Multilingual and Cross-cultural interactions for context-aware, and bias-controlled dialogue systems for safety-critical applications

ELOQUENCE is focused on the research and development of innovative technologies for collaborative voice/chat bots.

Voice assistant-powered dialogue engines have previously been deployed in a number of commercial and governmental technological pipelines, with a diverse level of complexity. In our concept, such a complexity can be understood as a problem of analysing unstructured dialogues.

ELOQUENCE’s key objective is to better comprehend those unstructured dialogues and translate them into explainable, safe, knowledge-grounded, trustworthy and bias-controlled language models.

We envision to develop a technology capable of learning by its own, by adapting from a very data-limited corpora to efficiently support most of the EU languages; from a sustainable computational framework to efficient and green-power architectures and, in essence, that may serve as a guidance for all European citizens whilst being respectful and showing the best of our European values, specifically supporting safety-critical applications by involving humans-in-the-loop.

Overall, ELOQUENCE’s project considers building on top and to improve of prior achievements in the domain of conversational agents, e.g. recently launched and public-domain Large Language Models (LLMs), such as chatGPT (e.g., more recent versions), or LaMDa most of them developed in non-EU countries.

While including key industrial enterprises from Europe (i.e., Omilia, Telefonica, Synelixis), ELOQUENCE will validate the developed technology through (i) safety-critical scenarios with human-in-the-loop for security-critical applications (i.e., emergency services in call centres) and (ii) smart home assistants via information retrieval and fact-checking against an online knowledge base for lesser risky autonomous systems (i.e., home-assistants).

ELOQUENCE will target the R&D of these novel conversational AI technologies in multilingual and multimodal environments and demonstrated in several pilots.