With increasing popularity of Deep Learning (DL) software development, code quality issues arise in DL software development. Code smell is one of the factors which reduce the quality of source code. Several previous studies investigated the prevalence of code smell in DL software systems to evaluate the quality of DL source code. However, there is still a lack of understanding of the awareness of individual DL developers in code smell. To more deeply understand the code smell risk in DL software development, it is needed to investigate the code smell awareness of DL developers. In this paper, we present an empirical study on code smell practices of DL developers. Specifically, we performed a quantitative analysis on code smell introduction and removal practices of DL developers. We collected a dataset of code smell introduction and removal history of DL developers from several open source DL software GitHub repositories. We then quantitatively analyzed the collected dataset. As a result of the quantitative analysis, we observed the following three findings on code smell introduction and removal practices of DL developers. First, DL developers tend to perform code smell introduction practice more than code smell removal practice. Second, DL developers have slightly broader code smell introduction scope than code smell removal scope. Third, regular and irregular DL developers have less difference in both code smell introduction and removal practices. The results indicate that DL developers have very poor awareness on code smell risk. Our findings suggest that DL software development project managers should provide a helpful guideline that makes DL developers actively participate code smell removal tasks.

JavaScript code blocks on Stack Overflow (SO) are often used in software projects. However, little is known about the issue of rule violation risk in SO JavaScript code blocks. Rule violation is one of the factors which degrades the quality of Java Script code. To prevent prevalence of rule violation by reusing SO JavaScript code blocks, it is needed to investigate how secure SO JavaScript code blocks are against rule violation. To examine the issue, we performed a quantitative analysis to investigate how many rule violations are, when first rule violation occurs and what is the trend of rule violations in evolution history of Stack Overflow JavaScript code blocks. We collected SO posts related to JavaScript and extracted the code blocks contained in the posts. By using ESLint, the most popular rule violation detection tool, we identified rule violations in the evolution history of our target code blocks. We then performed quantitative analyses on the identified rule violations. As the results of the analyses, we found that: (1) 60% of the studied code blocks evolve with any rule violations. (2) In the rule violated code blocks, 92% of the code blocks get first rule violation occurrence in the early phase of their evolution. (3) 80% of the rule violated code blocks never fix existing rule violations during their evolution. Our findings suggest that SO should provide a policy which can reduce rule violations in submitted JavaScript code blocks. The findings can also make SO users attend to rule violations when reusing SO JavaScript code blocks.

GitHub and Stack Overflow are often used together for software development. GH-SO users, who use both GitHub and Stack Overflow, contribute to the development of various software projects in GitHub and share their knowledge and experience on software development in Stack Overflow. To widely understand the interests and working habits of GH-SO users on software development, it is important to investigate how GH-SO users utilize GitHub and Stack Overflow. In this paper, we present an exploratory study on GitHub commit and Stack Overflow post activities of GH-SO users. Specifically, we investigate the working habits of GH-SO users on GitHub commit and Stack Overflow post activities. We randomly selected 19,756 of GH-SO users as our target sample and collected 2,819,483 and 2,147,317 of commit activity data and post activity data of the GH-SO users. We then categorized the collected commit and post activity datasets into specific categories on programming languages and statistically analyzed the categorized commit and post activity datasets. As the results of our analysis, we found the following: (1) The overall commit and post activities of the GH-SO users share some similarity. (2) The commit activities gradually change while the post activities drastically change over time. (3) The commit activities of the GH-SO users are broadly distributed while the post activities are narrowly distributed and the commit activity can be better predictor for post activity. (4) The commit activity of the GH-SO users tends to be performed prior post activity. We believe that our findings can contribute to finding the ways to better support commit and post activities of GitHub and Stack Overflow users.

Contentious news issues, such as the health care reform debate, draw much interest from the public; however, it is not simple for an ordinary user to search and contrast the opposing arguments and have a comprehensive understanding of the issues. Providing a classified view of the opposing views of the issues can help readers easily understand the issue from multiple perspectives. We present a disputant relation-based method for classifying news articles on contentious issues. We observe that the disputants of a contention are an important feature for understanding the discourse. It performs unsupervised classification on news articles based on disputant relations, and helps readers intuitively view the articles through the opponent-based frame and attain balanced understanding, free from a specific biased viewpoint. The method is performed in three stages: disputant extraction, disputant partitioning, and article classification. We apply a modified version of HITS algorithm and an SVM classifier trained with pseudorelevant data for article analysis. We conduct an accuracy analysis and an upper-bound analysis for the evaluation of the method.