Post-Doc Fellow
Dr George Karystianis
BSc in Computer Science and Medicine (2008) - University of Central Greece
MSc in Biohealth Informatics (2009) - University of Manchester
Ph.D in Epidemiology and Text Mining (2014) - University of Manchester
Medicine & Health
School of Population Health
Dr George Karystianis has a multi-disciplinary background (BSc in Computer Science and Medicine; MSc in Health Informatics) with a PhD in Text Mining and Epidemiology from the University of Manchester in 2014. With fifteen years of experience in designing text mining systems and understanding complex data needs, he has been able to create applications that identify information from large scale datasets (e.g., police records, epidemiological studies). Recently he has been leading a project that involves text mining half a million of domestic violence police records to extract key characteristics that can be used for predictive analytics and surveillance purposes - a world's first study - in collaboration with the NSW Police Force (NSWPF) resulting in several high profile publications (JMIR, The Gerontologist) and three federal government reports (Australian Institute of Criminology); and in 2020 got awarded a MRFF grant to investigate domestic violence in NSW during the COVID'19 state lockdowns on a population scale.
Some of his text mining systems have been used by hospitals (The Christie NHS Foundation trust, Manchester, UK), are publicly available for research (DOSES) or have significantly impacted the way research has been conducted. In particular, he automatically identified biases in animal based research resulting in the alteration of existing research protocols (ARRIVE guidelines) and receiving coverage by Nature. His work on domestic violence police narratives has been incorporated by the NSWPF into their administration infrastructure (called the CHIMERA system) to improve domestic violence surveillance and reporting.
George has been leading multi-disciplinary teams in prestigious international text mining competitions to solve various challenges in text mining resulting in a consistent ranking within the top ten teams worldwide. He forged and to this day maintains strong national and international research relationships with prestigious institutions (University of Manchester, UK; National Institute of Health, USA; University of Liverpool, UK; University of Sydney; Black Dog Institute).
E-mail
Location
University of New South Wales, Samuels Building, 2nd Floor, Kensington
- Publications
- Media
- Grants
- Awards
- Research Activities
- Engagement
- Teaching and Supervision
Key grants
1. Center for Research Excellence, $2,500,000 (January 2023-December 2028)
CI; Violence Perpetration: Profiling, Prediction and Prevention.
2. NHMRC 2020 MRFF COVID-19 Mental Health Research, $232,000 (January 2021-August 2022).
PI; A novel text mining and data linkage approach to investigate the mental health needs of the population during the COVID-19 period.
3. NSW Police Force, $100,000. (April 2019 – June 2019)
CI; Co-development of a mobile app to monitor mental health illness in perpetrators of crime in NSW.
4. Australian Institute of Criminology, $78,000 (June 2017 – June 2018)
CI; Fill-in the gaps: Text mining domestic violence narratives from the NSW Police Force COPS system.
Other
1. Seed Funding grant, $26,000 (July 2021 – November 2021)
CI; The Shifting Expectations, Attitudes, and Stereotypes towards Older persoNs Study (SEASONS)
2. Bureau of Crime Statistics and Research, $27,123 (January 2021 – May 2021)
PI; Text mining domestic violence police narratives to identify coercive control behaviours
Best Student Paper, Fourth International Symposium in Languages on Biology and Medicine, Singapore, 2011
Medical Research Council, UK studentship for PhD (2009-2012)
Medical Research Council, UK studentship for MSc (2008-2009)
1. Domestic violence: Investigating domestic violence in the state of NSW during the COVID'19 lockdowns.
2. Domestic violence: Applying text mining in domestic violence police narratives to improve surveillance and create predictive models.
3. Domestic violence: Investigating male victims of domestic violence and their attributes (e.g., mental health status, injuries, abuse types) through police narratives.
4. Epi-criminology: Investigating the area of epi-criminology by automatically detecting studied themes, affiliations and information related to incarcerated populations.
5. Ageism: Recognizing ageist views in Twitter data.
6. Sex offences: Automatic recognition of behavioural patterns on sex offence police records
George's research conclusions in using text mining on domestic violence police narratives highlighted that important information remains hidden under vast amounts of text with NSWPF initiating an upgrade of their administrative system to include his devised methods (now called CHIMERA). These results compelled the Bureau of Crime Statistics and Research in NSW to collaborate with CI Karystianis and expand his text mining approach to capture coercive control behaviours from domestic violence police narratives and present his results in a NSW parliamentary inquiry on the criminalization of coercive control.
Due to his extensive knowledge, CI Karystianis was contracted by NSWPF to work on their domestic violence narratives for the identification of self-harm, suicidal and substance abuse related behaviours during the COVID-19 pandemic and he has written an article in the Conversation along with CI Butler to raise awareness for mental health issues within the area of domestic violence. He is an honorary fellow at the Institute of Human Rights, UNSW and at the Ageing Futures Institute due to the nature of his work.
My Research Supervision
1. Dr Sharon Reutens (Secondary Supervisor)
2. Mr Wilson Lukmanjaya (Primary supervisor; starting May 2023)