STaRK: Benchmarking LLM Retrieval on Textual and Relational Knowledge Bases
This work addresses the gap in evaluating retrieval systems for complex, real-world queries that require blending textual and relational information, providing a comprehensive testbed for researchers and practitioners in information retrieval and AI.
The authors tackled the problem of retrieving information from semi-structured knowledge bases that combine textual and relational data by creating STaRK, a large-scale benchmark covering product search, academic paper search, and precision medicine queries. They developed a pipeline to synthesize realistic queries with ground-truth answers and found that current retrieval and LLM systems struggle with this benchmark, indicating a need for improved semi-structured retrieval methods.
Answering real-world complex queries, such as complex product search, often requires accurate retrieval from semi-structured knowledge bases that involve blend of unstructured (e.g., textual descriptions of products) and structured (e.g., entity relations of products) information. However, many previous works studied textual and relational retrieval tasks as separate topics. To address the gap, we develop STARK, a large-scale Semi-structure retrieval benchmark on Textual and Relational Knowledge Bases. Our benchmark covers three domains: product search, academic paper search, and queries in precision medicine. We design a novel pipeline to synthesize realistic user queries that integrate diverse relational information and complex textual properties, together with their ground-truth answers (items). We conduct rigorous human evaluation to validate the quality of our synthesized queries. We further enhance the benchmark with high-quality human-generated queries to provide an authentic reference. STARK serves as a comprehensive testbed for evaluating the performance of retrieval systems driven by large language models (LLMs). Our experiments suggest that STARK presents significant challenges to the current retrieval and LLM systems, highlighting the need for more capable semi-structured retrieval systems. The benchmark data and code are available on https://github.com/snap-stanford/STaRK.