一句话：提出 Attend and Represent，用注意力与表示学习改进软件验证算法选择，并关注可解释性。
软件验证工具种类繁多，开发者面对具体验证任务时需进行算法选择。现有软件验证算法选择器通常使用人工挑选的程序特征，并配合手工设计的启发式或机器学习策略；前者难以迁移到其他选择问题，后者则缺乏可解释性，无法说明为何选择某个工具。本文标题为 Attend and represent，摘要未展开具体技术路线，但问题定位指向结合注意力机制与表示学习来构建选择模型，以期兼顾迁移与解释。摘要也未报告评估基准、对比方法和实验结果，因此关键结果需参见论文正文。
Today, a plethora of different software verification tools exist. When having a concrete verification task at hand, software developers thus face the problem of algorithm selection. Existing algorithm selectors for software verification typically use handpicked program features together with (1) either manually designed selection heuristics or (2) machine learned strategies. While the first approach suffers from not being transferable to other selection problems, the second approach lacks interpretability, i.e., insights into reasons for choosing particular tools.
DOI 原文 ·
@inproceedings{paperbot5915,
title = {Attend and represent},
author = {Cedric Richter and Heike Wehrheim},
booktitle = {ASE},
year = {2020},
doi = {10.1145/3324884.3416633}
}