Reproduce “Retrieval-Augmented Generation with Knowledge Graphs for Customer Service Question Answering” on OpenResearch

Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.

Universal app for Apple silicon and Intel. Requires macOS 11 or later.

About this paper

Zhentao Xu, Mark Jerome Cruz, Matthew Guevara, Tie Wang, Manasi Deshpande, Xiaofeng Wang, Zheng Li

Researchers at LinkedIn developed a Retrieval-Augmented Generation (RAG) system integrated with Knowledge Graphs (KGs) to enhance question answering for customer service inquiries. This approach improved retrieval accuracy by 77.6% MRR and reduced median issue resolution time by 28.6% in a production environment.

Key insights

  • Explicitly modeling and leveraging the inherent structural and relational information in semi-structured enterprise data (like customer service tickets) through Knowledge Graphs is crucial for accurate retrieval.
  • Large Language Models can effectively bridge the gap between natural language queries and structured Knowledge Graph data by parsing user intents and dynamically generating graph traversal queries.
  • Combining the generative power of LLMs with the structured knowledge provided by KGs overcomes the limitations of purely text-based Retrieval-Augmented Generation, leading to more coherent and factually grounded answers.

Summary by alphaXiv.

Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.