- Advanced Text Analysis Techniques
- Text and Document Classification Technologies
- Topic Modeling
- Web Data Mining and Analysis
- Wikis in Education and Collaboration
- Educational Technology and Assessment
- Natural Language Processing Techniques
- Social Media in Health Education
- Biomedical Text Mining and Ontologies
- Open Education and E-Learning
- Health Literacy and Information Accessibility
- Health Sciences Research and Education
- Information Retrieval and Search Behavior
- Seismic Imaging and Inversion Techniques
- Modular Robots and Swarm Intelligence
- Student Assessment and Feedback
- Online and Blended Learning
- Reservoir Engineering and Simulation Methods
- Robotic Path Planning Algorithms
- Semantic Web and Ontologies
- Web and Library Services
- Multimodal Machine Learning Applications
- Intelligent Tutoring Systems and Adaptive Learning
- Machine Learning in Bioinformatics
- Expert finding and Q&A systems
University of Limerick
2011-2022
University College Dublin
2020
This article describes an unsupervised approach for automatic classification of scientific literature archived in digital libraries and repositories according to a standard library scheme. The method is based on identifying all the references cited document be classified and, using subject metadata extracted as catalogued existing conventional libraries, inferring most probable class itself with help weighting mechanism. We have demonstrated application proposed assessed its performance by...
This paper introduces a new approach to creating text representations and apply it standard classification collections. The is based on supplementing the well-known Bag-of-Words (BOW) representational scheme with concept-based representation that utilises Wikipedia as knowledge base. proposed are used generate Vector Space Model, which in turn fed into Support Machine classifier categorise collection of textual documents from two publically available datasets. Experimental results for...
The uncontrolled nature of user-assigned tags makes them prone to various inconsistencies caused by spelling variations, synonyms, acronyms and hyponyms. These in turn lead some the common problems associated with use folksonomies such as tag explosion phenomenon. Mapping user their corresponding Wikipedia articles, well-formed concepts, offers multifaceted benefits process subject metadata generation management a wide range online environments. include normalization inconsistencies,...
Purpose – The use of social media and in particular community Question Answering (Q & A) websites by learners has increased significantly recent years. vast amounts data posted on these sites provide an opportunity to investigate the topics under discussion those receiving most attention. purpose this paper is automatically analyse content a popular computer programming Q A website, StackOverflow (SO), determine exact As, narrow down their categories help subject difficulties learners....
Topical annotation of documents with keyphrases is a proven method for revealing the subject scientific and research to both human readers information retrieval systems. This article describes machine learning-based keyphrase that utilizes Wikipedia as thesaurus candidate selection from documents’ content. We have devised set 20 statistical, positional semantical features phrases capture reflect various properties those candidates highest keyphraseness probability. first introduce simple...
Topical indexing of documents with keyphrases is a common method used for revealing the subject scientific and research to both human readers information retrieval tools, such as search engines. However, that are manually indexed still in minority. This article describes new unsupervised automatic keyphrase extraction from which yields performance on par indexers. The based identifying references cited document be and, using assigned those references, generating set high-likelihood document....
This paper introduces a set of new approaches for text representation automatic classification Arabic textual documents. These are based on combining the well-known Bag-of-Words (BOW) and Bag-of-Concepts (BOC) schemes utilizing Wikipedia as knowledge base. The proposed representations used to generate vector space model, which in turn is fed into classifier categorize collection Three different machine learning classifiers have been utilized this work. Performance models evaluated comparison...
Purpose – This paper aims to report on the design and development of a new approach for automatic classification subject indexing research documents in scientific digital libraries repositories (DLR) according library controlled vocabularies such as DDC FAST. Design/methodology/approach The proposed concept matching-based (CMA) detects key Wikipedia concepts occurring document searches OPACs conventional via querying WorldCat database retrieve set MARC records which share one or more...
In this article, we first argue the importance and timely need of linking libraries Wikipedia for improving quality their services to information consumers, as such linkage will enrich articles at same time increase visibility library resources which are currently overlooked a large degree. We then describe development an automatic system subject indexing metadata records with concepts important step towards library–Wikipedia integration. The proposed is based on identifying all occurring in...
Purpose Linking libraries and Wikipedia can significantly improve the quality of services provided by these two major silos knowledge. Such linkage would enrich articles at same time increase visibility library resources. To this end, purpose paper is to describe design development a software system for automatic mapping FAST subject headings, used index materials, their corresponding in Wikipedia. Design/methodology/approach The proposed works first detecting all candidate concepts...
Medical information on English Wikipedia was accessed over 2 billion times in 2018. Our goal to develop an automated system assist volunteers improve articles with high-quality sources from journals such as The Cochrane Library. We created indexing by linking available reviews the library disease-related and evaluating relationship between quality importance of these number relevant cited reviews. first conducted a bibliometric analysis, identifying relevant/cited Citations were thematically...
In Natural Language Processing, automatic short answer grading remains a necessary launch-pad for the analysis of human responses in blended learning setting. This study presents pre-trained neural language models that use context dependent Sentence-Transformers to automatically grade student with two different input settings. It is found these achieves promising results when compared conventional Bidirectional Encoder Representation Transformer, (BERT), approaches applying various text...
With the exponential growth of Arabic text in digital form, need for efficient organization, navigation and browsing large amounts documents has increased. Text Classification (TC) is one important subfields data mining. The Bag-of-Words (BOW) representation model, which traditional way to represent TC, only takes into account frequency term occurrence within a document. Therefore, it ignores semantic relationships between terms treats synonymous words independently. In order address this...
Background The widespread availability of internet-connected smart devices in the health care setting has potential to improve delivery research evidence pathway and fulfill professionals’ information needs. Objective This study aims evaluate frequency with which physiotherapists experience needs, capacity digital resources these specific types they use do so. Methods A total 38 participants (all practicing physiotherapists; 19 females, males) were randomly assigned complete three...
Purpose With the significant growth in electronic education materials such as syllabus documents and lecture notes, available on internet intranets, there is a need for robust central repositories of to allow both educators learners conveniently share, search access them. The purpose this paper report work develop national repository course syllabi Ireland. Design/methodology/approach describes prototype system higher Ireland, which has been developed by utilising number information...
Measurement of words semantic relatedness plays an important role in a wide range natural language processing and information retrieval applications, such as full-text search, summarization, classification clustering. In this paper, we propose easy to implement low-cost method for estimating relatedness. The proposed is based on the utilization temporal footprints found publicly available corpora Google Books Ngrams (GBN), knowledge bases Wikipedia. extracted are represented time series,...
Appears in: ICERI2023 Proceedings Publication year: 2023Pages: 8362-8367ISBN: 978-84-09-55942-8ISSN: 2340-1095doi: 10.21125/iceri.2023.2138Conference name: 16th annual International Conference of Education, Research and InnovationDates: 13-15 November, 2023Location: Seville, Spain