一句话：Y-Map 用神经-符号回归离线预测 ORM schema 的插入延迟、查询延迟与内存，实现免基准测试的帕累托式选型。
选择高性能的对象关系映射（ORM）需在插入延迟、查询延迟和内存占用等多个冲突目标间权衡。实践中 ORM 框架通常采用固定启发式规则、只输出单一 schema；而基于先验学习的方法往往把选型简化为二元帕累托成员分类，难以进行细粒度权衡推理。本文提出 Y-Map，一个神经-符号框架：它离线通过规范驱动的 schema 合成与基准测试构建监督数据，在融合的 schema 表示上训练回归模型，该表示将可解释的结构描述符与 GraphCodeBERT、CodeT5+ 和 LLaMA 的嵌入相结合。推理时无需执行数据库负载，即可预测候选 schema 的连续插入延迟、查询延迟与内存占用，从而支持无基准的帕累托式选择——小空间可穷举打分，大空间可预算受限搜索。在九个基准对象模型上，Y-Map 相比代表性的基于基准测试的基线提升了权衡质量并降低了推理成本，展示了符号有效性保证与学习式性能预测相结合用于实用性能感知 schema 设计的前景。
Selecting high-performance object-relational mappings (ORMs) requires balancing competing objectives, such as insert and query latency and memory usage. In practice, ORM frameworks typically apply fixed heuristics and emit a single schema. At the same time, prior-learning-based approaches often reduce selection to binary Pareto membership classification, thereby limiting fine-grained tradeoff reasoning. We present Y-Map, a neural–symbolic framework that predicts continuous performance metrics for the ORM schema candidates without executing database workloads during inference. Y-Map constructs supervised training data offline via specification-driven schema synthesis and benchmarking, then learns a regression model over fused schema representations that combine interpretable structural descriptors with embeddings from GraphCodeBERT, CodeT5+, and LLaMA. By predicting continuous insert latency, query latency, and memory usage for candidate schemas, Y-Map enables benchmark-free Pareto-style selection, supporting exhaustive scoring in small spaces and budgeted search in larger ones. Across nine benchmark object models, Y-Map improves tradeoff quality and reduces inference cost relative to a representative benchmarking-based baseline, demonstrating the promise of combining symbolic validity guarantees with learned performance prediction for a practical performance-aware schema design.
@inproceedings{paperbot11579,
title = {Y-Map: A Multi-Encoder Y-Architecture for Predicting Performance-Aware Object-Relational Mappings},
author = {Sasan Azizian and Ayoub Hazrati and Artin Azizian and Elham Rastegari},
booktitle = {ICSE},
year = {2026},
doi = {10.1145/3774748.3795662}
}