Software developers must adhere to privacy regulations and apply privacy principles. However, developers may not be privacy experts and hence are likely to encounter issues when choosing, applying, and implementing the corresponding Privacy-Preserving Computation (PPC) techniques. As a result, these methods may be misunderstood or even not applied. In order to mitigate these issues, we must first identify and understand them in detail. To this end, we have conducted a study on the popular developer Q&A website Stack Overflow. Using manual coding, we have extracted themes and issues related to PPC techniques and discussed by developers. Additionally, we have analyzed and discussed use cases and how they align with current research. Our results confirm our assumption, showing (1) a lack of awareness and understanding, (2) complex or missing documentation, and (3) a lack of easy-to-use libraries. Based on our obtained results, we provide recommendations to educators and developers on how to address the identified issues.

错误:搜索内容不能为空,请输入英文关键词
错误:关键词超出字数限制,请精简
高级检索

“We’ve Met Some Problems”: Developers’ Issues with Privacy-Preserving Computation Techniques on Stack Overflow

  • Patrick Kühtreiber,
  • Sabrina Heimermann,
  • Sebastian Schillinger,
  • Delphine Reinhardt

摘要

Software developers must adhere to privacy regulations and apply privacy principles. However, developers may not be privacy experts and hence are likely to encounter issues when choosing, applying, and implementing the corresponding Privacy-Preserving Computation (PPC) techniques. As a result, these methods may be misunderstood or even not applied. In order to mitigate these issues, we must first identify and understand them in detail. To this end, we have conducted a study on the popular developer Q&A website Stack Overflow. Using manual coding, we have extracted themes and issues related to PPC techniques and discussed by developers. Additionally, we have analyzed and discussed use cases and how they align with current research. Our results confirm our assumption, showing (1) a lack of awareness and understanding, (2) complex or missing documentation, and (3) a lack of easy-to-use libraries. Based on our obtained results, we provide recommendations to educators and developers on how to address the identified issues.