Annex I: Application of AI in the public sector

• Author: Cepilovs, Aleksandrs
• Pages: 35-35

AI IN THE OPERATIONAL MANAGEMENT OF LARGE-SCALE IT SYSTEMS  35

Annex I: Application of AI in the public sector

AI application

AI function and value proposition

Use cases in public sector environments

AI-based knowledge

management (KM)

software

 Generation and systematisation of knowledge-

gathering, classifying, transforming, recording

and sharing knowledge;

 Expert systems supporting codification of

knowledge in KM systems;

 Neural networks for analysis of data and

production and sharing of knowledge.

 Clinical documentation powered by AI in

hospitals;

 Case documentation for law enforcement;

 Knowledge management, e.g. IT service desks.

AI process automation

systems

 Automation of standardised tasks using either

rules-based or machine learning based systems;

 Automated workflow processing; IoT and

intelligent sensor based technologies; case-

based reasoning;

 Robotic process automation, substituting

humans in routine or dangerous processes.

 Automated clinical diagnostics (e.g. cancer

diagnostics based on image recognition);

 Automated data entry and classification;

 Automated request processing and response.

Virtual agents

 AI-based agents for interaction with humans

using NLP and generation via text or sound;

 Closed-domain conversational agents as a first

layer of contact;

 Can be implemented with automated

translation.

 Chatbots for IT service desk;

 Systems for automated routing of requests with

NLP capability;

 Human-computer interaction in cases of

repetitive tasks with defined outcomes (closed-

domain systems with limited knowledge bases).

Predictive analytics

 Statistical analysis in cases of small data;

 Big-data analytics using machine learning

methods for decision making support;

 Predictive analytics for automated decision

making.

 Predictive analytics in policing (threat prediction

and prevention);

 Predictive modelling for weather emergencies

and seismic activity;

 Predictive modelling for infrastructure

maintenance.

Identity analytics

 Using machine learning for advanced analytics

of identity data in real time (including facial

recognition for identification and

authentication);

 Risk-based identity checks using big data and

machine learning.

 Facial recognition technology for crime

prevention;

 Facial recognition on borders for automated

traveller processing;

 Facial recognition for authentication in online

service environments;

 Fraud prevention in online service environments

to secure government data.

Cognitive robotics &

autonomous systems

 Systems with higher-level cognitive functions,

involving knowledge representation and ability

to learn and respond;

 Sometimes in connection with affective

computing to respond to human behaviour

including emotions.

 Autonomous vehicles in public transit;

 Autonomous vehicles in security and defence

applications;

 Robot-assisted surgery.

Recommendation

systems

 Information classification systems supporting

human decision making and action;

 Prediction of user preferences based on

historical data.

 Service personalisation in online service

environments based on historical data on service

consumption.

AI for cybersecurity and

threat intelligence

 Data mining and big data analytics for threat

landscape scanning, including NLP;

 Cyber threat identification, analysis, and

proactive response;

 Vulnerability monitoring.

 Cyber threat analysis, vulnerability identification

based on big data analytics and machine learning.

 Automated response to potential cyber threats.

Table 1: Opportunities for application of AI in the public sector (adapted by the authors based on Wirtz et al., 2019)