ear paper

publications.bib

@@ -1,10 +1 @@
-@ARTICLE{8743365,
-  author={An, Ya-Hui and Pan, Liangming and Kan, Min-Yen and Dong, Qiang and Fu, Yan},
-  journal={IEEE Access}, 
-  title={Resource Mention Extraction for MOOC Discussion Forums}, 
-  year={2019},
-  volume={7},
-  number={},
-  pages={87887-87900},
-  keywords={Context;Task analysis;Tagging;Message systems;Discussion forums;Context modeling;Semantics;Artificial intelligence;deep learning;hyperlinking;learning resources;MOOC discussion forums;name entity recognition},
-  doi={10.1109/ACCESS.2019.2924250}}
+@inproceedings{10.1145/3336191.3371769, author = {Lei, Wenqiang and He, Xiangnan and Miao, Yisong and Wu, Qingyun and Hong, Richang and Kan, Min-Yen and Chua, Tat-Seng}, title = {Estimation-Action-Reflection: Towards Deep Interaction Between Conversational and Recommender Systems}, year = {2020}, isbn = {9781450368223}, publisher = {Association for Computing Machinery}, address = {New York, NY, USA}, url = {https://doi.org/10.1145/3336191.3371769}, doi = {10.1145/3336191.3371769}, abstract = {Recommender systems are embracing conversational technologies to obtain user preferences dynamically, and to overcome inherent limitations of their static models. A successful Conversational Recommender System (CRS) requires proper handling of interactions between conversation and recommendation. We argue that three fundamental problems need to be solved: 1) what questions to ask regarding item attributes, 2) when to recommend items, and 3) how to adapt to the users' online feedback. To the best of our knowledge, there lacks a unified framework that addresses these problems. In this work, we fill this missing interaction framework gap by proposing a new CRS framework named Estimation"Action" Reflection, or EAR, which consists of three stages to better converse with users. (1) Estimation, which builds predictive models to estimate user preference on both items and item attributes; (2) Action, which learns a dialogue policy to determine whether to ask attributes or recommend items, based on Estimation stage and conversation history; and (3) Reflection, which updates the recommender model when a user rejects the recommendations made by the Action stage. We present two conversation scenarios on binary and enumerated questions, and conduct extensive experiments on two datasets from Yelp and LastFM, for each scenario, respectively. Our experiments demonstrate significant improvements over the state-of-the-art method CRM [32], corresponding to fewer conversation turns and a higher level of recommendation hits.}, booktitle = {Proceedings of the 13th International Conference on Web Search and Data Mining}, pages = {304–312}, numpages = {9}, keywords = {conversational recommendation, dialogue system, interactive recommendation, recommender system}, location = {Houston, TX, USA}, series = {WSDM '20} }