Richard G Clegg
richardclegg.bsky.social
Richard G Clegg
@richardclegg.bsky.social
Academic studying complex networks at Queen Mary University of London and amateur scuba instructor. I do a lot of research using the Raphtory software for temporal networks, a fast efficient way to analyse network data:
https://www.raphtory.com/

He/Him
Anyone else in UK academia feel "open access publishing" has made things worse? In 2010 I could download pretty most papers for free. Most journals did not charge to publish. Now many of the most obvious journals cost a few thousand quid article processing. Ability to download papers got no better.
December 4, 2025 at 2:28 PM
Watching a TV show and a character in law enforcement complained loudly that she had not got a gun. Later in the plot it would have been very useful for her to be able to fire a gun but she was unable to do so as there was no gun. Chekov's absence of gun.
December 4, 2025 at 12:33 AM
Lovely mathematical jargon: A Procrustes problem transforms points in embeddings until they look like each other. In legend the innkeeper Procrustes transformed his customers (by stretching or shrinking -- typically unpleasant for customer) to match his bed. This is literally embedding.
December 3, 2025 at 12:04 PM
Last day of lectures for UCL. Must have done something right as the students gave me a biscuit (well Jaffa cake). Good lecturer. Have biscuit.
December 2, 2025 at 5:11 PM
It was an absolutely baffling decision to ask for strike action - it was doomed to fail and it worsens the negotiating position. It is a genuine shame. Not that long ago the union looked strong. This is a real failure of leadership.
Unfortunately the HE 2025 ballot failed at 39% turnout. I have written about why and about what desperately needs to change in UCU's national structures off the back of this result. Some HEC members need to chi up and understand their role in all this, and how disconnected they are from members.
The HE Ballot failed, now what?
If you’re reading this it is because today the UCU’s 2025 national ballot over pay and conditions in universities has failed. The last time…
hitchcockian.medium.com
December 2, 2025 at 1:20 PM
Stable door shutting expert called in after horse bolting incident.
www.bbc.co.uk/news/article...
OBR calls in cyber expert over botched release of Budget analysis
Rachel Reeves's statement was thrown into chaos after journalists were able to access the document early.
www.bbc.co.uk
November 27, 2025 at 1:52 PM
I went to an event at Goodenough college. It was a great event but I could not shake the feeling I had wandered into a parallel parody academia.
November 25, 2025 at 8:40 PM
Nice afternoon lecturing at UCL. Bloomsbury is very attractive at Christmas.
November 25, 2025 at 5:26 PM
Konect aka Koblenz collection seems to have disappeared and been replaced by a cybersquatting website. Anyone know what happened there or have a working URL?
November 23, 2025 at 6:16 PM
Halloween papercraft. Fun model to make. @robives.bsky.social @threadsaplenty.bsky.social
November 2, 2025 at 12:21 PM
I really love this -- inferring Anglo Saxon history from linguistics, place names, geography and network science:
keithbriggs.info/AS_networks....
Keith Briggs: : Anglo-Saxon communication networks
keithbriggs.info
October 29, 2025 at 12:12 PM
Fascinated by how ChatGPT does at my exam questions. It managed to extract a correct transition matrix from a text description of a Markov chain "consider someone doing x with probability of doing y". It had a damn good go at drawing the chain -- in ASCII art! The later choice is wild but I love it.
October 28, 2025 at 2:04 PM
Maybe 2020 was somehow emotionally exhausting looking at the Markov chain exam question I set: "Sisyphus is pushing a boulder up a mountain which is 3000 metres tall. He will move every hour according to the following rules. With probability 1/2 he will go 1000 metres up the mountain..."
October 28, 2025 at 1:20 PM
Little bit of signal processing humour for a Monday morning.
October 13, 2025 at 11:08 AM
New addition to my teaching slides: "where LLM will mislead you". ChatGPT and FFT/DFT is interesting. It got a 4 point correct (I only tried one). On an 8 point it got it wrong (but showed working). When I said it was wrong it got it right (when pressed admiited to internal python fft routine).
October 12, 2025 at 2:57 PM
Editing colleague's powerpoint and thinking about typesetting crimes. Co-worker created a beautiful graphic but spelled temporal with no l. We superimposed a powerpoint leter l as it was last minute. Four years later I'm still using that graphic and adding the l every time. #feelsbadman
October 12, 2025 at 2:13 PM
I crack me up (nobody else but I find it funny). [Writing a presentation about tracing crypto dealings.]
October 11, 2025 at 2:37 PM
Always a pleasure to work with @computerphile.bsky.social - this week it's back to basics with the layers model of the Internet:

youtu.be/eelvWAURfdI?...
Network Layers Model (Networking Basics) - Computerphile
YouTube video by Computerphile
youtu.be
October 9, 2025 at 5:47 PM
Today's meeting
Me: "I have an early draft of the grant on topic X we discussed for previous six weeks. Do not be afraid to criticise."
Five other academics: "X is dull we want to do Y instead."
Me: "You bunch of buggers."
Now writing a draft of an grant on Y (which is > X).
October 9, 2025 at 10:32 AM
Reposted by Richard G Clegg
Here are your 10 -essential- AI prompts for academics ... make your life easy with help from @profserious.bsky.social profserious.substack.com/p/10-ai-prom...
10 AI Prompts for Academics
making those hard jobs a little easier ...
profserious.substack.com
October 5, 2025 at 7:56 AM
Honestly @bridgecommand.bsky.social is crazy good fun. Flying our own spaceship. Accidentally killing our senior officers.
Yet another excellent mission playing at spaceships @bridgecommand.bsky.social We achieved all of our objectives! Although sadly one of our Marine Colonels died heroically disabling the shields by boarding an enemy battleship, so we were able to nuke it… but sadly not before the Colonel was clear.
September 30, 2025 at 9:22 PM
After so many years of networks research a paragraph that begins "Let G = (V,E) be a simple, undirected network where V is the set of nodes" is like slipping under a warm blanket. So soothing.
September 30, 2025 at 2:08 PM
Writing a grant proposal and a co author told me that rigorous data analysis is not exciting or sexy. My head knows it is true but my heart cannot accept this.
September 30, 2025 at 12:10 PM
From a teams meeting I am in right now and I am honestly infuriated.
September 29, 2025 at 9:15 AM
Reposted by Richard G Clegg
Absolutely fantastic. The treatment is expensive and difficult but any progress against Huntington’s disease is good and this is amazing progress. Hooray!
September 24, 2025 at 2:42 PM