Computinghttp://hdl.handle.net/10547/6136972024-03-16T03:05:37Z2024-03-16T03:05:37ZEnhancing text comprehension via fusing pre-trained language model with knowledge graphQian, JingLi, GangminAtkinson, KatieYue, Yonghttp://hdl.handle.net/10547/6261982024-03-11T11:01:41Z2024-02-16T00:00:00ZEnhancing text comprehension via fusing pre-trained language model with knowledge graph
Qian, Jing; Li, Gangmin; Atkinson, Katie; Yue, Yong
Pre-trained language models (PLMs) such as BERT and GPTs capture rich linguistic and syntactic knowledge from pre-training over large-scale text corpora, which can be further fine-tuned for specific downstream tasks. However, these models still have limitations as they rely on knowledge gained from plain text and ignore structured knowledge such as knowledge graphs (KGs). Recently, there has been a growing trend of explicitly integrating KGs into PLMs to improve their performance. For instance, K-BERT incorporates KG triples as domain-specific supplements into input sentences. Nevertheless, we have observed that such methods do not consider the semantic relevance between the introduced knowledge and the original input sentence, leading to the issue of knowledge impurities. To address this issue, we propose a semantic matching-based approach that enriches the input text with knowledge extracted from an external KG. The architecture of our model comprises three components: the knowledge retriever (KR), the knowledge injector (KI), and the knowledge aggregator (KA). The KR, built upon the sentence representation learning model (i.e. CoSENT), retrieves triples with high semantic relevance to the input sentence from an external KG to alleviate the issue of knowledge impurities. The KI then integrates the retrieved triples from the KR into the input text by converting the original sentence into a knowledge tree with multiple branches, the knowledge tree is transformed into an accessible sequence of text that can be fed into the KA. Finally, the KA takes the flattened knowledge tree and passes it through an embedding layer and a masked Transformer encoder. We conducted extensive evaluations on eight datasets covering five text comprehension tasks, and the experimental results demonstrate that our approach exhibits competitive advantages over popular knowledge-enhanced PLMs such as K-BERT and ERNIE.
2024-02-16T00:00:00ZRevolutionising financial portfolio management: the non-stationary transformer's fusion of macroeconomic indicators and sentiment analysis in a deep reinforcement learning frameworkLiu, YuchenMikriukov, DaniilTjahyadi, Owen ChristopherLi, GangminPayne, Terry R.Yue, YongSiddique, KamranMan, Ka Lokhttp://hdl.handle.net/10547/6261832024-02-16T03:32:28Z2023-12-28T00:00:00ZRevolutionising financial portfolio management: the non-stationary transformer's fusion of macroeconomic indicators and sentiment analysis in a deep reinforcement learning framework
Liu, Yuchen; Mikriukov, Daniil; Tjahyadi, Owen Christopher; Li, Gangmin; Payne, Terry R.; Yue, Yong; Siddique, Kamran; Man, Ka Lok
In the evolving landscape of portfolio management (PM), the fusion of advanced machine learning techniques with traditional financial methodologies has opened new avenues for innovation. Our study introduces a cutting-edge model combining deep reinforcement learning (DRL) with a non-stationary transformer architecture. This model is designed to decode complex patterns in financial time-series data, enhancing portfolio management strategies with deeper insights and robustness. It effectively tackles the challenges of data heterogeneity and market uncertainty, key obstacles in PM. Our approach integrates key macroeconomic indicators and targeted news sentiment analysis into its framework, capturing a comprehensive picture of market dynamics. This amalgamation of varied data types addresses the multifaceted nature of financial markets, enhancing the model’s ability to navigate the complexities of asset management. Rigorous testing demonstrates the model’s efficacy, highlighting the benefits of blending diverse data sources and sophisticated algorithmic approaches in mastering the nuances of PM.
2023-12-28T00:00:00ZA Big Data maturity model for Electronic Health Records in hospitalsDaraghmeh, RaniaBrown, Raymondhttp://hdl.handle.net/10547/6261682024-02-05T13:41:04Z2021-07-26T00:00:00ZA Big Data maturity model for Electronic Health Records in hospitals
Daraghmeh, Rania; Brown, Raymond
The security vulnerabilities of Electronic Health Record (EHR) systems are unique due to their interoperable nature and their specific methods of use and storage. A critical review was undertaken of existing and generalized maturity models, addressing cyber and information security scenarios in the healthcare sector. The study investigated how hospitals address their EHR system security via a survey that was distributed to private hospitals in Amman, Jordan. As a result of the study of maturity models that target cybersecurity, information security, healthcare, and big data, it was found that maturity models in the healthcare sector do not provide tools for determining maturity for characteristics and processes specific to hospitals. A Design Science Research Methodology was adopted as a reliable and robust mechanism to develop an EHR maturity model (EHR-MM) for hospitals, to assess the security of the EHR specifically and effectively in their organizations. A case study was used for evaluation, employing a template that addresses the effectiveness of the maturity models through domain expert reviews, providing positive results for the proposed approach.
2021-07-26T00:00:00ZFlight delay prediction: data analysis and model developmentAnees, AzibHuang, Weihttp://hdl.handle.net/10547/6261672024-02-05T13:31:54Z2021-11-15T00:00:00ZFlight delay prediction: data analysis and model development
Anees, Azib; Huang, Wei
Flight delays in air transportation are a major concern that has adverse effects on the economy, the passengers, and the aviation industry. This matter critically requires an accurate estimation for future flight delays that can be implemented to improve airport operations and customer satisfaction. Having said that, a massive volume of data and an extreme number of parameters have restricted the way to build an accurate model. Many existing flight delay prediction methods are based on small samples and/or are complex to interpret with little or no opportunity for machine learning deployment. This paper develops a prediction model by analysing the data of domestic flights within the United States of America (USA). The proposed model gains insight into factors causing flight delays, cancellations and the relationship between departure and arrival delay using exploratory data analysis. In addition, Random Forest (RF) algorithm is used to train and test the big dataset to help the model development. A web application has also been developed to implement the model and the testing results are presented with the limitation discussed.
2021-11-15T00:00:00Z