LearnIR WSDM

WSDM 2018 Workshop on Learning from User Interactions

[Aldo Lipani @AldoLipani]

We have a vacancy for a #postdoc position at #UCL. The postdoc will pursue research projects on the intersection of #ML, #NLP, and #IR.

Follow the link to learn more about the position or share it to whomever may be interested!

https://t.co/XZt9TrzfKg

#NeurIPS2018 #nips2018

[Rishabh Mehrotra @erishabh]

The slides for our #www2018 (& CIKM'17) tutorial on "Inferring User Needs & Tasks" are finally online!

Brief outline of tutorial:
- Characterizing Tasks
- Task extraction algorithms
- Task based evaluation
- Applications of tasks

@TheWebConf #cikm2018

https://t.co/hScPIl2E7F

[Rishabh Mehrotra @erishabh]

IRJ Special Issue on *Learning from User Interactions* --

the Call for Papers for the Information Retrieval Journal @SpringerIRJ Special Issue on *Learning from User Interactions* is out.

Initial submission due: January 15, 2019

Details: https://t.co/SHpHmh7C6n

@learnIRWSDM

[Rishabh Mehrotra @erishabh]

"Market Design & Computerized Marketplaces" - Prof. Alvin Roth, Economics Nobel Laureate

#kdd2018 keynote talk on:
- traditional marketplaces: school matching
- computer aided markets: Uber, AirBnb & Tinder
- future markets: self-driving cars

Here's a brief overview: 1/n

[Rishabh Mehrotra @erishabh]

"Explore, Exploit, and Explain: Personalizing Explainable Recommendations with Bandits" accepted at #recsys2018

This full paper describes how we incorporate explanations with bandit based recommendations. https://t.co/iMQIOicjqb

[Ronny Kohavi @ronnyk]

Good warnings about the Tyranny of Metrics by @jerryzmuller, some relevant for #ABTesting and the Overall Evaluation Criterion (#OEC) discussed in https://t.co/4mUi2uCrSM .

[Spotify Research @SpotifyResearch]

Good morning #TheWebConf, Spotify London is looking for research scientists to work on laying the foundations of Spotify technologies, redefining state of the art & contributing to the research community by publishing papers. Talk to @mounialalmas & @erishabh here @TheWebConf

[Rishabh Mehrotra @erishabh]

"Hope overrides self-interest, self-interest overrides reason".

While designing a system requiring large coordination complexity, we need to provide the right incentives to encourage all agents to act in a coordinated manner.

#thewebconf keynote

[Rishabh Mehrotra @erishabh]

Outline for the tutorial on Understanding & Inferring User Tasks and Needs @TheWebConf
https://t.co/RoBQc8QXtS
The tutorial begins 9 AM on Tuesday in Salon Gratte-Ciel.
#thewebconf #www2018

[LearnIR WSDM @learnIRWSDM]

Deep PropDCG: counterfactual learning to rank for optimizing DCG - Aman Agarwal

#wsdm2018 LearnIR workshop

[Rishabh Mehrotra @erishabh]

"Counterfactual Learning to Rank for Optimizing DCG" by Aman Agarwal

#wsdm2018 LearnIR workshop

[LearnIR WSDM @learnIRWSDM]

"Counterfactual Learning to Rank for optimizing DCG" - Aman Agarwal

#wsdm2018 LearnIR workshop

[Rishabh Mehrotra @erishabh]

-- Squashing gates in GRUs hurts memory in long range sequential neural models.
-- Factored GRU structure is helpful due to block diagonal structure
-- Increased RNN memory & less computation

@alexbeutel summarizing the upcoming #aistats2018 paper

#wsdm2018 LearnIR keynote

[LearnIR WSDM @learnIRWSDM]

Movies dynamics over time -- (say) how movie popularity changes when it wins an award
-- @alexbeutel keynote #wsdm2018 LearnIR workshop

[LearnIR WSDM @learnIRWSDM]

Neural Recommender System, gives substantial gains without fine-tuning temporal features -- @alexbeutel
#wsdm2018 LearnIR keynote

[LearnIR WSDM @learnIRWSDM]

Delighted to host @alexbeutel from @googleresearch for a keynote talk on Dynamics and Context in Neural Recommender Systems

#wsdm2018 LearnIR workshop

[LearnIR WSDM @learnIRWSDM]

Post coffee break, we have @alexbeutel from @googleresearch giving a keynote talk about Neural Recommender systems
#wsdm2018 LearnIR workshop https://t.co/H39IO7UlST

[LearnIR WSDM @learnIRWSDM]

We're excited to announce that @alexbeutel (Senior Research Scientist @googleresearch) will give a keynote talk on "User Dynamics and Context in Neural Recommender Systems" at the #wsdm2018 Learn-IR workshop on Friday 9th Feb
Workshop schedule: https://t.co/coTaKXB5iH

[LearnIR WSDM @learnIRWSDM]

Which destinations to consider & how long to stay" are important questions that need to be addressed -- @Lynyus
#wsdm2018 LearnIR workshop