Write A Review of An Article In 500-600 Words

Local Government Unit Analytics for Program

Planning & Policy-Making in Caloocan City

Christine Diane Lim

Information Technology Department

De La Salle University

Taft Manila, Philippines

[email protected]

Gian Carlo Roxas

Information Technology Department

De La Salle University

Taft Manila, Philippines

[email protected]

Shermaine Sy

Information Technology Department

De La Salle University

Taft Manila, Philippines

[email protected]

Geraldine Atayan

Information Technology Department

De La Salle University

Taft Manila, Philippines

[email protected]

Abstract— The purpose of this paper is to demonstrate the

value of an information system for a Philippine local government

unit (LGU) namely, the Caloocan City Hall’s Planning

Department Social Division. With difficulties in identifying

relationships between various issues raised surrounding

education and health, appropriate interventions or programs are

not given. Thus, using IS benchmarking and agile methodology,

the researchers proposed a system design framework in

developing a data management and analytics system that is

comprised of modules and visualizations customized to the needs

of a Philippine LGU in uncovering relationships with their city

hall data to define root causes of issues and discover trends and

outliers useful in program planning for improvement of the

educational and social welfare of the city.

Index Terms— Local Government Unit; Planning; Data

Analytics ; E- Government; ICT4D (key words)

I. INTRODUCTION

More than research in the business field, ICT has also

influenced and improved public service and socio-economic

development in urban and rural cities. IS solutions had become

contributing factors in resolving community challenges by

presenting potentials of how technological applications can be

a medium to address issues defined by UN’s Millennium

Development goals [1]. ICT coupled with urban planning, can

provide an improved life quality as it has become a vital part

to assist government functions on their program planning for

designing what is needed for the city based on the data they

have surrounding schools, hospitals, and health centers.

This research explores the current situation of Caloocan

City’s Development and Planning division on their program

formulations towards health and education. Currently, multiple

data are being collected at three domains- demographics,

education and health. Demographics include population

perspective (including data on total population by age, group,

and sex and distribution), historical growth population,

household population of 5 years old and over by highest

grade/year level completed. Education domain would collect

data on total enrollments per school type, teacher and

classroom ratios, while health data would include distribution

of elementary school children at each district and total number

of doctors, nurses, and midwives to name a few. Faced with

budget constraints in acquiring data analytics software to assist

in finding meaning and extrapolations of results in terms of the

data collected and its relationships, this research contributes a

customized IS solution in developing a localized government

management system that would facilitate in determining the

appropriate program interventions based from the visualization

and correlations among the city data.

The motive for conducting a study in Caloocan City stems

from the fact that there has been no research in this sector

before. With the aid of this proposed system and its analytic

capabilities, not only will insight obtained from its

functionalities can better govern and provide more adequate

services to the community, but also, other local government

units may find value and importance in designing their city

planning division information systems in line with this system

development framework. With the dynamic nature of today’s

community needs, it is critical for government officials to

understand capabilities of technology to establish better

decision making from community issues that continue to

emerge.

II. REVIEW OF RELATED WORKS

The implementation of e-government in developing

nations face challenges in terms of its adoption and e-

leadership. Studies suggest this is influenced not just by

Proc. of the 2017 IEEE Region 10 Conference (TENCON), Malaysia, November 5-8, 2017

978-1-5090-1134-6/17/$31.00 ©2017 IEEE 2830

factors on human, social, cultural, and economic aspects but

may also root from its regulatory perspective [2]. Hence,

related works on urban planning were considered in the

development of the system design framework, specifically

concerning public officials and policy makers as the

primary stakeholders that appreciated analysis of public and

community data with the use of IS/ICT.

The case study conducted by Rahman in Bangladesh e-

government explores the need for an analytical hierarchy

process in which political leaders could easily conclude and

highlight that resource-related constraints were the most

critical factors in driving institutional determinants on their

e-government policy planning system. Data collected were

grouped into four main categories namely intuitional

(includes administrative leadership and commitment index),

resource (e.g. technical, human resources, and financial

data), access (education and awareness), and legal (security

and privacy) [3]. The Australian e- government case study

of Heera and Chang also discusses the importance of IT as

integral part of IT governance practice. This study

investigates the IT governance practices and principles in

the Australian legal aid sector by examining thoroughly two

legal aid commissions in two different states of Australia.

One of the key findings in their proposed framework was

the ability to create timely decisions and policies on

structures, processes, and communications formulated in

line with their organizational objective. This was also based

from data obtained and cross examined with related

documents and articles surrounding legal aid [4]. In sub-

Saharan Africa, data is being collected via e-participation

for increased government transparency and accountability.

Sub-Saharan Africa collects data surrounding its projects

and initiatives to measure effectiveness of e-participation

systems and services in places such as online deliberations,

service deliveries, collective action organization and

governance innovations. From these initiatives, situations

surrounding ICT infrastructure, inclusion, and political

leadership were contextualized. This provided insight in

terms of which projects and initiatives for e-Participation

should be prioritized in certain areas due to its low rate of e-

participation [5]. The government of Sri-Lanka also

advocates e-government in the advent of timely evaluation

of initiatives which provokes better policy making and

program planning. Similar to Sub-Saharan Africa, data

dimensions to assess public services include measuring

delivery indices of public services in terms of quality,

responsiveness, and impact. Thus, with this IS, the extent of

how much the government complies with public demands

are a major source of public value creation towards their

socially desirable outcomes [6]. Meanwhile in China, the e-

government website (EGWS) serves a facility for

government to obtain information contents surrounding its

residents. It also contains a function that serves as an

interactive platform between its provincial authorities and

residents. Using 31 provincial case studies, results indicated

the different performance evaluation among the provinces

in scope in terms of GDP and resident income among

others, which allowed provincial officers to determine

correlated variables significant in economic development

[7]. Developed countries such as local government in the

United States also explores strategies for better

management of public data analytics by outsourcing e-

government projects to IT/IS outsourcing firms. Multi-

regression analysis of 213 local governments were

examined gathering primary source of data from electronic

government surveys. This allowed local government units

to analyze key conceptual constructs in initiatives and

satisfaction levels towards select outsourcing vendors [8].

Korean government is also known for its use of its

advanced e-government systems for its reforms and

efficiencies in government transactions. In a case study in

the evaluation of strategic alignments of its e-government

G4C projects to various stakeholders, data dimensions such

as usage indices, productivity ratio are being collected and

measured. [9].

Building on previous research on e-government, policy

makers and government leaders may have collected data at

various dimensions, however data collected may still not

fully represent its social systems with the lack of analytical

techniques and capabilities which may further aid in

decision making [10]. More than e-participation interaction

for insight and operational efficiency, the analytics

capability is still lacking in the discussion of its pivotal role

in holistically describing and potentially predicting analytic

solutions towards a specific problem of the domain drawn

from historical trends and patterns [11]. One of the few

studies that have incorporated IS and analytics is the study

of Overbeek, et. al. which consisted of utilizing analytics

across e-government projects to map fundamental

dimension data such as project type, domain sector, and

administration level beneficiary to map against national

policy composition [12]. More than e-government to assist

in a government unit’s operational needs, the planning and

policy implementation must also be supported with e-

government knowledge obtained from data analytics to

formulate and evaluate applicability and practicability of

policies and programs [13]. A clear understanding of the

dimensions along which e-government data is classified,

clustered, and analyzed to form good implementation

policies and strategic planning capable of leveraging

government units to implement projects is still infeasible

[14].

III. METHODOLOGY

In this paper, the system design framework is derived for

the justification and implementation of adequate programs

and policies by studying trends and patterns from Caloocan

government data, and transforming its data to insight for

government initiatives in light with the specific needs of the

local government unit. The methodology was inspired by

Proc. of the 2017 IEEE Region 10 Conference (TENCON), Malaysia, November 5-8, 2017

2831

the meta-analysis survey conducted by Abareshi and Martin

which provided a review and comparative analysis of cross-

sectional studies on MIS journals between 1992-2006 in

terms of research design, units of analysis, and other

methodological attributes on system development and

implementation [15].

1) Government Data Collection and Interviews

The proponents collected community data surrounding

health, population demographics, and education obtained

from interview and survey reports ranging from 2013-2016.

Demographic data included tabulated results of household

population by age, group and sex, household population 5

years old and over by highest grade/year completed, age,

group and sex; and household population 10 years old and

over by age group sex, and marital status. Education data

included total enrollment in private schools, total

enrollment in public schools, no. of teachers in classrooms

and seats for public schools, and no. of teachers,

classrooms, and seats for private schools. Lastly, Health

data included percentage distribution of elementary school

children in each district in the division of Caloocan by

Nutritional status/by Gender, and total no. of doctors,

nurses and midwives. From the interview, these data are the

most prioritized for policy formulations and program

planning. These data are currently being analyzed with

traditional excel sheets from consolidation to visualization.

2) System Requirements Gathering

The proponents interviewed the target users of the

system, to create user stories/scenarios about their day-to-

day transactions. Interviews towards planning officers per

sector and external researchers (ie. Civil service

organization, non-governmental organizations, private

researchers, etc.) were conducted. Gap analysis and

Ishikawa Fishbone Technique were used in identifying the

main pain points of the current process of analytics for

social, health and education data. In this phase, the

proponents have identified the main problem to be difficulty

in providing adequate programs and policies given the

decentralized sources of information from multiple

spreadsheets and the inability to capture historical trends

and patterns from previous years. Based from user

interviews, by the time the intervention needed has been

identified, another issue would most likely to surface which

would mean for the sector leaders to derive and formulate ad

hoc programs as a reactive response.

3) Initial System Design Framework and Testing

Components of a data analytics software served as the

baseline in creating a system design framework which

comprises of modules such as the data warehouse, data

transformation, data visualization, and data analytics

features. However, based from the requirements gathering,

there is a standardized format input/output mandated by

Philippine government which meant for the research

proponents to hasten validation checks upon data upload

and conversion. As an additional value to the analytics

system, the proponents have also incorporated setting of

quantifiable metrics for thresholds. This will allow alerts

and rule induction for outliers.

A user acceptance test form was given to all 15 sector

leaders representing health, social, and education sub-

sectors to assess the system in terms of its system module

quality (objective measure) and user satisfaction (subjective

measure) derived from the study of Parikh and Fazlollahi on

system user satisfaction criteria. The evaluation uses a

satisfaction scale from 1 being least satisfied to 5 being

most satisfied [16].

4) Insight and Program Recommendations

Two data analytics techniques were applied namely trend

analysis and comparative analysis. Trend analysis is a

method of analysis done by comparing data over the years to

see if there is a consistent pattern or trend occurring. If the

trend is seen, then the organization will be able to create a

response strategy for it. It is important because it can help an

organization understand how well the business is doing; it

can also be used for predicting future trends that lie ahead.

Using trend analysis, an organization can find it where it is

performing well, underperforming, and provide evidence to

inform decision making. Meanwhile, Comparative analysis

is a method of analysis done by comparing two sets of data

to identify new trends and find differences or similarities

among the compared data.

IV. THE SYSTEM DESIGN FRAMEWORK AND RESULTS

With the main problem of difficulty in identifying

relationships, patterns, and trends among Caloocan City

data, a system design framework comprised of data feed,

data validation, reporting, visualization, outlier, pattern

detection and decision capability modules allowed faster

derivations of direct and indirect findings for the sector

leaders. The following tables below summarizes the UAT

results in terms on system module quality and user

satisfaction.

Rank System Module Quality Score

(Avg= 4.5 out of 5)

1 Data Feed and Validation 4.9

2 Reporting and Visualization 4.8

3 Outlier and Pattern Detection 4.5

4 Decision Capability 4.0

Rank User Satisfaction Score

(Avg=4.49 out of 5)

1 Relevance 5.0

2 Usefulness 4.8

Proc. of the 2017 IEEE Region 10 Conference (TENCON), Malaysia, November 5-8, 2017

2832

3 Ease of Use 4.5

4 Decision Time 4.5

5 Confidence 4.3

6 User Learning 4.3

7 Format 4.0

A platform of user-friendly search capabilities addressed

the problem on unsegregated and unstandardized naming

scheme of data for easy retrieval of the data. In order to

ensure credibility of the raw data uploads, the system can be

able to detect wrong file uploads that does not match with

the name of the upload facility title. Furthermore, data

integrity is protected as it also contains a feature to

highlight wrong data input due to missing fields, formatting

errors, or summation errors which the system will be able to

inform the user at which specific row among thousand

entries of the table uploaded. Figure 1 below reflects the

system validation rule of not accepting raw spreadsheet

feeds that have erroneous data or missing fields. Results for

this feature yielded a 4.8 out of 5.0 functionality score.

Fig 1. Data Validation for Missing Fields

The reporting and visualization also yielded a 4.8 out of 5.0

evaluation score. Users expressed satisfaction with the

flexibility of being able to drill down, filter pan and zoom

which to them was useful in identifying root causes of

visualization charts that flagged as outliers.

Furthermore, they could filter out unnecessary details and

include relevant data to see how much it affects the current

condition of the community. Figure 2 below allows year to

year comparison of school going age population vs

enrollment, which automatically flags 2015 as an outlier given

its significant increase compared to the previous years. With

the drill down feature, this allowed the user to identify that the

root cause of this outlier was due to the increase in household

population demographics of those 10 years old and over.

Fig 2. Highlighting Outliers in Data Trends

From the correlation visualizations, which uses the

Pearson correlation co-efficient algorithm, the relationship

of school - going age population enrollments to

demographics was a significant correlation to household

population data nutritional demographics of 10 years old

and above. This yielded a correlation score of 0.95 with p-

value 0.049. The system automatically generates charts for

variables yielding 0.90-1.0 interpreted as strong

correlations. Data charts can also be filtered based on

correlation size, 0.70 to <0.90 for high positive correlation,

0.50 to <0.70 for moderate positive correlation. Low and

negligible correlations are no longer produced as an output

in the system and are disregarded and those with p-values

<0.05 as seen in Figure 3 below.

Fig 3. Correlation Visualization for Nutritional Status and Enrollments

Other relationships and patterns were also identified

such as the relationship of health to demographics on the

other hand is determining the population size and its impact

on the health facility and health manpower requirements.

An example of this is using the birth and death rates as the

indicator on planning maternal and child health services.

The demographic profile included here gives the users

information about the background, population composition

and distribution of the city or municipality, and how the

population data is changed by different components such as

birth rates, death rates, and others.

From the reports, the educational sector was the sector

that needed most program and policy intervention. With this

system design framework, the sector leaders could easily

conclude four decision dimensions namely their direct

observations (from default visualizations), explanation

Proc. of the 2017 IEEE Region 10 Conference (TENCON), Malaysia, November 5-8, 2017

2833

(from the drill down, pan and zoom), implications (from the

correlations), and interventions (based on outliers).

Fig 4. Program Recommendation and Intervention Report

V. CONCLUSION

In a third world countries such as the Philippines where

ICT is not as widely advanced, prioritized, and explored in

local government units due to significant cost, manpower

efforts, and the like, this research presents a low-cost

alternative for local government units to consider

opportunities in investing with ICT in a form of a web-

based customized system that is easy to learn, and that

requires minimal manpower efforts since it has been tailor

fit as to how the users have been accustomed to performing

their process. With an analytics system in place, planning is

improving with expert judgements that are grounded with

more accurate data surrounding the sectors of health and

education in the community. For future research, the

incorporation of a predictive analytics model may further

improve this system in deriving possible outcomes and

significant findings to moderate potential issues or risks in

the community.

REFERENCES

[1] W. Wang, S. Elliot, and M.-A. Williams, “An IS

contribution to the UN Millennium Development Goals:

Next Generation Vaccination Management in the

Developing World,” ACIS 2010 Proc., 2010.

[2] B. Niehaves, “Public Sector Innovation and Diffusion

Processes : Decentralisation and Innovation Public Sector

Innovation and Diffusion Processes –,” AMCIS 2007 Proc.,

2007.

[3] S. Rahman, N. Rashid, and A. Yadlapalli, “Determining

Factors of E- Government Implementation : a Multi-Criteria

Decision – Making,” Pacific Asia Conf. Inf. Syst. 2014,

2014.

[4] D. Heera and V. Chang, “Practices and principles of IT

governance in Australian legal aid organizations,” Inf. Syst.

J., pp. 378–390, 2008.

[5] M. N. . b Ravishankar, “Implementing citizen-centric

strategic IS projects: An Indian case study,” PACIS 2010 -

14th Pacific Asia Conf. Inf. Syst., pp. 320–332, 2010.

[6] K. Karunasena and H. Deng, “Exploring the Public Value of

e-Government: An Empirical Study from Sri Lanka

Approach for Evaluating the Public Value of,” Proc. Bled

eConference eTrust Implic. Individ. Enterp. Soc. Bled., pp.

286–300, 2010.

[7] J. Zhang, J. Zhu, A. Ibrahim’s, and J. Fountain, “A Research

of E-Government Website Performance and Its Relations

With Economic Development,” Proc. 18th Pacific Asia

Conf. Inf. Syst. (PACIS 2014), pp. 1–11, 2014.

[8] Y. Chen and Y. Chen, “MANAGING E-GOVERNMENT

OUTSOURCING PROJECTS : LESSONS FROM U . S .

LOCAL GOVERNMENT,” AMCIS 2002 Proc., 2002.

[9] J. Alalwan, M. A. Thomas, and M. Thomas, “A Holistic

Framework to Evaluate E-government Systems A Holistic

Framework to Evaluate E-government Systems,” AMCIS

2011 Proc., 2011.

[10] L. Houghton, D. Mackrell, and L. Houghton, “The impact of

individual , collective and structural Sensemaking on the

usefulness of business intelligence data The impact of

Sensemaking on Business Intelligence : A work in progress

case study,” Mediterr. Conf. Inf. Syst. 2012, 2012.

[11] S. K. Mathew, “Business Analytics and Business Value : A

Case Study,” pp. 1–17.

[12] C. G. Reddick and S. Antonio, “Capability Challenges in

Transforming Government through Open and Big Data :

Tales of Two Cities,” Thirty Sixth Int. Conf. Inf. Syst., pp. 1–

21, 2015.

[13] B. Otto, “A MORPHOLOGY OF THE ORGANISATION

OF DATA GOVERNANCE,” Eur. Conf. Inf. Syst., 2011.

[14] S. J. Overbeek and M. Janssen, “Towards Integrated Public

Service Delivery Based On An Ontological Framework,”

Mediterr. Conf. Inf. Syst., 2009.

[15] A. Abareshi and B. Martin, “A Meta-Analysis of Survey-

Based Research in MIS Field from 1992-2006,” AISeL,

2008.

[16] M. Parikh and B. Fazlollahi, “An Empirical Examination of

The User Satisfaction / System,” AMCIS 2002 Proc., 2002.

Proc. of the 2017 IEEE Region 10 Conference (TENCON), Malaysia, November 5-8, 2017

2834