DIR 2021 1/2 is live at: https://t.co/23Jd8gDT9C.
Authors from NL and Belgium, consider submitting an abstract of your work for lightning talks at the workshop here: https://t.co/ZdF63kJvsE.
T-8 for #DIR2020! With over 100 registered participants, 10 confirmed speakers, an incredible keynote speaker (Thorsten Joachims) and 10+ posters we are now sure it will be a blast π Snatch up the last tickets before it's too late π https://t.co/9GtECPKZym
We have also launched our call for posters! Would you like to present your work at the #DIR2020 poster session? Poster sessions will be held in https://t.co/yO6gRUWFA2, so expect vibrant discussions and plenty of social interactions! π 2/n
Save the date and spread the word! π #DIR2020 will be taking place on December 3rd, online! @ngrams @jakubzavrel @dvdgrs @anneschuth @dodijk@arjenpdevries 1/n
PS we're adding links to slides to our program: https://t.co/NpjTJR9key (unfortunately did not come up with a way to add them to the paper booklets) #DIR2019
β‘οΈ#DIR2019 unique opporunity to meet local IR & #RecSys experts presenting their most recent work from top conferences! 5/5 on my evaluation metric: relevance+novelty+diversity+beautiful venue+stylish catering! Dank je @dir_2019 @dvdgrs @anneschuth @dodijk https://t.co/w1buD8DeRB
Had a great time @dir_2019 yesterday, diving back into the IR domain while listening to some great works from Dutch and Belgian IR researchers. You can now also find my two cents @ https://t.co/zCgoeIVD9u
choosing #RecSys algorithms in focus groups study protyping user control mechanisms for transparency in personalised content by FairNews project @HarambamJaron#DIR2019 https://t.co/Qh6oxBpX2S
The work presented "Designing for the better by taking users into account" revolves around focus groups with news readers to study the aspect of control in news recommenders.
Last talk of #DIR2019 by Jaron Harambam, who opens with a great question. I know the answer π! Interdisciplinary work between computer scientists and an anthropologist (Jaron), presented before at #RecSys2019!
Sorry, I think they already had the dataset, but study domain adaptation in this one (learning a model on dataset 1 (Enron) yields a bad model for dataset 2 (Avocado) due to domain specificity and jargon etc.). #DIR2019
Hello again @HAzarbonyad! Now presenting work done at @MSFTResearch, on commitment detection in emails! Where they crowd sourced a dataset for detecting when people express to commit to some task (think: "sure, I will [...]") #DIR2019
Next up in still a full house: @LiseStork on identifying scientific names in handwritten text, combining computer vision with entity linking, since entities play a central role in information needs of domain experts. Presented at #ECIR2019! Now at #DIR2019
sliding window&exponential decay estimators for bias correction in inverse propensity scoring average reward for system actions: off-policy evaluation in non-stationary environment for top-1 recommendation simulated on LastFM dataset @RolfJagerman#DIR2019 https://t.co/FQxgm1Rgpr