Tiny repository to hold the script to handle the web of science records
Do a search on Web of Science (www.webofknowledge.com) by going to the advanced search and using
GP="CMS Collaboration" AND AD="Ghent" AND PY="2019"
i.e. we search for a group author (GP) which matches the CMS Collaboration and for an address (AD) which matches Ghent, which helps to filter out a lot of proceedings from non-Ghent people. The PY is the year of publication which you want to adjust for the previous year you are in now.
Note: the web of science webpage only works from the UGent network, and is not scriptable (unless you can convince the central administration to pay a very expensive licence). Its lay-out changes all the time and the query for the advanced search has also changed in the past.
The resulting list can be exported ("Export...") by "Other File Formats", then you need to select the records (but there is a maximum of 500, so you can select 1-500, 501-1000, seperately etc... if needed), "Full Record" as record content and "plain text" as file format. Note I did this export on the search results webpage which takes way less time than going to all records individually.
I've written a script (see attachment) which breaks down this export file and divides them in separate records + downloads the pdf's for each of these records [possibly there might be a few special cases which have to be checked manually, just add it to the elif statements]. Also, try to open all pdf's to check if they were downloaded correctly. Journals like to change their website often, in order to make you spend time in updating your scripts. That's how bureaucracy works: "you shall not automatize!"
Zip the directory with wos-records and pdf's, as created by the script, and upload it to a filesharing link, and ask the Biblio people (biblio@ugent.be) to upload it.
(you probably need to use wetransfer or dropbox for the large zipfile)
(+tell them who are the UGent authors, even though this can actually be deduced from the records itself)