Integrating Issue Tracking Systems with Community-Based Question and Answering Websites

The Stack Overflow web site is an online community where programmers can ask and answer one another's questions, earning points and badges. The site offers guidance in the form of a Frequently Asked Questions (FAQ), beginning with “What kind of questions can I ask here?” The answer explains that “the best Stack Overflow questions have a bit of source code in them”. This paper explores the role of source code and non-source code text on Stack Overflow in both questions and answers. The primary contribution of this paper is to provide a more detailed understanding of whether the presence of source code (and how much) actually will produce the “best” Stack Overflow questions or answers. A second contribution of this paper is to determine how the non-code portions of the text might also contribute the “best” Stack Overflow postings.

Community question answering (CQA) services accumulate large amount of knowledge through the voluntary services of the community across the globe. In fact, CQA services gained much popularity recently compared to other Internet services in obtaining and exchanging information. Stack Overflow is an example of such a service that targets programmers and software developers. In general, most questions in Stack Overflow are usually ended up with an answer accepted by the askers. However, it is found that the number of unanswered or ignored questions has increased significantly in the past few years. Understanding the factors that contribute to questions being answered as well as questions remain ignored can help information seekers to improve the quality of their questions and increase their chances of getting answers from the community in Stack Overflow. In this study, we attempt to identify by data mining techniques the relevant features that will help predict the quality of questions, and validate the reliability of the features using some of the state-of-the-art classification algorithms. The features to be obtained have to be significant in the sense that they can help Stack Overflow to improve their existing CQA service in terms of user satisfaction in obtaining quality answers from their questions.

Community-based question answering sites such as Yahoo! Answers, Quora, and Stack Overflow have emerged as an effective means of information seeking on the Web. Anyone can obtain answers to their questions by posting them for other participants on these sites. However, not all questions get immediate answers from other users. Questions which are not interesting enough for the community may suffer from “starvation”. Such questions may take days/months to get satisfactory answers. This delay in response can be avoided by searching similar questions in an archive of previously answered questions or by forwarding to users who might potentially answer. Another approach would be to combine multiple valid answers into one and then generate a well-summarized answer. The proposed framework reduces starvation of new questions in Stack Overflow by profiling users based on interests and forwarding the questions to the right users to obtain trustful, complete, and relevant answers.

Bug reporting is a feature that is provided with the intent that the user can submit a bug report regarding software malfunction. Most of the sizeable open source software is equipped with this feature. Such reporting systems are generally termed “issue tracking systems.” One of the typical attributes of a bug report is severity, which is often marked as critical by users in most bug reports. However, not all bugs are critical, and the ones that are actually critical can be overlooked in such a situation. Likewise, there are some other types of classifications of bug reports, some of which require more critical attention in one context or the other. For example, the bug report might be labeled as one of the non-functional requirements like robustness, security, efficiency, etc., or might be in some other class domain like UI, Database, and Notifications, etc. Often, for large software projects, we are interested in two or more of such class labels for the consideration of bug reports, so as to determine the priority of assignment to the developer for resolution. This paper considers the Eclipse software database for the determination of multiclass labels for the purpose of the determination of priorities in changing contexts.

Bug fixing is a critical activity in the software development process. In issue tracking systems such as JIRA, each bug report is assigned a priority level to indicate the urgency and importance level of the bug. The priority may change during the bug fixing process, indicating that the urgency and importance level of the bug will change with the bug fixing. However, manually evaluating priority changes for bugs is a tedious process that heavily relies on the subjective judgment of developers and project managers, leading to incorrect priority changes and thus hindering timely bug fixes. Given the lack of research on bug priority change prediction, we propose a novel two-phase bug report priority change prediction method based on bug fixing evolution features and class imbalance handling strategy. Specifically, we divided the bug lifecycle into two phases: bug reporting and bug fixing, and constructed bug priority change prediction models for each phase. To evaluate the performance of our method, we conducted experiments on a bug dataset constructed from 32 non-trivial Apache projects. The experimental results show that our proposed bug fixing evolution features and the adopted class imbalance handling strategy can effectively improve the performance of prediction models. The F1-score of the prediction model constructed for the bug reporting phase reached 0.798, while the F1-weighted and F1-macro of the prediction model constructed for the bug fixing phase were 0.712 and 0.613, respectively. Furthermore, we explored the cross-project applicability of our prediction models and their performance at different priority levels. The findings indicate large variations in model performance across different projects, although the overall scores remain decent. Meanwhile, the predictive performance across various priority levels remained relatively consistently high.

Stack Overflow is a popular community-based Qaamp;A website that caters to technical needs of software developers. As of February 2015 -- Stack Overflow has more than 3.9M registered users, 8.8M questions, and 41M comments. Stack Overflow provides explicit and detailed guidelines on how to post questions but, some questions are very poor in quality. Such questions are deleted by the experienced community members and moderators. Deleted questions increase maintenance cost and have an adverse impact on the user experience. Therefore, predicting deleted questions is an important task. In this study, we propose a two stage hybrid approach -- DelPredictor -- which combines text processing and classification techniques to predict deleted questions. In the first stage, DelPredictor converts text in the title, body, and tag fields of questions into numerical textual features via text processing and classification techniques. In the second stage, it extracts meta features that can be categorized into: profile, community, content, and syntactic features. Next, it learns and combines two independent classifiers built on the textual and meta features. We evaluate DelPredictor on 5 years (2008 -- 2013) of deleted questions from Stack Overflow. Our experimental results show that DelPredictor improves the F1-scores over baseline prediction, a prior approach [12] and a text-based approach by 29.50%, 9.34%, and 28.11%, respectively.

Software developers are increasingly using cloud-based services that provide machine learning capabilities to implement `intelligent' features. Studies show that incorporating machine learning into an application increases technical debt, creates data dependencies, and introduces uncertainty due to their non-deterministic behaviour. We know very little about the emotional state of software developers who have to deal with such issues; and the impacts on productivity. This paper presents a preliminary effort to better understand the emotions of developers when experiencing issues with these services with the wider goal of discovering potential service improvements. We conducted a landscape analysis of emotions found in 1,425 Stack Overflow questions about a specific and mature subset of these cloud-based services, namely those that provide computer vision techniques. To speed up the emotion identification process, we trialled an automatic approach using a pre-trained emotion classifier that was specifically trained on Stack Overflow content, EmoTxt, and manually verified its classification results. We found that the identified emotions vary for different types of questions, and a discrepancy exists between automatic and manual emotion analysis due to subjectivity.

Many software comprehension tasks depend on how stakeholders textually describe their problems. These textual descriptions are leveraged by Text Retrieval (TR)-based solutions to more than 20 software engineering tasks, such as duplicate issue detection. The common assumption of such methods is that text describing the same issue in multiple places will have a common vocabulary. This paper presents an empirical study aimed at verifying this assumption and discusses the impact of the common vocabulary on duplicate issue detection. The study investigated 13K+ pairs of duplicate bug reports and Stack Overflow (SO) questions. We found that on average, more than 12.2% of the duplicate pairs do not have common terms. The other duplicate issue descriptions share, on average, 30% of their vocabulary. The good news is that these duplicates have significantly more terms in common than the non-duplicates. We also found that the difference between the lexical agreement of duplicate and non-duplicate pairs is a good predictor for the performance of TR-based duplicate detection.

When encountering an issue, technical users (e.g., developers) usually file the issue report to the issue tracking systems. But non-technical end-users are more likely to express their opinions on social network platforms, such as Twitter. For software systems (e.g., Firefox and Chrome) that have a high exposure to millions of non-technical end-users, it is important to monitor and solve issues observed by a large user base. The widely used micro-blogging site (i.e., Twitter) has millions of active users. Therefore, it can provide instant feedback on products to the developers. In this paper, we investigate whether social networks (i.e., Twitter) can improve the bug fixing process by analyzing the short messages posted by end-users on Twitter (i.e., tweets). We propose an approach to remove noisy tweets, and map the remaining tweets to bug reports. We conduct an empirical study to investigate the usefulness of Twitter in the bug fixing process. We choose two widely adopted browsers (i.e., Firefox and Chrome) that are also large and rapidly released software systems. We find that issue reports are not treated differently regardless whether users tweet about the issue or not, except that Firefox developers tend to label an issue as more severe if users tweet about it. The feedback from Firefox contributors confirms that the tweets are not currently leveraged in the bug fixing process, due to the challenges associated to discovering bugs through Twitter. Moreover, we observe that many issues are posted on Twitter earlier than on issue tracking systems. More specifically, at least one third of issues could have been reported to developers 8.2 days and 7.6 days earlier in Firefox and Chrome, respectively. In conclusion, tweets are useful in providing earlier acknowledgment of issues, which developers can potentially use to focus their efforts on the issues impacting a large user-base.

The sheer number of available technologies and the complex relationships among them make it challenging to choose the right technologies for software projects. Developers often turn to online resources (e.g., expert articles and community answers) to get a good understanding of the technology landscape. Such online resources are primarily opinion-based and are often out of date. Furthermore, information is often scattered in many online resources, which has to be aggregated to have a big picture of the technology landscape. In this paper, we exploit the fact that Stack Overflow users tag their questions with the main technologies that the questions revolve around, and develop association rule mining and community detection techniques to mine technology landscape from Stack Overflow question tags. The mined technology landscape is represented in a graphical Technology Associative Network (TAN). Our empirical study shows that the mined TAN captures a wide range of technologies, the complex relationships among the technologies, and the trend of the technologies in the developers' discussions on Stack Overflow. We develop a website (https://graphofknowledge.appspot.com/) for the community to access and evaluate the mined technology landscape. The website visit statistics by Google Analytics shows the developers' general interests in our technology landscape service. We also report a small-scale user study to evaluate the potential usefulness of our tool.

[Context and motivation] Traces between issues in issue tracking systems connect bug reports to software features, they connect competing implementation ideas for a software feature or they identify duplicate issues. However, the trace quality is usually very low. To improve the trace quality between requirements, features, and bugs, information retrieval algorithms for automated trace retrieval can be employed. Prevailing research focusses on structured and well-formed documents, such as natural language requirement descriptions. In contrast, the information in issue tracking systems is often poorly structured and contains digressing discussions or noise, such as code snippets, stack traces, and links. Since noise has a negative impact on algorithms for automated trace retrieval, this paper asks: [Question/Problem] Do information retrieval algorithms for automated traceability perform effectively on issue tracking system data? [Results] This paper presents an extensive evaluation of the performance of five information retrieval algorithms. Furthermore, it investigates different preprocessing stages e.g. stemming or differentiating code snippets from natural language and evaluates how to take advantage of an issue's structure e.g. title, description, and comments to improve the results. The results show that algorithms perform poorly without considering the nature of issue tracking data, but can be improved by project-specific preprocessing and term weighting. [Contribution] Our results show how automated trace retrieval on issue tracking system data can be improved. Our manually created gold standard and an open-source implementation based on the OpenTrace platform can be used by other researchers to further pursue this topic.

Stack Overflow and MedHelp are examples of domain-specific community-based question answering (CQA) systems. Different from CQA systems for general topics (e.g., Yahoo! Answers, Baidu Knows), questions and answers in domain-specific CQA systems are mostly in the same topical domain, enabling more comprehensive interaction between users on fine-grained topics. In such systems, users are more likely to ask questions on unfamiliar topics and to answer questions matching their expertise. Users can also vote answers based on their judgements. In this paper, we propose a Tri-Role Topic Model (TRTM) to model the tri-roles of users (i.e., as askers, answerers, and voters, respectively) and the activities of each role including composing question, selecting question to answer, contributing and voting answers. The proposed model can be used to enhance CQA systems from many perspectives. As a case study, we conducted experiments on ranking answers for questions on Stack Overflow, a CQA system for professional and enthusiast programmers. Experimental results show that TRTM is effective in facilitating users getting ideal rankings of answers, particularly for new and less popular questions. Evaluated on nDCG, TRTM outperforms state-of-the-art methods.

Quick, name the most popular online technical forum for chemistry. Having a hard time coming up with an answer? You’re not alone. Despite a few attempts, there is no widely-recognized place to go online to exchange technical information in chemistry. Some Examples If you look hard, you can find chemically-oriented forums scattered around the Web. For example: Chemical Forums Perhaps the best-known discussion board for chemistry.

Chapter 10 - Thread networks: Mapping message boards and email lists

Social collaborative platforms such as GitHub and Stack Overflow have been\nincreasingly used to improve work productivity via collaborative efforts. To\nimprove user experiences in these platforms, it is desirable to have a\nrecommender system that can suggest not only items (e.g., a GitHub repository)\nto a user, but also activities to be performed on the suggested items (e.g.,\nforking a repository). To this end, we propose a new approach dubbed Keen2Act,\nwhich decomposes the recommendation problem into two stages: the Keen and Act\nsteps. The Keen step identifies, for a given user, a (sub)set of items in which\nhe/she is likely to be interested. The Act step then recommends to the user\nwhich activities to perform on the identified set of items. This decomposition\nprovides a practical approach to tackling complex activity recommendation tasks\nwhile producing higher recommendation quality. We evaluate our proposed\napproach using two real-world datasets and obtain promising results whereby\nKeen2Act outperforms several baseline models.\n