Welcome to SUGO

Our research aims to drive transformation of public administrations with digital innovation.

Mission Statement

Digitization and innovation are the key catalysts for creating efficient and widely available public services for citizens and businesses. The goal of our work is to analyze application scenarios and to demonstrate how digital technologies can be applied as drivers of innovation in public spaces.

The research activities are conducted within the following key areas:

  • Digitization of Processes and Services in Public Settings
  • Artificial Intelligence in Public Administrations
  • Adoption of Digital Technologies in Public Healthcare

Supervision Offer

We are offering supervision for all projects within our scope, including seminar, internship, bachelor thesis, and master thesis projects. You are encouraged to either choose a project from our “current research topics page” or propose an own idea.

Current Research Topics

Data Analysis and Pattern Recognition in Business Processes

Bachelor Thesis Computer Science Project

The digitalization of business process is an enabler for increasing the efficiency in public organizations. With the increasing complexity in administrative processes considerable challenges arrise.

By analyzing the metadata of business processes or by detecting unusual states, the processes in such systems can be improved

Tasks include but are not limited to:

  • Evaluation of anomaly detection algorithms
  • Building or training models with data mining approaches based on process metadata

Approval Prediction in Business Processes with Machine Learning Approaches

Bachelor Thesis Computer Science Project

Approvals are an integral part of administrative processes which are often assessed by people with in-depth knowledge of the underlying domain. The available resources in an organization therefore determine the number of processes that can be completed.

Therefore, applying a machine learning approach based on historical data to predict or complete the approval step could significantly improve the effectiveness of the organization.

Tasks include but are not limited to:

  • Evaluation of prediction algorithms
  • Building or training models with historical data

Document Classification and Information Extraction for Business Process Selection

Bachelor Thesis Computer Science Project

Today's administrations face the challenge of increasing the efficiency and quality of their services while shortening the duration of their internal processes. Facilitating the communication process between citizens and administrations is a key success factor. In this context, the effective handling of unstructured data in the form of documents poses a major challenge.

Classification and extraction techniques can be used to increase efficiency and effectiveness by automatically analyzing documents. In addition, the extracted information can be used for business process selection.

Tasks include but are not limited to:

  • State of the art analysis
  • Analyse currently available products or libraries
  • Evaluation of classification and extraction algorithms

Raw translation tool for individual document sections

Bachelor Thesis Computer Science Project

Handling unstructured documents in foreign languages is a challenge in large eGovernment applications. The resources for translation tasks are typically limited, specially when dealing with multiple foreign languages.

Informal translation of document sections help save time and use resources more effectively. Translations can thus be used more frequently to find out whether particular sections are relevant and forwarding to an official translator is necessary.

Design and prototypical implementation of a translation tool for individual document sections.

Tasks include but are not limited to:

  • State of the art analysis
  • Design, conception and prototyping
  • Evaluation of machine learning techniques

Sentiment Analysis and Categorization of Customer Inquiries

Bachelor Thesis Computer Science Project

Efficient and effective communication processes with citizens play a central role in e-government applications.

Automated categorization of customer inquiries provides a means to simplify further processing. Text mining approaches can be used to anticipate the mood of the author in order to filter emotionally charged inquiries. This method can be used to assign such inquiries with a higher accuracy in order to better plan priorities or urgencies.

Tasks include but are not limited to:

  • State of the art analysis
  • Evaluation of classification algorithms
  • Evaluation of sentiment analysis approaches

Evaluation of Approaches for software-assisted Therapy Planning

Bachelor Thesis

Rehabilitation involves identifying a patient's health status and needs, linking problems to relevant patient and environmental factors, defining rehabilitation goals, planning and implementing interventions, and evaluating impact. Creating and adapting national rehabilitation plans is an essential aspect of continuously improving therapy results while optimizing the use of available resources.

The aim of the work is to investigate existing solutions for software-assisted therapy planning and develop concepts for possible improvements.

Tasks include but are not limited to:

  • Process analysis of therapy planning
  • Investigation of State of the Art Solutions
  • Design, conception and prototyping of subaspects
  • Assessing AI/ML approaches for therapy planning