Review
Engineering, Multidisciplinary
Haifeng Wang, Hua Wu, Zhongjun He, Liang Huang, Kenneth Ward Church
Summary: After more than 70 years of evolution, great achievements have been made in machine translation, especially with the emergence of neural machine translation (NMT) in recent years. This article reviews the history of machine translation and introduces NMT, including its frameworks and multilingual translation models. It also discusses cutting-edge simultaneous translation methods and various machine translation products and applications. Challenges and future research directions in this field are briefly discussed as well.
Article
Computer Science, Artificial Intelligence
Xiao Liu, Jing Zhao, Shiliang Sun, Huawen Liu, Hao Yang
Summary: This paper introduces a variational multimodal machine translation (VMMT) model that utilizes visual and textual information to simulate uncertainty in language. The model employs multitask learning to reduce the gap in semantic representation between different modes and applies the information bottleneck theory to filter redundancy.
INFORMATION FUSION
(2021)
Article
Computer Science, Artificial Intelligence
Abdur Razaq, Babar Shah, Gohar Khan, Omar Alfandi, Abrar Ullah, Zahid Halim, Atta Ur Rahman
Summary: Phrase generation involves changing a sentence in the natural language into a new one with a different syntactic structure but the same semantic meaning. The current sequence-to-sequence strategy focuses on recalling words and structures from the training dataset rather than learning word semantics. As a result, the generated statements are often grammatically accurate but linguistically incorrect. However, the proposed neural-based statistical machine translation (NSMT) model overcomes these limitations and achieves state-of-the-art performance on benchmark datasets.
NEURAL COMPUTING & APPLICATIONS
(2023)
Review
Public, Environmental & Occupational Health
Emma Quinn, Kai Hsun Hsiao, Isis Maitland-Scott, Maria Gomez, Melissa T. Baysari, Zeina Najjar, Leena Gupta
Summary: The study described literature on web-based apps for surveillance and response to acute communicable disease outbreaks in the community. Findings indicated that these apps are primarily designed to improve early detection of disease outbreaks and should have more features to support information exchange and outbreak response actions.
JMIR PUBLIC HEALTH AND SURVEILLANCE
(2021)
Article
Computer Science, Information Systems
Emre Satir, Hasan Bulut
Summary: This paper proposes a hybrid system-based method that uses SMT predictions to prevent quality deterioration in the beam search algorithm used in NMT decoding, and presents two different algorithms for reranking NMT n-best lists. Experimental results show that the method prevents the decrease in translation quality and produces gains in BLEU and METEOR scores compared to baseline results for different beam sizes.
INFORMATION SCIENCES
(2021)
Article
Computer Science, Software Engineering
Vahid Garousi, David Cutting, Michael Felderer
Summary: The study showed that most users are dissatisfied with contact-tracing apps, with major issues including high battery drainage and doubts on the effectiveness of the apps.
JOURNAL OF SYSTEMS AND SOFTWARE
(2022)
Article
Computer Science, Software Engineering
Eduardo Blazquez, Juan Tapiador
Summary: This paper introduces Kunai, a C++ library for static analysis of Android apps. It focuses on providing advanced static analysis features for Dalvik code, such as parsing, disassembling, and code analysis. Kunai is written in C++ and emphasizes efficiency and extensibility for easy integration of new analysis modules. It is particularly useful for developing analysis pipelines for processing large app datasets.
Article
Computer Science, Software Engineering
Hourieh Khalajzadeh, Mojtaba Shahin, Humphrey O. Obie, Pragya Agrawal, John Grundy
Summary: Failure to consider the characteristics, limitations, and abilities of diverse end-users during mobile app development may lead to human-centric issues for end-users. This paper examines the human-centric issues reported by end-users through app reviews and discussed by developers on GitHub. It also investigates the feasibility and usefulness of an automated tool for detecting and classifying human-centric issues. The findings highlight the importance of addressing these issues and suggest possible future work to improve mobile app development.
IEEE TRANSACTIONS ON SOFTWARE ENGINEERING
(2023)
Article
Education & Educational Research
Hind Alotaibi, Dania Salamah
Summary: The study aimed to assess the impact of mobile translation apps on the performance of trainee translators. The results indicated that students who used the Reverso Context app performed better in terms of translation quality and performance, with fewer errors in lexical, cohesion, omission, and text-type.
EDUCATION AND INFORMATION TECHNOLOGIES
(2023)
Article
Mathematics
Jani Dugonik, Mirjam Sepesy Maucec, Domen Verber, Janez Brest
Summary: This paper proposes a hybrid machine translation system that combines statistical machine translation with neural machine translation to improve translation quality. Two NMT systems and two SMT systems were built for Slovenian-English translation. A multilingual language model was used to embed the source sentence and translations into the same vector space, and features were extracted based on distances and similarities. Well-known classifiers were used to predict the best translation, and the proposed method achieved notable improvements in BLEU score.
Article
Computer Science, Artificial Intelligence
Tong Li, Yong Li, Mingyang Zhang, Sasu Tarkoma, Pan Hui
Summary: Mobile apps are an essential part of people's daily lives, and this article focuses on inferring user profiles from their spatiotemporal mobile app usage behavior. The authors propose a multi-relational heterogeneous graph attention network (MRel-HGAN) to achieve this task, which effectively utilizes the rich semantic information of the multi-relational structure in the mobile app usage graph. Experimental results demonstrate the effectiveness and superiority of MRel-HGAN in user profiling for gender and age attributes.
ACM TRANSACTIONS ON INTELLIGENT SYSTEMS AND TECHNOLOGY
(2023)
Article
Computer Science, Theory & Methods
Sameen Maruf, Fahimeh Saleh, Gholamreza Haffari
Summary: Machine translation is a vital task in natural language processing that automates the translation process and decreases reliance on human translators. With the advancement of neural networks, translation quality has improved, surpassing that of statistical techniques. This survey article focuses on the advancements in document-level machine translation post-neural revolution, highlighting the current state and future directions of the field.
ACM COMPUTING SURVEYS
(2021)
Article
Computer Science, Information Systems
Bakheet Aljedaani, Aakash Ahmad, Mansooreh Zahedi, Muhammad Ali Babar
Summary: This article presents a study on mobile health apps, where attack simulation scenarios were used to monitor user actions. The findings show that most users have negative views on access permissions, and many users do not carefully review privacy policies before granting permissions, leading to undesired or malicious access to health data.
INFORMATION AND SOFTWARE TECHNOLOGY
(2023)
Article
Computer Science, Artificial Intelligence
Xueqing Wu, Yingce Xia, Jinhua Zhu, Lijun Wu, Shufang Xie, Tao Qin
Summary: The study investigates utilizing BERT to encode contextual information and finds that concatenating all contextual sequences into a longer one and then encoding it with BERT achieves the best translation results. The approach led to state-of-the-art BLEU scores on various translation tasks.
Article
Health Care Sciences & Services
James Benoit, Lisa Hartling, Michelle Chan, Shannon Scott
Summary: The study aimed to assess the quality of acute childhood illness apps available to North American parents and caregivers. While there were hundreds of apps identified, only a small percentage received high ratings, and there was a lack of evidence-based Canadian content in the app marketplaces.
JOURNAL OF MEDICAL INTERNET RESEARCH
(2021)
Article
Computer Science, Software Engineering
Zhe Liu, Chunyang Chen, Junjie Wang, Yuekai Huang, Jun Hu, Qing Wang
Summary: Graphical User Interface (GUI) serves as a visual link between software application and users, enabling interaction between them. However, the complexity of GUI poses challenges to its implementation, as display issues often occur due to software or hardware compatibility. To address this, a fully automated approach called Nighthawk is proposed, which uses deep learning to detect and locate display issues in GUI screenshots for developers to fix. Additionally, a heuristic-based training data auto-generation method is introduced to generate labeled training data. Evaluation shows that Nighthawk achieves high precision and recall in detecting UI display issues and successfully uncovers previously-undetected issues in popular Android apps.
IEEE TRANSACTIONS ON SOFTWARE ENGINEERING
(2023)
Article
Computer Science, Software Engineering
Mojtaba Shahin, Christabel Gonsalvez, Jon Whittle, Chunyang Chen, Li Li, Xin Xia
Summary: This study investigates how secondary school girls perceive computational thinking practices when using the micro:bit device in a collaborative setting. It identifies challenges the girls face and best practices they adopt while working on computational solutions.
JOURNAL OF SYSTEMS AND SOFTWARE
(2022)
Article
Computer Science, Information Systems
Mengxi Zhang, Huaxiao Liu, Chunyang Chen, Yuzhou Liu, Shuotong Bai
Summary: The study found that only 1.2% of GitHub repositories contain the PRT, mainly in high popularity and with a large number of PRs. Contributors are willing to accept the PRT that requires key information, while it also helps to manage repositories, leading to shorter review time, fewer duplicated pull requests, and minimal invalid comments.
INFORMATION AND SOFTWARE TECHNOLOGY
(2022)
Article
Computer Science, Software Engineering
Junjie Wang, Ye Yang, Song Wang, Chunyang Chen, Dandan Wang, Qing Wang
Summary: Crowdsourced software testing, also known as crowdtesting, is a specialized form of crowdsourcing that requires skilled and dedicated crowdworkers. This paper addresses the issue of inappropriate task selection in crowdtesting, which leads to unpaid and wasted effort. The authors propose a context-aware personalized task recommendation approach called PTRec, which leverages a testing context model and a learning-based recommendation model to help crowdworkers make informed decisions. The evaluation of PTRec on a large crowdtesting platform demonstrates its potential in improving bug detection efficiency and increasing crowdworkers' earnings.
IEEE TRANSACTIONS ON SOFTWARE ENGINEERING
(2022)
Article
Computer Science, Software Engineering
Sen Chen, Chunyang Chen, Lingling Fan, Mingming Fan, Xian Zhan, Yang Liu
Summary: Mobile apps provide new opportunities for people with disabilities to act independently, but there are still accessibility issues that need to be addressed. This study proposes an automated app page exploration tool to collect a comprehensive dataset of accessibility issues and investigates the characteristics of these issues. The findings highlight the importance of maintaining mobile app accessibility for users, especially the elderly and disabled.
IEEE TRANSACTIONS ON SOFTWARE ENGINEERING
(2022)
Proceedings Paper
Computer Science, Information Systems
Zhe Liu, Chunyang Chen, Junjie Wang, Yuekai Huang, Jun Hu, Qing Wang
Summary: Mobile apps are essential for daily life, and manual testing plays a crucial role in ensuring app quality. However, manual testing can be time-consuming and inefficient due to repeated actions and missed functionalities. Inspired by the game candy crush, NaviDroid proposes an approach that guides testers with highlighted next operations for more effective and efficient testing.
PROCEEDINGS OF THE 2022 CHI CONFERENCE ON HUMAN FACTORS IN COMPUTING SYSTEMS (CHI' 22)
(2022)
Article
Computer Science, Information Systems
Xinyan Li, Han Wang, Chunyang Chen, John Grundy
Summary: The ongoing COVID-19 pandemic has highlighted the importance of dashboards in providing real-time information. However, there is often a gap between the information needs of the public and the supply provided by existing COVID-19 dashboards. Through an empirical study comparing people's needs on Twitter with existing information suppliers, we found that people are interested in various aspects beyond COVID-19, such as the relationship between COVID-19 and other viruses, its origin, vaccine development, fake news, and its impact on women, schools/universities, and businesses. We also identified common visualization and interaction patterns used in dashboards, which can help developers optimize their designs to meet people's needs and improve future crisis management dashboard development.
IEEE TRANSACTIONS ON SERVICES COMPUTING
(2022)
Article
Computer Science, Hardware & Architecture
Sen Chen, Lingling Fan, Chunyang Chen, Minhui Xue, Yang Liu, Lihua Xu
Summary: Mobile phishing attacks using disguise techniques have raised security concerns, with current detection methods potentially vulnerable. A new attack technique, GUI-Squatting attack, can automatically generate phishing apps on the Android platform using deep learning algorithms. Experimental results suggest existing phishing defenses are ineffective against emergent attacks, stimulating the need for more efficient detection techniques.
IEEE TRANSACTIONS ON DEPENDABLE AND SECURE COMPUTING
(2021)
Proceedings Paper
Computer Science, Software Engineering
Bo Yang, Zhenchang Xing, Xin Xia, Chunyang Chen, Deheng Ye, Shanping Li
Summary: Visual design smells in UI design indicate violations of good design guidelines. By following a design system, developers can avoid common design issues. An automated UI design smell detector helps identify and address UI design problems.
2021 IEEE/ACM 43RD INTERNATIONAL CONFERENCE ON SOFTWARE ENGINEERING: COMPANION PROCEEDINGS (ICSE-COMPANION 2021)
(2021)
Article
Computer Science, Software Engineering
Xiang Chen, Chunyang Chen, Dun Zhang, Zhenchang Xing
Summary: This paper proposes an automatic unsupervised approach to build a thesaurus for software engineering text, utilizing software-specific and general corpora to identify terms, infer morphological forms, and perform graph analysis. Experimental results show high coverage and accuracy of the approach, confirmed through manual verification of abbreviations and synonyms in the thesaurus.
IEEE TRANSACTIONS ON SOFTWARE ENGINEERING
(2021)
Proceedings Paper
Computer Science, Software Engineering
Bo Yang, Zhenchang Xing, Xin Xia, Chunyang Chen, Deheng Ye, Shanping Li
Summary: The study revealed that Material Design guidelines extend beyond UI aesthetics, covering seven general design dimensions and four component design aspects. Violating these guidelines leads to visual design smells in UIs. The automated UI design smell detector UIS-Hunter has high detection accuracy and helps developers learn best practices for Material Design.
2021 IEEE/ACM 43RD INTERNATIONAL CONFERENCE ON SOFTWARE ENGINEERING (ICSE 2021)
(2021)
Proceedings Paper
Computer Science, Software Engineering
Tianming Zhao, Chunyang Chen, Yuanning Liu, Xiaodong Zhu
Summary: Graphical User Interface (GUI) is essential in modern software, and a good GUI design is crucial for software success. Automated generated GUIs can enhance design personalization, and a model called GUIGAN has been developed to automatically generate GUI designs similar to natural language generation.
2021 IEEE/ACM 43RD INTERNATIONAL CONFERENCE ON SOFTWARE ENGINEERING (ICSE 2021)
(2021)
Proceedings Paper
Computer Science, Software Engineering
Yuanchun Li, Liayi Hua, Haoyu Wang, Chunyang Chen, Yunxin Liu
Summary: This paper introduces a highly practical backdoor attack achieved with reverse-engineering techniques over compiled deep learning models, showing its effectiveness and vulnerability of real-world mobile deep learning apps.
2021 IEEE/ACM 43RD INTERNATIONAL CONFERENCE ON SOFTWARE ENGINEERING (ICSE 2021)
(2021)
Proceedings Paper
Computer Science, Software Engineering
Kaibo Cao, Chunyang Chen, Sebastian Baltes, Christoph Treude, Xiang Chen
Summary: The research found that the difficulty for developers to efficiently search for the information they need on Stack Overflow mainly stems from the gap between user intentions and text meanings, as well as the semantic gap between queries and post content. To address this issue, an automated software-specific query reformulation approach based on deep learning was proposed, which can generate candidate reformulated queries when given the user's original query. Experimental results demonstrated significant improvements in terms of ExactMatch and GLEU.
2021 IEEE/ACM 43RD INTERNATIONAL CONFERENCE ON SOFTWARE ENGINEERING (ICSE 2021)
(2021)
Proceedings Paper
Computer Science, Software Engineering
Yujin Huang, Han Hu, Chunyang Chen
Summary: The study shows that embedding deep learning models in mobile applications, such as Android apps, may be vulnerable to adversarial attacks. The experiment demonstrates that hackers can successfully attack real-world Android apps by identifying pre-trained models.
2021 IEEE/ACM 43RD INTERNATIONAL CONFERENCE ON SOFTWARE ENGINEERING: SOFTWARE ENGINEERING IN PRACTICE (ICSE-SEIP 2021)
(2021)