You might also like
The Forensics Stream, which is a part of the Speech and Language Lab—a joint initiative between the School of Literature, Languages and Linguistics (College of Arts & Social Sciences) and the School of Culture, History and Language (College of Asia & the Pacific) —has a clear research goal: improving the evaluation of linguistic evidence (e.g., voice recordings and text messages) as part of forensic science. The team’s research has direct applications in two broad areas: assisting the courts to incorporate linguistic evidence in the legal process in a scientifically rigorous way; and improving efficiency and accuracy in police and national security investigations.
Technological development over the last few decades has really proved beneficial for people in many walks of life. For instance, we all enjoy technologies today with which we can communicate with others instantly and inexpensively. However, these technologies have not been embraced just for social pleasure and convenience. The growing urgency and importance of this field of research is apparent for much more serious issues. The high degree of anonymity that has come hand in hand with rising online communication has led to rising malicious cyber activity both nationally and globally. Some activities by terrorist organisations and on the dark web are good examples of this trend. These are newly emerging serious risks to national and global security.
So, how does the team’s research intersect with these challenges? Here is an example scenario from forensic text comparison, in which the Forensic Stream team’s research contribution is invaluable.
"The police have been monitoring underground chatlog sites for trading child pornography and are tracking text communications between a target dealer (the offender) and his customers. They also have a suspect and have confiscated his laptop under a search warrant. The laptop contains lawful chatlog messages between the suspect and his friends, for which the suspect confirms authorship. The police would like to know whether the suspect’s and offender’s texts are written by the same person."
In forensic voice comparison, recorded voice samples are analysed as forensic evidence, instead of text samples.
Because of the complexity and multidisciplinary nature of the field, the researchers at this lab come with varied and diverse expertise: linguistics and acoustic phonetics, forensic voice/text comparison, speech science, and forensic science. They have significant experience in collaborating with national stakeholders, such as law-enforcement agencies and legal professionals, as well as international partner institutions, such as the National Research Institute of Police Science in Japan.
Dr Yuko Kinoshita is an expert in forensic voice comparison and has worked as a consultant to a number of law enforcement agencies, evaluating the voice evidence in which the speakers of recordings are disputed. She is one of the first researchers who applied the same evaluative framework to bring voice evidence evaluation to the gold standard of forensic science, as practice in DNA analysis.
Dr Frantz Clermont is currently visiting the lab as Honorary Associate Professor. He is engaged in international collaborations with speech science colleagues in Australia, America and England. He brings multidisciplinary expertise in formant dynamics of vowels and diphthongs, cepstral characterisation of speaker variability, numerical computation, and scientific writing.
Dr Kimiya Akita is a forensic scientist from the National Research Institute of Police Science, Japan. He is visiting the lab for training purposes as he has taken up the task of forensic voice comparison at his home institute.
Michael Carne is a new HDR student who will be working on an intersection of forensic voice and text comparisons as his PhD project. He was awarded first-class honours in forensic voice comparison in Vietnamese.
Caroline Hendy is a summer scholar under a scholarship from the Australian Signals Directorate. She was awarded first-class honours in the acoustic analysis of Light Warlpiri fricatives, and is currently working on forensic voice comparison in Vietnamese-accented English.
Leading this stream of research as a co-director of the Speech and Language Lab is Dr Shun Ishihara. He successfully applied the likelihood ratio-based evaluative method to linguistic text evidence for the first time in the field and demonstrated that forensic text evidence can also satisfy the strictest requirements for the admissibility of scientific evidence.
The goal of this team is to lead the field of forensic voice/text comparison by providing solutions to real life problems. They have strong capacity in education and training in this field, as well as researching and assisting with case analysis. They are keen to extend their collaborations and relationships to institutions in other countries in the Asian and Pacific region and are actively seeking opportunities.
According to Shun, “We believe in, and work towards fair, dependable and transparent evaluation of linguistic evidence, as the trust in the judicial system is the bedrock of any civil society. Forensic voice/text comparison is a new area of study, having so many fundamental studies to be done. If you would like to be a part of our ambitions, please come and talk to us!”