Data science approaches to extract knowledge from the internet

Description

Due to the worldwide popularity of its services and the active involvement of users in the content generation process, Internet is the biggest source of data, with hundreds of exabytes being exchanged on a monthly basis and several billions of connected devices.
While the value of this data is huge, handling its ever growing volume is challenging and the ungoverned process with which data is injected in the network is constantly generating cyberthreats. Specific data science approaches and solutions are needed.
Network management and security involves the study of data traffic, its characterization and monitoring with approaches that dynamically and continuously adapt to new conditions as well as new threats. Anomaly detection algorithms have to be combined with the design of appropriate countermeasures to cybersecurity threats.
Similar approaches are needed for monitoring the quality of experience perceived by users running an application or receiving a service so as to actuate some reactive mechanisms to cope undesired quality deterioration.
With contents being generated by users’ themselves, network and service monitoring techniques have to preserve privacy and to treat information in a fair and inclusive way.

ERC sectors 

  • PE6_1 Computer architecture, pervasive computing, ubiquitous computing
  • PE6_5 Cryptology, security, privacy, quantum cryptography
  • PE6_7 Artificial intelligence, intelligent systems, multi agent systems
  • PE6_11 Machine learning, statistical data processing and applications using signal processing (e.g. speech, image, video)
  • PE7_8 Networks (communication networks, sensor networks, networks of robots, etc.)

Keywords 

  • Machine learning
  • Anomaly detection
  • Network monitoring
  • Quality of experience
  • Cybersecurity
  • Privacy