Institute for Data Engineering and Sciences

Specialisation in Information Systems

DOCIS-2021-01 Research on healthcare wearable data streams intelligent analysis and biocompatible communications in healthcare wearables

The elderly care solutions revolve around the home safety of the elderly, life care, abnormality monitoring, and timely treatment. Regarding medical treatment, health, security, and life enhancement for the elderly, wearable technology has stepped forward immensely into smart elderly care, being interactive with the elder in every aspect, and shaping the elder’s new lifestyle that technology brings. Wearable technology for healthcare can be defined as one that has intelligence and can give and receive input and provide meaningful output.  It provides a valuable monitoring tool for healthcare. For some time, it has brought us unique challenges, since wearable devices and electronics for healthcare present a unique interface of technology and person that needs to account both for technological and personal aspects of the problem. The following aspects may be considered as some of the challenges facing the field of healthcare wearables:

  1. The challenge of modern wearable electronics in health and wellness needs to be supported by extensions of sensing and data analytics capabilities. Thus, the fundamental research topic is how to efficiently and informatively interpret the data generated by wearable devices. In other words, the research topic focuses on the interpretation of such data streams and connect with health outcomes, using sensor data to guide behavioral interventions, and using medical data mining and machine learning techniques to obtain valuable insights into, and patterns of, elderly patient data for disease prediction, emergency detection and regulation of elderly health systems.
  2. The second research topic lies in biocompatible communications. Communications from the body to the outside world and communications among multiple wearables on the body demand new solutions, as traditional radio methods experience challenges due to absorption by body tissue.  The related challenges include the development of efficient methods for communicating through or on the body, including the organization of wearables in body sensor networks and their integration into the Internet of Things. We can explore the following issues: 1) Can a future wearable communicate with other smartphones and tablets? 2) Can we consider the elder’s body itself as the communication medium to exchange information with another wearable?

Supervisors: Professor Zhang Shuguang (, Professor George Du, Associate Professor Jacky Ho (, Shirley Weng In Siu (

Keywords: Healthcare wearable data streams: Intelligent analysis; Biocompatible communications; Healthcare wearables; Smart elderly healthcare

DOCIS-2021-02 Information and intelligence technologies-based smart learning

Information and intelligence technologies have become an irreversible force driving the transformation of teaching and learning practices. Cloud computing, learning analytics, big data, and artificial intelligence are being adopted in today’s teaching and learning at varying degrees.

  1. We all understand that the traditional system of higher education is often insufficiently effective; it must change its concepts of the educational process. There is a necessity to include in the system of higher education the key elements of smart technologies. In the last decade, innovations in higher education have emerged in teaching and learning practices at an ever-accelerating rate. The latest advances in pedagogies and technologies have brought new opportunities for the development of smart learning environments. At the same time, emerging information technology and innovations also have provided us with many valuable research topics which Ph.D. students can choose as their Ph.D. thesis research project. These research topics are listed as follows:
  2. Research on intelligent-driven pedagogical approaches: to answer how to integrate technologies into the curriculum in a smart learning environment to improve the effectiveness and efficacy of students’ learning.
  3. Research on personalized adaptive learning: to promote the development of personalized learning and adaptive learning for students, the research focuses on how to design smart learning  ecosystems aimed at realizing the integration of smart learning to personalize self-regulated learning, for example, to employ big data and data analysis technologies in monitoring learners’ differences and changes in individual characteristics, individual performance, personal development, and adaptive and adapting teaching strategies;
  4. Research on a comprehensive and dynamic learner model construction. Research topics cover: a) how to use multimedia, Internet, and agency technologies to enhance, enrich, and accelerate the learning process; b) how to open educational resources intelligent information technology and international norms, flexible learning that enhances learners’ ability to change behaviors; 3) how to use smart devices and social networks, for learners to develop a learning path for self-initiated creative learning, etc.
  5. Research on the integration of formal learning and informal learning to create an autonomous learning environment to support individual learners;
  6. Research on assessment methods.  Facing a smart learning surrounding, how design to use new methods to evaluate the effectiveness of the smart learning environment; for example, how accurately evaluate students’ learning performance in artificial intelligence smart classrooms through artificial intelligence technologies. Another example is how timely to carry out feedback and intervention with the help of learning analytics and specifically, with the application of machine learning and predictive modeling techniques, collection and analysis data, such as student course performance, learning behavior, previous academic history, learner characteristics, etc., to realize real-time prediction of a course; and computer-adaptive assessment and testing

Supervisors: Professor Keith Morrison (, Professor George Du (, Associate Professor Ana Maria Pires Correia (

Key Words: Intelligent-driven pedagogical approaches; Smart education, Smart learning environment; Smart learning ecosystems design, Artificial intelligence technologies-based assessment methods.


The development of technology-based applications to support elderly care is a vast area of research and with multiple perspectives, from creating new products to support the patient to smart systems to classify, recognize patterns, and cluster data from biological signals monitoring, including integration with patient’s clinical data.  The research may consider online and offline data processing and the generation of intelligent diagnostic support tools.

Supervisor: Alexandre Lobo (

Keywords: Elderly care; Artificial Intelligence; Machine Learning.

DOCIS-2021-04 Research on Social Media Sentiment Analysis

The rapid development of the Internet and mobile devices, especially the upsurge of Web 2.0 has enabled the emergence of social media platforms that led to the creation of the field of research called  ‘sentiment analysis’ to analyze this large-scale online resource of unstructured opinions. Sentiment analysis makes use of text analytics to capture public opinion. It is one of the most appealing use cases of Natural Language Processing (NLP), with interest from both industry and academia. 

Research nowadays embraces sentence level subjectivity detection, aspect-based sentiment analysis, sentiment analysis on figurative language, topic-based polarity classification, classification of the implicit polarity of events i.e. recognizing “pleasant” or “unpleasant” events when no explicit polarity marker is mentioned, emotion classification, and more recently, stance detection and argument mining, all on a wide variety of languages and diversified media. Task definitions have subsequently evolved towards more complex challenges where subjectivity, polarity recognition, and opinion mining have been enriched with fine-grained aspects and topic-level predictions.  

Existing research has produced numerous techniques for different tasks, including supervised and unsupervised methods. In recent years the success and popularity of deep learning in other domains has led to its use in sentiment analysis. For sentiment classification, Recurrent Neural Networks in particular Long- and Short-Term Memory networks, and their ability to capture long-distance dependencies have obtained state-of-the-art results in polarity classification.

However, despite the growing body of research and the amount of progress made in sentiment analysis, dealing with ‘affective phenomena’ in text such as subjectivity, opinions, emotions, mood, affect, attitude, and feelings have proven to be a complex, interdisciplinary problem that is far from being solved. That said, some inherent challenges should be understood when exploring the use of sentiment analysis. These challenges provide research fields for PhD students’ theses and research. The more challenging sentiment analysis tasks are listed as follows:

  1. Aspect Based Sentiment Analysis:   To capture the sentiments expressed on different aspects of entities such as products, movies, companies, etc. on user-generated comments.
  2. Emotion Analysis:  To detect and recognize types of feelings in texts, such as anger, disgust, fear, happiness, sadness, and surprise.
  3. Spam and Fake Detection: Fake reviews and fake news are closely related phenomena, as both consist of writing and spreading false information or beliefs. The biggest challenge here is the lack of an efficient way to tell the difference between a real review and a fake one.
  4. Multimodal Sentiment Analysis: With the proliferation of social multimedia, multi-modal sentiment analysis will bring new opportunities by integrating other, complementary data streams such as facial and vocal displays and expressions that express sentiment often in a very powerful way.

Supervisor: Professor George Du (

Keywords: sentiment analysis; deep machine learning; opining mining; sentiment classification, aspect-based sentiment analysis; natural language processing.

DOCIS-2021-04 Research on Data Science-based Smart Macau Tourism Destination

Tourism today is characterized by tourists being heavy users of social networks, producing a huge amount of data, an internet trail, or a digital footprint. Both companies and the tourism sector frequently use those data to create, explore and exploit valuable information. Smart tourism has attracted the minds of tourism researchers, developers, administrators, and providers.

Smart tourism is a new movement to boost the development of tourism through current emerged technology, such as Machine learning, Deep learning, Big Data, the Internet of Things (IoT), RFID, and NFC, etc. The development of smart tourism aims to improve information management and governance, promote the innovation of tourism services and products, improve the traveler tour experience, and, eventually, enhance and keep the competitiveness edge of tourism firms. The process of Smart tourism is aided by integrated efforts at a destination to collect and aggregate data from physical infrastructure, social connections, governments, businesses, and other organizations and humans and to transform this data with advanced analytical tools into meaningful experiences and business value-propositions.

The terms smart tourism and smart tourism destination are becoming ever more pervasive and they, therefore,  have received great academic attention in recent years and an increasing number of destinations and hospitality establishments are investing in smart tourism initiatives. Smart tourism would be not only a technology and data issue but also a management and governance issue of organizations and cities.

However, while governments and businesses around the world are aggressively pushing the smart tourism agenda forward, smart tourism research initiatives are still in their infancy and seem to not fully cover the whole spectrum of smart tourism-related issues and questions. For example, it is a general academic agreement that tourism companies in Macau often lag in adopting technological innovations. For tourism management authority, it is needed to develop to a level that benefits certain stakeholders such as the residents. Thus, several valuable research topics need to be resolved.

Topic 1 How does the tourism company develop the smart of smart technological utilities as an integrated part of the tourist experience?

Topic 2 How innovations should be integrated and correlated with smart tourism destination public agents?

Topic 3 How the social networks, smartphones, and IoT affect on the development of Smart tourism?

Topic 4 How deal with these issues of information governance, data privacy, internet security, and extreme technology dependence?

Topic 5 How to employ the tourist social media sentiment analysis methods to improve service quality and precision marketing?

Supervisors: Professor George Du (, Professor Jenny Phillips (

Keywords: Smart tourism; Smart tourism destination; Smart technological utilities, tourism Data analysis, Tourists sentiment analysis, Information governance: Data privacy; Precision marketing

DOCIS-2021-06 Study of Data Science-based Environmental Sciences

Data science is not only having a profound impact on several areas including commerce, health, and smart cities but also offering a rich tapestry of new techniques to support both a deeper understanding of the natural environment in all its complexities, as well as the development of well-founded mitigation and adaptation strategies in the face of climate change. Data science, unquestionably,  for the natural environment brings data science researchers about new challenges for data science, particularly around complexity, spatial and temporal reasoning, and managing uncertainty. We are experiencing studying the natural environment are increasingly data-rich with a pressing need for new techniques to make sense of the accelerating amount of data being captured about environmental facets and processes. Therefore, the potential for environmental data science is enormous, and indeed understanding, and managing the impact of, environmental change is a grand challenge for the emerging subject of data science. the data science-based environmental sciences research approach is distinctive with a set of challenges.

Topic 1. Study of underlying data sources and types of data

The data used for based environmental sciences is entirely different from the term “big data” in the data science field. The most striking factor in environmental sciences is the heterogeneity of the underlying data sources and types of data. Environmental data comes from a wide variety of sources and this is increasing rapidly with innovations in data capture. There is a question of how to manage the variety and heterogeneity in underlying sources of data, including achieving interoperability across data sets to seek a wide variety of natural phenomena.

Topic 2. Study of multi-dimensional modeling techniques

Modeling enables us to make sense of the data through data-driven model monitoring techniques. Modeling also is the principal tool for understanding the environment and forecasting or projecting environmental change. Monitoring techniques in data mining and data analysis are different from those in the environmental science field where models are often heavily parameterized or a range of possible ensemble models. Different modeling techniques bring data science scholars a big challenge.

Topic 3 Study of environment Complexity

The environment of our life with economic and social concerns is a complex system. Dealing with this inherent complexity is a major challenge for data science. For example, data science also offering interesting perspectives on how to handle complexity, for example, how to employ machine learning in dealing with and responding to emergent phenomena and in dealing with surprises in complex systems.

Topic 4 Study of spatial-temporal data

Research on the environment is often related to reasoning about natural phenomena across space and time. Reasoning about natural phenomena across space and time lies in estimating spatial or temporal patterns in data and deciphering the effects of covariates across the domain of interest. This has led to a deluge of spatially and temporally referenced data. Traditional methodological approaches are not well-suited for handling large spatial-temporal data sets. There is therefore an increasing challenge to develop approaches that can handle large Spatio-temporal data sets and to estimate the fine-scale structure within.

Supervisors: Professor George Du (, Professor David Gonçalves (, Associate Professor Raquel Vasconcelos (

Keywords: Data science; Spatial-temporal data analysis; Environment Complexity; Multi-dimensional modeling techniques; Data Science-based environmental analysis.


 A variety of technologies are available today to convert brain activity into signals or images. Electroencephalograms are the best cost-benefit solution, although the technology still suffers several challenges during data collection. Other technologies such as EMG or fMRI are quite mature, but the costs are usually prohibitive for non-medical applications. A potential solution is the brain fNIRS imaging monitoring systems, especially with lower-cost solutions presented by the industry. The proposed research topic aims to work on several challenges that can be found when developing BCI – Brain-Computer Interface systems, not only from the hardware perspective, including sensors and equipment, but also from the software side, including data collection and privacy, security, and processing. Artificial Intelligence methods and nonlinear signal processing tools will be strongly considered for the data analysis step.

 Supervisor: Alexandre Lobo (

Keywords: Brain-Computer Interfaces; Neuromarketing; Artificial Intelligence


The generation of a large amount of healthcare data for elderly care is a common evolution of smart systems, including HIoT solutions. The development of data analytics solutions for visualization, classification, decision support, and forecasting are key areas of research that can be covered here. Big data requirements and related processes can also be an interesting possibility of research given the fact that healthcare units are not technology-oriented and existing systems are not ready to follow these requirements.

Supervisor: Alexandre Lobo (

Keywords: Health Analytics; Big Data; Cloud computing


The main objective of this area is to develop applied research on the adoption of automated solutions and integrated networks of IoT (Internet of Things) devices to create intelligent and sustainable facilities, including buildings and open areas. Environmental and energy sensors and actuators will be focused on the studies, considering the use of Long-Range communication technologies, such as LORAWAN and other open technologies.

Supervisor: Alexandre Lobo (

Keywords: Sustainable Facilities; Smart Facilities; IoT; LoraWAN


The integration of the concept of the Internet of Things (IoT) and the area of Healthcare has been the object of extensive research recently. A significant area of research in developing new applications to make use of these data is wide open for new products. This research area includes computer-based and mobile applications, including biosensors networks and the integration of multiple parallel sensors.

Supervisor: Alexandre Lobo (

Keywords: HIoT (Healthcare Internet of Things); Mobile Development; Sensors networks; Cloud technologies.

Supervisor: Gerald Estadieu (

 Keywords: Geographical Information Systems; Artificial Intelligence; Smart Cities


Mental health problems are common in older people and can include anxiety and dementia. Many older people also suffer from sleep and behavioral problems, cognitive decline, and confusion. Often, patients cannot recognize the problem themselves, resulting in the condition being discovered very late. The term “digital phenotype” refers to the use of data from personal digital devices to quantify social, physical, cognitive, emotional, and behavioral phenotypes at the individual level. It is a powerful tool to better understand patients for scientific or clinical purposes. This project will explore the use of smartphones and other wearable devices (such as smartwatches) to monitor and diagnose users’ mental health. Multimodal data will be collected from various sources (sensors, GPS, cameras, text, and speech, etc.) and their relationship to the user’s psychological, emotional, and behavioral conditions will be studied. In this project, statistical, machine learning, deep learning algorithms will be developed to facilitate disease detection, monitoring, and clinical support of user mental health. The goal of this project is to enable accurate and early detection of mental health problems, especially in seniors, to provide better medical support and improve their quality of life.

Supervisor: Shirley Weng In Siu (

Keywords: Mental Health Monitoring: Wearables; Digital Phenotyping; Elderly Care


The development of healthcare data analytics systems has the potential to transform the current state of the art. Nevertheless, it also brings new challenges and key research areas to be considered. The lack of data governance and data handling processes is a key challenge to attend the CIA security triad Confidentiality, Integrity, and Availability. Several research applications can be considered in this topic, such as new approaches for real-time data encryption, the adoption of blockchain technologies for data integrity, and strategies for data protection.

Supervisor: Alexandre Lobo (

Keywords: Data Protection; Blockchain; Encryption.


 Image processing is one of the most promising areas in Information Systems and Data Sciences. The research possibilities are significant, from intelligent computational vision applications to medical imaging, including healthcare applications, while the variety of techniques and approaches is increasing daily. The trends are related to image acquisition and high-quality image processing; 2D capture evolving to 3D and AI supplemental systems for image analysis. Deep Learning algorithms and models are a core area of research. Supervisors:

Shirley Weng In Siu (

Keywords: Image Processing; Medical Imaging; Artificial Intelligence; Deep Learning


The rationale for this project is two-fold. On the one hand, we propose thoroughly describing the Soundscape in a territory rich in diversity and has huge potential for citizen participation. This will include efforts in noise mapping, sound mapping, soundwalks, grounded theory efforts on rich descriptions of the environment, and the use of alternative objective metrics. With these components, we intend to provide an original and substantial insight into the qualitative dimensions of soundscapes, attending to a wide variety of geographical, and cultural facets, following a multimodal sensing strategy. On the other hand, we intend to leverage the richness of the gathered data by developing artificial intelligence algorithms to autonomously assess and predict the evaluation of a given Soundscape based on recordings alone.

Supervisor: Gerald Estadieu (

Keywords: Soundscape; Artificial Intelligence; Multimodal Mapping; Sound Design


The number of Geographical Information Systems applications is significantly growing, with several industries becoming increasingly dependent on them. Areas such as decision support, market segment targeting, distribution network planning, and emergency responses. Traditional models were focused on storage, management, and analysis of data referenced spatially to the Earth. The new research trends are to use geospatial technology as a tool for organizations to collate different maps and remote sensing data and generate location-specific business models. Geographic prediction models are also a core area of study.


New technologies (e.g., electric autonomous vehicles, internet of things ‘IoT’ products, smart home systems, wearable devices, blockchain applications, etc.) are reshaping everyday life. But why are some technologies spreading faster than others? Some innovations have struggled to gain acceptance in society even if they are arguably better than the current ones. Diffusion of Innovations (Rogers, 1962), Technology Acceptance Model (TAM; Davis, 1986), and Unified Theory of Acceptance and Use of Technology (UTAUT; Venkatesh et al., 2003) are well-accepted frameworks that have attempted to explain why individuals accept and use innovations, however many areas of research are still lacking. By using established models as the basis for the study, this broad-area project proposes to research the extent to which individuals or organizations accept and use emerging technologies. The project is quantitative and uses statistical tools (i.e., structural equation modeling) to analyze the data.

Supervisor: Alessandro Lampo (

Keywords: Technology Acceptance; Diffusion of Innovation; Mathematical Modelling

Last Updated: December 16, 2022 at 4:00 pm