arXiv subjects

Hourieh Khalajzadeh

Publications and source records attributed to Hourieh Khalajzadeh.

29 records · Page 2Linked to original sources

arXiv2023

A Study of Gender Discussions in Mobile Apps

Mobile software apps ("apps") are one of the prevailing digital technologies that our modern life heavily depends on. A key issue in the development of apps is how to design gender-inclusive apps. Apps that do not consider gender inclusion, diversity, and equality in their design can create barriers (e.g., excluding some of the users because of their gender) for their diverse users. While there have been some efforts to develop gender-inclusive apps, a lack of deep understanding regarding user perspectives on gender may prevent app developers and owners from identifying issues related to gender and proposing solutions for improvement. Users express many different opinions about apps in their reviews, from sharing their experiences, and reporting bugs, to requesting new features. In this study, we aim at unpacking gender discussions about apps from the user perspective by analysing app reviews. We first develop and evaluate several Machine Learning (ML) and Deep Learning (DL) classifiers that automatically detect gender reviews (i.e., reviews that contain discussions about gender). We apply our ML and DL classifiers on a manually constructed dataset of 1,440 app reviews from the Google App Store, composing 620 gender reviews and 820 non-gender reviews. Our best classifier achieves an F1-score of 90.77%. Second, our qualitative analysis of a randomly selected 388 out of 620 gender reviews shows that gender discussions in app reviews revolve around six topics: App Features, Appearance, Content, Company Policy and Censorship, Advertisement, and Community. Finally, we provide some practical implications and recommendations for developing gender-inclusive apps.

cs.SE↗

arXiv2022

Automated Detection, Categorisation and Developers' Experience with the Violations of Honesty in Mobile Apps

Human values such as honesty, social responsibility, fairness, privacy, and the like are things considered important by individuals and society. Software systems, including mobile software applications (apps), may ignore or violate such values, leading to negative effects in various ways for individuals and society. While some works have investigated different aspects of human values in software engineering, this mixed-methods study focuses on honesty as a critical human value. In particular, we studied (i) how to detect honesty violations in mobile apps, (ii) the types of honesty violations in mobile apps, and (iii) the perspectives of app developers on these detected honesty violations. We first develop and evaluate 7 machine learning (ML) models to automatically detect violations of the value of honesty in app reviews from an end user perspective. The most promising was a Deep Neural Network model with F1 score of 0.921. We then conducted a manual analysis of 401 reviews containing honesty violations and characterised honest violations in mobile apps into 10 categories: unfair cancellation and refund policies; false advertisements; delusive subscriptions; cheating systems; inaccurate information; unfair fees; no service; deletion of reviews; impersonation; and fraudulent looking apps. A developer survey and interview study with mobile developers then identified 7 key causes behind honesty violations in mobile apps and 8 strategies to avoid or fix such violations. The findings of our developer study also articulate the negative consequences that honesty violations might bring for businesses, developers, and users. Finally, the app developers' feedback shows that our prototype ML-based models can have promising benefits in practice.

cs.SE↗

arXiv2022

Diverse End User Requirements

As part of our larger research effort to improve support for diverse end user human-centric aspects during software development, we wanted to better understand how developers currently go about addressing these challenging human-centric aspects of their end users in contemporary software development projects. We wanted to find out which are the key end user human-centric aspects that software developers currently find challenging to address, and how they currently go about trying to address diverse end user human-centric aspects. We wanted to find out what sorts of end user human-centric aspects they tend to encounter, which ones they view as more important and which more challenging to address, what techniques (if any) they currently use to address (some of) them, and where they perceive further research in this area could be done to provide them practical support. To this end we carried out a detailed online survey of developers and development team managers, receiving 60 usable responses. We interviewed 12 developers and managers from a range of different practice domains, role specialisations and experience levels to explore further details about issues.

cs.SE↗

arXiv2022

Supporting Developers in Addressing Human-centric Issues in Mobile Apps

Failure to consider the characteristics, limitations, and abilities of diverse end-users during mobile apps development may lead to problems for end-users such as accessibility and usability issues. We refer to this class of problems as human-centric issues. Despite their importance, there is a limited understanding of the types of human-centric issues that are encountered by end-users and taken into account by the developers of mobile apps. In this paper, we examine what human-centric issues end-users report through Google App Store reviews, which human-centric issues are a topic of discussion for developers on GitHub, and whether end-users and developers discuss the same human-centric issues. We then investigate whether an automated tool might help detect such human-centric issues and whether developers would find such a tool useful. To do this, we conducted an empirical study by extracting and manually analysing a random sample of 1,200 app reviews and 1,200 issue comments from 12 diverse projects that exist on both Google App Store and GitHub. Our analysis led to a taxonomy of human-centric issues that categorises human-centric issues into three-high levels: App Usage, Inclusiveness, and User Reaction. We then developed machine learning and deep learning models that are promising in automatically identifying and classifying human-centric issues from app reviews and developer discussions. A survey of mobile app developers shows that the automated detection of human-centric issues has practical applications. Guided by our findings, we highlight some implications and possible future work to further understand and incorporate human-centric issues in mobile apps development.

cs.SE↗

arXiv2022

Human Values Violations in Stack Overflow: An Exploratory Study

A growing number of software-intensive systems are being accused of violating or ignoring human values (e.g., privacy, inclusion, and social responsibility), and this poses great difficulties to individuals and society. Such violations often occur due to the solutions employed and decisions made by developers of such systems that are misaligned with user values. Stack Overflow is the most popular QA website among developers to share their issues, solutions (e.g., code snippets), and decisions during software development. We conducted an exploratory study to investigate the occurrence of human values violations in Stack Overflow posts. As comments under posts are often used to point out the possible issues and weaknesses of the posts, we analyzed 2000 Stack Overflow comments and their corresponding posts (1980 unique questions or answers) to identify the types of human values violations and the reactions of Stack Overflow users to such violations. Our study finds that 315 out of 2000 comments contain concerns indicating their associated posts (313 unique posts) violate human values. Leveraging Schwartz's theory of basic human values as the most widely used values model, we show that hedonism and benevolence are the most violated value categories. We also find the reaction of Stack Overflow commenters to perceived human values violations is very quick, yet the majority of posts (76.35%) accused of human values violation do not get downvoted at all. Finally, we find that the original posters rarely react to the concerns of potential human values violations by editing their posts. At the same time, they usually are receptive when responding to these comments in follow-up comments of their own.

cs.SE↗

arXiv2022

How are Diverse End-user Human-centric Issues Discussed on GitHub?

Many software systems fail to meet the needs of the diverse end-users in society and are prone to pose problems, such as accessibility and usability issues. Some of these problems (partially) stem from the failure to consider the characteristics, limitations, and abilities of diverse end-users during software development. We refer to this class of problems as human-centric issues. Despite their importance, there is a limited understanding of the types of human-centric issues encountered by developers. In-depth knowledge of these human-centric issues is needed to design software systems that better meet their diverse end-users' needs. This paper aims to provide insights for the software development and research communities on which human-centric issues are a topic of discussion for developers on GitHub. We conducted an empirical study by extracting and manually analysing 1,691 issue comments from 12 diverse projects, ranging from small to large-scale projects, including projects designed for challenged end-users, e.g., visually impaired and dyslexic users. Our analysis shows that eight categories of human-centric issues are discussed by developers. These include Inclusiveness, Privacy & Security, Compatibility, Location & Language, Preference, Satisfaction, Emotional Aspects, and Accessibility. Guided by our findings, we highlight some implications and possible future paths to further understand and incorporate human-centric issues in software development to be able to design software that meets the needs of diverse end users in society.

cs.SE↗

arXiv2021

A First Step Towards Detecting Values-violating Defects in Android APIs

Human values are an important aspect of life and should be supported in ubiquitous technologies such as mobile applications (apps). There has been a lot of focus on fixing certain kinds of violation of human values, especially privacy, accessibility, and security while other values such as pleasure, tradition, and humility have received little focus. In this paper, we investigate the relationship between human values and Android API services and developed algorithms to detect potential violation of these values. We evaluated our algorithms with a manually curated ground truthset resulting in a high performance, and applied the algorithms to 10,000 apps. Our results show a correlation between violation of values and the presence of viruses. Our results also show that apps with the lowest number of installations contain more violation of values and the frequency of the violation of values was highest in social apps.

cs.SE↗

arXiv2021

Latexify Math: Mathematical Formula Markup Revision to Assist Collaborative Editing in Math Q&A Sites

Collaborative editing questions and answers plays an important role in quality control of Mathematics Stack Exchange which is a math Q&A Site. Our study of post edits in Mathematics Stack Exchange shows that there is a large number of math-related edits about latexifying formulas, revising LaTeX and converting the blurred math formula screenshots to LaTeX sequence. Despite its importance, manually editing one math-related post especially those with complex mathematical formulas is time-consuming and error-prone even for experienced users. To assist post owners and editors to do this editing, we have developed an edit-assistance tool, MathLatexEdit for formula latexification, LaTeX revision and screenshot transcription. We formulate this formula editing task as a translation problem, in which an original post is translated to a revised post. MathLatexEdit implements a deep learning based approach including two encoder-decoder models for textual and visual LaTeX edit recommendation with math-specific inference. The two models are trained on large-scale historical original-edited post pairs and synthesized screenshot-formula pairs. Our evaluation of MathLatexEdit not only demonstrates the accuracy of our model, but also the usefulness of MathLatexEdit in editing real-world posts which are accepted in Mathematics Stack Exchange.

cs.HC↗

arXiv2021

Automated Identification of Security Discussions in Microservices Systems: Industrial Surveys and Experiments

Lack of awareness and knowledge of microservices-specific security challenges and solutions often leads to ill-informed security decisions in microservices system development. We claim that identifying and leveraging security discussions scattered in existing microservices systems can partially close this gap. We define security discussion as "a paragraph from developer discussions that includes design decisions, challenges, or solutions relating to security". We first surveyed 67 practitioners and found that securing microservices systems is a unique challenge and that having access to security discussions is useful for making security decisions. The survey also confirms the usefulness of potential tools that can automatically identify such security discussions. We developed fifteen machine/deep learning models to automatically identify security discussions. We applied these models on a manually constructed dataset consisting of 4,813 security discussions and 12,464 non-security discussions. We found that all the models can effectively identify security discussions: an average precision of 84.86%, recall of 72.80%, F1-score of 77.89%, AUC of 83.75% and G-mean 82.77%. DeepM1, a deep learning model, performs the best, achieving above 84% in all metrics and significantly outperforms three baselines. Finally, the practitioners' feedback collected from a validation survey reveals that security discussions identified by DeepM1 have promising applications in practice.

cs.SE↗

arXiv2021

Human-Centric Issues in eHealth App Development and Usage: A Preliminary Assessment

Health-related mobile applications are known as eHealth apps. These apps make people more aware of their health, help during critical situations, provide home-based disease management, and monitor/support personalized care through sensing/interaction. eHealth app usage is rapidly increasing with a large number of new apps being developed. Unfortunately, many eHealth apps do not successfully adopt Human-Centric Issues (HCI) in the app development process and its deployment stages, leading them to become ineffective and not inclusive of diverse end-users. This paper provides an initial assessment of key human factors related to eHealth apps by literature review, existing guidelines analysis, and user studies. Preliminary results suggest that Usability, Accessibility, Reliability, Versatility, and User Experience are essential HCIs for eHealth apps, and need further attention from researchers and practitioners. Therefore, outcomes of this research will look to amend support for users, developers, and stakeholders of eHealth apps in the form of improved actionable guidelines, best practice examples, and evaluation techniques. The research also aims to trial the proposed solutions on real-world projects.

cs.SE↗

arXiv2021

Checking App Behavior Against App Descriptions: What If There are No App Descriptions?

Classifying mobile apps based on their description is beneficial for several purposes. However, many app descriptions do not reflect app functionalities, whether accidentally or on purpose. Most importantly, these app classification methods do not work if the app description is unavailable. This paper investigates a Reverse Engineering-based Approach to Classify mobile apps using The data that exists in the app, called REACT. To validate the proposed REACT method, we use a large set of Android apps (24,652 apps in total). We also show REACTs' extendibility for malware/anomaly detection and prove its reliability and scalability. However, our analysis shows some limitations in REACT procedure and implementation, especially for similar feature based app grouping. We discuss the root cause of these failures, our key lessons learned, and some future enhancement ideas. We also share our REACT tools and reproduced datasets for the app market analyst, mobile app developers and software engineering research communities for further research purposes.

cs.SE↗