Unmediated AI-Assisted Scholarly Citations

2025-12-22

Open Conference Proceedings

AI Summary

This work addresses the limitations of traditional literature databases, which rely on manual querying, and the risk of hallucinated citations inherent in large language models (LLMs) despite their natural language interaction capabilities. To reconcile these issues, the authors propose a novel architecture based on the Model Context Protocol (MCP), which integrates the conversational strengths of LLMs with direct access to authoritative bibliographic databases such as DBLP. The system enables fuzzy matching and interactive retrieval through natural language queries, yet bypasses the model during citation export by fetching reference data directly from trusted sources. This approach preserves user-friendly interactivity while fundamentally eliminating citation hallucination, thereby ensuring the authenticity and completeness of academic references. The resulting MCP-DBLP system effectively transforms conventional bibliographic services into a high-accuracy, conversational scholarly assistant, with clear potential for extension to other academic data repositories.

Technology Category

• Data Mining & Knowledge Management: Conversational Systems for Recommendation & Retrieval
• Natural Language Processing: Conversational AI/Dialog Systems
• Cognitive Modeling & Cognitive Systems: Conceptual Inference and Reasoning

Application Category

• Search and Retrieval-Augmented AI: Large language models for search
• User Modeling, Personalization and Recommendation: User modeling and simulation for interactive and conversational systems
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions

Abstract

Traditional bibliography databases require users to navigate search forms and manually copy citation data. Language models offer an alternative: a natural-language interface where researchers can write text with informal citation fragments and have them automatically resolved to proper references. However, language models generate fabricated (hallucinated) citations at substantial rates, making them unreliable for scholarly work. We present an architectural approach that combines the natural language interface of LLM chatbots with the accuracy of direct database access, implemented through the Model Context Protocol. Our system enables language models to search bibliographic databases, perform fuzzy matching, and export verified entries, all through conversational interaction. A key architectural principle bypasses the language model during final data export by fetching entries directly from authoritative sources, with timeout protection, to guarantee accuracy. We demonstrate this approach with MCP-DBLP, a server providing access to the DBLP computer science bibliography. The system transforms form-based bibliographic services into conversational assistants that maintain scholarly integrity. This architecture is adaptable to other bibliographic databases and scholarly data sources.

Problem

• hallucinated citations
• AI-assisted citations
• scholarly integrity
• bibliographic accuracy
• language model reliability

Innovation

• Model Context Protocol
• hallucination mitigation
• conversational bibliographic assistant
• direct database access
• scholarly citation integrity