Source author record

Guangyuan Hu

Guangyuan Hu appears in the imported research catalog. Authorship, coauthor and topic links are available while profile ownership is still unclaimed.

ResearcherUnclaimed source record

Digital Libraries Cryptography and Security eess.SP

Catalog footprint

What is connected

3works

3topics

4close collaborators

Actions

Connect this record

Open graph Browse works

Inspect adjacent papers, topics, institutions and collaborators without losing the researcher page.

Building this map preview

BZPEER is loading the nearby papers, people, topics and institutions for this page.

preprint2020arXiv

Funding information in Web of Science: An updated overview

Despite the limitations of funding acknowledgment (FA) data in Web of Science (WoS), studies using FA information have increased rapidly over the last several years. Considering this WoS'recent practice of updating funding data, this paper further investigates the characteristics and distribution of FA data in four WoS journal citation indexes. The research reveals that FA information coverage variances persist cross all four citation indexes by time coverage, language and document type. Our evidence suggests an improvement in FA information collection in humanity and social science research. Departing from previous studies, we argue that FA text (FT) alone no longer seems an appropriate field to retrieve and analyze funding information, since a substantial number of documents only report funding agency or grant number information in respective fields. Articles written in Chinese have a higher FA presence rate than other non-English WoS publications. This updated study concludes with a discussion of new findings and practical guidance for the future retrieval and analysis of funded research.

preprint2020arXiv

Smartphone Impostor Detection with Built-in Sensors and Deep Learning

In this paper, we show that sensor-based impostor detection with deep learning can achieve excellent impostor detection accuracy at lower hardware cost compared to past work on sensor-based user authentication (the inverse problem) which used more conventional machine learning algorithms. While these methods use other smartphone users' sensor data to build the (user, non-user) classification models, we go further to show that using only the legitimate user's sensor data can still achieve very good accuracy while preserving the privacy of the user's sensor data (behavioral biometrics). For this use case, a key contribution is showing that the detection accuracy of a Recurrent Neural Network (RNN) deep learning model can be significantly improved by comparing prediction error distributions. This requires generating and comparing empirical probability distributions, which we show in an efficient hardware design. Another novel contribution is in the design of SID (Smartphone impostor Detection), a minimalist hardware accelerator that can be integrated into future smartphones for efficient impostor detection for different scenarios. Our SID module can implement many common Machine Learning and Deep Learning algorithms. SID is also scalable in parallelism and performance and easy to program. We show an FPGA prototype of SID, which can provide more than enough performance for real-time impostor detection, with very low hardware complexity and power consumption (one to two orders of magnitude less than related performance-oriented FPGA accelerators). We also show that the FPGA implementation of SID consumes 64.41X less energy than an implementation using the CPU with a GPU.

preprint2016arXiv

Funding acknowledgment analysis:Queries and Caveats

Thomson Reuters' Web of Science (WoS) began systematically collecting acknowledgment information in August 2008. Since then, bibliometric analysis of funding acknowledgment (FA) has been growing and has aroused intense interest and attention from both academia and policy makers. Examining the distribution of FA by citation index database, by language, and by acknowledgment type, we noted coverage limitations and potential biases in each analysis. We argue that in spite of its great value, bibliometric analysis of FA should be used with caution.