https://t.co/0FsQxpVnyP - Dis·cover the news, a Portuguese media watchdog created by Flávio Martins and André Mourão received an award from @ArquivoWeb_PT in the CIÊNCIA 2019 Science Summit #ciencia2019pt
Link: https://t.co/ID0S4qausU
\cc @FCTNOVA@iNOVAmedialab@NovaUnl
FFmpeg is moving to Rust 🦀
Our use of C and Assembly in FFmpeg has been an unacceptable violation of safety.
FFmpeg will be running 10x slower - but we're doing it for your safety.
All your videos will appear green - safety first, working software later.
We are hiring!
If you are a #Psychology#Sociology#Neuro student, BSc or MSc level, and care about the role of human biases in the spread of #disinformation, please consider applying
If you are not but happen to know who is, please share
https://t.co/Py1XlW4OGI
@SpacLab@lip_pt
A proposta mais criticada de sempre #ChatControl tem votação marcada para a reunião de Ministros da Justiça no próximos dias 9 e 10. A proposta do conselho ataca a segurança das comunicações. Como votará Portugal @justica_pt@RPPortugalUE@govpt?
https://t.co/io8sahxs9U
@lauradietz99@ShirKi @gonendh Did I install a 2fa app on my Mac to get codes right on my system tray? Yes I did! Use https://t.co/0yERIgNC4X and the app for iOS https://t.co/Kli1BMGJhD
For studying news topics, a tool like https://t.co/cAuuLkJGch is needed. This screenshot is a slide from my PhD candidacy presentation displaying one of the many reasons why this concept is invaluable. Excellent work @flaviomartins and @andmou! #webarchiving#webarchives
@yoavgo I built a smaller collection that contains approximately 800k English tweets from Twitter verified accounts (mostly news sources) https://t.co/qDqHr14mp5
@yoavgo Your best bet seems to be to select lang=en from https://t.co/HNNZQYO54v the archival scheme is kinda awkward though. I would modify this script to do it fast https://t.co/ebQc2zVDCW
@srchvrs @diazf_acm @ACMSIGIR My experiments on the Tweets2013 dataset show that you need 10000 on the 2013 queries. For 2014, I did not reach the same level of map even at 10000. Its great on more standard collections though.
@srchvrs @diazf_acm @ACMSIGIR My recent experience is that re-ranking RM wasn't as effective as the re-retrieval implementation for very short documents (microblog). To get similar MAP I had to retrieve and re-rank a lot more than 1000 candidate documents.