Robert Turnbull

Building on a previous research project (Histoire des Traductions en Langue Française, XVe- XVIe siècles, ANR project - equivalent to ARC - finished in 2015), chapters published (Duché, 2015, 2016, 2017) and papers given at international conferences (RSA, SRS, 2018), this project will further investigate Early modern Women translators in Europe. It seeks to promote cultural awareness by charting the contours and transnational connections of early modern women translators through the use of a sophisticated, interactive visualisation tool. In particular, the project aims to:

1. identify the early modern women translators (not only in France, but also in other European countries);
2. map the intellectual as well as personal and cultural networks to which these women belong (author, publisher, reader etc.);
3. create a repository for metadata on early-modern scholarship.

In 2020, I worked as the Project Co-ordinator of the Living with COVID social media campaign. This campaign, led by Prof Rachel Fensham, communicated reliable information about the pandemic to young people. It comprised six brief videos featuring a secondary-school student who learns to ‘deal with’ a character named COVID. Each episode was based on a theme with a companion article on the campaign’s website, certified by an expert at the University. As project coordinator, I arranged meetings with academics and external partners, arranged for the spending of the $45,000 funding we received, helped generate and review the content of the episodes and articles, liaised with the University’s branding, social media, and COVID response teams and met with representatives from the Victorian Education Department and the Department of Premier and Cabinet. The campaign was very successful and generated the most-viewed post in 2020 on the University’s new TikTok channel. The Dean of the Arts Faculty awarded the project the ‘Most Significant Contribution to Building Partnerships’ in 2020.

See the six episodes in this playlist:

Microbial eukaryotes (protists) play key roles in global element cycles, ecosystem functioning, food production and disease. There are two major impediments to our understanding of their evolution and genome diversity. First, the majority of microbial eukaryotes have not been discovered yet. Second, large molecular datasets are needed to resolve their ancient evolutionary relationships, but genomic data is very sparse. The uncertainty created by these problems are hindering clear communication and have led to a sense of scientific paralysis in many branches of protist biology. With this project, we aim to break through this paralysis using big data. Specifically, I aim to use publicly available metagenomes (data obtained from shotgun sequencing of total environmental DNA) to discover new lineages and place them in an evolutionary context. Metagenomes are typically generated to answer a specific ecological question, but the raw data are a treasure trove of information for different purposes. In bacteriology, which was facing similar problems, bacterial genomes assembled from metagenome data have shaken our knowledge to its core. It has resulted in hundreds of entirely new lineages and genome-scale phylogenies providing unprecedented insight into bacterial evolution and comparative genomics. I wish to unlock the potential of metagenomes to target the organelle genomes (mitochondria & chloroplasts) of microbial eukaryotes. These genomes are relatively easy targets (small, high coverage, recognisable gene set) but will require a specific set of criteria to be extracted from metagenome data and accurately binned into species-specific sets of contigs. That is what this project sets out to achieve.

The intended outcomes are:

1. A working pipeline for automated download of metagenome reads, their assembly and calculation of contig statistics
2. A large library of assembled metagenomes generated with this pipeline, permitting testing of alternative strategies for organelle genome binning
3. Advanced proof of concept of the method for project proposals.

Since their discovery in the 1930s, the mysterious collections of giant stone jars scattered throughout central Laos have remained one of the great prehistoric puzzles of south-east (SE) Asia. It is thought that the jars represent the mortuary remains of an extensive and powerful Iron Age culture. This project seeks to determine the true nature of these sites, which date to a dynamic period of increasing complexity in SE Asia (c.500BCE–500CE). The project entails extensive reconnaissance, precision mapping, archaeological excavation and analysis of associated burial material. Using a suite of cutting-edge archaeological technologies, it is expected to have far-reaching benefits for archaeology, science, Laos and World Heritage.

This project seeks to visualise the movements and migrations of architects within and across the British Empire from the mid-nineteenth to the mid-twentieth century. At the heart of this project are the questions: ‘Where did architects come from and go to?’ and ‘Where did they work?’

To date, architectural histories have largely been grounded in a single place, disregarding architects’ movements across jurisdictions. But architects have long been highly mobile professionals. Their careers, then as now, could span the globe. Nor were their journeys simply a trip from centre to periphery. Indeed, a good number of colonial architects had careers which spanned Australasia, East and Southeast Asia, Africa, and the United Kingdom. And in these places, they could work for multiple entities – themselves, other firms, public agencies – making their careers complex journeys.

The project uses disparate, primarily textual, sources – trade directories, newspapers, and other archival material – to trace the movements of hundreds of architects through various architectural firms, and through the colonies and concessions of the British Empire over the course of a century.

By using database and visualisation tools, the project will enable new ways of understanding architectural history, new methods for synthesis and analysis of large and disparate data sets in design histories and new interfaces for the presentation of complex historical datasets that involve different geographical locations and movements, over long time frames in varying professional configurations.

Bioeconomic simulation models are a critical tool used to estimate both the impacts that might be caused by pests or diseases and the relative value of the various interventions that we deploy to manage them.

Typically, these models describe the impacts of a single species on a single asset at relatively small spatial and temporal scales. Management agencies, however, are required to take an ‘all hazards’ approach when determining how they might best protect assets from the negative impacts of pests and disease at state and national scales.

The ‘value model,’ developed at the University of Melbourne, is the only model globally that is capable of simultaneously modelling the arrival, spread and impact of multiple biological hazards on multiple assets over time. However, in order to be effectively deployed within management agencies, or expanded for additional research use, the core architecture of the model needs to be faster.

This project will explore a range of options for improving the performance of the model spanning: how the dispersal of organisms is modelled, how the model is encoded, how compute resources are utilised and how the result data is stored. Improvements in any (or all) of these areas will enable uptake of the model into real-world decision-making contexts ultimately delivering improved biosecurity outcomes for society.

Professor Thieberger is working on several Text Encoding Projects and would benefit from extending work begun with Robert Turnbull and Daniel Russo-Batterham. The pages at https://bates.org.au are an example, and he now has another project (with Simon Musgrave at Monash) working to prepare Heath’s grammar, texts, dictionary, and media of the Aboriginal language Nunggubuyu and put it online.

In SCIP/Digital Studio he is working with Birgit Lang (German) on a manuscript that she has transcribed, a diary of a Freudian analyst that has not been published before. With a DS intern the transcript has been encoded as a first draft TEI document and it is now ready for final editing. However, each of these projects needs a IIIF server for the images, and some assistance with the TEI, that Robert and Daniel can provide.

The proposed research intends to analyse trends in building construction projects in the State of Victoria from 1996 to 2019. The researchers have sourced historical building permit data from the Victorian Building Authority (VBA), which includes a broad range of information relating to over 1.4 million building projects issued with building permits between 1996 and 2019 (in Excel format).

This includes information on project location, cost, type, builder, architect, materials used etc.. In collaboration with MDAP, this data will be visualised within an interactive location-based tool and the data analysed to identify trends over time, including material use, construction type, prevalence of solar hot water systems and rainwater tanks.

This will help to understand the nature of housing stock across different municipalities in the state of Victoria and also identify relevant correlations between; building project characteristics, such as the prevalence of rainwater tanks by site area or local government area, or solar hot water systems by number of storeys.

This project aims to convert low-resolution CT images of large samples to super-resolution by using artificial intelligence (AI)-based image processing algorithms.

AI algorithms are supposed to be used to split connected particles in CT images. Then, the particle volume, particle surface area and interparticle contact areas will be computed from the enhanced super-resolution CT images. The values will be compared with their counterparts derived from the low-resolution CT images before enhancement.

This collaboration will increase access to herbarium-derived biodiversity data by creating tools to capture, analyse, and share herbarium specimen associated data.

Biodiversity data, such as those that accompany museum and herbarium specimens, comprise a massive global data resource. This data can be used for rapid biodiversity analyses and to inform conservation and policy decision making. To do this it needs to be accessible within an integrated biodiversity infrastructure, that retains links among spatial, trait, and evolutionary data for taxa.

This project will provide tools to support efficient mobilisation of high-quality specimen associated primary (e.g. taxon, location) and secondary (e.g. molecular) biodiversity data. It will generate a pipeline for collation and phylogenetic analyses of molecular data generated by University of Melbourne researchers and students. It will also create an online phylogenetic browser to enable exploration of native Australian plant and algal evolutionary trees. This will be hosted on the University of Melbourne Herbarium Collection Online for access by students, researchers, and the global public.

This collaboration aims to screen whole genome sequencing datasets of globally distributed kelp species for host-associated eukaryotic taxa, filling a considerable gap in kelp holobiome biology.

Bacterial diversity of seaweed microbiomes has received considerable attention over the years. But far less is known about their eukaryotic component such as algae, fungi, and animals living in/on organisms – collectively called the Eukaryome. We need new diagnostic frameworks for characterising eukaryotic microbial diversity. This will allow us to meet increasing industry demands for certified seaweed products, and better understand ecologically significant kelp forest communities.

The collaboration will help establish a robust workflow for future genomic analyses of entire microbial communities. It will employ state-of-the art machine learning algorithms to classify sequences and enhance knowledge about kelp holobiomes.

The ongoing SARS-CoV-2 pandemic has seen an unprecedented amount of genome data generated, with over 2 million complete genomes in the GISAID platform. These data have been pivotal to track the evolution of the virus on a global scale and to detect mutations of epidemiological importance, such as those in variants of concern.

Many important biological questions could be addressed with these data. However, there can be issues assessing data quality and suitability. The main limitation in accessing sequence data from public databases is that many genomes have substantial stretches of missing data and are thus not sufficiently reliable to infer mutational events. Others can lack important metadata, such as travel history or vaccination status.

This project has two key aims:

1. Develop a system where one can select thousands of sequences that meet certain criteria for quality control or the presence of metadata that are fit-for-purpose for epidemiological questions
2. Develop tools to specify Bayesian hierarchical models in an intuitive format that will use the above fit-for-purpose data.

The genomics revolution has spurred movements to sequence the genomes of all life on earth. This has led to a flurry of genomes in the past five years – everything from redwood trees to gorillas. This revolution has also piqued the interest of conservation biologists who are fighting the ever-increasing extinction crisis.

Genomic information can be used for species management to ensure the best approaches are used for maintaining or increasing genetic diversity. More recently this information is being used for more directed approaches to help promote characteristics that allow species to persist with intractable threats like climate change and disease.

This MDAP collaboration will be used to build computational and genomics resources for amphibians. Amphibians are the most imperiled group of vertebrates and face many threats such as the devastating disease, chytridiomycosis. Despite their conservation value, the development of genomic resources for amphibians has lagged behind other vertebrates. This is because their unusually large genomes are difficult to assemble and work with. This collaboration, between Dr Kosch’s team at One Health Research Group and MDAP, will help build the critical resources to identify gene targets for genetic intervention to help restore amphibians threatened by chytridiomycosis to the wild.

Audio recordings provide powerful forensic evidence in criminal trials. The problem is they are often of poor quality, to the extent the jury needs a transcript to understand the content.

The Research Hub for Language in Forensic Evidence (The Hub) aims to develop evidence-based scientific methods for creating reliable transcripts of indistinct forensic audio. This requires collecting and analysing large amounts of empirical data regarding how listeners interpret indistinct audio under various conditions.

The current project aims to develop tailored infrastructure and methodological protocols to enable ongoing experimental research on this topic.

Over the past ten years, many experiments have been conducted, and methods have evolved. We want to consolidate these methods to streamline and expand our ongoing research. To do this, this collaboration will focus on two major requirements:

1. Infrastructure to deploy audio to participants, allow participants to input their transcripts and enable researchers to add metadata to responses. Existing platforms are too restrictive for our purposes. Dedicated infrastructure will enable us to collect and analyse data more efficiently, with a better experience for participants.
2. Streamlined methodology. Our experimental approach has evolved over the last decade. While this work has produced valuable results, the methods now need to be refined and rationalised into efficient protocols.

Recently, the Hub has been working with the Language Testing Research Centre to develop our capabilities for scoring transcripts. This collaboration would further benefit from the increased interdisciplinarity offered by MDAP, for example in platform development, human sciences and data security. There is scope for outputs of the proposed project to be extended to other branches of linguistic research.

Audio recordings provide powerful forensic evidence in criminal trials. The problem is they are often of poor quality, to the extent the jury needs a transcript to understand the content.

The Research Hub for Language in Forensic Evidence (The Hub) aims to develop evidence-based scientific methods for creating reliable transcripts of indistinct forensic audio. This requires collecting and analysing large amounts of empirical data regarding how listeners interpret indistinct audio under various conditions.

The current project aims to develop tailored infrastructure and methodological protocols to enable ongoing experimental research on this topic.

Over the past ten years, many experiments have been conducted, and methods have evolved. We want to consolidate these methods to streamline and expand our ongoing research. To do this, this collaboration will focus on two major requirements:

1. Infrastructure to deploy audio to participants, allow participants to input their transcripts and enable researchers to add metadata to responses. Existing platforms are too restrictive for our purposes. Dedicated infrastructure will enable us to collect and analyse data more efficiently, with a better experience for participants.
2. Streamlined methodology. Our experimental approach has evolved over the last decade. While this work has produced valuable results, the methods now need to be refined and rationalised into efficient protocols.

Recently, the Hub has been working with the Language Testing Research Centre to develop our capabilities for scoring transcripts. This collaboration would further benefit from the increased interdisciplinarity offered by MDAP, for example in platform development, human sciences and data security. There is scope for outputs of the proposed project to be extended to other branches of linguistic research.

The task of bacterial taxonomic classification plays a pivotal role in the public health management of infectious disease outbreaks. Accurate identification of bacterial strains is essential for effective treatment protocols and subsequent genomic analyses.

With the growing prevalence of culture-independent methodologies, such as metagenomics, there is a pressing need for more rapid and precise approaches to bacterial classification. Current leading methods, while proficient, often require significant computational resources and extended processing times.

Considering recent advancements in artificial intelligence, particularly in processing DNA sequences, this project proposes the application of geometric deep learning techniques. These techniques are anticipated to surpass existing sequence homology-based methods in efficiency and accuracy for bacterial classification.

Our goal is to optimise these methods for computational resource management and processing speed. We will work with the MDAP team to apply expertise in deep learning and artificial intelligence.

The outcome of this project will be the development of open-source tools, designed for seamless integration into existing public health genomic workflows.

Breast cancer is the most common cancer in women worldwide with more than two million new cases diagnosed annually. It is also one of the two leading causes of cancer-related deaths in women.

Digital Mammography (DM) is the prevalent imaging method for breast screening. Yet significant limitations persist despite recent innovations such as digital breast tomosynthesis (DBT). These limitations include low sensitivity and specificity, radiation doses and painful compression of the breast. Unlike DM, DBT is capable of 3D imaging of the breast, albeit with limited depth resolution.

In contrast, computed tomography (CT) based methods, including phase-contrast CT (PCT) technique employed by our team, are superior to both DM and DBT in their ability to create high-fidelity 3D representations of the tissue density distribution in the breast.

We propose to investigate AI and ML based methods for denoising CT images of the human breast. Initially, we will analyse PCT scans of mastectomy samples collected by our team over the last 10 years at the Imaging and Medical beamline of the Australian Synchrotron.

While image denoising is a well-established area of information technology, denoising images without loss of spatial resolution, as required in medical CT, is a challenging problem. This has been the subject of recent research in Australia and overseas using AI and DL, with promising results published in scientific literature in the last few years.

We will draw on a significant existing dataset of PCT scans of mastectomy samples, collected under different imaging conditions, including different doses / noise levels. This can be used for effective training and fine-tuning of AI models.

Working with MDAP will provide expertise in working with complex GPU infrastructure, AI, algorithm and software development, and data analysis.

The outcomes of this project will greatly benefit our PCT research and will be applied to the world-first synchrotron-based breast cancer PCT scans of live patients scheduled at the Australian Synchrotron for 2024.

In the longer term, the methods developed as a part of this project could be expanded further to imaging of other organs which have been already demonstrated to benefit from X-ray phase contrast, such as joints (in connection with osteoarthritis), the heart and brain.

Linear A is a script that was used on Crete and its surrounding islands by the Minoan civilization during the Bronze Age (ca. 1800–1450 BCE). It was first discovered in 1900 AD at the site of Knossos, alongside another related, but slightly later script called Linear B, which was associated with Mycenaean culture and attested on both Crete and mainland Greece (ca. 1400–1200 BC).

While Linear B was shown to represent an early form of ancient Greek in 1952 AD, Linear A remains undeciphered and appears to encode an unrelated language. Attempts at decipherment using conventional, manual methods have proven unsuccessful to date, owing in no small part to the restricted size of the current corpus of texts.

However, in recent years, deep learning techniques have shown great success as a tool in deciphering other ancient scripts. By combining expertise in Aegean scripts from the University of Melbourne’s Classics and Archaeology department with the technical capabilities of the MDAP team, we are hoping to investigate the nature of Linear A through language models trained via deep learning, working towards an effort at machine decipherment.

A key part of this project will involve pre-training models in potentially related languages, before refining the models on Linear A, to test the level of cognancy between them. The identification of a cognate language for Linear A would lay the groundwork for machine decipherment – a method already successfully applied to other ancient languages with known cognates, including Ugaritic and Linear B.

This project uses modern scientific methodologies, based on phylogenetic analysis, to examine mediaeval Latin texts and answer key questions about the manuscripts in which they are found. Traditional methods of examining these texts require hundreds of hours of painstaking examination and double-checking of raw data contained in manuscripts and published editions—it is anticipated that this method of analysis will reduce this time considerably.

Livyʼs History of Rome is a seminal historical work from the ancient world, and its text has been studied intensively, but the English mediaeval scholar Nicholas Trevet wrote an influential Latin commentary on Livy which has never been published. There are fundamental scholarly questions about Trevetʼs sources and the subsequent influence of his commentary which have been impossible to address because of the absence of a modern critical edition of this work. As a first step in a new examination, the team aims to look at Trevet’s sources.

Trevet followed a set practice of the period, citing words of the work which he was explaining in an abbreviated form known as lemmata. These lemmata offer us an important window into Trevetʼs manuscript sources. The project sets out to identify the manuscript sources for the lemmata as precisely as we can using published material for Livy. Knowing what general family of manuscript he used can give us more precise information on where he wrote his own commentary and when, key pieces of information for understanding the audience of his study which are still unknown.

Specific problems complicate this investigation. The information on textual variants contained in editions of Livy is highly condensed using traditional methods of manuscript analysis. This needs to be unpacked before it can be plotted against sample data derived directly from the manuscripts by the CI. The other main problem which requires attention is the way in which Trevet frequently exploited the flexibility of Latin word order to restructure the text derived from Livy.

It is anticipated that the expertise provided by MDAP in data analysis, software development to aid in textual analysis, and the digital humanities will considerably shorten and simplify the examination of this material.