About Us

The Knowledge Engineering Group (KEG) addresses problems in three main research domains: Data Engineering, Natural Language Processing and Neuroscience.

In Data Engineering, the group develops techniques and prototypes for managing and extracting knowledge from structured and unstructured datasets. Key areas include time series forecasting, real-time sensor data processing, IoT usage characterization, community detection and opinion mining, and medical decision support. We address challenges such as incomplete records, imbalanced datasets, and optimizing models for big data.

In Natural Language Processing, the group applies machine learning methods to tasks such as text classification, named entity recognition, and multilingual text generation, focusing on low-resource settings and cross-lingual transfer. Also, the group addresses challenges related to the social impact of NLP, such as bias detection in uni- and multimodal systems, knowledge distillation, and enhancing explainability in AI models. Notable applications include AI assistants for cooking and home automation, and sentiment analysis tailored to the Romanian language.

The group’s Neuroscience research involves developing machine learning techniques for spike sorting, burst detection, and EEG data analysis and mapping functional brain networks using graph neural networks. In collaboration with the Transylvanian Institute of Neuroscience, the group aims to develop computational tools for analyzing neuronal and brain activity.