The aim of the project is to prototype and evaluate algorithms capable of using NLP and machine learning to detect malicious transactions associated with the registration of domain names. Particular emphasis is to be placed on detecting registrations of .nl domain names under false pretences, e.g. by providing false Whois data, with a view to running C&C servers or phishing scams. Knowledge and tooling resulting from this research is expected to contribute to further enhancement of the security and stability of .nl.
The project is expected to yield the following results:
- NLP algorithms capable of detecting malicious domain names at the time of registration
- Prototype algorithms and evaluations of them
- A study report
We are looking for a highly motivated MSc student with a passion for internet technology, data analysis and machine learning. You are energised by investigating exactly how mechanisms and systems work and you are able to devise and evaluate solutions. The ideal candidate is a self-starter with an enquiring mind, capable of developing proofs of concept.
If you are interested in this project, e-mail email@example.com explaining why you think you are the right person to carry it out. Questions can be mailed to the same address.
About SIDN Labs
SIDN manages the Dutch national internet domain, .nl. We are an innovative domain name registry with our own DNS anycast network and more DNSSEC-secured domain names than any other top-level domain in the world (2.4 million of the 5.5 million .nl domain names). SIDN Labs is our R&D arm, which develops and evaluates new technologies and systems with a view to further enhancing the stability and security of .nl, the DNS and the wider internet. On our ENTRADA big data platform, we retain more than a year's DNS data, including about 65 billion unique DNS queries, of which we receive 150,000,000 a day. More information on www.sidnlabs.nl.