Tao Zhang

Prof. Tao Zhang

School of Computer Science and Engineering, Macau University of Science and Technology, Macao, China
orcid
Professor of Computer Science and Engineering

Research Focus & Metrics

Academic Career Length

23Years (2004-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
1 February 2019 in Information and Software Technology
Verified
Journal Article
Closed access
ZXZhou Xu
Zhou Xu
JLJin Liu
Jin Liu
XLXiapu Luo
Xiapu Luo
ZYZijiang Yang
Zijiang Yang
YZYifeng Zhang
Yifeng Zhang
Tao ZhangTao Zhang
Tao Zhang

1 July 2016 in Journal of Systems and Software
Verified
Journal Article
Closed access
Tao ZhangTao Zhang
Tao Zhang
JCJiachi Chen
Jiachi Chen
GYGeunseok Yang
Geunseok Yang
BLByungjeong Lee
Byungjeong Lee
DLDaniel Luo
Daniel Luo

10 October 2022
Verified
Conference Paper
Open access
HYHe Ye
He Ye
MMMatias Martinez
Matias Martinez
DLDaniel Luo
Daniel Luo
Tao ZhangTao Zhang
Tao Zhang
MMMartin Monperrus
Martin Monperrus
Learning-based program repair has achieved good results in a recent series of papers. Yet, we observe that the related work fails to repair some bugs because of a lack of knowledge about 1) the application domain of the program being repaired, and 2) the fault type being repaired. In this paper, we solve both problems by changing the learning paradigm from supervised training to self-supervised training in an approach called SelfAPR. First, SelfAPR generates training samples on disk by perturbing a previous version of the program being repaired, enforcing the neural model to capture project-specific knowledge. This is different from the previous work based on mined past commits. Second, SelfAPR executes all training samples and extracts and encodes test execution diagnostics into the input representation, steering the neural model to fix the kind of fault. This is different from the existing studies that only consider static source code as input. We implement SelfAPR and evaluate it in a systematic manner. We generate 1 039 873 training samples obtained by perturbing 17 open-source projects. We evaluate SelfAPR on 818 bugs from Defects4J, SelfAPR correctly repairs 110 of them, outperforming all the supervised learning repair approaches.

31 January 2018 in IEEE Transactions on Dependable and Secure Computing
Verified
Journal Article
Closed access
YSYiran Shen
Yiran Shen
HWHongkai Wen
Hongkai Wen
Chengwen LuoChengwen Luo
Chengwen Luo
WXWeitao Xu
Weitao Xu
Tao ZhangTao Zhang
Tao Zhang
DRDaniela Rus
Daniela Rus
With the fast penetration of commercial Virtual Reality (VR) and Augmented Reality (AR) systems into our daily life, the security issues of those devices have attracted significant interests from both academia and industry. Modern VR/AR systems typically use head-mounted devices (i.e. headsets) to interact with users, and often store private user data, e.g. social network accounts, online transactions or even payment information. This poses significant security threats, since in practice the headset can be potentially obtained and accessed by unauthenticated parties, e.g. identity thieves, and thus cause catastrophic breach. In this paper, we propose a novel GaitLock system, which can reliably authenticate users using their gait signatures. Our system doesn't require extra hardware, e.g. fingerprint sensors or retina scanners, but only uses the on-board inertial measurement units (IMUs) equipped in almost all mainstream VR/AR headsets to authenticate the legitimate users from intruders, by simply asking them to walk a few steps. To achieve that, we propose a new gait recognition model Dynamic-SRC, which combines the strength of Dynamic Time Warping (DTW) and Sparse Representation Classifier (SRC), to extract unique gait patterns from the inertial signals during walking. We implement GaitLock on Google Glass (a typical AR headset), and extensive experiments show that GaitLock outperforms the state-of-the-art systems significantly in recognition accuracy ($>98\%$ success in 5 steps), and is able to run in-situ on the resource-constrained VR/AR headsets without incurring high energy cost.

1 June 2016
Verified
Conference Paper
Closed access
LYLe Yu
Le Yu
DLDaniel Luo
Daniel Luo
XLXule Liu
Xule Liu
Tao ZhangTao Zhang
Tao Zhang
Recent years have witnessed the sharp increase of malicious apps that steal users' personal information. To address users' concerns about privacy risks, more and more apps are accompanied with privacy policies written in natural language because it is difficult for users to infer an app's behaviors according to the required permissions. However, little is known whether these privacy policies are trustworthy or not. It is worth noting that a questionable privacy policy may result from careless preparation by an app developer or intentional deception by an attacker. In this paper, we conduct the first systematic study on privacy policy by proposing a novel approach to automatically identify three kinds of problems in privacy policy. After tackling several challenging issues, we realize our approach in a system, named PPChecker, and evaluate it with real apps and privacy policies. The experimental results show that PPChecker can effectively identify questionable privacy policies with high precision. Moreover, applying PPChecker to 1,197 popular apps, we found that 282 apps (i.e., 23.6%) have at least one kind of problems. This study sheds light on the research of improving and regulating apps' privacy policies.

1 July 2014
Verified
Conference Paper
Closed access
GYGeunseok Yang
Geunseok Yang
Tao ZhangTao Zhang
Tao Zhang
BLByungjeong Lee
Byungjeong Lee
Bug fixing is an essential activity in the software maintenance, because most of the software systems have unavoidable defects. When new bugs are submitted, triagers have to find and assign appropriate developers to fix the bugs. However, if the bugs are at first assigned to inappropriate developers, they may later have to be reassigned to other developers. That increases the time and cost for fixing bugs. Therefore, finding appropriate developers becomes a key to bug resolution. When triagers assign a new bug report, it is necessary to decide how quickly the bug report should be addressed. Thus, the bug severity is an important factor in bug fixing. In this paper, we propose a novel method for the bug triage and bug severity prediction. First, we extract topic(s) from historical bug reports in the bug repository and find bug reports related to each topic. When a new bug report arrives, we decide the topic(s) to which the report belongs. Then we utilize multi-feature to identify corresponding reports that have the same multi-feature (e.g., Component, product, priority and severity) with the new bug report. Thus, given a new bug report, we are able to recommend the most appropriate developer to fix each bug and predict its severity. To evaluate our approach, we not only measured the effectiveness of our study by using about 30,000 golden bug reports extracted from three open source projects (Eclipse, Mozilla, and Net beans), but also compared some related studies. The results show that our approach is likely to effectively recommend the appropriate developer to fix the given bug and predict its severity.