ICSE 2018 论文

仅简单分析自己感兴趣的论文。

 

9.6 Million Links in Source Code Comments: Purpose, Evolution, and Decay

分析了 source code comments中的link问题。发现其存在decay, insufficient versioning, and lack of bidirectional traceability问题。属于Empirical study类型的论文。

 

A Framework for Checking Regression Test Selection Tools

对回归测试选择工具进行了测试。

 

A Neural Model for Generating Natural Language Summaries of Program Subroutines

分析了source code summarization问题。主要基于neural machine translation方法。

 

A Novel Neural Source Code Representation based on Abstract Syntax Tree

已有的方法将代码视为自然语言文本，因此会遗漏重要的语义信息。

已有研究一般基于AST，但AST存在size大的问题，并且存在long-term dependency的问题。论文提出了AST-based neural network。

 

Active Inductive Logic Programming for Code Search

代码搜索问题。论文提出交互式代码搜索方法ALICE

 

Analyzing and Supporting Adaptation of Online Code Examples

代码搜索。从online code角度入手。

 

Adversarial Sample Detection for Deep Neural Network through Model Mutation Testing

基于变异测试检测Adversarial Sample

 

Attack Surface Identification for Vulnerability Assessment through Program Metrics

安全漏洞预测

 

CRADLE: Cross-Backend Validation to Detect and Localize Bugs in Deep Learning Libraries

深度学习库中检测和定位bugs

 

Class Imbalance Evolution and Verification Latency in Just-in-Time Software Defect Prediction

Just-in-time缺陷预测问题

 

Exposing Library API Misuses via Mutation Analysis

API误用问题

 

FOCUS: A Recommender System for Mining API Function Calls and Usage Patterns

API函数使用问题

 

Graph-based Mining of In-the-Wild, Fine-grained, Semantic Code Change Patterns

针对代码变更的建模

 

Guiding Deep Learning System Testing using Surprise Adequacy

深度学习系统的测试

 

Hunting for Bugs in Code Coverage Tools via Randomized Differential Testing

对代码覆盖工具

 

Natural Software Revisited

NL2Type: Inferring JavaScript Function Types from Natural Language Information

Mining and Continuously Growing a Dataset of Reproducible Real-World Software Faults and Fixes

Learning to Spot and Refactor Inconsistent Method Names

 

On Learning Meaningful Code Changes via Neural Machine Translation

 

Pattern-based Mining of Opinions in Q&A Websites

 

ReCDroid: Automatically Reproducing Android Application Crashes from Bug Reports

 

Recovering Variable Names for Minified Code with Usage Contexts

 

 Search-Based Energy Testing of Android

APP能耗问题

 

When Code Completion Fails: a Case Study on Real-World Completions

代码补全