A recent study led by computer scientists from QUT suggests that machine-based deep learning could expedite the repatriation of ancestral human remains of Australian and Indigenous peoples that are currently held in anthropological collections.
In a paper published in The review of computer use in social science.In their article, “A Knowledgeable Neural Network for Uncovering Historical Records to Aid in the Return of Indigenous Ancestral Human Remains,” a group of researchers led by Dr. Md Abul Bashar and Professor Richi Nayak, who are affiliated with the School of Computer Science at QUT, describe their approach.
Spotlight on Women in Science: Dr. Rose Oelkers and Her Tree-Ring Research
According to Dr. Bashar, the process of bringing back the bodies of the ancestors of Australian and other First Nations peoples from Western scientific institutions mostly relies on finding information in scientific and other publications from 1790 to 1970. These documents would detail how their remains were taken, given away, sold, or traded between different institutions.
Since the mid-1990s, national and major research libraries have implemented digital library initiatives that allow us to systematically search through extensive collections of historical books, journals, and newspapers. Nevertheless, it still poses a challenge to locate pertinent information within these vast collections.
Women in STEM: Meet Rose Oelkers, a Tree-Ring Researcher Making a Difference
Dr. Bashar and Professor Nayak, with the assistance of academics from the Australian National University and University of Tasmania, were funded by the Australian Research Council. They also collaborated with data scientists and social science researchers from the Research, Reconcile, Renew Network (RRR) who specialize in text mining techniques. Together, they have developed a machine learning-driven solution that automates the task of locating and analyzing applicable texts in a semantic manner.
According to Professor Nayak, the researchers at RRR encounter a major obstacle in their endeavor to help indigenous communities reclaim the remains of their ancestors from Western scientific collections.
The researchers at RRR make an effort to verify the identity, present whereabouts, and other required details in order to properly bury the remains. Achieving this involves the researchers dedicating significant time and mental effort to track the journey of the stolen remains. The remains have gone through various networks, which include collectors, donors, private sellers, commercial dealers, and scientists associated with museums and other establishments with anthropological collections, although not all of these networks have been fully identified yet.
Rose Oelkers: A Trailblazer in Tree-Ring Research and an Inspiration to Women in STEM
Presently, museum catalogues offer certain data that can be analyzed. However, solely depending on the accumulated information in these catalogues carries the possibility of committing harmful errors.
According to Dr. Bashar, text mining has made notable advancements in the research field, specifically in the development of methods to discover and extract valuable information from extensive document collections.
Dr. Bashar stated that text mining has primarily been used on already existing written collections. However, in the fields of humanities and social sciences, there are numerous research inquiries regarding previous human ideas and actions that lack machine-readable historical document archives.
researchers are required to rely on search engines to find, collect, and examine written material related to individuals, locations, and historical occurrences. The difficulty lies in the fact that search engine results Algorithms and indexing practices have the potential to introduce bias into search results, which can lead to incomplete representation of all relevant information.
The Impact of Rose Oelkers’ Tree-Ring Research on Climate Change and Environmental Science
Researchers from RRR find the value of keyword-based search services offered by corpus creators highly reliant on how search queries are constructed. If the search terms used are very precise, only a small number or even no relevant documents will be retrieved. On the other hand, if the terms used are too broad and general, a significant amount of irrelevant documents will be retrieved.
According to Professor Nayak, when classification models, especially those based on deep learning, are trained with only a small number of labeled documents (that indicate relevance or non-relevance), their accuracy can be significantly reduced.
Professor Nayak explained that they investigated the utilization of an Informed Neural Network (INN) model, which utilizes expert-derived contextual knowledge to present a comprehensive understanding of documentary content, with the aim of enhancing the precision of our detection model.
The RRR researchers have given a specific group of words that are likely to be found in relevant documents.
“This served as the foundation for training a deep learning classification model using a limited number of labeled documents.”
Our search results validate the effectiveness of employing an INN network model in efficiently and accurately finding relevant documents pertaining to the investigation of the global commercial trade in Indigenous human remains.
Also Read: Celebrating Women in Science: Rose Oelkers, Tree-Ring Researcher
Dr. Rose Oelkers: A Role Model for Women in Science and a Leader in Tree-Ring Research
In 2014, it was created with the purpose of increasing knowledge and awareness about repatriation. Its goal is to support professionals and researchers dedicated to bringing elderly individuals back to their original homes.
Dr. Bashar, Professor Nayak, Professor Cressida Fforde, Dr. Gareth Knapman, and Professor Paul Turnbull are part of a team called the Research, Reconcile, Renew Network. This team was formed in 2014 with the goal of increasing awareness and understanding of repatriation. They aim to assist both practitioners and researchers in their endeavors to bring elderly individuals back to their homeland.
Dr. Knapman works as a Research Fellow at the Australian National University’s Center for Heritage and Museum Studies. She is currently involved in the ARC Discovery Project titled “Profit and Loss, The commercial trade in Indigenous human remains.”
Professor Fforde has primarily focused his research on the history of European institutions removing and studying Indigenous human remains during the nineteenth century. This research also examines efforts to retrieve and return these remains, all within the context of the “race” paradigm.
Professor Turnbull specializes in the study of anthropology and the biomedical sciences during colonial times. His field of expertise revolves around the theft and scientific exploitation of the bodily remains of Australian and other Indigenous populations. His research on this topic has gained significant recognition worldwide.
