Resources for paper: Estimation-Action-Reflection: Towards Deep Interaction Between Conversational and Recommender Systems.
Our vision is to let recommender systems embrace the conversational technologies. Such system is able to dynamically obtain user's preference and make more accurate recommendations.
We have published one work in WSDM'2020.
Paper: Estimation-Action-Reflection: Towards Deep Interaction Between Conversational and Recommender Systems. DOI: https://dl.acm.org/doi/pdf/10.1145/3336191.3371769
Slides: https://ear-conv-rec.github.io/EAR-slides-wsdm.pdf
Poster: https://ear-conv-rec.github.io/EAR-poster-wsdm.pdf
Shortcuts:
Manual (READ ME!):
xxxxxxxxxxhttps://ear-conv-rec.github.io/manual.html
Code:
xxxxxxxxxxwget https://yisong.me/publications/EAR.zip
Dataset:
xxxxxxxxxxhttps://drive.google.com/file/d/1JHLwgASA6XlFTqLUTTO7qCPV5QJKG2dA/view?usp=drive_web
We are currently applying for a patent, and looking for a license to release it. Stay tuned for our update!
[Update 28Feb. 2020] Our patent is approved! Code will be released in one month. (Yisong is currently trapped in US, his air plane is canceled dut to the COVID-19 epidemic 😣)
[Update 8th April 2020] 🎉 Yisong has managed to document our code up! We now release our Codes, User Manual to our EAR System and datasets. This page contains everything:
https://ear-conv-rec.github.io/manual.html
We have 2 demos on Yelp and LastFM, representing Enumerated Questions and Binary Questions.
We have successfully presented the demo to one potential partner which raised great interest.
Please check this link to have a first impression to this interesting application!
Coming soon ...