I’ve made something, but I’m not too sure what it really is or if it’s of any use what-so-ever!
I started playing around with the “relationships” code (briefly mentioned here) to see what would happen if you selected multiple books and ranked the aggregated suggestions by frequency of appearance.
For example, here are 4 O’Reilly books that I’ve selected. The code pulls in the list of suggestions for each book, then (underneath the line) displays the books that appear in all 4 lists, followed by the ones that are only common to 3 lists, then 2, then those that only appear in one of the lists (with up to 100 being displayed in total).
So, in theory, if I’d picked up those 4 books and was looking for a couple more to borrow, then “PHP & MySQL Web Development” and “HTML & XHTML” should be safe bets.
You can click on any of the books above the line to remove it from the selection. If you click on “HTML” and “Web Based Applications” books to remove them, then you should be left with the “PHP” and “MySQL” books — in that case, a book on e-commerce becomes the first suggestion.
To start with a random book, use the following link and reload the page until you find one that interests you and that has several suggestions…
https://library.hud.ac.uk/perl/relationships2.pl
…then try adding one of the suggested books to the selection (by clicking on it) to see what the revised suggestions are.
Month: March 2007
CILIP: “Re-imagining the Library” Executive Briefing
I’ve got the great honour of speaking at the CILIP “Re-imagining the Library” Executive Briefing next week, although I’ll admit that I’m not looking forward to the journey down to London — getting up at 5am always disagrees with me, plus my body will think it’s lunch time at 9am!
I remember reading a librarian’s blog post where she said that short presentations are the hardest and I’ve got just 15 minutes …that’s only 900 seconds! Plus I’m the last on before lunch, and I know how fidgety I get at conferences if the speaker looks like he might start eating into the precious lunch break 😀
Looking at the final version of the presentation, I can’t help feeling I’m trying to cram too much in… but there’s so much I want to talk about! Rest assured, I won’t be stood there talking in a slow boring monotone — I’ll be gabbling 10 to the dozen like a demented chipmunk on helium.[1]
After timing myself this morning, there should be exactly 9 seconds spare for quick questions at the end. I did consider adding subliminal images of sandwiches and cakes to the “any questions?” page to try and ensure that no-one asks me anything that I can’t answer in 3 words or less (preferably “Q: Is it time for lunch now?”, “A: Yes!”).
[1] I did originally type “chipmonk”, which presumably would be a deep fat friar?[2] Also, if you do a Google search for “ten to the dozen“, it proudly tells you that…
ten to the dozen = one trillion
…yet it does understand what a bakers dozen is 😀
[2] this is why I never tell jokes during presentations
“I’m in the mood for watching a film about…”
Via this post on Lorcan Dempsey’s blog, I noticed that the Internet Movie Database now has genre keywords and tag clouds.
For example, library brings up a cloud where murder appears to be the largest tag. The list of films matching the chosen facet(s) is ordered by rating, so you get instant recommendations for whatever mood you’re in.
It’s also fun trying to guess which films might match multiple facets — 10 points if you can name a film that matches both “library” and “flame thrower” without looking!