2020年IFC报告：金融区块链在大数据与AI计算平台的应用前瞻
需积分: 2 178 浏览量
更新于2024-07-16
收藏 365KB PDF 举报
本篇报告名为《IFC Report on Computing Platforms for Big Data Analytics and Artificial Intelligence》（2020年4月），是由国际货币基金组织Irving Fisher Committee on Central Bank Statistics（IFC）与瑞士银行清算银行（BIS）共同编撰，由意大利银行的Giuseppe Bruno等人参与，旨在探讨在金融系统中应用区块链技术的最新进展和案例研究。这份报告是银行业界发布的首份区块链白皮书，对于银行同行和相关领域的从业者来说，具有显著的参考价值和指导意义。
在报告中，作者们基于2019年1月在罗马举办的由意大利银行主办、IFC支持的关于“大数据和机器学习计算平台”的研讨会成果。研讨会详细讨论了如何利用先进的云计算平台来处理海量数据，提升金融行业的效率和安全性，特别是在处理涉及金融交易、客户行为分析、风险管理等领域的数据方面。报告强调，区块链作为一种分布式账本技术，正被用于增强金融系统的透明度、减少欺诈和降低成本，这表明它正在逐渐成为金融业转型的重要工具。
值得注意的是，尽管报告的观点主要代表了作者，但并不一定反映IFC、其成员、意大利银行、BIS或参会机构的具体立场。报告全文可以在BIS的官方网站上获取，所有引用和翻译须注明原出处，并尊重版权，该报告的在线版本的国际标准连续出版物编号(ISBN)和在线发行识别号(ISNN)也一并列出。
通过阅读这份报告，读者将了解到区块链在金融领域的具体应用场景，包括但不限于支付结算、证券交易、供应链金融以及身份验证等方面的技术挑战和解决方案。此外，报告还可能涉及对区块链技术在隐私保护、监管合规性和标准化等方面的考量，这些都是当前和未来金融行业采用区块链技术时所面临的关键议题。整体而言，这份报告为金融机构和相关从业者提供了深入了解区块链在金融领域潜力和实践的重要参考资料。
Computing platforms for big data analytics and artificial intelligence
3
underlying computing platforms. Section 5 draws some general conclusions looking
forward.
1.2 Increasing interest in big data information…
The term big data is relatively vague and refers in general to the huge proliferation
of information brought by the so-called data revolution and generated by social
media, web-based activities, machine sensors, financial, administrative or business
operations, etc. One well known characterisation of this concept relates to the “Vs” of
big data, such as volume (ie number of records and attributes), velocity (speed of data
production eg tick data) and variety (eg structure and format of the data set) (Laney
(2001)).
The financial big data sets central banks and international financial institutions
more specifically are facing typically comprise four main sorts of sources: internet-
based indicators, commercial records, financial market indicators and administrative
registers. Their analysis can provide many opportunities, such as: the filling of existing
data gaps; the collection of new data that can flexibly complement “traditional”
survey- and census-based official statistics; the availability of almost real-time data,
opening up the possibility of getting more timely economic signals and bridging time
lags in statistical publication; and the possibility to apply brand new methodologies
to extract original knowledge (Bholat (2015)). These opportunities can greatly
facilitate the conduct of central bank policies, for instance by enhancing economic
forecasts, assisting micro supervision tasks, conducting new types of financial stability
assessments and obtaining more rapid feedback on policy decisions (IFC (2017)).
In addition to capturing structured transactional data that sit at the core of
central banks‘ work, various and innovative big data sets – such as granular data
collected from the web (eg page views, online ads, search indicators), sentiment
indicators derived from social media (eg Facebook, LinkedIn, Twitter), payment
transactions and traditional textual information processed with new big data tools
(eg policy announcements, speeches, emails, press articles) – can be mobilised in
practice to facilitate central banks’ day-to-day work in various areas, including the
production of statistical information, economic analysis and research supporting
monetary and financial stability policies, and supervisory tasks (Tissot (2019)).
4
As regards statistical work, big data can enhance the quality of existing, “more
conventional” statistics, or even complement them with new types of indicators. One
good example, particularly important in view of central banks’ price stability
mandates, relates to the measurement and analysis of inflation patterns. The use of
new big data-type collection methods has expanded significantly, allowing more
detailed price information, greater product variability and a larger number of
4
Compared with counterpart institutions like NSOs, central banks’ interest has so far been rather
limited for the wider range of unstructured data comprising geospatial data (eg mobile phones,
satellite images), smart grid data (eg information derived on the users connected to electricity
networks) and more generally all types of sensor data that result from the recording of some type of
input from the environment (eg temperature); see Meeting of the Expert Group on International
Statistical Classifications (2015) for a general review of the wide range of what can constitute a big
data source. However, central banks’ interest may well expand for these types of data too as new
technologies/ideas emerge.
剩余21页未读，继续阅读
2022-09-23 上传
156 浏览量
455 浏览量
241 浏览量
2022-09-24 上传
115 浏览量
105 浏览量
2022-09-20 上传
