ArXiv Agent Implementation

[jravenel]

ArXiv Agent Implementation

Overview

Add an ArXiv agent to enable searching and analyzing research papers through the ArXiv API, with ontology mapping capabilities.

Components to Implement

1. Ontology

• Create src/ontologies/application-level/ArXivOntology.ttl
  • ArXivPaper (subclass of Document)
  • ArXivAuthor (subclass of User)
  • ArXivCategory (subclass of Category)
  • Object properties (hasAuthor, hasCategory)
  • Data properties (arxivId, abstract, publishedDate)

2. Integration

• Create src/integrations/ArXivIntegration.py
  • ArXivIntegrationConfiguration
  • ArXivIntegration class
  • Methods:
    • search_papers()
    • get_paper()
    • as_tools()

3. Pipeline

• Create src/pipelines/arxiv/ArXivPaperPipeline.py
  • ArXivPaperPipelineConfiguration
  • ArXivPaperPipelineParameters
  • Methods:
    • run()
    • as_tools()
    • as_api()

4. Assistant

• Create src/assistants/expert/domain/ArXivAssistant.py
  • ArXivAssistant class
  • create_arxiv_agent factory function
  • System prompt
  • Tool configuration

5. Integration with Terminal App

• Update src/apps/terminal_agent/main.py
  • Add run_arxiv_agent function
  • Import assistant factory

6. Build System

• Update Makefile
  • Add chat-arxiv-agent target
• Update pyproject.toml
  • Add chat-arxiv-agent script entry
  • Add arxiv package dependency

Dependencies

• arxiv package: make add dep=arxiv

Testing Checklist

• Search for papers by topic
• Get detailed paper information
• Add papers to knowledge graph
• Query paper relationships
• Analyze research trends

Example Usage

# Search papers
Find recent papers about large language models

# Get paper details
Get details about paper 2303.08774

# Add to knowledge graph
Add the paper with ID 2303.08774 to the knowledge graph

Related Components

• GitHub agent (reference implementation)
• Document ontology (parent classes)
• ABI core framework

Notes

• Uses arXiv API through arxiv package
• Implements ontology mapping for papers, authors, and categories
• Follows ABI framework patterns for integration and pipeline implementation