We are constantly improving the capabilities of existing digital solutions using artificial intelligence. Here we highlight the currently used AI solutions.
e-Land registry data entries are in Estonian language (free text) and foreign users have a hard time understanding the content and registry decisions sent by-email. So, to solve this problem we offer the opportunity of machine translation in the e-Land Register and e-Business Register information systems.
It is a neural machine translation service where the translation is done sentence by sentence. Professional users can give recommendations to change the translation.
This is a private company service provided by TILDE SIA.
The transcription process in court is time-consuming and takes a lot of effort, so to make the process quicker we are using automated transcription service.
It is a speech and voice recognition software, which allows to transcribe the shorthand report of the hearing, which is used to prepare the protocol, based on the real-time and digitally stored audio recordings of the court session.
Prisoners need to access their personal data and digital Court file regularly and whenever they need to file appeals they use paper. To make these processes easier, quicker, and more efficient prisons can use hand-held computer devices with face recognition software that inmates can use to access the necessary services.
Legal transactions require identity verification, and it is not always convenient or possible for citizens to meet with the notary face to face. To solve this issue the e-Notary portal offers the opportunity verify your identity and validate your document using face recognition.
It is possible, by agreement with the notary, to carry out transactions by means of remote verification at a place suitable for the client (e.g., at home). All notarial acts except marriage and divorce can be performed by means of remote verification.
This is a private company service provided by Veriff's face recognition.
In order to publish court decisions, it is necessary to remove personal data. To make this process quick and efficient we are using automatic processing engine that recognizes personal data and makes it anonymous.
To improve data quality in court documents, we have started using automated labeling and anonymizing the data (removing personal data), can also be used in open data format.
To improve the quality of the analysis with court documents, we work with data labeling. It is possible to analyze all texts of crime events based on qualifications, subjects, objects and addresses. This helps to increase the quality of judgments and simplifies the compilation of legal statistics. Users can find similar cases and improve the quality of decisions and work efficiency in general.
Automatization possibilities:
- Analyzation of crime descriptions’ texts: named entities, more precise classifications, object recognitions, addresses, amounts of damage etc.
- Indexing all proceeding documents: documents can be queried (with metadata classifications); automatic summarization; lemmatization, anonymization etc.
- Natural Language to SQL Queries: Ask questions natural language to get information from justice data warehouse. E.g. What is the average amount for civil claims in 2023?
- Chatbot "Bürokratt": government agents, registries, services and virtual assistants all working together in one network to solve the problem of Government service users, enabling them access to a spectrum of services including personal data and e-services.
- Cloud solutions: OpenAI Azure services with publicly available data (AI assistants, deep analysis etc).