一句话:利用主题模型对 IDE 交互数据建模,以预测开发者未来行为。
论文关注 IDE 中开发者交互数据的建模与未来行为预测。作者指出,从日常点击和按键收集的交互数据已被用于经验研究与软件工程推荐系统,并观察到这些数据在各类 IDE 中具有若干共同特征。针对预测开发者未来行为,论文采用主题模型(topic models)对 IDE 交互数据进行建模。由于提供的英文摘要在此处截断,未包含具体方法细节、实验设置与量化结果,故无法确认其预测准确率、对比基线或跨 IDE 泛化能力。总体而言,该工作属于面向 IDE 的开发者行为预测与工具支持方向。
Interaction data, gathered from developers' daily clicks and key presses in the IDE, has found use in both empirical studies and in recommendation systems for software engineering. We observe that this data has several characteristics, common across IDEs:
DOI 原文 ·
@inproceedings{paperbot4453,
title = {Predicting future developer behavior in the IDE using topic models},
author = {Kostadin Damevski and Hui Chen and David C. Shepherd and Nicholas A. Kraft and Lori Pollock},
booktitle = {ICSE},
year = {2018},
doi = {10.1145/3180155.3182541}
}