Associate Professor Richi Nayak

Find Richi Nayak on

Associate Professor, School of Computer Science

PhD (Queensland University of Technology)

Research theme: Information

Research discipline: Data Science

She is the Applied Data Science Program Leader of the University Centre for Data Science (CDS). She is an internationally recognised expert in data mining, text mining and web intelligence. She has combined knowledge in these areas very successfully with diverse disciplines such as Social Science, Science, and Engineering for technology transfer to real-world problems to change their practices and methodologies. Her particular research interests are machine learning and in recent years she has concentrated her work on text mining, personalization, automation, and social network analysis. She has published high-quality conference and journal articles and highly cited in her research field. She has received a number of awards and nominations for teaching, research and service activities.

Research areas Text Mining for data organization and understanding With the advancements in computing resources and digitalization, an increasing amount of data is generated in text format. In this research stream, my research group is engaged in developing sophisticated and novel Text Clustering and Information Extraction methods based on the concepts of ranking-centered, hubs, density-based and matrix/tensor factorization. These innovative methods are suitable for big and complex data to provide accurate and scalable solutions. We have applied these methods to several applications such as robotic marketer, social media mining, community discovery, information harvesting, robot navigation, trend detection, concept mining, abuse detection, spam review detection, recommendation, and personalization.

Applications of data mining and machine learning into solving real-world problems Developing real-time data mining systems by utilising the hidden patterns and rules behind the complex sets of data sets such as School education; Energy Bills; Intestine Bowel Disease; Road environmental and accidents data set; Anaesthetic time series data set; Active aging survey dataset and Structural health monitoring. In this research stream, my research group is engaged in developing machine learning and data mining algorithms and systems that can be deployed in practice by various industries and used in data-driven intelligent decision making. In an industry-funded project, we are using these techniques to extract useful information and build an information repository so the client can quiz the related information easily.

Algorithms for Automation, Personalisation and Pattern Mining With the Internet of Things and Digital Twinning, algorithms are required to mine patterns and trends from the complex data and develop applications based on these patterns and trends. In this project stream, we have developed algorithms to understand the spatiotemporal context for anomaly detection and location-based navaigation recommendations. We have also developed Deep learning algorithms to automatically generate marketing reports based on past reports and extensive online information scrapped from the related data sources.

Additional information

Her research expertise spans multiple domains. She is actively engaged in and leading transdisciplinary research. Some of the past projects that she led are listed below with their impact on ICT industries, government and communities.

Applications of data mining into solving real-world problems– developing real-time data mining systems by utilising the hidden patterns and rules behind the complex sets of data sets such as Road environmental and accidents data set; Anaesthetic time series data set; Active ageing survey dataset and Structural health monitoring. o Her research conducted with Queensland Dept of Transport and Main Roads (QDTMR) has developed the followings:

  • Risk-based decision support modelling for managing skid-resistance
  • Decision Support modelling based on data mining for skid resistance
  • Pavement Deterioration models with Data Mining

These projects have contributed new knowledge to Main Road domain and have influenced decision making processes at QDTMR. o Her research conducted with Queensland Dept of Public Works (QDPWD) resulted  into  a  computer  software  package  that  interactively  supports building designers, owners and maintainers by predicting service life outcomes for a range of metal building components, in different locations and susceptible to corrosions from a variety of sources. o Her research conducted with Anaesthetists at Royal Brisbane Hospital  (RBH) resulted  into  identifying  patterns  for  monitoring  and  assessing  patient’s health   during   surgery,   and   helping   anaesthetists   monitor   anaesthetic medicine admission with the aim to control cardiac arrest during surgery.

Personalisation  of  Web  Services  using  Link  and  Network  Mining  and  Product Recommendation – this research focused on the development of techniques and tools to better identify customer behaviour and need, and to apply these tools and techniques directly to improve products and services. It developed novel data mining techniques for user profiling and market segmentation, and also develops recommendations and information filtering techniques for improving the relevance of content in services such as personalised search and matching. o She has developed a number of innovative recommendation algorithms that have been trialed by a leading Australian Dating Network and Car seller website. She has developed a number of innovative algorithms that take the structural and semantic features embedded in data into consideration to identify patterns in the data and utilize them in decision-making.

Clustering algorithms: She has significantly contributed to the field of big data by focusing on the data problem of variety in which data appears in multiple formats, and the data analytics algorithms should take advantage of this added information.

  • She has proposed several robust, accurate and efficient multi-view clustering algorithms that can combine several types of features simultaneously in the clustering process.
  • Recently, she has proposed the novel concept of ranking and indexing infrastructure used in information retrieval in designing the clustering methods. In the MediaEval forum, one of these methods outperformed the other participants in the task of event detection. These algorithms are fast and efficient.
  • One of the techniques that she developed introduced a novel level structure; using this structure, she proposed an incremental clustering algorithm that could group a very large collection within a very short time. This work was ranked in the top 5 papers of the leading international conference – The 10th Pacific-Asia Conference on Knowledge Discovery and Data Mining (PAKDD 2006).

Tensor Space Modelling and Analysis: She has very successfully developed a number of methods based on Tensor Space (multi-dimensional) models.

  • Proposed a number of clustering methods that innovatively use Tensor Space model to include the structural and content features effectively and efficiently to find subgroups.
  • Proposed a number of learning-to-rank algorithms that are able to provide high-quality recommendations in social media systems.

She has successfully developed data mining solutions to real-world applications. In doing so, she proposed a number of customized data mining algorithms. She has developed methods for a number of application domains such as social networks, social media, information retrieval, Web services, e-commerce, and m-commerce. Some of them are:

  • A constrained clustering algorithm to deal with the nature and scale of the social networks’ users, utilising both the implicit and explicit information on these networks.
  • A bi-partite graph mining technique and a co-clustering method that are suitable for grouping the collections that involve two different types of entities or objects. This method takes advantage of both types of information and uses them efficiently in clustering.
Type
Editorial Role for an Academic Journal
Reference year
2020
Details
The Journal of Data Mining & Digital Humanities is concerned with the intersection of computing and the disciplines of the humanities, with tools provided by computing such as data visualization, information retrieval, statistics, text mining by publishing scholarly work beyond the traditional humanities.
Type
Committee Role/Editor or Chair of an Academic Conference
Reference year
2018
Details
Steering Committee Member of the Australasian Data Mining Conference - the only leading association in this region - since 2012. General Co-chair of the International Conference on Data Science, Intelligent Computing and Cyber Security (ICDIC 2020) and the Australasian Conference on Data Mining (AusDM 2020, 2015, 2014). Program Committee Member of prestigious conferences, e.g., KDD 2016-19, CIKM 2018-19, WSDM 2017; PAKDD 07-19, AusDM12-08, ICDM14, ACM SAC 2014-20, EDBT11, AI20-09 and many others. She was program chair of the 2014 AusDM (Australian conference on Data Mining) to be held in Brisbane, December 2014. She was publicity chair of the 2013 ACM SIGIR to be held in GodCoast, July 2014. She was track chair of Database and Data Mining in the 3rd International Conference on Computer Science and its Applications 2011 (CSA-11). She was local chair of the 2012 joint WCCI World Conference on Computational Intelligence. She was local chair of the 2013 PAKDD 13.
Type
Advisor/Consultant for Community
Reference year
2018
Details
2018-2021: Board of Studies and Advisory Committee: International expert member, SNS College of Technology, Anna University, Chennai, India. 2016-2018: Board member of the Ascot Primary State School Council. Advised the school on their data management, analytics, and digitalization practices.
Type
Editorial Role for an Academic Journal
Reference year
2017
Details
Founder and Editor-in-Chief of the International Journal of Knowledge and Web Intelligence (2008-10) and now serve on the Editorial Board (2011-). Associate Editor of International Journal of Knowledge-Based and Intelligent Engineering Systems (2010-14) and International Journal of Data Mining, Modelling and Management (2015-18). Editorial advisory reviewer board member of the International Journal of Knowledge-Based & Intelligent Engineering Systems (KES).
Type
Academic Honours, Prestigious Awards or Prizes
Reference year
2016
Details
Dr Nayak was the recipient of the 2016 WiT Infotech Outstanding Achievement Award for her exemplary services to the field of Data Analytics in the Queensland state and outside. Dr Nayak was the ICT ambassador appointed by the WiT on 2017.
Title
Profit and Loss: The Commercial Trade in Indigenous Human Remains
Primary fund type
CAT 1 - Australian Competitive Grant
Project ID
DP200101814
Start year
2020
Keywords
Title
Improving the ability of the Australian cotton industry to report its sustainability performance
Primary fund type
CAT 1 - Australian Competitive Grant
Project ID
QUT1705
Start year
2016
Keywords
Agriculture; Agroecosystem Health; Natural Resource Management
Title
Human Cues for Robot Navigation
Primary fund type
CAT 1 - Australian Competitive Grant
Project ID
DP140103216
Start year
2014
Keywords
Autonomous Robots; Mapping and Navigation; Spatial Cognition
Title
The Neglected Dimension Of Community Liveability: Impact On Social Connectedness And Active Ageing
Primary fund type
CAT 1 - Australian Competitive Grant
Project ID
LP0883447
Start year
2009
Keywords
Community Liveability; Social Engagement; Community Well Being; Social Isolation; Population Ageing
  • Deep Learning for Longitudinal Classification: Applying Incremental Learning to Longitudinal Classification for Early Prediction Problems
    PhD, Principal Supervisor
    Other supervisors: Professor Yuefeng Li
  • A Context-aware Spatial-temporal Anomaly Detection Framework for Data Streams
    PhD, Principal Supervisor
    Other supervisors: Associate Professor Shlomo Geva
  • Mining Ontology from Text Document for Question Answering
    PhD, Associate Supervisor
    Other supervisors: Professor Yuefeng Li
  • Transformation of Lithium Mining Waste Material into Zeolites using Machine Learning Principles
    PhD, Associate Supervisor
    Other supervisors: Professor Graeme Millar, Dr John Outram
  • Research on big data for public electricity utility's smart meterings
    Professional Doctorate, Principal Supervisor
    Other supervisors: Dr Maolin Tang
  • Dynamic Demand Estimation for Online Simulation of Large-Scale Urban Network
    PhD, Associate Supervisor
    Other supervisors: Associate Professor Ashish Bhaskar
  • Controlling data mining driven risk profiles and applying them as triggers in service delivery in complex data environments.
    PhD, Principal Supervisor
    Other supervisors: Professor Alistair Barros
  • Spatial multimodal sentiment analysis to establish relationship between consumer preferences and farming practices in local agri-food networks
    PhD, Mentoring Supervisor
    Other supervisors: Dr Alan Woodley
  • Tensor Modelling for Fine-grained Type Entity Inference in Knowledge Graphs
    MPhil, Principal Supervisor
    Other supervisors: Associate Professor Yue Xu