Wei Wu

Wei Wu

Auto-generated profile, based on publicly available information

Wei Wu

School of Artificial Intelligence and Data Science, University of Science and Technology of China, Hefei, China

Research Focus & Metrics

Academic Career Length

3Years (2024-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
10 July 2024
Verified
Conference Paper
Open access
WWWei Wu
Wei Wu
CWChao Wang
Chao Wang
DSDazhong Shen
Dazhong Shen
CQChuan Qin
Chuan Qin
LCLiyi Chen
Liyi Chen
HXHui Xiong
Hui Xiong
Collaborative filtering methods based on graph neural networks (GNNs) have witnessed significant success in recommender systems (RS), capitalizing on their ability to capture collaborative signals within intricate user-item relationships via message-passing mechanisms. However, these GNN-based RS inadvertently introduce excess linear correlation between user and item embeddings, contradicting the goal of providing personalized recommendations. While existing research predominantly ascribes this flaw to the over-smoothing problem, this paper underscores the critical, often overlooked role of the over-correlation issue in diminishing the effectiveness of GNN representations and subsequent recommendation performance. Up to now, the over-correlation issue remains unexplored in RS. Meanwhile, how to mitigate the impact of over-correlation while preserving collaborative filtering signals is a significant challenge. To this end, this paper aims to address the aforementioned gap by undertaking a comprehensive study of the over-correlation issue in graph collaborative filtering models. Firstly, we present empirical evidence to demonstrate the widespread prevalence of over-correlation in these models. Subsequently, we dive into a theoretical analysis which establishes a pivotal connection between the over-correlation and over-smoothing issues. Leveraging these insights, we introduce the Adaptive Feature De-correlation Graph Collaborative Filtering (AFDGCF) framework, which dynamically applies correlation penalties to the feature dimensions of the representation matrix, effectively alleviating both over-correlation and over-smoothing issues. The efficacy of the proposed framework is corroborated through extensive experiments conducted with four representative graph collaborative filtering models across four publicly available datasets. Our results show the superiority of AFDGCF in enhancing the performance landscape of graph collaborative filtering models.

5 June 2025 in bioRxiv
Verified
Preprint
Open access
PDF file
QLQiuyi Li
Qiuyi Li
WWWei Wu
Wei Wu
YZYiheng Zhu
Yiheng Zhu
FFFuli Feng
Fuli Feng
JYJieping Ye
Jieping Ye
ZWZheng Wang
Zheng Wang
The rapid growth of genomic and metagenomic data has underscored the pressing need for advanced computational tools capable of deciphering complex biological sequences. In this study, we introduce Gener anno , a compact yet powerful genomic foundation model (GFM) specifically optimized for metagenomic annotation. Trained on an extensive dataset comprising 715 billion base pairs (bp) of prokaryotic DNA, Gener anno employs a transformer encoder architecture with 500 million parameters, enabling bidirectional attention over sequences up to 8192 bp at single-nucleotide resolution. This design addresses key limitations of existing methods, including the inability of traditional Hidden Markov Models (HMMs) to handle fragmented DNA sequences from multi-species microbial communities, as well as the suboptimal tokenization schemes of existing GFMs that compromise fine-grained analysis. At its core, Gener anno excels in identifying coding regions from fragmented and mixed DNA sequences—a hallmark of metagenomic analysis. It achieves superior accuracy compared to traditional HMM-based methods (e.g., GLIMMER3, GeneMarkS2, Prodigal) and recent LLM-based approaches (e.g., GeneLM), while demonstrating robust generalization ability on archaeal genomes. Leveraging its advanced contextual understanding capability, Gener anno further enables two essential functions: pseudogene prediction and taxonomic classification—both performed based solely on raw sequence data, without reliance on reference databases or comparative genomics. These functionalities collectively streamline the metagenomic analysis pipeline, significantly reducing preprocessing requirements and enabling end-to-end interpretation of sequencing data. Beyond its primary role in metagenomic annotation, Gener anno also serves as a powerful GFM. To evaluate its broader utility, we curated the Prokaryotic Gener Tasks—a comprehensive benchmark suite specifically tailored for prokaryotic genomic analysis. It includes gene fitness prediction, antibiotic resistance identification, gene classification, and taxonomic classification, reflecting diverse aspects of functional genomics. On this benchmark, Gener anno consistently outperforms existing GFMs such as DNABERT-2, NT-v2, and GenomeOcean, demonstrating strong generalization capabilities across a wide range of genomic tasks. Overall, Gener anno provides a unified framework that integrates multiple critical functions for metagenomic annotation and beyond. By eliminating dependencies on external resources and offering rich contextual understanding of genomic sequences, this work delivers a foundational tool for advancing functional genomics in complex microbial communities. Implementation details and supplementary resources are available at https://github.com/GenerTeam/GENERanno .


8 August 2026
Verified
Conference Paper
Open access
WWWei Wu
Wei Wu
PZPeilun Zhou
Peilun Zhou
LCLiyi Chen
Liyi Chen
QWQimeng Wang
Qimeng Wang
CLChengqiang Lu
Chengqiang Lu
HXHui Xiong
Hui Xiong

20 April 2026 in ACM Transactions on Knowledge Discovery From Data
Verified
Journal Article
Open access
YBYunchu Bai
Yunchu Bai
CWChao Wang
Chao Wang
YSYing Sun
Ying Sun
CQChuan Qin
Chuan Qin
WWWei Wu
Wei Wu
HXHui Xiong
Hui Xiong
Corporate profiling serves as a critical analytical tool for modern enterprises, enabling data-driven decision-making in investment strategies, risk assessment, and strategic planning. It requires integrating quantitative metrics, qualitative insights, and network relationships to capture a company’s role in the business ecosystem. However, traditional methods struggle to synthesize heterogeneous data and model complex interdependencies among corporations, news, and market dynamics, often addressing these aspects in isolation. To address these challenges, this article introduces Financial Graph-based Mixture of Experts Prompt Learning (FGMPL), an innovative framework that unifies graph prompt learning with a multi-task paradigm for corporate profile modeling. The proposed framework reformulates node- and edge-level tasks into a coherent graph-level representation and employs multi-view contrastive learning to effectively integrate textual details with relational structures. Moreover, a novel Financial Multi-Experts Prompting mechanism—with learnable tokens coupled with a Mixture of Experts (MoE) design—is presented to enhance the processing of heterogeneous graph data and bridge the gap between pre-training and downstream tasks. To further improve adaptability, a meta-learning-based prompt tuning strategy is incorporated, enabling rapid transition to various downstream applications. Extensive experiments on real-world financial graphs show that FGMPL consistently outperforms strong pre-training and graph-prompting baselines across corporate performance prediction, relationship prediction, and news classification in both full-data and few-shot settings. In addition, cross-market transfer on a NASDAQ dataset and interpretability/efficiency analyses further demonstrate its robustness and practical applicability.