Physical Sciences and Mathematics Commons^™

Using Chatgpt To Generate Gendered Language, Shweta Soundararajan, Manuela Nayantara Jeyaraj, Sarah Jane Delany

Conference papers

Gendered language is the use of words that denote an individual's gender. This can be explicit where the gender is evident in the actual word used, e.g. mother, she, man, but it can also be implicit where social roles or behaviours can signal an individual's gender - for example, expectations that women display communal traits (e.g., affectionate, caring, gentle) and men display agentic traits (e.g., assertive, competitive, decisive). The use of gendered language in NLP systems can perpetuate gender stereotypes and bias. This paper proposes an approach to generating gendered language datasets using ChatGPT which will provide data for data-driven …

Exploring The Design Of Low-End Technology To Increase Patient Connectivity To Electronic Health Records, Rens Kievit, Abdullahi Abubakar Kawu, Mirjam Van Reisen, Dympna O'Sullivan, Lucy Hederman

Conference papers

The tracking of the vitals of patients with long term health problems is essential for clinicians to determine proper care. Using Patient Generated Health Data (PGHD) communicated remotely allows patients to be monitored without requiring frequent hospital visits. Issues might arise when the communication of data digitally is difficult or impossible due to a lack of access to internet or a low level of digital literacy as is the case in many African countries. The VODAN-Africa project (van Reisen et al., 2021) started in 2020 and has greatly increased the capabilities of clinics in different countries in both Africa and …

Exploring Wav2vec 2.0 Model For Heart Murmur Detection, Davoud Shariat Panah, Andrew Hines, Susan Mckeever

Conference papers

The lack of access to cardiology resources in many regions of the world has motivated the development of automatic diagnostic systems based on cardiac signals. In recent years, a wide range of supervised learning models have been proposed that can make an initial diagnosis of heart disease from heart sounds. To achieve high accuracy, however, such supervised learning models generally require a large amount of labeled data, which can be costly to obtain. In this regard, self-supervised learning has been recently employed to reduce the over-reliance on annotated data. Wav2vec 2.0 is an audio self-supervised learning model that has shown …

An Investigation Into The Application Of The Meijering Filter For Document Recapture Detection, John Magee, Stephen Sheridan Phd, Christina Thorpe Phd

Conference papers

The proliferation of mobile devices allows financial institutions to offer remote customer services, such as remote account opening. Manipulation of identity documents using image processing software is a low-cost, high-risk threat to modern financial systems, opening these institutions to fraud through crimes related to identity theft. In this paper we describe our exploratory research into the application of biomedical image algorithms to the domain of document recapture detection. We perform a statistical analysis to compare different types of recaptured documents and train a support vector machine classifier on the raw histogram data generated using the Meijering filter. The results show …

Interpretable Timbre Synthesis Using Variational Autoencoders Regularized On Timbre Descriptors, Anastasia Natsiou, Luca Longo, Sean O'Leary

Conference papers

Controllable timbre synthesis has been a subject of research for several decades, and deep neural networks have been the most successful in this area. Deep generative models such as Variational Autoencoders (VAEs) have the ability to generate a high-level representation of audio while providing a structured latent space. Despite their advantages, the interpretability of these latent spaces in terms of human perception is often limited. To address this limitation and enhance the control over timbre generation, we propose a regularized VAE-based latent space that incorporates timbre descriptors. Moreover, we suggest a more concise representation of sound by utilizing its harmonic …

An Exploration Of The Latent Space Of A Convolutional Variational Autoencoder For The Generation Of Musical Instrument Tones, Anastasia Natsiou, Sean O'Leary, Luca Longo

Conference papers

Variational Autoencoders (VAEs) constitute one of the most significant deep generative models for the creation of synthetic samples. In the field of audio synthesis, VAEs have been widely used for the generation of natural and expressive sounds, such as music or speech. However, VAEs are often considered black boxes and the attributes that contribute to the synthesis of a sound are yet unsolved. Existing research focused on the way input data can influence the generation of latent space, and how this latent space can create synthetic data, is still insufficient. In this manuscript, we investigate the interpretability of the latent …

Co-Designing Assistive Technology With And For Persons Living With Dementia, Dympna O'Sullivan, Jonathan Turner, Siobhan O'Neill, Micheal Wilson, Julie Doyle

Conference papers

Dementia is a chronic and progressive neurodegenerative illness, which can lead to significant difficulties in a person’s capacity to perform activities of daily living (ADLs) and engage in meaningful activities. There is an acute need, which digital health technologies can potentially fulfil, to provide proactive support for persons living with dementia (PLwD) and their caregivers. However, there is limited involvement of PLwD in the design of technology that could be used to support their personal plans for independent living at home. In this paper, we describe how we are employing a co-design methodology to support engagement in an assistive technology …

Wifi-Based Human Activity Recognition Using Attention-Based Bilstm, Amany Elkelany, Robert J. Ross, Susan Mckeever

Conference papers

Recently, significant efforts have been made to explore human activity recognition (HAR) techniques that use information gathered by existing indoor wireless infrastructures through WiFi signals without demanding the monitored subject to carry a dedicated device. The key intuition is that different activities introduce different multi-paths in WiFi signals and generate different patterns in the time series of channel state information (CSI). In this paper, we propose and evaluate a full pipeline for a CSI-based human activity recognition framework for 12 activities in three different spatial environments using two deep learning models: ABiLSTM and CNN-ABiLSTM. Evaluation experiments have demonstrated that the …

Applying Positional Encoding To Enhance Vision-Language Transformers, Xuehao Liu, Sarah Jane Delany, Susan Mckeever

Conference papers

Positional encoding is used in both natural language and computer vision transformers. It provides information on sequence order and relative position of input tokens (such as of words in a sentence) for higher performance. Unlike the pure language and vision transformers, vision-language transformers do not currently exploit positional encoding schemes to enrich input information. We show that capturing location information of visual features can help vision-language transformers improve their performance. We take Oscar, one of the state-of-the-art (SOTA) vision-language transformers as an example transformer for implanting positional encoding. We use image captioning as a downstream task to test performance. We …

Towards A Framework For Privacy-Preserving Pedestrian Analysis, Anil Kunchala, Mélanie Bouroche, Bianca Schoen-Phelan

Conference papers

The design of pedestrian-friendly infrastructures plays a crucial role in creating sustainable transportation in urban environments. Analyzing pedestrian behaviour in response to existing infrastructure is pivotal to planning, maintaining, and creating more pedestrian-friendly facilities. Many approaches have been proposed to extract such behaviour by applying deep learning models to video data. Video data, however, includes an broad spectrum of privacy-sensitive information about individuals, such as their location at a given time or who they are with. Most of the existing models use privacy-invasive methodologies to track, detect, and analyse individual or group pedestrian behaviour patterns. As a step towards privacy-preserving …

Exploring The Integration Of Patient Generated Health Data In A Fair Digital Health System In Low-Resourced Settings: A User-Centered Approach, Abdullahi Abubakar Kawu, Rens Kievit, Adamu Abubakar, Mirjam Van Reisen, Dympna O'Sullivan, Lucy Hederman

Conference papers

This article presents the initial user-centered research exploring the opportunities in the collection of Patient-Generated Health Data (PGHD) within the context of a project aimed at improving health management and outcomes among residents in African countries. Through interviews with a doctor, a patient and two data managers, the local status and opinions regarding PGHD collection, integration and use are investigated. The findings suggest that PGHD have only been encountered in paper forms - and are mostly patient driven, however opportunities for PGHD for the facility and patient were identified and included supporting the treatment of whitecollar hypertension, treatment planning and …

Determining Child Sexual Abuse Posts Based On Artificial Intelligence, Susan Mckeever, Christina Thorpe, Vuong Ngo

Conference papers

The volume of child sexual abuse materials (CSAM) created and shared daily both surface web platforms such as Twitter and dark web forums is very high. Based on volume, it is not viable for human experts to intercept or identify CSAM manually. However, automatically detecting and analysing child sexual abusive language in online text is challenging and time-intensive, mostly due to the variety of data formats and privacy constraints of hosting platforms. We propose a CSAM detection intelligence algorithm based on natural language processing and machine learning techniques. Our CSAM detection model is not only used to remove CSAM on …

Investigation, Detection And Prevention Of Online Child Sexual Abuse Material: A Comprehensive Survey, Vuong Ngo, Christina Thorpe, Cach N. Dang, Susan Mckeever

Conference papers

Child sexual abuse inflicts lifelong devastating consequences for victims and is a growing social concern. In most countries, child sexual abuse material (CSAM) distribution is illegal. As a result, there are many research papers in the literature which proposed technologies to detect and investigate CSAM. In this survey, a comprehensive search of the peer reviewed journal and conference paper databases (including preprints) is conducted to identify high-quality literature. We use the PRISMA methodology to refine our search space to 2,761 papers published by Springer, Elsevier, IEEE and ACM. After iterative reviews of title, abstract and full text for relevance to …

The Ukicer 2022 Conference Poster: Techmate: A Best Practice Toolkit For Driving Sustainable Acceleration Towards Gender Equality In Technology Disciplines In Heis., Alina Berry

Conference papers

TechMate is a research project that is being developed to enhance gender balance in technology disciplines, in particular computing higher education in Ireland and beyond. Gender imbalance in computing education is a well-known issue: in Ireland, less than 15% of the student population in computer science, ICT and related disciplines are women. Despite a significant amount of research and practical work conducted in the recent decades, the problem still persists and this research initiative aims to improve the situation.

Among the main aims of this project, there is a development of a toolkit to drive sustainable acceleration towards gender equality …

An Ontological Approach For Recommending A Feature Selection Algorithm, Aparna Nayak, Bojan Bozic, Luca Longo

Conference papers

Feature selection plays an important role in machine learning or data mining problems. Removing irrelevant features increases model accuracy and reduces the computational cost. However, selecting important features is not a simple task as one feature selection algorithm does not perform well on all the datasets that are of interest. This paper tries to address the recommendation of a feature selection algorithm based on dataset characteristics and quality. The research uses three types of dataset characteristics along with data quality metrics. The main contribution of the work is the utilization of Semantic Web techniques to develop a novel system that …

Learning Fruit Class From Short Wave Near Infrared Spectral Features, An Ai Approach Towards Determining Fruit Type, Ayesha Zeb, Waqar Shahid Qureshi, Abdul Ghafoor, Dympna O'Sullivan

Conference papers

This paper analyzes the potential of using shortwave NIRS (near-infrared spectroscopy) for fruit classification problems. The research focuses on O-H and C-H overtone features of fruit and its correlation with NIRS and therefore opens a new dimension of fruit classification problems using NIRS. Eleven fruits, which include apple, cherry, hass, kiwi, grapes, mango, melon, orange, loquat, plum, and apricot, were used in this study to cover physical characteristics such as peel thinness, pulp, seed thickness, and size. NIR spectral data is collected using the industry-standard F-750 fruit quality meter (wavelength range 300-1100nm) for all fruit mentioned above. Different shallow machine …

Linked Data Quality Assessment: A Survey, Aparna Nayak, Bojan Bozic, Luca Longo

Conference papers

Data is of high quality if it is fit for its intended use in operations, decision-making, and planning. There is a colossal amount of linked data available on the web. However, it is difficult to understand how well the linked data fits into the modeling tasks due to the defects present in the data. Faults emerged in the linked data, spreading far and wide, affecting all the services designed for it. Addressing linked data quality deficiencies requires identifying quality problems, quality assessment, and the refinement of data to improve its quality. This study aims to identify existing end-to-end frameworks for …

Extending R2rml-F To Support Dynamic Datatype And Language Tags, Aparna Nayak, Bojan Bozic, Luca Longo

Conference papers

Linked data is often generated from raw data with the help of mapping languages. Complex data transformation is one of the essential parts while uplifting data which either can be implemented as custom solutions or separated from the mapping process. In this paper, we propose an approach of separating complex data transformations from the mapping process that can still be reusable across the systems. In the proposed method, complex data transformations include the entailment of (i) language tag and (ii) datatype present at the data source. The proposed method also includes inferring missing datatype information. We extended R2RML-F to handle …

Check Your Tech - The Ethics Of Deepfakes In A Political Context, Dympna O'Sullivan, Damian Gordon, Ioannis Stavrakakis, Michael Collins

Conference papers

No abstract provided.

Check Your Tech - The Ethics Of Gamification In Education, Dympna O'Sullivan, Ioannis Stavrakakis, Damian Gordon, Anna Becevel

Conference papers

No abstract provided.

The Future Of Medicine Is Digital: Developing Educational Materials To Explore The Ethics Of Digital Pills., Dympna O'Sullivan, J. Paul Gibson, Yael Jacob, Ioannis Stavrakakis, Damian Gordon

Conference papers

Digital Pills are a drug-device technology that permit to combine traditional medications with a monitoring system that automatically records data about medication adherence and patients’ physiological data. They are a promising innovation in digital medicine, however their use has raised a number of ethical concerns. In this paper, we outline some of the main Digital Pills technologies and explore key ethical challenges surrounding their use. In this paper, we introduce educational materials we have developed that provide an insight into the technologies and ethical aspects that underpin Digital Pills.

(Linked) Data Quality Assessment: An Ontological Approach, Aparna Nayak, Bojan Bozic, Luca Longo

Conference papers

The effective functioning of data-intensive applications usually requires that the dataset should be of high quality. The quality depends on the task they will be used for. However, it is possible to identify task-independent data quality dimensions which are solely related to data themselves and can be extracted with the help of rule mining/pattern mining. In order to assess and improve data quality, we propose an ontological approach to report data quality violated triples. Our goal is to provide data stakeholders with a set of methods and techniques to guide them in assessing and improving data quality

Human Or Robot?: Investigating Voice, Appearance And Gesture Motion Realism Of Conversational Social Agents, Ylva Ferstl, Sean Thomas, Cédric Guiard, Cathy Ennis, Rachel Mcdonnell

Conference papers

Research on creation of virtual humans enables increasing automatization of their behavior, including synthesis of verbal and nonverbal behavior. As the achievable realism of different aspects of agent design evolves asynchronously, it is important to understand if and how divergence in realism between behavioral channels can elicit negative user responses. Specifically, in this work, we investigate the question of whether autonomous virtual agents relying on synthetic text-to-speech voices should portray a corresponding level of realism in the non-verbal channels of motion and visual appearance, or if, alternatively, the best available realism of each channel should be used. In two perceptual …

The Development Of Teaching Case Studies To Explore Ethical Issues Associated With Computer Programming, Michael Collins, Damian Gordon, Dympna O'Sullivan

Conference papers

In the past decade software products have become pervasive in many aspects of people’s lives around the world. Unfortunately, the quality of the experience an individual has interacting with that software is dependent on the quality of the software itself, and it is becoming more and more evident that many large software products contain a range of issues and errors, and these issues are not known to the developers of these systems, and they are unaware of the deleterious impacts of those issues on the individuals who use these systems. The authors of this paper are developing a new digital …

Exploring The Personality Of Virtual Tutors In Conversational Foreign Language Practice, Johanna Dobbriner, Cathy Ennis, Robert J. Ross

Conference papers

Fluid interaction between virtual agents and humans requires the understanding of many issues of conversational pragmatics. One such issue is the interaction between communication strategy and personality. As a step towards developing models of personality driven pragmatics policies, in this paper, we present our initial experiment to explore differences in user interaction with two contrasting avatar personalities. Each user saw a single personality in a video-call setting and gave feedback on the interaction. Our expectations, that a more extroverted outgoing positive personality would be a more successful tutor, were only partially confirmed. While this personality did induce longer conversations in …

Adaptable And Reusable Educational ‘Bricks’ For Teaching Computer Science Ethics, Andrea Curley, Damian Gordon, Ioannis Stavrakakis, Anna Becevel, J.P. Gibson, Dympna O'Sullivan

Conference papers

No abstract provided.

Multi-Modal Self-Supervised Representation Learning For Earth Observation, Pallavi Jain, Bianca Schoen Phelan, Robert J. Ross

Conference papers

Self-Supervised learning (SSL) has reduced the performance gap between supervised and unsupervised learning, due to its ability to learn invariant representations. This is a boon to the domains like Earth Observation (EO), where labelled data availability is scarce but unlabelled data is freely available. While Transfer Learning from generic RGB pre-trained models is still common-place in EO, we argue that, it is essential to have good EO domain specific pre-trained model in order to use with downstream tasks with limited labelled data. Hence, we explored the applicability of SSL with multi-modal satellite imagery for downstream tasks. For this we utilised …

Check Your Tech, Whose Responsibility Is It When Cyberharassment Occurs?, Dympna O'Sullivan, Damian Gordon, Michael Collins, Emma Murphy

Conference papers

Social media has become a dominant aspect of many people’s lives in many countries. Unfortunately that resulted in widespread issues of bullying and harassment. While frequently this harrassment is intentional, there have been occasions where automated processes have been inadvertently responsible for this sort of harassment. The software tools that allow people to harass others could have further features added to them to reduce the amount of harassment that occurs, but more often than not, where programmers are developing these systems then don’t anticipate the range of ways that these technologies will be used (this is called “consequence scanning”). The …

Monitoring Quality Of Life Indicators At Home From Sparse And Low-Cost Sensor Data., Dympna O'Sullivan, Rilwan Basaru, Simone Stumpf, Neil Maiden

Conference papers

Supporting older people, many of whom live with chronic conditions or cognitive and physical impairments, to live independently at home is of increasing importance due to ageing demographics. To aid independent living at home, much effort is being directed at reliably detecting activities from sensor data to monitor people’s quality of life or to enhance self-management of their own health. Current efforts typically leverage smart homes which have large numbers of sensors installed to overcome challenges in the accurate detection of activities. In this work, we report on the results of machine learning models based on data collected with a …

Is Twitter A Bad Place? The Responsibility That Social Media May Have Had In The 2021 Storming Of Capitol Hill., Ioannis Stavrakakis, Damian Gordon, Dympna O'Sullivan, Andrea Curley

Conference papers

The events of 6 th January 2021 in the United States of America, where rioters stormed the heart of their democracy, the US Capitol Complex (which houses their bicameral parliament) were shocking to see. The reasons for this riot were myriad, including to protest the outcomes of the presidential elections and two senate elections, as well as to prevent the counting that day of the electoral votes that formally certify the election result. These events will be analysed and reflected upon for years to come, and blame will be placed at many people’s doors, and inevitability one that has already …