Jie Yang
I am currently an applied scientist at Amazon Research. Before, I obtained my PhD from Delft University of Technology and spent ~one year at eXascale Infolab.
I work on human-centric AI, a class of intelligent systems where humans are an integral component of the computational process. This is not only because that humans are often the data providers, but also because that they can play an active role in enhancing the system with respect to issues such as accuracy, robustness, transparency, and fairness. My research is both theoretical and experimental, as it requires to understand human characteristics (e.g., reliability and bias) while designing more powerful human-in-the-loop models and systems.
My recent work focuses on leveraging crowdsourcing and deep probabilistic modeling for debugging noisy training data (Scalpel-CD) and active learning (DALC). Currently, I am heavily involved in building human-AI systems for search and recommendation in Amazon's Choice and in Alexa.
The best way to reach me is by email, jiy at amazon dot com.
What's new?
- [Upcoming] - Will attend HCOMP 2019 and be a mentor for the doctoral consortium, see you in Portland!
- 08/2019 - Paper on training data augmentation accepted at EMNLP 2019.
- 01/2019 - Three papers on active learning, crowdsourcing, and recommendation accepted at WWW 2019.
Selected publications
- Sepideh Mesbah, Jie Yang, Robert-Jan Sips, Manuel Valle Torre, Christoph Lofi, Alessandro Bozzon, Geert-Jan Houben. Training Data Augmentation for Detecting Adverse Drug Reactions in User-Generated Content. Empirical Methods in Natural Language Processing (EMNLP), 2019
- Natalia Ostapuk*, Jie Yang, Philippe Cudré-Mauroux. ActiveLink: Deep Active Learning for Link Prediction in Knowledge Graphs. The Web Conference (WWW), 2019
- Jie Yang, Alisa Smirnova, Dingqi Yang, Gianluca Demartini, Yuan Lu, Philippe Cudré-Mauroux. Scalpel-CD: Leveraging Crowdsourcing and Deep Probabilistic Modeling for Debugging Noisy Training Data. The Web Conference (WWW), 2019
- Dingqi Yang, Bingqing Qu, Jie Yang, Philippe Cudré-Mauroux. Revisiting User Mobility and Social Relationships in LBSNs: A Hypergraph Embedding Approach. The Web Conference (WWW), 2019
- Jie Yang, Thomas Drake, Andreas Damianou, Yoelle Maarek. Leveraging Crowdsourcing Data for Deep Active Learning - An Application: Learning Intents in Alexa. The Web Conference (WWW), 2018
- Jie Yang, Carlo van der Valk, Tobias Hoßfeld, Judith Redi, Alessandro Bozzon. How do Crowdworker Communities and Microtask Markets Influence Each Other? A Data-Driven Study on Amazon Mechanical Turk. AAAI Conference on Human Computation and Crowdsourcing (HCOMP), 2018
- Zhu Sun*, Jie Yang*, Jie Zhang, Alessandro Bozzon, LongKai Huang, Chi Xu. Recurrent Knowledge Graph Embedding for Effective Recommendation. ACM Conference on Recommender Systems (RecSys), 2018
- Guanliang Chen, Jie Yang, Claudia Hauff, Geert-Jan Houben. LearningQ: A Large-scale Dataset for Educational Question Generation. AAAI Conference on Web and Social Media (ICWSM), 2018
- Gregory Afentoulidi, Zoltán Szlávik, Jie Yang, Alessandro Bozzon. Social Gamification in Enterprise Crowdsourcing. ACM Conference on Web Science (WebSci), 2018
- Wenjie Pei*, Jie Yang*, Zhu Sun, Jie Zhang, Alessandro Bozzon, David Tax. Interacting Attention-Gated Recurrent Networks for Recommendation. ACM International Conference on Information and Knowledge Management (CIKM), 2017
- Zhu Sun*, Jie Yang*, Jie Zhang, Alessandro Bozzon, Yu Chen, Chi Xu. MRLR: Multi-level Representation Learning for Personalized Ranking in Recommendation. International Joint Conference on Artificial Intelligence (IJCAI), 2017
- Ujwal Gadiraju, Jie Yang, Alessandro Bozzon. Clarity is a Worthwhile Quality - On the Role of Task Clarity in Microtask Crowdsourcing. ACM Conference on Hypertext and Social Media (Hypertext), 2017
Douglas Engelbart Best Paper Award - Zhu Sun, Jie Yang, Jie Zhang and Alessandro Bozzon. Exploiting both Vertical and Horizontal Dimensions of Feature Hierarchy for Effective Recommendation. AAAI Conference on Artificial Intelligence (AAAI), 2017
- Jie Yang, Judith Redi, Gianluca Demartini, Alessandro Bozzon. Modeling Task Complexity in Crowdsourcing. AAAI Conference on Human Computation and Crowdsourcing (HCOMP), 2016
- Jie Yang, Zhu Sun, Alessandro Bozzon, Jie Zhang. Learning Hierarchical Feature Influence for Recommendation by Recursive Regularization. ACM Conference on Recommender Systems (RecSys), 2016
- Jie Yang, Alessandro Bozzon, Geert-Jan Houben. Harnessing Engagement for Knowledge Creation Acceleration in Collaborative Q&A Systems. User Modeling, Adaption and Personalization (UMAP), 2015
- Jie Yang, Claudia Hauff, Alessandro Bozzon, Geert-Jan Houben. Asking the Right Question in Collaborative Q&A Systems. ACM Conference on Hypertext and Social Media (Hypertext), 2014
- Jie Yang, Ke Tao, Alessandro Bozzon, and Geert-Jan Houben. Sparrows and Owls: Characterisation of Expert Behaviour in StackOverflow. User Modeling, Adaption and Personalization (UMAP) , 2014
* indicates a student that I supervised or an equal contributor.
Services
I actively serve the research community. My recent acitivties include services on the program committee of AAAI (2020), AAAI HCOMP (2018-2019), AAAI ICWSM (2018-2020), ACM IUI (2018-2020), ACM CHIIR (2019-2020), ESWC (2019), and WWW (2018). In addition, I serve on the editorial board of the Journal of Human Computation and review for transactions and journals such as TKDE and IJCHS.