Uncertainty in big data analytics: survey, opportunities, and challenges
Uncertainty in big data analytics: survey, opportunities, and challenges
复制标题
大数据分析中的不确定性：调查、机遇与挑战
DOI：
10.1186/s40537-019-0206-3
复制
发表时间：
2019-06-04
影响因子：
8.1
通讯作者：
Bowers, Kate M.
中科院分区：
文献类型：
--
作者：
Hariri, Reihaneh H.;Fredericks, Erik M.;Bowers, Kate M.
大数据分析已经得到了学术界和工业界的广泛关注，因为对了解大规模数据集趋势的需求不断增加。传感器网络、网络物理系统和物联网（IoT）的普及使数据收集（包括医疗保健、社交媒体、智慧城市、农业、金融、教育等）的规模大大增加。然而，从传感器、社交媒体、财务记录等收集的数据由于噪声、不完整性和不一致性而具有固有的不确定性。分析如此大量的数据需要先进的分析技术，以高精度和先进的决策战略有效地审查和/或预测未来的行动方针。随着数据的数量、种类和速度的增加，内部固有的不确定性也在增加，导致对结果分析过程和由此做出的决策缺乏信心。与传统的数据技术和平台相比，人工智能技术（包括机器学习、自然语言处理和计算智能）在大数据分析中提供了更准确、更快和可扩展的结果。以前对大数据分析进行的研究和调查往往集中在一两种技术或特定的应用领域。然而，在应用于大数据分析的不确定性领域以及应用于数据集的人工智能技术方面，几乎没有做过什么工作。本文回顾了以前在大数据分析方面的工作，并讨论了识别和减轻该领域不确定性的开放挑战和未来方向。
Big data analytics has gained wide attention from both academia and industry as the demand for understanding trends in massive datasets increases. Recent developments in sensor networks, cyber-physical systems, and the ubiquity of the Internet of Things (IoT) have increased the collection of data (including health care, social media, smart cities, agriculture, finance, education, and more) to an enormous scale. However, the data collected from sensors, social media, financial records, etc. is inherently uncertain due to noise, incompleteness, and inconsistency. The analysis of such massive amounts of data requires advanced analytical techniques for efficiently reviewing and/or predicting future courses of action with high precision and advanced decision-making strategies. As the amount, variety, and speed of data increases, so too does the uncertainty inherent within, leading to a lack of confidence in the resulting analytics process and decisions made thereof. In comparison to traditional data techniques and platforms, artificial intelligence techniques (including machine learning, natural language processing, and computational intelligence) provide more accurate, faster, and scalable results in big data analytics. Previous research and surveys conducted on big data analytics tend to focus on one or two techniques or specific application domains. However, little work has been done in the field of uncertainty when applied to big data analytics as well as in the artificial intelligence techniques applied to the datasets. This article reviews previous work in big data analytics and presents a discussion of open challenges and future directions for recognizing and mitigating uncertainty in this domain.