Research Symposium
Please connect to our Q&A platform to take part in today's discussions.
-
Morning Session 09:55 - 12:00 [Hybrid]
09.55 - 10:00
Prof Karen Rafferty
Welcome10:00 - 10:30
Ulster University Presentations
Prof Chris Nugent
Introductions
Dr Federico Cruciani
Risk Language Models (RLM)
Dr Federico Cruciani
Privacy Enhancement Technologies (PET)
Dr Samuel Moore
Decision Support System
Dr Zeeshan Tariq
Intelligent Process Analytics
Dr Federico Cruciani
Semi Automatic Data Labelling
Dr Ian Cleland
Synthetic Data Generation
Muhammad Imad
Deep Learning for Anomaly Detection
Muhammad Asim Ali
Explainable and Transparent Imbalance Learning
Kevin Nolan
Predicting Chaos with Machine Learning
Majid Liaquat
Generating Privacy-preserving Synthetic Data
Prof Chris Nugent
Close10:30 - 10:40
Ulster University Q&A
10:40 - 11:10
Queen's University Presentations
Ethan Bradley
Table Extraction
Dr Muhammad Roman
Retrieval Augmented Generation
Gyan Wickremasinghe
Blockchain
Dr Vimal Dwivedi
Digital Twin
Dr Ayesha Abdul Majeed
Risk Assessment
Dr Olaide Oyelade
Idea Generation11:10 - 11:20
Queen's University Q&A
11:20 - 11:50
PwC Presentations
11:50 - 12:00
PwC Q&A
Prof Karen Rafferty
Close - Lunch Demos 12:00 - 13:00
Lunch will be served in the Allstate Software Studio.
Live demos and posters by Ulster University, Queen's University Belfast and PwC will be on display at the rear of the room.
Please take your seats at 12.55PM if you are staying for the afternoon research forum.
- Afternoon Session 13:00 - 16:00
13:00 - 13:35
Data Preparation
Chair: Dr Samuel Moore
Dr Ian Cleland, Majid Liaquat & Dr Min Jing
Synthetic Data Generation
Dr Ian Cleland, Majid Liaquat & Dr Min Jing
Generating Privacy-preserving Synthetic Data
Dr Federico Cruciani
Semi Automatic Data Labelling
Muhammad Asim Ali
Explainable and Transparent Imbalance Learning13:35 - 14:20
Data Analytics & Visualisation
Chair: Dr Vishal Sharma
Dr Zeeshan Tariq & Dr Ian McChesney
Intelligent Process Analytics
Callum Smith
Secure Multi-modal VR for Corporate Systems
Dr Samuel Moore
Decision Support System
Kevin Nolan
Predicting Chaos with Machine Learning
Yongsheng Dai
Few Shot Anomaly Detection in Financial Time Series
Muhammad Imad
Deep Learning for Anomaly Detection14:20 - 14:30
Tea/Coffee Break
14:30 - 15:15Intelligent Process
Chair: Dr Ian McChesney
Dr Federico Cruciani
Risk Language Models (RLM)
Dr Ayesha Abdul Majeed
Risk Assessment
Dr Vimal Dwivedi
Digital Twin
Gyan Wickremasinghe
Blockchain
Dr Olaide Oyelade
Idea Generation
Philip Neill
AI Driven Digital Workers15:15 - 15:55
Intelligent Analytics
Chair: Dr Barry Devereux
Dr Federico Cruciani
Privacy Enhancement Technologies (PET)
Ethan Bradley
Table Extraction
Dr Muhammad Roman
Retrieval Augmented Generation
Andrew Brown
Document Understanding and NLP
Tianyu Ren
Question Answering for Testing15:55 - 16:00
Prof Karen Rafferty & Prof Chris Nugent
Close
Analysis and Evaluation of Model features for Digital Workers and their correlation with adoptability.
Distributed ledger technology for applications beyond cryptocurrency (e.g. supply chain management, auditing management).
An AI powered DSS used to simplify and speed up beneficial ownership decision making with human in the loop involvement.
Development of a digital twin to enhance supply chain resilience, ensuring that operations can withstand and quickly recover from disruptions.
Accurate and intelligent financial market anomalies detection based on financial time series data and few-shot learning.
This project is focused on automatic idea generation using large language models (LLMs) and knowledge graph reasoning.
Optimisation of processes within the client onboarding.
Creation of PET asset will enable safe use of confidential data to generate new insights and enhance business services.
A language model-based question answering system to assess the user's understanding of the provided compliance report and generate a score and feedback.
Generating up-to-date knowledge from company documents using retrieval augmented generation (RAG).
This project aims to develop a risk modelling toolchain for complex financial systems, incorporating link prediction, automated rule generation, risk calculation and confidence level analysis.
Identification of the relevant detail within risk and control descriptions in an automated manner.
Developing an effective mixed reality environment for visualisation of complex audit data.
Automation of generating labelled data for training NLP models.
Generation of synthetic data will enable training of models for specific business domains with limited use of client data.
Using synthetic data and deep learning methods to improve the extraction of information from table images in the financial domain.