Explore influential papers, citation networks, citation contexts, and knowledge graphs across multidisciplinary scientific domains.
https://citation-hub-website.vercel.app/
https://citationdatabase.streamlit.app
https://huggingface.co/spaces/Daniel0315/cithub_website
https://huggingface.co/datasets/Daniel0315/IDCite
CitationHub is a large-scale citation context database and interactive exploration platform designed to support:
- Citation Intent Classification
- Citation Recommendation
- Scholarly Retrieval
- Knowledge Graph Construction
- Contextual Citation Evaluation
- Research Trend Analysis
- Scientific Discovery Support
Unlike traditional citation databases that treat citations as simple links between papers, CitationHub preserves:
- citation context (actual citing sentence)
- citation intent labels (why the citation was made)
- multidisciplinary field information
- co-citation relationships
- citation event structures
- knowledge graph representation
This enables more fine-grained and explainable scholarly analysis.
The video above demonstrates the main workflow of the CitationHub system, including citation context exploration, citation network analysis, knowledge graph visualization, and interactive dashboard navigation. Users can efficiently explore citation relationships, intent-aware citation behaviors, and multidisciplinary scholarly connections through the platform. The platform provides several interactive modules for citation exploration and knowledge graph analysis.
🎥 See CitationHub in action: Watch the interactive demo
Users can explore:
- highly cited seed papers
- citation event statistics
- citation intent distributions
- field distributions
- related citing papers
- co-cited seed papers
Supported filters include:
- Title or DOI
- Field
- Country
- Journal
- Citation Year
This enables efficient exploration of citation behaviors across scientific disciplines.
Visual exploration of relationships among:
- seed papers
- citing papers
- co-cited papers
This helps identify citation flow and scholarly influence propagation.
CitationHub transforms citation events into structured scholarly knowledge graphs for:
- semantic querying
- graph-based reasoning
- citation event understanding
- explainable scholarly discovery
Global visualization of citation distributions by:
- country
- institution
- affiliation
This supports international collaboration analysis.
Advanced analytics for:
- citation intent comparison
- field-level analysis
- journal-level patterns
- temporal citation trends
CitationHub currently contains:
| Category | Count |
|---|---|
| Seed Papers | 23,479 |
| Citation Events | 1.83M+ |
| Citing Papers | 1.44M+ |
| Authors | 16,839 |
| Countries | 108 |
| Scientific Fields | 21 |
This makes CitationHub one of the largest multidisciplinary citation-context-aware resources.
CitationHub supports 7 major citation intent categories:
- Background
- Uses
- Similarities
- Motivation
- Differences
- Future Work
- Extends
These intent labels provide controllable and interpretable signals for:
- intent-aware citation retrieval
- reranking systems
- citation recommendation
- scholarly evaluation
Intent-conditioned citation recommendation and selective reranking.
Citation events → RDF triples → semantic scholarly graphs.
Moving beyond simple citation counts toward semantic impact measurement.
Field-specific citation behaviors and knowledge evolution analysis.
CitationHub/
├── app.py
├── requirements.txt
└── README.md
We are expanding CitationHub toward:
- full citation event ontology
- LLM-based citation reasoning
- agentic scholarly discovery systems
- explainable citation recommendation
- benchmark datasets for top-tier citation retrieval research