SKA-Bench: A Fine-Grained Benchmark for Evaluating Structured Knowledge Understanding of LLMs

TL;DR

SKA-Bench is a comprehensive benchmark designed to evaluate the fine-grained structured knowledge understanding capabilities of large language models across various knowledge forms and challenge factors.

Contribution

The paper introduces SKA-Bench, a novel, rigorous benchmark that assesses multiple structured knowledge forms and fundamental abilities of LLMs, addressing limitations of previous evaluations.

Findings

Existing LLMs struggle with structured knowledge understanding.

Performance is affected by noise, order, and hallucinations.

DeepSeek-R1 shows notable but still limited capabilities.

Abstract

Although large language models (LLMs) have made significant progress in understanding Structured Knowledge (SK) like KG and Table, existing evaluations for SK understanding are non-rigorous (i.e., lacking evaluations of specific capabilities) and focus on a single type of SK. Therefore, we aim to propose a more comprehensive and rigorous structured knowledge understanding benchmark to diagnose the shortcomings of LLMs. In this paper, we introduce SKA-Bench, a Structured Knowledge Augmented QA Benchmark that encompasses four widely used structured knowledge forms: KG, Table, KG+Text, and Table+Text. We utilize a three-stage pipeline to construct SKA-Bench instances, which includes a question, an answer, positive knowledge units, and noisy knowledge units. To evaluate the SK understanding capabilities of LLMs in a fine-grained manner, we expand the instances into four fundamental ability testbeds: Noise Robustness, Order Insensitivity, Information Integration, and Negative Rejection. Empirical evaluations on 8 representative LLMs, including the advanced DeepSeek-R1, indicate that existing LLMs still face significant challenges in understanding structured knowledge, and their performance is influenced by factors such as the amount of noise, the order of knowledge units, and hallucination phenomenon. Our dataset and code are available at https://github.com/zjukg/SKA-Bench.