Flemish Infoline - Automatic classification of incoming phone calls

Natural Language Processing (understanding spoken language) is used to automatically classify incoming telephone questions at the 1700 (Flemish Infoline), to integrally transcribe and document the conversations, and to suggest answers. With this project, Informatie Vlaanderen wants to optimise the operation of the Flemish Information Line with the help of Artificial Intelligence (AI). Concretely, the project wants to detect and categorise the incoming questions faster and more 'finely-meshed', and also integrate this 'finely-meshed' into the editorial management. In this way, the answers can be suggested more quickly, more adequately and also more 'automatically' to the answer provider, regardless of the chosen communication channel (oral or written). To this end, the project aims to harness the broad potential of AI, with emphasis on Natural Language Processing (NLP), Machine Learning (ML) and Deep Learning (DL), and language and speech technology (Speech-to-Text [STT] and Text-to-Speech [TTS]). By using AI modules, the project wants to interpret and also 'manipulate' the question/answer process of the Flemish Infoline in order to make the whole process more efficient and automatic, using all the implicit knowledge present in the process. In essence, the project aims at two objectives: (1) the quicker and more efficient capturing and categorising of the question, and (2) the quicker and more efficient forwarding of the 'correctly captioned' question to the 'correct' answer provider with the 'correct' skills, whereby a valid and standardised answer (suggestion) is immediately sent along with it (so that the answer provider does not have to look up the intended answer script).