may arise in practice. We recommend familiarising yourself with the structure of the guide before applying the methodology and completing the accompanying worksheet. Several sections of the guide describe real-world systems in which algorithmic bias or biased AI components have been identified. These case studies help readers better understand, identify and prevent such risks.
The Algorithmic Bias Risk Management Methodology serves as a practical tool for implementing the principles described in the guide. While the guide explains the key concepts, underlying principles and relationships associated with bias in AI and algorithmic systems, the methodology provides concrete activities and steps for analysing algorithmic bias and managing related risks in AI systems and systems that use algorithmic decision-making. The methodology is flexible and can be used alongside existing risk management standards and practices within an organisation. Its application is supported by a fillable checklist provided in spreadsheet format. Report on the use of the ADM systems and methodology can be found here.
The AI System Bias Assessment Checklist is designed to support a simplified analysis of bias in an AI system within a single file. The worksheet is intended to be used together with the methodology and guide.
The materials were developed as part of the EquiTech project (Project 101144709 — Improving response to risks of discrimination, bias and intolerance in automated decision-making systems to promote equality).
Study on privacy-enhancing technologies
31.03.2023
Among other things, privacy-enhancing technologies (PETs) help to ensure data protection, for example, at the service level, and avoid unnecessary data processing (minimal processing) or potential leaks. The implementation of PETs in Estonia has begun, albeit at a slow pace. It would, therefore, be beneficial to have a broader understanding of the potential applications and risks of PETs, considering where and which purpose they can be used for.
The findings of the study are applicable not only in the public sector but are also valuable for fostering the development of expertise in this field within the private sector. The study helps to provide clarity for private enterprises interested in offering solutions based on PETs. The study thus has a broader impact, including on data processing between the private and public sectors.
You can find the PET study and supplementary materials below (in Estonian):
Study on privacy-enhancing technologies
31.03.2023
Among other things, privacy-enhancing technologies (PETs) help to ensure data protection, for example, at the service level, and avoid unnecessary data processing (minimal processing) or potential leaks. The implementation of PETs in Estonia has begun, albeit at a slow pace. It would, therefore, be beneficial to have a broader understanding of the potential applications and risks of PETs, considering where and which purpose they can be used for.
The findings of the study are applicable not only in the public sector but are also valuable for fostering the development of expertise in this field within the private sector. The study helps to provide clarity for private enterprises interested in offering solutions based on PETs. The study thus has a broader impact, including on data processing between the private and public sectors.
You can find the PET study and supplementary materials below (in Estonian):
Consent service analysis
The objective of this analysis was to analyse the consent service across different sectors (general part), develop a comprehensive concept for the service based on health data (Health Information System, Digital Prescription Centre) and evaluate its potential applicability to other national databases. The report (in Estonian) can be found here.
Analysis of the data stored in national databases and registers and their machine-readable output capacity
The purpose of this analysis was to provide an overview of the data stored in national databases and registers. The analysis was needed for the development of new services and to determine whether and in what form this data is currently available for the development of such services.
The report (in Estonian) can be found here.
Analysis of the data stored in national databases and registers and their machine-readable output capacity
The purpose of this analysis was to provide an overview of the data stored in national databases and registers. The analysis was needed for the development of new services and to determine whether and in what form this data is currently available for the development of such services.
The report (in Estonian) can be found here.
03.02.2023
The survey allows us to gain insights into the expectations and needs of Bürokratt users with the focus on using services. The survey was carried out with the intention to establish specific objectives for the development of Bürokratt. It is important to note that the prerequisite for using Bürokratt is the concentration of certain critical services in the environment, to which less critical or lower priority services may be added in the future. Consequently, the Bürokratt development team has recognised the need to commission a user survey on Bürokratt to address the following questions:
-
What are the priority services?
-
Which technical solutions ensure high usability of Bürokratt?
-
What are users’ fears and prejudices?
-
What are the prerequisites for using Bürokratt?
The user survey was commissioned by the Ministry of Economic Affairs and Communications in co-operation with the Information System Authority. The survey was carried out by Civitta Eesti AS. The survey was financed from the sources of the EU Recovery and Resilience Facility (RRF).
The user survey on Bürokratt’s functionalities (in Estonian) can be found here.
2021
The objective of this project was to assess the feasibility and viability of classifying customer enquiries at the level of the state, ministry and institution, and to determine the steps for doing this in the most effective way possible. The project also outlines the aspects that need to be taken into consideration in the future, along with the risks involved, the viability of the project and necessary activities for further development. Additionally, we developed a proof of concept (PoC) for customer enquiry classification for technological testing, a description of the PoC and a guide for institutions on how to tag customer requests.
The analysis (in Estonian) can be found here.
04.12.2020
The purpose of the analysis and piloting of the use of alternative channels is to analyse the possibilities of using modern communication channels and the opportunities they offer to streamline communication between the state and its citizens. In this context, channels refer to chat-based applications such as Facebook Messenger, WhatsApp, Apple iMessages, Signal and others. When implementing new channels, it is essential to improve the user experience, reduce the administrative burden on the state and increase the number of citizens involved. The project was carried out in co-operation with experts from the Ministry of Economic Affairs and Communications, the Information System Authority, the Agricultural Registers and Information Board, and Solita during the period from September 2020 to December 2020.
The analysis (in Estonian) can be found here.
2019-2020
The pre-analysis of the Bürokratt chatbot maps the wants, needs and requirements that must be taken into consideration in the development of a chatbot that would assist citizens across the state agencies of Estonia. The document covers topics related to the development of the PoC solution and, to a lesser extent, also the requirements for follow-up projects.
The post-analysis of the Bürokratt chatbot offers an overview of the Bürokratt prototype developed during the project and provides input for the development of a production-ready Bürokratt.
The analysis (in Estonian) can be found here.