A5007984753- Recommender Systems and Techniques
- Information Retrieval and Search Behavior
- Web Data Mining and Analysis
- Advanced Bandit Algorithms Research
- Face and Expression Recognition
- Data Mining Algorithms and Applications
- Image Retrieval and Classification Techniques
- Advanced Text Analysis Techniques
- Stochastic Gradient Optimization Techniques
- Advanced Graph Neural Networks
Naver (South Korea)
2023
Korea Advanced Institute of Science and Technology
2021
Feature selection, which is a technique to select key features in recommender systems, has received increasing research attention. Recently, Adaptive Selection (AdaFS) shown remarkable performance by adaptively selecting for each data instance, considering that the importance of given feature field can vary significantly across data. However, this method still limitations its selection process could be easily biased major frequently occur. To address these problems, we propose Multi-view...
Knowledge Distillation (KD), which transfers the knowledge of a well-trained large model (teacher) to small (student), has become an important area research for practical deployment recommender systems. Recently, Relaxed Ranking (RRD) shown that distilling ranking information in recommendation list significantly improves performance. However, method still limitations 1) it does not fully utilize prediction errors student model, makes training efficient, and 2) only distills user-side...
Large Language Models (LLMs) excel at understanding the semantic relationships between queries and documents, even with lengthy complex long-tail queries. These are challenging for feedback-based rankings due to sparse user engagement limited feedback, making LLMs' ranking ability highly valuable. However, large size slow inference of LLMs necessitate development smaller, more efficient models (sLLMs). Recently, integrating label generation into distillation techniques has become crucial,...