Alun Davies

Revealing Bias in Internet Measurements

Alun Davies
0

Internet measurement platforms give us a clearer understanding of the state of the Internet. But none of these platforms gives us a ‘view from everywhere’, so that understanding will always be based on a partial and potentially biased view of things. In this episode, Pavlos Sermpezis talks about the importance of revealing bias in Internet measurements.



The RIPE NCC is not liable for the views and perspectives expressed by speakers. This podcast is brought to you via third party hosting platform Acast. Check out acast.com/privacy for privacy and opt-out information.

You can also follow our podcast on Apple Podcasts and Spotify.

The analysis Pavlos and his colleagues are carrying out is part of the AI4NetMon project that aims to deliver methods for reducing biased findings from Internet measurement infrastructures. The platforms under investigation include RIPE NCC operated services RIPE RIS, RIPE Atlas, RIPEstat, and also RouteViews.

01:20 - Pavlos's winning article on Bias in Internet Measurement Infrastructures

01:37 - About the Data and Science Laboratory (DataLab)

02:53 - More on the RIPE Routing Information Service (RIS)

04:10 - Robert talking about RIPE Atlas in Episode 3

19:46 - Watch Pavlos's presentation at RIPE 84

20:20 - Other data sources Pavlos and his colleagues are using: CAIDA’s AS-rank, PeeringDB, Internet Health Report (AS-hegemony), and bgp.tools.

22:15 - Read more about RACI support for AI4NetMon

0

You may also like

View more

About the author

Alun Davies Based in Amsterdam

Hailing from a research background in philosophy, linguistics and computer science, I came to the RIPE NCC back in 2016 and took on my role as the RIPE Labs Editor in 2020.

Comments 0