Page 233 - Contributed Paper Session (CPS) - Volume 7

CPS2069 Pamela Kaye A. T.
4. The Challenges Ahead for Big Data Application in the Philippine
Central Bank
In a Harvard Business Review (HBR) article entitled “Big Data at Work:
Dispelling the Myths, Uncovering the Opportunities”, one of the key
recommendations is to adapt the DELLTA Framework in incorporating Big
17
Data in business operations. The framework suggests the two (2) phases of
Discovery and Production, emphasizing that Big Data initiatives must be
unified across the entity through effective leadership. Targets must be well-
defined, technological infrastructures must be suited for Big Data, and data
scientists must be hired and empowered.
For the Philippine central bank, the extensive benefits on the use of Big
Data range from its potential applications as innovative approach for
economic and financial analysis to unconventional leading economic and
financial indicator that will support its core mandates. However, institutions
exploring the possible applications of Big Data will have to face the
concomitant challenges that include the following:
1. External Data Access: One of the major sources of Big Data are
proprietary information from the private sector. Given the central bank’s
limited clout on gathering data, it needs to enter into bilateral agreements
with its identified data sources, or solicit the support of the national
statistics agency in accessing the needed information. Some proprietary
information are made available at a cost and is fast becoming a highly-
valuable asset to the private sector. The cost of accessing Big Data may
increase over time and negotiations in establishing public-private
partnership agreements is crucial in the near-term.
2. Data Quality: Ensuring and verifying the reliability of the
statistics/indicators derived and obtained from Big Data is important to
minimize, if not totally eliminate, the risks associated with the use of Big
Data. From a statistical standpoint, Big Data may not necessarily cover
and represent random samples of the target population. Hence,
thorough examination on the soundness of the methodology and
metadata must be undertaken to ensure data quality. Continuity of the
data series may also be a concern since Big Data are mostly sourced from
the private sector (as a by-product of their daily business operations) and
they operate in a continually changing competitive environment. Hence,
statistical comparability of time series could potentially be affected.
Moreover, outliers and missing information in the Big Data time series
must be clearly detected in order to be statistically resolved with
imputations or sound estimates.

17 DELLTA Framework: Data, Enterprise, Leadership, Targets, Technology, and Analytics
220 | I S I W S C 2 0 1 9