This is the KG & DU Team, Official R&D entity under the 360 AI Research Institute. We’ve been focusing on "Knowledge Graph" and "Document Understanding", conducting research and developing below:
1. Knowlege Graph Construction,Represenation & NLP Application
1)🏆Competition：2022. Ranked 1st in the Competition (2022.08): Ogbl-wikikg2 leadboard
2)🏆Competition：2022. Ranked 2nd in the Competition: CCKS2022-multi-modal entity alignment leadboard
3)🔬paper: 2022. Multimodal Entity Alignment Model Based on Interactive Encoder and Knowledge Distillation,(https://sigkg.cn/ccks2022/?page_id=600)
4)🏆Competition：2022. Ranked 2nd in the Competition: CAIL2022 leadboard
5)🔬Paper: 2022. Triplere: Knowledge graph embeddings via tripled relation vectors:paper
2. Document Analysis, Layout Parser, Document Understanding & Multi-modal Large Language Models
1)🛞Opesource: 2024.Chiense oriented document layout models:360LayoutAnalysis Github, Huggingface
2)🔬Paper: 2024.FlowCE:First Multi-Dimensional Evaluation of Flowchart Comprehensionfor Multimodal Large Language Models paper, Github
3)🏆Competition: 2024.Ranked 1st in the Competition: ICPR 2024 Competition on Multi-line Mathematical Expressions Recognition leadboard
4)🔬Paper: 2024.Enhancing Complex Formula Recognition with Hierarchical Detail-Focused Network paper
5)🔬Paper: 2024.First Multi-Dimensional Evaluation of Flowchart Comprehension for Multimodal Large Language Modelspaper
6)🏆Competition：2025. Ranked 2nd Competition on End-to-End Document Image Machine Translation Towards Complex Layouts（VLM-OCR-Free）
7)🏆Competition：2025. Ranked 3rd Competition on Understanding Chinese College Entrance Exam Papers
3. Retrieval-augmented Generation & Application
1)🔬Paper: 2024.CRUD-RAG: A Comprehensive Chinese Benchmark for Retrieval-Augmented Generation of Large Language Models: paper