WSDM 2026

119 papers
The 2026 edition of the WSDM conference featured research focused on large language models (LLMs), data valuation, and multilingual information retrieval. A significant portion of the work examined the capabilities and limitations of LLMs, particularly in specialized tasks such as algorithmic code generation and named entity recognition. Research also addressed the evaluation of training data through neural modeling and the analysis of user-generated content on news platforms and blogs. Specific research included an evaluation of LLMs' computational thinking skills compared to human programmers and the introduction of self-training methods for entity recognition using LLM-based judgment. The conference also hosted the WSDM CUP, which focused on multilingual retrieval challenges. Prolific contributors to this year's proceedings included authors such as Yanghua Xiao, Zhixu Li, and Chengyi Yang. Other research explored advancements in topic modeling for comment detection and efficient data valuation techniques like Fast-DataShapley.

Top Research Directions

Research directions most represented by WSDM 2026 papers.

Learnable Manifold Curvature in Geometric Deep Learning 1
Deep Learning on Irregular 3D Meshes 1
Tangent Space Linearization for Non-Euclidean Data 1
Graph Neural Networks in Mixed Curvature Spaces 1

Top Authors

AuthorPapers
Xin Lin2
Andrew Yates1
Chiranjib Bhattacharyya1
Chul-Ho Lee1
Eugene Yang1
Jingchao Wang1
Lan Zhang1
Luca Soldaini1
Minh Phu Vuong1
Mrinal Das1
Trapit Bansal1
Yang Shi1

Papers

Title
Sharpness-aware Federated Graph Learning
SoftHist: Teaching Graph Injection Attackers to Camouflage with Memory
STORM: A Spatio-Temporal Factor Model Based on Dual Vector Quantized Variational Autoencoders for Financial Trading
TableMind: An Autonomous Programmatic Agent for Tool-Augmented Table Reasoning
TaxoDiff: Improving Taxonomy Completion with Diffusion Guided Dynamic Negative Sampling
TempRetriever: Fusion-based Temporal Dense Passage Retrieval for Time-Sensitive Questions
The Future of Personalized Universal Assistant
Third Workshop on Generative AI for Recommender Systems and Personalization
Time-Aware Adaptive Side Information Fusion for Sequential Recommendation
TimeMAE: Self-Supervised Representations of Time Series with Decoupled Masked Autoencoders
Towards Multifaceted Graph Condensation in Discrete Realm
Towards Next-Generation Recommender Systems: A Benchmark for Personalized Recommendation Assistant with LLMs
TRUE: A Reproducible Framework for LLM-Driven Relevance Judgment in Information Retrieval
Unlocking Scaling Law in Industrial Recommendation Systems with a Three-step Paradigm based Large User Model
Unsupervised Dense Retrieval with Conterfactual Contrastive Learning
Using Brand Knowledge Bases and LLM Agents to Enhance E-commerce Retailers' Catalog Quality
Waterlogformer: A Multimodal Model for Waterlogging Prediction
Workshop on Benchmarking Causal Models (CausalBench)
WSDM CUP 2026: Multilingual Retrieval

« prev Page 3/3