ROSE Scoring Cascade Workflow

Paper context

Paper title: ROSE: An Intent-Centered Evaluation Metric for NL2SQL Abstract: Execution Accuracy (EX), the widely used metric for evaluating the effectiveness of Natural Language to SQL (NL2SQL) solutions, is becoming increasingly unreliable. It is sensitive to syntactic variation, ignores that questions may admit multiple interpretations, and is easily misled by erroneous ground-truth SQL. To address this, we introduce ROSE, an intent-centered metric that focuses on whether the predicted SQL answers the question, rather than consistency with the ground-truth SQL under the reference-dependent paradigm. ROSE employs an adversarial Prover-Refuter cascade: SQL Prover assesses the semantic correctness of a predicted SQL against the user's intent independently, while Adversarial Refuter uses the ground-truth SQL as evidence to challenge and refine this judgment. On our expert-aligned validation set ROSE-VEC, ROSE achieves the best agreement with human experts, outperforming the next-best metric by nearly 24% in Cohen's Kappa. We also conduct a largescale re-evaluation of 19 NL2SQL methods, revealing four valuable insights. We release ROSE and ROSE-VEC to facilitate more reliable NL2SQL research. Passages referencing this figure: true, uphold for false) on S P S_{P} and a diagnostic label L q L_{q} for Q Q . The function is defined as follows: R ​ e ​ f ​ ( Q , S p , S g ; E p ∗ , E g ∗ , R ∗ | D , C ) → ( j r , L q ) Ref(Q,S_{p},S_{g};E_{p}^{*},E_{g}^{*},R^{*}\ |D,C)\rightarrow(j_{r},L_{q}) The conditional arguments ( E p ∗ , E g ∗ , R ∗ E_{p}^{*},E_{g}^{*},R^{*} ) are invoked only when E p ≢ E g E_{p}\not\equiv E_{g} . Figure 1: ROSE Scoring Cascade. The red path indicates the workflow when execution results do not match, while the blue path is followed when they do. 4.3 ROSE ROSE is determined by the workflow illustrated in Figure 1 . A predicted SQL must first be executable. If its execution result differs from the ground-truth SQL, it must pass SQL Prover’s independent evaluation. Finally, it must withstand t ;E_{p}^{*},E_{g}^{*},R^{*}\ |D,C)\rightarrow(j_{r},L_{q}) The conditional arguments ( E p ∗ , E g ∗ , R ∗ E_{p}^{*},E_{g}^{*},R^{*} ) are invoked only when E p ≢ E g E_{p}\not\equiv E_{g} . Figure 1: ROSE Scoring Cascade. The red path indicates the workflow when execution results do not match, while the blue path is followed when they do. 4.3 ROSE ROSE is determined by the workflow illustrated in Figure 1 . A predicted SQL must first be executable. If its execution result differs from the ground-truth SQL, it must pass SQL Prover’s independent evaluation. Finally, it must withstand the adversarial challenge from Adversarial Refuter, which uses the ground-truth as counter-evidence. Failure at any of these stages results in a score of 0. Only a query that successfully navigates this entire c