Big Data and Data-Driven Healthcare Systems

Alexander, Cheryl Ann; Wang, Lidong

doi:10.12691/jbms-6-3-7

Cited by 3 publications

(2 citation statements)

References 26 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…It is possible to find specific applications of big data to almost any industry: health-care, supply chain, agribusiness, biology, environmental, transportation, public administration, telecommunications, logistics, safety, energy, travel, legal, among other (HURWITZ;KAUFMAN;BOWLES, 2015;GARTNER, 2015;XU;HE;LI, 2014;LOKERS et al, 2016;BECHHOFER et al, 2013;BALLIU et al, 2016;CHEN et al, 2017;ALEXANDER;WANG, 2018;LI, 2017;SEELE, 2017). Knowledge developed during a data science initiative for a particular industry may be replicated across different industries.…”

Section: Interdisciplinarity Of Resulting Applications From Data Sciementioning

confidence: 99%

Iinterdisciplinarity in Data Science over Big Data: findings for mining industry

Pinto¹,

Cardoso²,

Pinheiro³

et al. 2019

I&S

View full text Add to dashboard Cite

Data Science and Big Data are leveraged by businesses in many ways to improve operational and strategic capabilities, and ultimately, to positively impact corporate financial performance. However, there are challenges related to Big Data, such as modelling, new paradigms and novel architectures that require original approaches to address data complexities. In the specific case of iron ore mining industry, there is a considerable pressure at present to reduce costs due to the recent major fall in iron ore prices. This study discusses if an interdisciplinary approach could help mining industries to extract the most of data science initiatives over big data. In this study we applied a narrative literature review method to briefly present a chronological review of disciplines and interdisciplinarity as well as the evolution of data science over big data. Then we discussed: 1) the importance of involving people from different profiles; 2) the relevance of technology transfer inside computing research field; 3) the requirements for integrating so many different people and technologies in such initiative. We concluded that achieving results with Data Science initiative over big data is not related to a single knowledge area, especially in mining industries.

show abstract

Section: Interdisciplinarity Of Resulting Applications From Data Sciementioning

confidence: 99%

Iinterdisciplinarity in Data Science over Big Data: findings for mining industry

Pinto¹,

Cardoso²,

Pinheiro³

et al. 2019

I&S

View full text Add to dashboard Cite

show abstract

“…While the majority of clinical data is made up of structured information (Jee and Kim, 2013), which can often be readily integrated into data models for research, there is a significant amount of semi-structured and unstructured data which is increasingly being targeted by machine learning practitioners for analysis. As a general rule, this unstructured data is more difficult to analyse due to an absence of a standardised data model (Ann Alexander and Wang, 2018). Unstructured clinical data includes a variety of media, such as video, audio, image and text-based data, with the majority of such data being made up of text and images.…”

Section: Introductionmentioning

confidence: 99%

Analysing Representations of Memory Impairment in a Clinical Notes Classification Model

Ormerod¹,

Rincón²,

Robertson³

et al. 2019

Proceedings of the 18th BioNLP Workshop and Shared Task

View full text Add to dashboard Cite

Despite recent advances in the application of deep neural networks to various kinds of medical data, extracting information from unstructured textual sources remains a challenging task. The challenges of training and interpreting document classification models are amplified when dealing with small and highly technical datasets, as are common in the clinical domain. Using a dataset of de-identified clinical letters gathered at a memory clinic, we construct several recurrent neural network models for letter classification, and evaluate them on their ability to build meaningful representations of the documents and predict patients' diagnoses. Additionally, we probe sentence embedding models in order to build a humaninterpretable representation of the neural network's features, using a simple and intuitive technique based on perturbative approaches to sentence importance. In addition to showing which sentences in a document are most informative about the patient's condition, this method reveals the types of sentences that lead the model to make incorrect diagnoses. Furthermore, we identify clusters of sentences in the embedding space that correlate strongly with importance scores for each clinical diagnosis class.

show abstract