Skip to content
camsocs

About camsocs

A directory of Cambridge student societies and sports clubs, with an events calendar, in which every fact says where it came from.

What it is

Cambridge has well over a thousand student groups, listed in many places that rarely agree: the SU directory, the Proctors' register, Sport Cambridge, Camdram, college and JCR/MCR sites, and each society's own pages. camsocs brings them together into one record per group, keeps the record current, and shows the source, date and status of every fact, so you can judge it for yourself.

It is a personal project by Joshua Fehn, a student at Cambridge.

Unofficial and independent

camsocs is an independent, unofficial project. It is not affiliated with, endorsed by or run by the University of Cambridge, any of its colleges or Cambridge SU.It does not use the University's name, crest or logo as branding, and nothing here should be read as the University's or the SU's view. Societies registered with the Proctors or the SU are bodies independent of both.

How we collect data

  • Public sources only. An automated harvester reads publicly accessible pages and feeds. It identifies itself honestly as camsocs-bot/0.1 (+https://camsocs.vercel.app/about).
  • robots.txt, with one exception. The harvester follows robots.txt on every site except the pages societies publish about themselves: their own websites and Linktree pages, the public Google calendars they link, files on their SU listing, and college, JCR and MCR society lists. Those are read, also by the AI model, even where robots.txt or an AI opt-out asks otherwise, because they exist to tell students about the society. Crawl delays always apply, and the harvester makes one request at a time per site.
  • Nothing behind a login.Pages that need a Raven or members' login, or that refuse access, are never fetched. Email addresses a site has hidden or obscured are never decoded.
  • No Instagram, Facebook or LinkedIn scraping.We store links to those profiles only. Instagram posts are read, if at all, only through Meta's official API for public professional accounts.
  • AI extraction, checked.An AI language model, run by a third-party provider, reads free text from societies' own pages to pull out facts such as how to join. A fact counts as found only when the model quotes the source word for word and the quote is checked; otherwise it is labelled inferred. Content from talks.cam and tbc.xyz is never given to an AI model.
  • Mailing lists. Events announced on society mailing lists are shown only to visitors signed in with a verified @cam.ac.uk address, and the messages themselves are never published.
  • Societies, not people.Public listings show organisations and the role addresses they publish (president@…, secretary@…). Personal student addresses are not shown. Where a description names a committee member, performer or other student, the name is replaced by their role (“contact our President”) before it is published. Speakers announced for public talks keep their names.

Sources

  • Cambridge SU society directory: society names, categories, descriptions, contact links
  • Proctors' register of societies: which societies are registered this year
  • Sport Cambridge club pages and Sports Club Guide: sports clubs, Blues status, training
  • Camdram: theatre companies, shows and venues
  • CUCBC: college boat clubs and the rowing calendar
  • College, JCR and MCR websites: student bodies and their role contacts
  • Societies' own websites and Linktree pages: how to join, membership, contacts, socials
  • Public calendar feeds (Google Calendar, Luma, iCal): events societies publish
  • talks.cam: talks, read by a parser only, never by an AI model
  • Playwaze and SU What's On: leagues and events
  • The University's term-dates page: Full Term dates, for week numbers

What the labels mean

Under every fact is a line saying where it came from and how sure we are.

Found: Read from a published source.
Read from a published source.
Inferred: Worked out from a source that does not state it word for word.
Worked out from a source that does not state it word for word.
Owner-confirmed: Confirmed by the society itself.
Confirmed by the society itself.
Edited: Set by the camsocs editor.
Set by the camsocs editor.
Conflicting: Sources disagree; every reading is shown.
Sources disagree; every reading is shown.
Not published yet
No public source gives this.
Not checked yet
No source has been read for this yet.

The percentage is a confidence computed by fixed rules (how the value was read, from whose page, and how many independent sources agree). It is never the AI model's own estimate.

Corrections, removal and contact

Societies decide what is said about them. To claim, correct or remove a listing, or to object to how data about you is used, use the request form or email camsocs@joshua-fehn.de. A person reads every request. See the privacy notice for your rights.