As stated above, the scientific exchanges will serve a two-fold objective. The first objective is a generic scientific exchange between the research groups, to foster collaboration on topics such as the creation of multilingual dialogue corpora or the development of socially-competent spoken dialogue systems. The four partners to the project share a common scientific interest in the computational study of human-human and human-machine dialogues. In particular, the creation of multilingual dialogue corpora is a research topic of special interest to the project partners. While large multilingual text corpora have become a common resource in the last years, very few of these resources focus on conversations. This scarcity remains an important bottleneck both for foundational research and the development of practical applications.
The second project objective will focus on writing a project proposal to Horizon 2020 on the automatic analysis of emergency calls. Medical emergencies are often characterised by numerous uncertainties, as the emergency operators need to assess a complex medical situation based on a single information source, namely the account of patients or eye witnesses on the phone. These callers are often under high levels of psychological stress and typically have no medical training. Based on the collected information, the operators must quickly decide how to handle the situation and dispatch medical resources (such as doctors or ambulances) accordingly. Harnessing this combination of high uncertainty and strong time constraints is a demanding task which is subject to considerable cognitive and emotional load. The proposed exchange will allow us to prepare in 2019 a concrete project proposal that will investigate whether novel technological solutions (drawing upon speech and language technology, machine learning and dialogue modelling) can help make emergency response services more reliable and effective.