Filip Bednárik has worked for Essential Data since 2014. He is a key expert in projects in which the intelligent search in the Slovak language is being implemented. He enjoys working with big data and its analysis, evaluation, and presentation. Since 2018 Filip has been working as an AI expert on the project “Anonymization Services for the Ministry of Justice”. He has extensive experience in information systems design, automation, process optimization and app development. Filip prefers a technological stack based on open-source solutions, microservices, and a cloud-based infrastructure.
In a world where artificial intelligence is available to general public and able to extract information from text in a matter of seconds, there arises a greater need to protect personal data when publishing documents online. Anonymization is an important process that retains the nature of the original document while removing personally identifiable information from it. The Ministry of Justice of the Slovak Republic in collaboration with Essential Data are currently deploying an IT solution that combines human and artificial intelligence to speed up and improve the anonymization process of judicial decisions in Slovakia. The innovation lies in the combination of the use of modern web technologies, UX (User experience) principles, knowledge in the form of rules for the recognition of personal information, NLP (Natural language processing) tools and an AI model trained on inputs from court staff. From the production operation we can see that more than 55% of the documents are anonymized without the need for correction, the average time of anonymization is 3.5 minutes, and the success rate at which the AI correctly identifies the data is around 90%.
- What are micro-services? Basic principle of micro-service functioning
- Benefits and advantages of micro-services in project, analysis, architecture and operation
- The need for micro-services in the eGovernment environment
- Current problems related to the implementation of architecture of micro-services
- Methods of solving micro-service problems in ED projects
Public sector generates large quantities of data, that can be used as an input to decision making by government institutions. What’s more, not only government, but many companies from private sector can utilize this information and base their decisions on a publicly available information. However, in order to deliver useful information to consumer, there is a technical challenge in collecting, extracting, transforming and presenting this data as a relevant statistical output. Tasks of choosing the correct source, preparing the data, data cleanup and report designing for a specific view are contributing to difficulty of this publication process, which makes it tedious and non-intuitive.
The judicial and court statistics are significant sources of information for justice department. Improvement and development of these statistics were ones of the main goals of the project RESS. As a part of our presentation we will introduce implemented solution in form of an application, that produces required statistical outputs over millions of entries in real time within a few seconds.
During the design phase of the application, we put great emphasis on user accessibility and experience. Besides that, inexperienced users can easily get statistical information from justice department on a single page. There is an active help in form of tutorial available mainly for new users. Created statistical reports can be easily printed out or exported to standard exchange formats. This implemented solution can generate desired reports in real time and is used not only for purposes of analytical center of Ministry of justice, but also as a support in decision making of public experts and common people.