Declarative Networks: Towards Robust and Explainable Deep Learning. The aim of this project is to develop declarative machine learning techniques that exploit inherent structure and models of the world. Deep learning has become the dominant approach for machine learning with many products and promises built on this technology. But deep learning is expensive, opaque, brittle and relies solely on human labelled data. This project intends to make deep learning more reliable by establishing theory a ....Declarative Networks: Towards Robust and Explainable Deep Learning. The aim of this project is to develop declarative machine learning techniques that exploit inherent structure and models of the world. Deep learning has become the dominant approach for machine learning with many products and promises built on this technology. But deep learning is expensive, opaque, brittle and relies solely on human labelled data. This project intends to make deep learning more reliable by establishing theory and algorithms that allow physical and mathematical models to be embedded within a deep learning framework, providing performance guarantees and interpretability. This would likely benefit machine learning based products that can understand the world and interact with humans naturally through vision and language.Read moreRead less
Information access through web-scale question-answer pair finding, ranking and matching. This project will aim to take web search to a new level of sophistication in accepting queries in the form of complex natural language questions, and returning a ranked list of natural language answers automatically extracted from a broad range of web user forums.
Exploiting Context in Multilingual Understanding and Generation. Automatic translation technologies produce incoherent and incorrect outputs in critical areas, such as health, finance, and law. This is due to translating sentences independently, without regard to the global extra-sentential context and rich linguistic structures inherent in the wider document context. This project aims to exploit global linguistic structures, capitalising on recent advances in deep neural networks, in order to g ....Exploiting Context in Multilingual Understanding and Generation. Automatic translation technologies produce incoherent and incorrect outputs in critical areas, such as health, finance, and law. This is due to translating sentences independently, without regard to the global extra-sentential context and rich linguistic structures inherent in the wider document context. This project aims to exploit global linguistic structures, capitalising on recent advances in deep neural networks, in order to generate coherent and faithful text. Expected outcome include next-generation computational technologies for language understanding and generation. This should significantly benefit document-based language technologies and increase their applications in a range of cultural, industrial, and health settings.Read moreRead less
Adaptive Context-Dependent Machine Translation for Heterogeneous Text. While automatic machine translation technologies are undoubtedly useful to a wide range of users, they often produce incoherent outputs for many types of input, for example, medical, literature, or even conversational text. This project will develop new adaptive machine translation systems to handle many domains and text styles, including heterogeneous mixed-domain inputs. It will develop multi-task machine learning methods f ....Adaptive Context-Dependent Machine Translation for Heterogeneous Text. While automatic machine translation technologies are undoubtedly useful to a wide range of users, they often produce incoherent outputs for many types of input, for example, medical, literature, or even conversational text. This project will develop new adaptive machine translation systems to handle many domains and text styles, including heterogeneous mixed-domain inputs. It will develop multi-task machine learning methods for training collections of domain-specific translation systems while leveraging correlations between domains. This approach will reduce the big data requirements of current translation systems, and improve translation quality across a wide range of different language pairs and application domains.Read moreRead less
The Language of Knowledge. Knowledge is central to our lives. The way we attribute knowledge to others governs whom we trust, how we reason, what we do, and whether we succeed. The significance of this project consists in a deeper understanding of our attributions of knowledge, situated within the context of questions. This project offers the benefits of quality foundational research in an interdisciplinary venue. It connects to a range of areas of National Research Priority including An Environ ....The Language of Knowledge. Knowledge is central to our lives. The way we attribute knowledge to others governs whom we trust, how we reason, what we do, and whether we succeed. The significance of this project consists in a deeper understanding of our attributions of knowledge, situated within the context of questions. This project offers the benefits of quality foundational research in an interdisciplinary venue. It connects to a range of areas of National Research Priority including An Environmentally Sustainable Australia, Promoting and Managing Good Health, and Frontier Technologies for Building and Transforming Australian Industries, insofar as all of these areas of National Research Priority are concerned with the acquisition and transfer of knowledge.Read moreRead less
Using past climate extremes to guide infrastructure planning for the future. This project aims to analyse a 2000-year palaeoclimate record of single event and complex climate extremes to provide a long-term context for observed changes in climate extremes over recent decades. This project expects to generate new knowledge about long-term variability in the frequency and magnitude of climate extremes that occur on seasonal - decades time-scales. It also expects to provide information about com ....Using past climate extremes to guide infrastructure planning for the future. This project aims to analyse a 2000-year palaeoclimate record of single event and complex climate extremes to provide a long-term context for observed changes in climate extremes over recent decades. This project expects to generate new knowledge about long-term variability in the frequency and magnitude of climate extremes that occur on seasonal - decades time-scales. It also expects to provide information about complex extremes that involve multiple types of impacts (e.g. drought followed by flood, simultaneous drought and fire). Expected benefits of the project include improved understanding of climate extremes and improved risk estimates for the impacts of climate extremes on Australian government and industry infrastructure.Read moreRead less
Small stones can break big canoes: Securing records of the world's indigenous languages. Fragile records of the world’s Indigenous languages are at risk of being lost. Better research methods that can benefit not only academics but also the general public aim to be developed in this project and used to train new researchers and community members in creating better records. Collaborating across Australia and the Pacific in building tools that will result in better research practices, it will incr ....Small stones can break big canoes: Securing records of the world's indigenous languages. Fragile records of the world’s Indigenous languages are at risk of being lost. Better research methods that can benefit not only academics but also the general public aim to be developed in this project and used to train new researchers and community members in creating better records. Collaborating across Australia and the Pacific in building tools that will result in better research practices, it will increase knowledge of what research has been done, and will target areas for future focus. Modelling new fieldwork methods, building reusable datasets, and curating long-term collections of language records will all be part of this project, as will outreach to support similar work both in linguistics and in the broader community.Read moreRead less
Hearing the future: supporting Indigenous linguistic diversity. This project aims to find new ways to support the extraordinary diversity of Indigenous languages spoken in Australia. In Arnhem Land the ability to understand but not speak a language is widespread and plays a crucial role supporting linguistic diversity. This ability, receptive multilingualism, will be examined using an innovative interdisciplinary methodology, generating new understandings about the relationship between multiling ....Hearing the future: supporting Indigenous linguistic diversity. This project aims to find new ways to support the extraordinary diversity of Indigenous languages spoken in Australia. In Arnhem Land the ability to understand but not speak a language is widespread and plays a crucial role supporting linguistic diversity. This ability, receptive multilingualism, will be examined using an innovative interdisciplinary methodology, generating new understandings about the relationship between multilingualism and linguistic diversity that are crucial to tackling the global decline in Indigenous languages. The findings will help communities, educators and policymakers develop new strategies to support Australia’s Indigenous languages which are vital to Indigenous health and wellbeing.Read moreRead less
Semantic categories: Exploring the history of the Baining languages of Island Melanesia. The project adds to our knowledge of the historical, linguistic and ethnic relationships within Island Melanesia, thus contributing directly to our understanding of the complexities of this region. This regional focus expands Australia's expertise in Melanesian research, and reaffirms Australia's leading position in this area. Such expertise attracts outstanding international students and researchers to Aust ....Semantic categories: Exploring the history of the Baining languages of Island Melanesia. The project adds to our knowledge of the historical, linguistic and ethnic relationships within Island Melanesia, thus contributing directly to our understanding of the complexities of this region. This regional focus expands Australia's expertise in Melanesian research, and reaffirms Australia's leading position in this area. Such expertise attracts outstanding international students and researchers to Australia, strengthens the Australian research community, and increases the international visibility of Australian-based research. Furthermore, studying semantic categories provides a unique insight into how speakers of different languages categorise the world around and within them, thereby facilitating intercultural understanding.Read moreRead less
Mega spatial-scale, multi time-scale, ensemble assessment of climate change driven coastal change in South Eastern Australia. Climate change driven variations in mean sea level, storm surges, and waves will change the world's coastline. This project will, for the first time, develop innovative modelling methods to quantify the integrated impact of these climate drivers on coastal erosion along Australia's most developed and populated coastline: Sydney to Brisbane.