Postdoctoral Research Fellow at
NExT++ Research Center and National Cybersecurity R&D Lab
School of Computing
National University of Singapore

Email: yunshan dot ma at u dot nus dot edu dot com or mysbupt at gmail.com
Google Scholar PageGitHub Page

Biography

I am a Postdoctoral Research Fellow in National University of Singapore, where I am a member of NExT++ Research Center, supervised by Prof. Chua Tat-Seng. And I am also under National Cybersecurity R&D Lab and work with Prof. Chang Ee-Chien and Prof. Liang Zhenkai. My research focuses on the emerging and practical problem of multimodal event forecasting, which aims to understand multimodal information and make forecasting in various domains, such as international politics, macro economy, finance, cyber threats, fashion trend, etc. I am also working on computational fashion, including personalized mix-and-match, fashion outfit generation, fashion report generation, etc. In addition, I study a specialized branch of recommender system, i.e., bundle recommendation, including multimodal bundle construction and recommendation. In summary, I am interested in investigating computational (e.g., data mining, information retrieval, and multimedia) approaches to solve realworld problems. I am enthusiatic in discovering, formulating, and solving new and crucial problems, which have been not or rarely touched by the current academic and industrial community.

Prospective Ph.D., Master, and Undergraduate Students

I am looking for highly motivated students (PhD, master, undergraduate students) to work together on multimodal event forecating, computational fashion, and recommender system. Please feel free to send me your CV and transcripts, if you have interest. We are also actively looking for opportunities in research, partnership and commercialization in exciting data science projects.

News

  • [New!] 2024/05 Three papers are accepted by ACL'24 main track, KDD'24 research track, and KSEM'24!
         ACL'24: Analyzing Temporal Complex Events with Large Language Models? A Benchmark towards Temporal, Long Context Understanding.  [pdf coming soon]
         KDD'24: LARP: Language Audio Relational Pre-training for Cold-Start Playlist Continuation.  [pdf coming soon]
         KSEM'24: EE-LCE: An Event Extraction Framework Based on LLM-Generated CoT Explanation.  [pdf coming soon]

  • [New!] 2024/04 One paper is accepted by ICMR'24!
         Smart Fitting Room: A One-stop Framework for Matching-aware Virtual Try-On.  [pdf]

  • [New!] 2024/03 One demo paper is accepted by The Web Conf'24 and one paper is accepted by SIGIR'24!
         WWW'24 Demo: FashionReGen: LLM-Empowered Fashion Report Generation.  [pdf]
         SIGIR'24: DiFashion: Towards Personalized Outfit Generation.  [pdf]

  • [New!] 2024/02 One paper is accepted by TOIS!
         Filter-based Stance Network for Rumor Verification.  [pdf]

  • [New!] 2024/01 One paper is accepted by The Web Conf'24 and one paper is accepted by TOIS!
         WWW'24: Learning to Generate Explainable Stock Predictions using Self-Reflective Large Language Models.  [pdf]
         TOIS: MultiCBR: Multi-view Contrastive Learning for Bundle Recommendation.  [pdf]

  • 2023/11 One paper is accepted by TORS!
         Enhancing Item-level Bundle Representation for Bundle Recommendation.  [pdf]

  • 2023/10 One paper is accepted by WSDM'24 and one paper gets accepted by TKDE!
         WSDM'24: Leveraging Multimodal Features and Item-level User Feedback for Bundle Construction.  [pdf]
         TKDE: Rule-guided Counterfactual Explainable Recommendation.  [pdf]

  • 2023/08 One paper gets accepted by CIKM'23!
         Diffusion Variational Autoencoder for Tackling Stochasticity in Multi-Step Regression Stock Price Prediction.  [pdf]

  • 2023/06 One paper gets accepted by IEEE RA-L and one paper is accepted by IPM!
         RA-L: A Learning-Based Approach for Estimating Inertial Properties of Unknown Objects from Encoder Discrepancies.  [pdf]
         IPM: Personalized Fashion Outfit Generation with User Coordination Preference Learning.  [pdf]

  • 2023/05 Two papers are accepted by KDD'23!
         Context-aware Event Forecasting via Graph Disentanglement.  [pdf]
         FLOOD: A Flexible Invariant Learning Framework for Out-of-Distribution Generalization on Graphs.  [pdf]

  • 2023/04 One paper gets accepted by SIGIR'23!
         Strategy-aware Bundle Recommender System.  [pdf]

  • 2023/03 One paper gets accepted by TOIS!
         Causal Disentangled Recommendation Against User Preference Shifts.  [pdf]

  • 2022/05 One paper gets accepted by KDD'22!
         CrossCBR: Cross-view Contrastive Learning for Bundle Recommendation.  [pdf]
  • Professional Services

  • Organizing Committee
         - I will be a volunteers co-chair in The Web Conf 2024.
         - I chair the session of Recommendations and Ads of SIRIP in SIGIR 2023.
         - I serve as an area co-chair of Information Retrieval, Text Classification, and Question Answering in CCL 2023.
  • PC Memebers
         - KDD, WWW, SIGIR, IJCAI, ACMMM, WSDM, AAAI, etc.
  • Invited Reviewers
         - TKDE, TOIS, TMM, ToMM, TWeb, TIST, etc.
  • Honors and Awards

  • Best Student Paper Award, ICMR 2021.
  • Research Achievement Award, National University of Singapore 2019.
  • Best Paper Finallist, ACMMM 2018.
  • Background

  • 2022.04-present: Postdoctoral Research Fellow, NExT++ Research Center, National University of Singapore
         Supervisor: Prof Tat-Seng Chua.
  • 2017.08-2022.03: PhD in Computer Science, NExT++, National University of Singapore
         Supervisor: Prof Tat-Seng Chua.
  • 2011-2015: Bachelor in Communication Engineering, Beijing University of Posts and Telecommunications (BUPT).