しましまのSIGIR2019まとめ

Gabriella Kazai @gkazai

Marco Wrzalik presenting his PhD proposal on document similarity for interactive exploration, complete with demo. #SIGIR2019 DC pic.twitter.com/L32ZRE4hBr

拡大

拡大

拡大

拡大

ACM SIGIR 2019 @sigir2019

"Deep chit-chat: deep learning for chatbots" Room Louis Armand Ouest pic.twitter.com/YMh4t8T86N

拡大

拡大

ACM SIGIR 2019 @sigir2019

"Building economic models and measures of search" Sale 4 pic.twitter.com/4AEsHQieLS

拡大

拡大

Amal Alharbi @alharbi_amal

Learning to Rank in Theory and Practice: From Gradient Boosting to Neural Networks and Unbiased Learning. Interesting tutorial @sigir2019 pic.twitter.com/ApXVtNFBSm

拡大

ACM SIGIR 2019 @sigir2019

"Effective online evaluation for web search" Salle 3 pic.twitter.com/9dWZo9j87Z

拡大

拡大

Harrie Oosterhuis @HarrieOos

Walking into the room you have to present in and finding out it's the large auditorium and you are presenting for 200+ people. Woops. @sigir2019 ltr-tutorial-sigir19.isti.cnr.it With @mdr @RolfJagerman @bemikelive @mr_courage @RamK1729 @claudiolucchese @fmnardini pic.twitter.com/uj5xxnPCRz

拡大

ACM SIGIR 2019 @sigir2019

« Learning to Rank in theory and practice : From Gradient Boosting to Neural Networks and Unbiased Learning » Room Gaston Berger pic.twitter.com/x1b1NpQqaE

拡大

拡大

ACM SIGIR 2019 @sigir2019

« Doctoral Consortium » Room A&B pic.twitter.com/gmwUplRAsM

拡大

拡大

Mounia Lalmas @mounialalmas

Explanation in recommender systems via knowledge graphs #sigir2019 tutorial on explanation in search and recsys pic.twitter.com/VDleLWiuui

拡大

拡大

Mounia Lalmas @mounialalmas

Unified view of RecSys and Search ... still hold #sigir2019 tutorial on explanation in search and recsys pic.twitter.com/vATbreg7Ut

拡大

Min Zhang @mzhang_bj

Seven #sigir2019 Tutorials are running now: Learning to rank, Explainable recommendation and search, Deep Chit-Chat, Online evaluation, Table extraction, Building economic models and measures for search, Example based search. There will be 4 more in the afternoon. 👏🏻👏🏻👏🏻 pic.twitter.com/gXXPYTr66A

拡大

Mounia Lalmas @mounialalmas

RecSys ... the 5W .... good to remember #sigir2019 tutorial on explanation in search and recsys pic.twitter.com/EoE2UvxB1j

拡大

Maarten de Rijke @mdr

@HarrieOos @sigir2019 @claudiolucchese @fmnardini @bemikelive @mr_courage @RamK1729 @RolfJagerman Franco has taken over. pic.twitter.com/3XgjKy1bu9

拡大

Gabriella Kazai @gkazai

Alfan F Wicaksono is looking at user interaction models in job search. #SIGIR2019 DC. Clicks underestimate prob to continue. @leifos @pt_ir pic.twitter.com/fHDpOTRmG0

拡大

拡大

拡大

拡大

Mounia Lalmas @mounialalmas

Various approaches based on deep models for explainable recommendation #sigir2019 tutorial on explanation in search and recsys pic.twitter.com/GEHfnPmppK

拡大

Mounia Lalmas @mounialalmas

Summaries of three main types of approaches for explainable recommendation, including attention models, deep models, KGs and post-hoc #sigir2019 tutorial on explanation in search and recsys pic.twitter.com/cvLVUFZTPK

拡大

拡大

拡大

Gabriella Kazai @gkazai

Binsheng Liu on using query variants generated by humans and by random walk on click graph to improve search performance. #SIGIR2019 DC pic.twitter.com/K74aqfzcyP

拡大

拡大

拡大

拡大

Mounia Lalmas @mounialalmas

Challenges and directions ... still lots to do for fully explainable recommendation in RecSys #sigir2019 tutorial on explanation in search and recsys pic.twitter.com/fndZYFXMyT

拡大

Leif Azzopardi @leifos

@guidozuc explained the Probability Ranking Principle and its limitations - fixed costs and fixed interactions. He now explains the Interactive PRP as way to overcome these. #bemms @sigir2019 pic.twitter.com/GL5ZBsXXWX

拡大

ACM SIGIR 2019 @sigir2019

Coffee is here! pic.twitter.com/EtyE4Fn51d

拡大

Maarten de Rijke @mdr

@HarrieOos @sigir2019 @claudiolucchese @fmnardini @bemikelive @mr_courage @RamK1729 @RolfJagerman Claudio is back on stage with a hands-on segment pic.twitter.com/uA1LfybEZ7

拡大

Leif Azzopardi @leifos

Ranking choices using the iPRP involves knowing the cost of evaluating choices (ei), the probability of taking a choice (pi) and the average benefit of the choice (ai). #bemms @sigir2019 🤯 time for coffee ☕️ pic.twitter.com/zpLzKpAgN6

拡大

拡大

拡大

Maarten de Rijke @mdr

@HarrieOos @sigir2019 @claudiolucchese @fmnardini @bemikelive @mr_courage @RamK1729 @RolfJagerman Coffee break! pic.twitter.com/GTMPmZRHL2

拡大

Gabriella Kazai @gkazai

Rashmi Sankepally on Event Information Retrieval: event detection, diversification over event types, event impact prediction. #SIGIR2019 DC pic.twitter.com/ijouHW0PHU

拡大

拡大

拡大

拡大

Leif Azzopardi @leifos

@guidozuc calculates the expected benefit of choices 🎱given the IPRP to a packed and HOT 🥵 room #bemms @sigir2019 pic.twitter.com/PrplaUZwDQ

拡大