A5101523508- Recommender Systems and Techniques
- Advanced Graph Neural Networks
- Topic Modeling
- Machine Learning in Healthcare
- Privacy-Preserving Technologies in Data
- Opinion Dynamics and Social Influence
Tencent (China)
2024-2025
Venus Medtech (China)
2024
Recent advances in Large Language Models (LLMs) have been changing the paradigm of Recommender Systems (RS). However, when items recommendation scenarios contain rich textual information, such as product descriptions online shopping or news headlines on social media, LLMs require longer texts to comprehensively depict historical user behavior sequence. This poses significant challenges LLM-based recommenders, over-length limitations, extensive time and space overheads, suboptimal model...
Recent years have witnessed the rapid development of online recruitment platforms, which provide a convenient way for matching job seekers and recruiters by leveraging recommendation systems. Indeed, this is <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">reciprocal recommendation</i> problem needs to consider preferences both simultaneously, making it different from traditional uni-directional user-item problems. Existing studies mainly...
Medication recommendation is one of the most critical health-related applications, which has attracted extensive research interest recently. Most existing works focus on a single hospital with abundant medical data. However, many small hospitals only have few records, hinders applying medication to real world. Thus, we seek explore more practical setting, i.e. , multi-center recommendation. In this but total number records large. Though may benefit from affluent it also faced challenge that...
In real clinics, the medical data are scattered over multiple hospitals. Due to security and privacy concerns, it is almost impossible gather all together train a unified model. Therefore, multi-node machine learning systems currently mainstream form of model training in healthcare systems. Nevertheless, distributed relies on exchange gradients, which has been proved under risk leakage. That means malicious attackers can restore user's sensitive by utilizing publicly shared serious problem...