Baig Mansoor Ali, Alzahrani Somayah J
King Faisal Specialist Hospital & Research Center, Riyadh, Saudi Arabia.
Department of Biostatistics Epidemiology & Scientific Computing, KFSHRC, Riyadh, Saudi Arabia.
Stud Health Technol Inform. 2019 Jul 4;262:43-46. doi: 10.3233/SHTI190012.
The buzz words 'Data Science' and 'Data Scientist' are trending high in this age of information. The boundaries are still undefined, the exact skill sets are unclear, and the job description is still murky. This is an attempt to identify some mandatory or desired skills based on what data science demands from a data scientist. A very generic job description for a data scientist is 'A person who can perform advanced analytics on the institutional data', this gives a very unclear picture to the decision maker to identify the right resources within their data science activity. Practically the data scientist should be the one who can understand and moreover be involved with the data life cycle starting from inception > collection > operation > extraction > observation > preparation > description > prediction > prescription > Archival. Each of these aspects of data has a science behind it. An old team 'Jack of all trades' briefly defines this job description. A good data scientist essentially needs to be a good programmer, a good business/system/data analyst, a good statistician, one who can seamlessly visualize data, and is empowered with a vision to use and apply the necessary tools, techniques and methodologies in a scientific and applicable realistic way. Healthcare/Research environment is a complicated vertical when it comes to data, hence having domain knowledge is almost critical, complying with aspects of data governance such as patient privacy, consent, ethics etc.
在这个信息时代，流行语“数据科学”和“数据科学家”热度极高。其边界仍未明确界定，确切的技能组合尚不清晰，工作描述也依然模糊。本文旨在根据数据科学对数据科学家的要求，确定一些必备或理想的技能。数据科学家一个非常笼统的工作描述是“能够对机构数据进行高级分析的人”，这让决策者在确定其数据科学活动中的合适资源时，很难有清晰的概念。实际上，数据科学家应该是这样一个人，他能够理解并参与从初始>收集>操作>提取>观察>准备>描述>预测>规定>存档的数据生命周期。数据的每一个方面背后都有一门科学。一个老掉牙的说法“万事通”能简要定义这个工作描述。一个优秀的数据科学家本质上需要是一名优秀的程序员、一名优秀的业务/系统/数据分析师、一名优秀的统计学家，能够无缝地可视化数据，并具备以科学且适用的现实方式使用和应用必要工具、技术和方法的眼光。在数据方面，医疗保健/研究环境是一个复杂的领域，因此具备领域知识几乎至关重要，要遵守数据治理的各个方面，如患者隐私、同意、伦理等。
