Towards Demystifying the Impact of Dependency Structures on Bug Locations in Deep Learning Libraries

Abstract

Background: Many safety-critical industrial applications have turned to deep learning systems as a fundamental component. Most of these systems rely on deep learning libraries, and bugs of such libraries can have irreparable consequences. Aims: Over the years, dependency structure has shown to be a practical indicator of software quality, widely used in numerous bug prediction techniques. The problem is that when analyzing bugs in deep learning libraries, researchers are unclear whether dependency structures still have a high correlation and which forms of dependency structures perform the best. Method: In this paper, we present a systematic investigation of the above question and implement a dependency structure-centric bug analysis tool: Depend4BL, capturing the interaction between dependency structures and bug locations in deep learning libraries. Results: We employ Depend4BL to analyze the top 5 open-source deep learning libraries on Github in terms of stars and forks, with 279,788 revision commits and 8,715 bug fixes. The results demonstrate the significant differences among syntactic, history, and semantic structures, and their vastly different impacts on bug locations. Their combinations have the potential to further improve bug prediction for deep learning libraries. Conclusions: In summary, our work provides a new perspective regarding to the correlation between dependency structures and bug locations in deep learning libraries. We release a large set of benchmarks and a prototype toolkit to automatically detect various forms of dependency structures for deep learning libraries. Our study also unveils useful findings based on quantitative and qualitative analysis that benefit bug prediction techniques for deep learning libraries.