Presented by:
Prof Jonathan ZHU
Chair Professor, Department of Media and Communication;
Chair Professor, School of Data Science, CityU

Project Introduction:
The ongoing COVID-19 pandemic has once again demonstrated the complexity, vulnerability, and unpredictability of the contemporary international relations. Of various research traditions on global studies, network analysis has been an increasingly popular approach to help uncover hidden, indirect, and multilateral dynamics underlying the global structure. In the current study, we apply ego-network analysis, a set of useful but less known tools, to explore the ongoing reconfiguration of the global politico-economic ecosystems based on crossnational flow of goods, capital and information. While our findings are preliminary, the ego-network approach appears to be able to offer a variety of unexpected insights with important policy implications.

Presented by:
Prof Hong YAN
Chair Professor, Department of Electrical Engineering, CityU

Project Introduction:
This project started in April this year. We aim to investigate big data processing techniques for performance analysis, prediction and control of smart factories. We will study effective methods for manufacturing data filtering and imputation to improve the data quality. Statistical models, machine learning algorithms and optimization techniques will be proposed for data classification and analysis. Computer software will be developed for automated and intelligent control of smart factories.

Presented by:
Dr Yuner ZHU
Postdoc, Department of Public Policy, CityU

Project Introduction:
Digital communication technologies have played a critical role in modern-day protests around the globe in the last decade. Labeled as “crowd-enabled connective actions,” the mobilizing structure of networked social movements may differ fundamentally from professional social movements led by movement organizations or political parties. These networked social movements can resolve free-rider problems of collective action by providing information networks and frame alignment to its participants, which undermines the need for formal leadership.

The Hong Kong protests have operated across diverse traditional media and social media platforms (a Reddit-like forum, Telegram channels, Instagram graphics, and Facebook pages) in a cosmopolitan city with an extremely high level of social media and smartphone penetration rate. This setting is ideal for big data analysis. In this presentation, we provided a preliminary topic and network analysis of 20 million comments over 44 weeks from 2019 to 2020 on LIHKG.com. We analyze all posts with those in the public affairs forum to understand the relationship between different topics and drill down to unpack post volumes, trend, pattern, centrality rank, degree distribution, and core-periphery comments. We identify that the pattern of user activity is highly correlated to offline events. We then reveal how specific topics and social networks form frame alignment. We also identified a number of hyper-active users who contributed frequently to the discussions related to the protests and found that over time a few dozen user accounts gained in-group and out-group recognition in different communities. In other words, while there were no authoritative individuals or organizations in mobilizing and framing the protests, a "networked and leaderful structure" helped distribute the role of information dissemination and frame coordination. In conclusion, we outline the value of big data analysis of social phenomena and point to the challenges it presents for social science research. 

Presented by:
Dr Ka Chun WONG
Associate Professor, Department of Computer Science, CityU

Project Introduction:
The early detection of cancers has the potential to save many lives. A recent attempt has been demonstrated successful. However, we note several critical limitations. Given the central importance and broad impact of early cancer detection, we aspire to address those limitations. We explore different supervised learning approaches for multiple cancer type detection and observe significant improvements; for instance, one of our approaches (i.e., CancerA1DE) can double the existing sensitivity from 38% to 77% for the earliest cancer detection (i.e., Stage I) at the 99% specificity level. For Stage II, it can even reach up to about 90% across multiple cancer types. In addition, CancerA1DE can also double the existing sensitivity from 30% to 70% for detecting breast cancers at the 99% specificity level. Data and model analysis are conducted to reveal the underlying reasons. A website is built at http://cancer.cs.cityu.edu.hk/

Presented by:
Dr Yanni SUN
Associate Professor, Department of Electrical Engineering, CityU

Project Introduction:
Microbial communities, which contain microbes sharing the same living space, ubiquitously exist anywhere from various human body sites to different environmental niches. Elucidating the composition and functional profiles of the microbiome (i.e. the ensemble of all the microbes in a niche) is essential in many fields including pathogen discovery, ecology, epidemiology etc. Applying modern sequencing technologies to sequence all the genetic materials in various samples has become the most popular method for studying microbial communities. Massive amount of sequencing data has been accumulated, posing significant computational challenges for converting the data into knowledge. In this short talk, we will share the ongoing researches about this topic at CityU.

Presented by:
Dr John LEE
Associate Professor, Department of Linguistics and Translation, CityU

Project Introduction:
Text readability assessment aims to analyze the difficulty of a document, and predict the school grade for which the document is most suitable. Classic approaches rely largely on surface cues, such as word frequency and sentence length. Syntactic and semantic complexity, however, also have significant impact on text difficulty. In this talk, we present a level-annotated corpus of Chinese textbook materials, and report preliminary results of a readability assessment system that takes not only lexical, but also syntactic and semantic complexity into account.