SmartDATA, ICT for a connected world

The emergence of artificial intelligence (AI) and streaming data hold great promise for improving the performance of machine-learning methods. The main obstacle to the successful introduction of AI is the lack of algorithm transparency. The introduction of AI for streaming data requires technical as well as organisational changes. As the established telecommunications operator in the Netherlands, KPN connects millions of business and consumer customers through communications and ICT services. Being a trusted service provider is key to KPN's strategy. To deliver the best possible service to customers and maximize commercial effectiveness and profitability, KPN is already applying basic machine learning to many aspects of its business. More advanced AI algorithms require more data and privacy. Currently, not all data can be used at KPN, due to internal privacy guidelines and regulation.
The proliferation of AI and machine learning technologies offers enormous potential to develop predictive tools. The research project investigates four main questions:
(1) How to deal with privacy and transparency? (2) How to deal with streaming data? (3) How to realise accurately targeted personal offering? (4) How to apply representational learning to the Dutch language? The overall goal of the SmartDATA project is to address the four problems to facilitate the use of modern AI technology by the telecom sector. To achieve this goal, an interdisciplinary team of JADS-affiliated experts is working together with KPN.