Connecting GenAI to Trusted Scholarly Resources through MCP: From Literature Discovery to Research Analytics

GenAI is increasingly used in academic research, but the value of AI-generated responses depends on the quality, transparency, and trustworthiness of the sources behind them. This workshop introduces how Model Context Protocol (MCP) can connect GenAI tools with trusted scholarly resources and support more effective research workflows.

Through practical examples using library-supported resources such as Scite and Consensus, participants will explore how connected AI can assist with literature discovery, evidence checking, citation-context analysis, and simple research analytics.

Outline:

• The big picture: GenAI in research
• What MCP is and why it matters
• Live demos:
  • Literature discovery with Claude/ChatGPT and Scite/Consensus MCP
  • Prototyping a simple research analytics dashboard with Cursor and Scite/Consensus MCP
• Good practices for responsible use

Note: Some third-party GenAI services may require a VPN and/or an individual account. You can still follow the concepts, demonstrations, and workflows without using these services directly during the session. No technical setup or programming experience is required.

All students and staff are welcome.

Note: Zoom URL and passcode will be sent to registrants at 5pm, 1 day before the event.