Dialogue Systems

"I need to gather the latest research findings on AI and its applications in medicine for a presentation at an international conference. How can I find the most relevant scientific articles and integrate them effectively into my presentation?" For scientific library staff, this may be a challenging but manageable request, whereas for Artificial Intelligence, it represents a true technological challenge. The "Dialogue Systems" group is working on advanced systems capable of assisting with such complex tasks using natural language.

The "Dialogue Systems" group specializes in the development and enhancement of AI-powered human-computer dialogue systems that can perform in-depth semantic analysis of spoken language. Advanced statistical modeling techniques are employed, enabling the system to search large datasets and extract relevant information. The example of scientific research represents one of the more complex application fields for Artificial Intelligence: systems that not only guide users through simple informational queries but also assist in solving complex problems, identifying pertinent data, and processing it effectively.

Spoken language has long been the primary medium of human communication. It's entirely part of our daily lives that we rarely pause to consider it — we simply understand the words and sentences effortlessly. Yet, when interacting with computers, users still rely on basic tools like keyboards and mice, despite the fact that modern AI and machine learning technologies have far surpassed human capabilities in many areas. Currently, computers don’t fully comprehend spoken language in a way that allows it to serve as the sole input method. However, advancements in AI and natural language processing (NLP) have recently achieved product-level functionality in certain domains. A growing number of speech recognition and natural language dialogue systems are now commercially available, enabling the automatic conversion of speech into text, which is useful for both professionals and everyday users.

Nevertheless, more advanced levels of speech processing are still largely confined to research. These include robust semantic analysis, which allows systems to understand the meaning behind spoken word sequences, and the development of conversational abilities, enabling systems to ask users questions and provide relevant responses. The application of Explainable AI (XAI) and Argumentative Dialogue Technologies to these higher-level speech processing tasks is opening the door to entirely new forms of interaction. By modeling these sophisticated elements, AI systems can not only interact naturally through speech but also offer more intuitive, transparent, and context-aware user experiences, paving the way for more seamless integration of spoken language technologies in diverse fields and applications.

The Dialogue Systems Group has placed its general research focus on the development and evaluation of user-friendly Spoken Language Dialogue Systems (SLDS). This objective is based on the following major aspects: Sentiment Analysis, Stress and Emotion Processing, Explainable AI and Argumentative Dialogue.