Patient Record De-duplications in Health Information Systems
Funding Source: eHealth – UL Hospitals Group
Investigators:Hussain Mahdi (UL), Arash Joorabchi (UL), Brian McKeon (ULHG)
A duplicate medical record occurs when a single patient is associated with more than one medical record, resulting in partial duplications that only capture a portion of a patient's medical history. Treating patients based on incomplete medical history or profile, can cause serious errors and complications. Duplicate medical records can also negatively affect communications between healthcare providers and their patients. Healthcare organisations often use multiple information systems for clinical and administrative services, which increase their vulnerability to patient matching errors and duplicate medical records. Many duplicate medical records occur due to small errors and inconsistencies introduced in patient registration process, due to administrative existing protocols and procedures. Simple mistakes, like misspelling a patient's name can easily create duplicate records, resulting in inconsistent medical histories and information. Identifying individual patients becomes even more difficult when multiple patients share the same name and other personal information. This collaborative research project aims to assess the level and extent of patient record duplication in existing information systems of ULHG, and investigate the design and development of a automatic and robust record linkage system.
Fostering Library-Wikipedia Integration: Automatic Mapping of FAST Subject Headings to Wikipedia Articles
Funding Source: OCLC (Online Computer Library Center), Inc., OH, USA & ALISE (Association for Library and Information Science Education), WA, USA.
A full Wikipedia-Library integration would create a bi-directional link and flow of information and users between the Wikipedia and libraries. In such environment, information seekers may start their search activities from either of these sources and traverse back and forth as needed. Linking Wikipedia articles to the records of relevant library materials will give information seekers the option to readily acquire lists of library resources which provide them with more in-depth knowledge on their subjects of interest. In this paradigm each Wikipedia article will be linked to the records of relevant materials in a global union catalogue of libraries around the world, WorldCat.org, which in turn provides bibliographic information on the materials of interest and directs information seekers to their local libraries, where they can access those materials. Introduction of this new Wikipedia-Library information seeking paradigm will consequently improve the visibility of library resources which are currently overlooked to a large extend by those information consumers with lower information literacy skills. This project investigates the development of a mapping algorithm for automatic mapping of FAST subject headings to their equivalent Wikipedia articles/topics. The mapping algorithm will deploy various text mining techniques such as string matching, explicit semantic analysis, and citation analysis to find the best matching article for a given FAST term.
KEYSTONE (Semantic Keyword-based Search on Structured Data Sources)
The main goal of this Action project is to launch and establish a cooperative network of researchers, practitioners, and application domain specialists working in fields related to semantic data management, the Semantic Web, information retrieval, artificial intelligence, machine learning and natural language processing, to enable research activities and technology transfer in the area of keyword-based search over structured data sources. The coordination effort will promote the development of a new paradigm that provides users with keyword-based search capabilities for structured data sources as they currently do with documents. Furthermore, it will exploit the structured nature of data sources in defining complex query execution plans by combining partial contributions from different sources.
Automatic Arabic Text Classification using Bag of Words and Bag of Concepts
Automatic Text Classification (ATC) is one of the most important tasks in data mining for organizing information and knowledge discovery. The goal of ATC is to alleviate the need of manually organizing large collections of text documents, which is done by assigning one or more predefined categories to a given textual document via applying appropriate natural language processing techniques. The classification process involves three components: text pre-processing, text representation and the classifier which is built using one of the Machine Learning (ML) algorithms. In general, all existing text representations are based on the Bag-of-Words and Bag-of-Concepts models and their variations. This on-going project focuses on investigating combining words and concepts for text representations for Arabic Automatic Text Classification (ATC) and the impact of these representation models on the accuracy of the classification, when used with various stemming methods and classifiers. It aims to design and develop an Arabic ATC prototype system and assess the roles of its main components on the classification accuracy.
Remote Human Presence Detection System Using a Novel Electrostatically-Enhanced Displacement Current Sensor
Funding Source: Institute for Protection & Security of Citizen (IPSC) - Joint Research Centre (JRC) of the EU, Ispra, Italy
Investigators: Hussain Mahdi (UL), Lorenzo Faggion (IPSC), Constantin Coutsomitros (IPSC)
Over the last few years, increasing attention has been paid to the research field of remote detection of human electrophysiological and other bio activity related signals. Such an interest is mainly due to a growing need to develop new systems for the contact-less monitoring of human vital parameters for clinical and healthcare purposes. At the same time, military and humanitarian activities have renewed the desire of having human presence identification systems based on biological parameters, both for security and rescue activities. The aim of this collaborative research project is to provide a potential solution to the latter needs by investigating the development of a low cost displacement current sensor for non contact, non invasive detection of electrophysiological and other bio-activity related signals, well suited for applications requiring remote and portable means for detection of human presence.
An Investigation of the Impact of Learner Support Initiatives on Retention in ICT Programmes
Funding Source: National Forum for the Enhancement of Teaching and Learning in Higher Education, Dublin, Ireland
Investigators: Hussain Mahdi (UL), Michael English (UL), Arash Joorabchi (UL), Clem O'Donnell (UL), James Murphy (UL)
Student retention is of significant importance within higher education (HE) and ranks as one of the most studied areas in HE. Non-progression of students through their programme of study can have significant emotional and financial impacts on the individual themselves. Non-progression also impacts the level and quality of services provided by third level institutions as well as having significant financial implications for these institutions. The aim of this research project is to investigate the relationship between students’ participation in and engagement with the learner support services and progression of first-year students in undergraduate ICT programmes in partner institutions. The research focuses on examining how these services align with the characteristics of best practice in student retention, and evaluating the relationship and, hence, the impact of our SALS services on the progression of students on ICT programmes, as a major indicator of retention in the field. This is to be achieved via gathering data pertaining to first year students of core ICT programmes at partner institutions over a number academic years, and conducting appropriate learning analytics. This is complemented with gathering and analysing data regarding the current first-year students' choices and expectations of their study programmes/career, whether they contemplated leaving, and factors that helped their continuing.
Unsupervised Automatic Topical Indexing of Scientific Documents According to Library Controlled Vocabularies
Funding Source: OCLC (Online Computer Library Center), Inc., OH, USA & ALISE (Association for Library and Information Science Education), WA, USA.
The goal of this project was to develop a new unsupervised approach for automatic topical indexing of scientific literature according to library classification resources, with an emphasis on providing an easy-to-implement and efficient alternative to the machine learning-based approach for practitioners in the digital library community. The unsupervised nature of this approach allows practitioners to develop effective Automatic Text Classification (ATC) systems for scientific digital libraries without encountering the obstacles associated with the machine learning based approach. In effect, the project investigated the development of a new trend of ATC systems that are based on leveraging: (a) the intellectual work that has been put into developing and maintaining extensive resources and systems for classifying and organizing the vast amount of materials archived in conventional libraries and; (b) the intellectual effort of expert library cataloguers who have used the above classification resources and systems to manually classify and index millions of books and other materials in libraries around the world over the last century.
Completed Research Projects
Improving Primary Science Learning and Teaching through Inquiry and Tutorial Dialogues
Funding Source: HEA Strategic Innovation Fund Cycle II + IRCHSS
Investigators: Hussain Mahdi (UL), Jonathan Dunne (UL), John O’Reilly (UL)
DSP Based Controller for Multi-Rail DC-DC Converter Systems
Funding Source: IRCSET
Investigators: Hussain Mahdi (UL), James Mooney (UL), Mark Halton (UL)
Skills SuperStore: A Learning Content Management System For the Development of Students' Study Skills
