Search results

The purpose of this paper is to describe a new approach to sentence representation learning leading to text classification using Bidirectional Encoder Representations from Transformers (BERT) embeddings. This work proposes a novel BERT-convolutional neural network (CNN)-based model for sentence representation learning and text classification. The proposed model can be used by industries that work in the area of classification of similarity scores between the texts and sentiments and opinion analysis.

The approach developed is based on the use of the BERT model to provide distinct features from its transformer encoder layers to the CNNs to achieve multi-layer feature fusion. To achieve multi-layer feature fusion, the distinct feature vectors of the last three layers of the BERT are passed to three separate CNN layers to generate a rich feature representation that can be used for extracting the keywords in the sentences. For sentence representation learning and text classification, the proposed model is trained and tested on the Stanford Sentiment Treebank-2 (SST-2) data set for sentiment analysis and the Quora Question Pair (QQP) data set for sentence classification. To obtain benchmark results, a selective training approach has been applied with the proposed model.

On the SST-2 data set, the proposed model achieved an accuracy of 92.90%, whereas, on the QQP data set, it achieved an accuracy of 91.51%. For other evaluation metrics such as precision, recall and F1 Score, the results obtained are overwhelming. The results with the proposed model are 1.17%–1.2% better as compared to the original BERT model on the SST-2 and QQP data sets.

The novelty of the proposed model lies in the multi-layer feature fusion between the last three layers of the BERT model with CNN layers and the selective training approach based on gated pruning to achieve benchmark results.

Through harnessing the benefits of the internet, e-learning systems provide flexible learning opportunities that can be delivered at a fixed cost at a time and place to suit the user. As such, e-learning systems can allow students to learn at their own pace while also being suitable for both distance and classroom-based learning activities. Adaptive educational hypermedia systems are e-learning systems that employ artificial intelligence. They deliver personalised online learning interventions that extend electronic learning experiences beyond a mere computerised book through the use of intelligence that adapts the content presented to a user according to a range of factors including individual needs, learning styles and existing knowledge. The purpose of this paper is to describe a novel adaptive e-learning system called dyslexia adaptive e-learning management system (DAELMS). For the purpose of this paper, the term DAELMS will be employed to describe the overall e-learning system that incorporates the required functionality to adapt to students’ learning styles and dyslexia type.

The DAELMS is a complex system that will require a significant amount of time and expertise in knowledge engineering and formatting (i.e. dyslexia type, learning styles, domain knowledge) to develop. One of the most effective methods of approaching this complex task is to formalise the development of a DAELMS that can be applied to different learning styles models and education domains. Four distinct phases of development are proposed for creating the DAELMS. In this paper, we will discuss Phase 3 which is the implementation and some adaption algorithms while in future papers will discuss the other phases.

An experimental study was conducted to validate the proposed generic methodology and the architecture of the DAELMS. The system has been evaluated by group of university students studying a Computer Science related majors. The evaluation results proves that when the system provide the user with learning materials matches their learning style or dyslexia type it enhances their learning outcomes.

The DAELMS correlates each given dyslexia type with its associated preferred learning style and subsequently adapts the learning material presented to the student. The DAELMS represents an adaptive e-learning system that incorporates several personalisation options including navigation, structure of curriculum, presentation, guidance and assistive technologies that are designed to ensure the learning experience is directly aligned with the user's dyslexia type and associated preferred learning style.

The purpose of this study is to introduce new tools to develop a more precise and focused bibliometric analysis on the field of digitalization in healthcare management. Furthermore, this study aims to provide an overview of the existing resources in healthcare management and education and other developing interdisciplinary fields.

This work uses bibliometric analysis to conduct a comprehensive review to map the use of the unified theory of acceptance and use of technology (UTAUT) and the unified theory of acceptance and use of technology 2 (UTAUT2) research models in healthcare academic studies. Bibliometric studies are considered an important tool to evaluate research studies and to gain a comprehensive view of the state of the art.

Although UTAUT dates to 2003, our bibliometric analysis reveals that only since 2016 has the model, together with UTAUT2 (2012), had relevant application in the literature. Nonetheless, studies have shown that UTAUT and UTAUT2 are particularly suitable for understanding the reasons that underlie the adoption and non-adoption choices of eHealth services. Further, this study highlights the lack of a multidisciplinary approach in the implementation of eHealth services. Equally significant is the fact that many studies have focused on the acceptance and the adoption of eHealth services by end users, whereas very few have focused on the level of acceptance of healthcare professionals.

To the best of the authors’ knowledge, this is the first study to conduct a bibliometric analysis of technology acceptance and adoption by using advanced tools that were conceived specifically for this purpose. In addition, the examination was not limited to a certain era and aimed to give a worldwide overview of eHealth service acceptance and adoption.