Next Article in Journal
Unveiling International New Ventures’ Success: Employee’s Entrepreneurial Behavior
Previous Article in Journal
Cross-Country Differences in Entrepreneurial Internationalization Tendencies: Evidence from Germany and Pakistan
 
 
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

An Intelligent Advisory System to Support Managerial Decisions for A Social Safety Net

1
Centre for Information Technology and Systems, University of Lagos, Lagos 100213, Nigeria
2
Department of Electrical and Information Engineering, Covenant University, Ota 112233, Nigeria
3
Department of Computer Engineering, Atilim University, 06830 Ankara, Turkey
4
Department of Creative and Visual Art, Federal University, Lafia P.M.B. 146, Nigeria
5
Faculty of Informatics, Kaunas University of Technology, 51368 Kaunas, Lithuania
*
Author to whom correspondence should be addressed.
Adm. Sci. 2019, 9(3), 55; https://doi.org/10.3390/admsci9030055
Submission received: 25 June 2019 / Revised: 31 July 2019 / Accepted: 1 August 2019 / Published: 7 August 2019

Abstract

:
Social investment programs are designed to provide opportunities to the less privileged so that they can contribute to the socioeconomic development of society. Stakeholders in social safety net programs (SSNPs) target vulnerable groups, such as the urban poor, women, the unemployed, and the elderly, with initiatives that have a transformative impact. Inadequate policy awareness remains a challenge, resulting in low participation rates in SSNPs. To achieve all-inclusive development, deliberate policies and programs that target this population have to be initiated by government, corporate bodies, and public-minded individuals. Artificial intelligence (AI) techniques could play an important role in improving the managerial decision support and policy-making process of SSNPs and increasing the social resilience of urban populations. To enhance managerial decision-making in social investment programs, we used a Bayesian network to develop an intelligent decision support system called the Social Safety Net Expert System (SSNES). Using the SSNES, we provide an advisory system to stakeholders who make management decisions, which clearly demonstrates the efficacy of SSNPs and inclusive development.

1. Introduction

In any society, poor, vulnerable, and disabled people exist. They often form the bottom of the socioeconomic pyramid and require deliberate socioeconomic intervention to harness their potential and move them into mainstream socioeconomic and political spheres (Mohanty et al. 2014). A large number of people are restricted or excluded from the socioeconomic development process due to ethnicity, gender, sexual orientation, age, poverty, or disability (Hernández and Pérez 2016). As a result, inequality across the globe is increasing. Of total global assets, 85% is owned by the richest 10% of people, while a paltry 1% is owned by the poorest 50% (United Nations 2017). The effects of such exclusion are staggering, deepening inequality across the globe.
The process of integrating these vulnerable groups into the sustainable development drive of society is called inclusive development (Grover 2014; United Nations 2017). Evidence has shown that successfully implemented social programs correlate well with sustainable development goals to eliminate poverty and provide peace and prosperity for all (Giribabu et al. 2018). However, the aims of inclusive development cannot be achieved without addressing the needs of disadvantaged people (Noga and Wollbring 2013). Social instability, injustice, and social hazards can undermine sustainable development aiming to achieve a balance of economic, environmental, and social values and goals (Bibri and Krogstie 2017). The explosive growth of urbanization and population in developing countries such as Nigeria has introduced new burdens on the development of modern infrastructures (Silva et al. 2018). Deliberate targeting of the economically disadvantaged segment of society often involves policies, programs, and projects that are poverty-related, which are collectively referred to as social safety net programs (SSNPs) (Amornsiriphong and Piemyat 2012), a metaphor for alleviating the challenges of the target group.
Inclusive development incorporates human rights standards and principles, such as nondiscrimination, participation, and accountability (Grover 2014). To accomplish inclusive development, nations must consider factors such as creating gainful and productive employment. This should be accompanied by robust SSNPs to shield those who earn little or cannot work. Specifically, developing countries need to increase public services in social sectors such as health, education, water, transportation, and power. This requires prudent public expenditure and well-designed and -implemented fiscal policies that stimulate growth and mitigate poverty. Based on global experiences, valuable policy advisories can be offered in areas such as SSNPs, employment strategies, and job creation. In addition, such collaboration will reduce poverty, stimulate poverty-related growth, and achieve sustainable development goals. These efforts will also enhance the ability of governments to formulate fiscal policies and strategies. Vital public services must be made accessible to everyone. In this regard, governments must improve economic governance and public investment (Mohanty et al. 2014).
As part of efforts to reach all segments of the less privileged and isolated for their integrated development into society, research studies have intensified in various socioeconomic spheres where perceived or existential threats of alienation exist. Different types of inclusive development have emerged, such as inclusive education (Kricke and Neubert 2017), smart classrooms (Wogu et al. 2019), inclusive innovation (Amaro-Rosales and Gortari-Rabiela 2016), inclusive growth (Raheem et al. 2018), inclusive transformation (Liu et al. 2018), inclusive industrial robotics (Wogu et al. 2018), inclusive conflict prevention (Okewu et al. 2019), and all-inclusiveness (Klaufus et al. 2017). However, drawing the attention of stakeholders, such as governments, corporate bodies, and public-minded individuals, to the quest of harnessing the potential of these underprivileged groups is required for meaningful and productive economic activities. Efforts to promote SSNPs and inclusive development have included vocational education and training for the unemployed in India (Ahmed 2016); promoting innovation for inclusive development (Hernández and Pérez 2016); a big data analytics framework to enhance and support decision-making in organizations (Elgendy and Elragal 2016); the use of geographic information systems, demographic data, and data analytics techniques to calculate potential food insecurity and to assess food assistance distribution locations (Bacon and Baker 2017); and a sustainable development capacity measurement model based on information entropy that identifies changes in sustainable development levels (Liang et al. 2017). In summary, none of the above studies focused on using data analytics to design a framework for tackling problems associated with SSNPs to promote inclusive development. To achieve all-inclusive development, governments and public institutions must enhance their processes of decision-making and optimize social benefits schemes.
We assert that artificial intelligence (AI) and data analytics can contribute to the improvement of SSNP schemes. Poverty-related data can be used to understand the behavior and activities of the less privileged to guide the formulation of a robust response system using data analytics techniques such as statistics, data mining, knowledge discovery in databases, data warehouses, and machine learning (Wu et al. 2016). The ability to elicit patterns in data is useful for efficient and effective decision-making (Martin-Sanchez and Aguiar-Pulido 2017). Learning patterns in data provide knowledge or information, offering stakeholders support for optimal decision-making. Decisions based on mere conjecture or boardroom politics can lead to suboptimal applications of scarce resources, and social investment programs would be no less affected. Only an unbiased and nonprejudiced data-driven decision process can address this issue (Kowalczyk and Buxman 2015).
In this paper, we describe the development and use of an intelligent decision support system (DSS) called the Social Safety Net Expert System (SSNES) that uses poverty-related data and a learning algorithm to produce knowledge for assisting SSNP stakeholders with decisions and initiatives. The information provided includes the identification of genuine beneficiaries of social welfare packages and advice on the social investment intervention measures that would best serve the community and prevent the misuse of funds by political elites, which in some cases (e.g., in Bangladesh (Kundo 2018)) have made SSNPs ineffective in reducing the poverty level. The paper is oriented toward researchers and practitioners in management sciences interested in the applications of AI.

2. Related Work

2.1. Social Safety Net Programs

SSNPs are a set of services offered by the state, corporate bodies, or public-minded individuals to assist the poor and vulnerable to prevent them from falling further into poverty or disadvantage (Ahmed 2016). Services include unemployment benefits (Das 2016), welfare (Stranz et al. 2017), basic healthcare (Oberlin and Pizmony-Levy 2016), conditional cash transfers (Tovar and Urrutia 2017), social investment microcredit (Rahman 2014), food and nutrition security (Porter and Goyal 2016), child protection (Peterman et al. 2017), support in case of debt insolvency (Spooner 2017), shelters for the homeless (Bender et al. 2018), support for aging populations (Mutchler et al. 2018), and widow empowerment programs (Pradhan and Afrin 2015), among others. SSNPs are carefully designed with the aim of empowering the less privileged so that their abilities can be harnessed for socioeconomic growth and development (United Nations 2017). Hence, initiatives such as conditional cash transfers from the government, universal healthcare offers, and free education could assist the economically challenged in being more successful members of society, while the level of poverty is reduced in the process. In developed countries, SSNPs have been institutionalized. The USA and Canada are good examples of countries with structured SSNPs (Spross 2015; Hoynes and Stabile 2019). Advocates of SSNPs believe that these programs lower crime rates and reduce poverty levels. However, critics argue that economic growth is inhibited by the taxes committed to the support of SSNPs (Conning and Kevane 2002). They are also of the view that these initiatives are a barrier to socioeconomic advancement because SSNPs encourage people to be unproductive and poor (Spross 2015). Conversely, many individuals who have lost their sources of income and livelihood in times of economic troubles are often unwilling to use available SSNPs because of the fear of social stigmatization (Sherman 2013). Regardless, in any society, a segment of the population is outside the mainstream of development and needs help being integrated. Deliberate social intervention policies and programs should be targeted at these population groups (Amornsiriphong and Piemyat 2012).
Developing nations have started taking action toward the creation of SSNPs. In South Africa, people who are unable to support themselves can obtain grants, with the majority of grants administered by social services targeting children (Government of South Africa 2017). In Nigeria, social investment programs include conditional cash transfers to the elderly and weak, a home-grown school feeding program, soft loans for small businesses, and monthly stipends for the unemployed (Umukoro 2013). In Niger, assistance through SSNPs has increased household resilience against unpredictable climate events (Gambo Boukary et al. 2016). In Ethiopia, the Productive Safety Net Program provides cash transfers to improve child nutrition (Porter and Goyal 2016). In Zambia, Kenya, Ghana, and Lesotho, social cash transfers lead toward and facilitate graduation from poverty (Daidone et al. 2015). Providing opportunities for everyone to contribute to development can broaden the global economic base and reduce poverty (Mohanty et al. 2014). Hence, all groups of people should have a platform for receiving developmental benefits, creating opportunities, and taking part in decision-making. The importance of supporting decision-making in managerial decisions has been emphasized in many areas, such as infrastructure asset management (Chen and Bai 2019), business process information management (Zavadskas et al. 2019), environmentally sustainable and ecofriendly automotive production (Iuga et al. 2019), and corporate philanthropy (Wu et al. 2018). Moreover, improved social service management contributes to reducing income inequality and achieving social justice (Dearing 2017).

2.2. Applications of Data Analytics

Data analytics could be used to understand the behavior and activities of the poor so that SSNP programs are not hijacked by fake beneficiaries. Data analytics refers to data-assisted problem-solving and decision-making that is used to unveil patterns in data, which are invaluable for knowledge and information generation for the purpose of rational and informed decision-making (Li et al. 2016), leading to optimal resource allocation. Decision support systems (DSSes) can be used for trade-off and uncertainty analyses for the improved planning and management of cities and society (Chichernea 2014). Data mined and knowledge extracted from social support networks can be used to enable a mechanism for the protection and promotion of resilience among poor populations of cities in developing countries (Juliano and Yunes 2014). The resilience of individuals, families, and communities as a whole can be increased with patterns uncovered from SSNP data and knowledge that is learned and used effectively. Maintaining the resilience of urban systems requires interaction between social, psychological, physical, structural, and environmental factors, while monitoring systems such as the one presented in this paper are crucial for high awareness and shared decision-making for resilient people in a resilient city (Pitrėnaitė-Žilėnienė and Torresi 2014). Data analytics draws techniques from a number of communities (Cooper 2012), including statistics, business intelligence, web analytics, operations research, AI, social network analysis (Behera et al. 2017), and information visualization. Here, data mining and AI were used to design an intelligent DSS to serve as an advisory framework for stakeholders in the SSNP sector, confirming the social role and importance of software systems for communities (Damaševičius 2009). The systematic use of AI methods could lead to the development of new social welfare institutions that are more responsive to the needs of people and more development-oriented (Patel 2018).

3. Methodology

3.1. Context

We used Nigeria as a case study given ongoing deliberate government innovations and public policies targeted at enhancing and unleashing the potential of its vulnerable population through inclusive development. The increase in the population of those less privileged in Nigeria has caused insecurity challenges. Insurgency and ethnoreligious crises are the main humanitarian issues that have increased the proportion of those who are impoverished, resulting in large numbers of internally displaced persons (IDPs). IDPs are particularly found in Northeast Nigeria, where the Boko Haram insurgency has continued for years (Hamid and Baba 2014), resulting in a serious humanitarian crisis. Included in IDPs are children, orphans, poor women, the disabled, and jobless youths and men. In addition to IDPs, the elderly, the rural and urban poor, unemployed and underemployed youths, and persons with disabilities exist country-wide. These are all people who need one or more forms of an SSNP. However, Nigeria has abundant human and material resources. Since gaining independence in 1960, insecurity and corruption have led to an impoverished population (Nwagboso 2012; Tom and Attai 2014). Nigeria’s economic indicators are weak: massive unemployment, a harsh business environment, low productivity, a highly import-dependent economy, an unstable exchange rate, high susceptibility of the economy to external dynamics, low GDP, a high crime rate, a monolithic economy, and a double-digit interest rate. Given these factors, it is no surprise that a large portion of the population is unemployed, poor, disabled, and vulnerable.
For the effective targeting and identifying of SSNP beneficiaries in Nigeria, the following criteria were formulated: an elderly person 65 years old or above, a widow without any visible means of livelihood, a person with a disability, or an unemployed youth. However, the decision to offer social welfare packages involved a combination of these factors. In most cases, the decision was based on evidence that had multivariate and multiconditional requirements. A compelling need exists to direct the attention of governments, corporate bodies, and public-minded individuals to the plight of this vulnerable segment of society. The aim is to empower those within these groups through social investment program (SIP) initiatives to unleash their potential for sustainable development (Okewu et al. 2017; Okewu et al. 2018). Ongoing SIP initiatives in Nigeria targeted at the economic empowerment and emancipation of economically challenged segments of the population are analyzed in Table 1.

3.2. Instruments

The criteria specified by social investment and communications art experts provide the stage for multicriteria decision modeling. We developed the Social Safety Net Expert System (SSNES), which uses a learning algorithm to understand patterns in poverty-related data to generate unbiased information that serves as the input in decision-making for the implementation of SIPs.
The layered architecture of SSNES is shown in Figure 1, with four layers: presentation, services, data, and external systems. At the presentation layer, the user interfaces include fixed and mobile devices such as desktops, laptops, tablets, and phones, and users include social welfare beneficiaries and social investors.
The services and data layers use the Hadoop MapReduce framework (Dean and Ghemawat 2004), a NoSQL technology known for handling big data, including both structured and unstructured data. It uses parallel and distributed processing so that the SSNES can respond quickly. The volume of national poverty-related social data is huge, complex, and sometimes unstructured: as a result, traditional data handling technologies, including statistical packages and SQL databases, would be inadequate. To create a properly functioning hybrid intelligent system for enhanced social investments decisions, components such as data management, predictive analytics, security, a knowledge base, an inference engine, explanation facilities, and relationship mining are integrated and have the ability to interoperate with external systems such as social media, SMS applications, and emailing systems.
To ensure that the SSNES could reason and learn in a social investment environment characterized by uncertainty and imprecision and to include selective perception and divergent and contradicting opinions on what should constitute SSNP initiatives, intelligent technologies such as probabilistic reasoning and an evolutionary algorithm were integrated. Specifically, naïve Bayes and a decision tree learning algorithm were applied to the poverty-related data to generate knowledge that could guide optimal decision-making and problem-solving in the social welfare subsector.
Naïve Bayes was used to determine the effectiveness of an SSNP advocacy strategy as a part of inclusive development. A decision tree learning algorithm was applied to predict if a person would qualify as a beneficiary of an SSNP. Decision tree learning relies on the mapping of observations for an item to draw conclusions about the target value of the item using the decision tree as a predictive model (Song and Kim 2010). The Bayesian network (Heckerman 1996) is a schematic diagram of a set of random variables and their conditional independence using a directed acyclic graph. For instance, probabilistic relationships between income levels and poverty levels could be illustrated using a Bayesian network such that when poverty levels are specified, the network computes the probabilities of income levels. A high-level overview of the SSNES intelligent decision system is shown in Figure 2.

3.3. Procedure

To provide a hybrid intelligent platform for problem-solving and decision-making in the social investment space, 17 SSNP advocacy campaign strategies and 11 examples of past recipients/nonrecipients of SSNP packages were analyzed. To start, examples of multivariate observations of the SSNP advocacy campaign strategies, which were past methods of creating awareness and raising the consciousness of social investors on the need to harness the potential of the less privileged for inclusive development, were compiled. As outlined in Table 2, the training data were extracted from an SSNP advocacy database (only selected variables and records are shown). The data were anonymized to ensure privacy protection and adhere to ethical standards in handling information concerning vulnerable groups.
Expert opinions on the appropriateness of the variables or attributes (Advocacy Strategy, Attention-Catching, and Message Retention) and their values, including Advocacy Strategy (Outdoor, Indoor, or Mixed), Attention-Catching (Yes or No), and Message Retention (High, Fair, or Low), were sought from visual communications artists (as domain experts). The possible impact of these advocacy multivariate observations on social investment stakeholders was investigated to determine their effectiveness. The perceived advocacy campaign impact was expressed as a class attribute or variable (Impact_Rating) and its values (Effective, Not Effective), along with classification and soft computing standards.
A sample of the data used and elicited from a database of vulnerable people is presented in Table 3 (only selected variables and records are shown). These are examples of the less privileged that were classified as beneficiaries or otherwise in previous poverty-related socioeconomic empowerment initiatives.

4. Results

Using the training data (see sample in Table 1; Advocacy Strategy, Attention-Catching, and Message Retention are attributes; Impact_Rating is the class label attribute), naïve Bayesian probability was applied to determine if an instance of a proposed advocacy campaign strategy would be effective at drawing attention to the situation of the underprivileged. Such a prediction can guide SSNP investors about the potential return on an advocacy investment.
Given the above training data, the aim was to predict the class of the tuple X to determine if an advocacy campaign strategy would effectively draw the attention of critical stakeholders to the economic empowerment of the vulnerable segments of society. The information gain of each attribute is tabulated in Table 4.
Since age had the highest information gain among the attributes, it was selected as the splitting attribute. Node N was labeled with age, and branches were grown from each of the attribute’s values. The tuples were then partitioned accordingly. The attribute with the highest information gain among the remaining attributes was then selected, and the process continued until a complete decision tree was obtained. The final decision tree returned by the algorithm is shown in Figure 3.
Based on the above tree, only two rules were generated, as follows:
  • Rule 1: (Age = “65 and Above”) → (Genuine_Beneficiary = “Yes”);
  • Rule 2: (Age = “Below 65” ^ Physical State = “Not Disabled” ^ Marital Status = “Married”) → (Genuine_Beneficiary = “No”).

5. Evaluation

We evaluated the social effectiveness of the SSNES system using the methodology suggested by the World Bank (IEG 2011) for impact evaluations of SSNPs. We used the difference-in-differences method, which performs a before/after comparison with a comparison of means. The differences in the growth rates of outcomes between the beneficiaries who were selected using the SSNES system and nonbeneficiaries were interpreted as impacts attributable to the system. The SSNP interventions applied were unconditional cash transfers of 5000 naira per month to a bank account set up specifically to receive the assistance payments. The aim was to lift poor and vulnerable households out of poverty or protect them from falling into poverty. The intervention was applied in the poor informal neighborhood of Ota, Ogun State, Nigeria. The city itself has an estimated 163,783 residents and has the third largest concentration of industry in Nigeria. The intervention was applied to 314 households selected by the SSNES program and was funded by the World Bank program. The control population included the same number of households, who did not receive the SSNP intervention.
The monitored indicators of the intervention included healthcare usage (number of preventive healthcare procedures, such as parasite treatments, attended childbirths, and health check-ups), school attendance (ratio of children attending school classes), consumption (household expenditures), poverty headcount (ratio of households with their income or expenditure falling under the poverty line), and child labor (hours of work among children). The SSNP intervention resulted in an increase in school enrollment by 13%, an increase in health checks by 15%, an increase in consumption by 7%, a reduction in extreme poverty (headcount ratio) by 11%, and a decrease in child labor by 6%. The results demonstrate the effectiveness of the developed decision support system for reducing poverty and increasing the social resilience of the population.

6. Discussion

The best methods for creating awareness and sensitivity and educating communities about integrating the less fortunate into inclusive development can be learned from data patterns. Both data mining and machine learning techniques were used to design the proposed system, including a Naive Bayes and a decision tree classifier.
The rules defined in Section 3 essentially form the initial knowledge base of the DSS and were designed and proposed to enhance SSNP initiatives and support inclusive development. The SSNES performs “data-to-information-to-decision” conversion using a learning algorithm to understand multivariate and multicriteria rules to generate further rules or knowledge. The SSNES was built to enable the identification of patterns in poverty-related data to generate knowledge or information that will guide critical stakeholders to implement social welfare schemes that economically empower vulnerable populations. As such, the SSNES offers support to governments, corporate bodies, and public-minded individuals committed to social investment.
The knowledge base component of the SSNES applies metaheuristics to poverty-related training data to find patterns in data and elicit the best rules. This component also contains expert-generated rules obtained from applying data mining techniques such as the decision tree algorithm, which is a classification algorithm. The component called inference engine handles reasoning and is based on the rules in the knowledge base. To enhance machine reasoning, it interprets the rules. As the name suggests, the knowledge refining system component streamlines the rules generated for rational and informed decision-making. The explanation facilities component clarifies the actions taken by the SSNES. The user refers to persons that rely on the system for the purposes of decision-making, and software actions are communicated via the user interface.
The integration of the naïve Bayesian probability and decision tree algorithm data mining techniques allows SSNES to be an intelligent expert system that can learn and reason in imprecise and uncertain environments characterized by selective perception, sentiment analysis, and divergent opinions of stakeholders in the SSNP ecosystem. The SSNES uses a metaheuristic search and a multilayered approach to learn and generate the best rules from SSNP data for the purposes of problem-solving and informed and rational decision-making.
Our outcomes indicate that a hybrid intelligent expert system can be designed using the limited amount of data available on the beneficiaries of SSNPs. Since the core objective was to demonstrate if data analytics could be used to build a DSS for SSNPs and inclusive development, the result is considered to be positive. Hence, despite the constraints of using a limited number of training data, sufficient grounds exist to infer that the expert system could predict the efficacy of an advocacy campaign strategy for drawing the attention of stakeholders toward the poor.
Note that here we did not evaluate the success of the SSNP itself, as there are mixed opinions on whether SSNPs contribute to reducing poverty levels (Gilligan et al. 2009), but rather we evaluated the eligibility of candidates to receive transfers. The SSNES had the capability to determine if a person qualified as a beneficiary for the SSNP packages. Data analytics can be used to resolve problems, enhance decision-making (Elgendy and Elragal 2016), and increase efficacy in the SSNP subsector and promote inclusive and socially responsible development through informed decision-making (Li et al. 2016). The lack of sample size (17 SSNP advocacy campaign strategies and 11 examples of past recipients/nonrecipients of the SSNP packages) could have constrained the statistical significance of the outcome (Turner et al. 2006), as the larger the training dataset, the more representative it would be of an SSNP database.

7. Conclusions

This study addressed the dual issues of advocacy and identification of genuine beneficiaries of social welfare schemes. The poor and vulnerable in society can be empowered to broaden the national economic base through inclusive and sustainable development. To achieve the aims of all-inclusive sustainable development, deliberate policies and programs that target this population have to be initiated by government, corporate bodies, and public-minded individuals. To address problems associated with social safety net program (SSNP) advocacy and to enhance managerial decision-making in social investment programs, intelligent technologies (Bayesian network, decision tree learning) were used to develop a hybrid intelligent decision support system called the Social Safety Net Expert System (SSNES).
The SSNES has the ability to reason and learn patterns in poverty-related social data by generating a set of rules (knowledge base). Generated knowledge guides managerial decision-making for the social welfare of vulnerable populations, including the rural and urban poor, widows, the disabled, the unemployed, and the elderly, among others. The SSNES produces knowledge that could guide stakeholders and SSNP managers when making decisions to introduce economic empowerment programs for the less privileged to increase their social resilience. Using the SSNES system, we provide an advisory system for stakeholders and managers that clearly demonstrates the efficacy of SSNPs in inclusive development.

Author Contributions

Conceptualization, S.M.; Data curation, E.O.; Funding acquisition, R.D.; Investigation, J.O.; Methodology, S.M.; Resources, J.O.; Software, E.O. and J.O.; Supervision, S.M.; Validation, S.M.; Visualization, E.O.; Writing – original draft, E.O., S.M. and J.O.; Writing—review & editing, R.D. and R.M..

Funding

This research received no external funding.

Conflicts of Interest

The authors declare no conflicts of interest.

References

  1. Ahmed, Tutan. 2016. Labour market outcome for formal vocational education and training in India: Safety net and beyond. IIMB Management Review 28: 98–110. [Google Scholar] [CrossRef] [Green Version]
  2. Amaro-Rosales, Marcela, and Rebecade Gortari-Rabiela. 2016. Inclusive innovation in the Mexican agricultural sector: coffee producers in Veracruz. Economía Informa 400: 86–104. [Google Scholar] [CrossRef]
  3. Amornsiriphong, Somsak, and S. Piemyat. 2012. Social Capital: Means of Social Safety Net and Social Protection in Thai Communities. Procedia Engineering 32: 1152–56. [Google Scholar] [CrossRef]
  4. Bacon, Christopher M., and Gregory A. Baker. 2017. The rise of food banks and the challenge of matching food assistance with potential need: towards a spatially specific, rapid assessment approach. Agriculture and Human Values 34: 899–919. [Google Scholar] [CrossRef]
  5. Behera, Ranjan Kumar, Santanu Kumar Rath, Sanjay Misra, Robertas Damaševičius, and Rytis Maskeliūnas. 2017. Large Scale Community Detection Using a Small World Model. Applied Science 7: 1173. [Google Scholar] [CrossRef]
  6. Bender, Kimberly, Stephanie Begun, Kaite Dunn, Emily Mackay, and Jonah DeChants. 2018. Homeless Youths’ Interests in Social Action via Photovoice. Journal of Community Practice 26: 107–20. [Google Scholar] [CrossRef]
  7. Bibri, Simon Elias, and John Krogstie. 2017. Smart sustainable cities of the future: An extensive interdisciplinary literature review. Sustainable Cities and Society 31: 183–12. [Google Scholar] [CrossRef]
  8. Chen, Lin, and Qiang Bai. 2019. Optimization in Decision Making in Infrastructure Asset Management: A Review. Applied Sciences 9: 1380. [Google Scholar] [CrossRef]
  9. Chichernea, Virgil. 2014. The Use of Decision Support Systems (Dss) in Smart City Planning and Management. Romanian Economic Business Review, Romanian-American University 8: 238–51. [Google Scholar]
  10. Conning, Jonathan, and Michael Kevane. 2002. Community-based targeting mechanisms for social safety nets: A critical review. World Development 30: 375–94. [Google Scholar] [CrossRef]
  11. Cooper, Adams. 2012. A Brief History of Analytics A Briefing Paper. JISC CETIS Analytics Series 1: 1–21. [Google Scholar]
  12. Daidone, Silvio, Luca Pellerano, Sudhanshu Handa, and Benjamin Davis. 2015. Is graduation from social safety nets possible? Evidence from sub-saharan Africa. IDS Bulletin 46: 93–102. [Google Scholar] [CrossRef]
  13. Damaševičius, Robertas. 2009. On the human, organizational, and technical aspects of software development and analysis. Information Systems Development: Towards a Service Provision Society, 11–19. [Google Scholar] [CrossRef]
  14. Das, Tushar Kanti. 2016. Mahatma Gandhi National Rural Employment Guarantee Act (MGNREGA) as Social Safety Net: Analysis of Public Works in Odisha, India. Review of Economic Perspectives 16: 337–60. [Google Scholar] [CrossRef]
  15. Dean, Jeffrey, and Sanjay Ghemawat. 2004. MapReduce: Simplified data processing on large clusters. Paper presented at 6th Conference on Symposium on Operating Systems Design & Implementation (OSDI’04), Berkeley, CA, USA, December 6–8; Volume 6. [Google Scholar]
  16. Dearing, Tiziana C. 2017. Social Services, Social Justice, and Social Innovations: Lessons for Addressing Income Inequality. Religions 8: 89. [Google Scholar] [CrossRef]
  17. Elgendy, Nada, and Ahmed Elragal. 2016. Big Data Analytics in Support of the Decision Making Process. Procedia Computer Science 100: 1071–84. [Google Scholar] [CrossRef]
  18. Gambo Boukary, Aboubakr, Adama Diaw, and Tobias Wünscher. 2016. Factors Affecting Rural Households’ Resilience to Food Insecurity in Niger. Sustainability 8: 181. [Google Scholar] [CrossRef]
  19. Gilligan, Daniel O., John Hoddinott, and Alemayehu Seyoum Taffesse. 2009. The Impact of Ethiopia’s Productive Safety Net Programme and its Linkages. Journal of Development Studies 45: 1684–1706. [Google Scholar] [CrossRef]
  20. Giribabu Dandabathula, Chandrika Mohapatra, Sudhakar Reddy C., and P. V. V. Prasada Rao. 2018. Holistic correlation of world’s largest social safety net and its outcomes with sustainable development goals. International Journal of Sustainable Development and World Ecology 26: 113–28. [Google Scholar] [CrossRef]
  21. Government of South Africa. 2017. Services: Social Relief of Distress Programme; Cape Town: Government of South Africa.
  22. Grover, Adit. 2014. Importance of CSR in Inclusive Development. In International Relations Conference on India and Development Partnerships in Asia and Africa: Towards a New Paradigm (IRC-2013). Procedia—Social and Behavioral Sciences 157: 103–8. [Google Scholar] [CrossRef]
  23. Hamid, Adamu Muhammad, and Ibrahim M. Baba. 2014. Resolving Nigeria’s ‘Boko Haram’ Insurgence: What Role for the Media? Procedia—Social and Behavioral Sciences 155: 14–20. [Google Scholar] [CrossRef]
  24. Heckerman, David. 1996. A tutorial on Learning with Bayesian Network. Redmond: Microsoft Research. [Google Scholar]
  25. Hernández, José Luis Sampedro, and Claudia Díaz Pérez. 2016. Innovation for Inclusive Development: A proposal for analysis. Economía Informa 396: 34–48. [Google Scholar] [CrossRef]
  26. Hoynes, Hilary, and Mark Stabile. 2019. How do the US and Canadian social safety nets compare for women and children? Journal of Labor Economics 37: S253–S288. [Google Scholar] [CrossRef]
  27. IEG (Independent Evaluation Group). 2011. Evidence and Lessons Learned from Impact Evaluations on Social Safety Nets. Washington: World Bank. [Google Scholar]
  28. Iuga (Butnariu), Anca N., Vasile N. Popa, and Luminita I. Popa. 2019. Comparative Analysis of Automotive Products Regarding the Influence of Eco-Friendly Methods to Emissions’ Reduction. Energies 12: 6. [Google Scholar] [CrossRef]
  29. Juliano, Maria Cristina Carvalho, and Maria Angela Mattar Yunes. 2014. Reflections on the social support network as a mechanism for the protection and promotion of resilience. Ambiente & Sociedade 17: 135–54. [Google Scholar] [CrossRef]
  30. Klaufus, Christien, Paulus van Lindert, Femke van Noorloos, and Griet Steel. 2017. All-Inclusiveness versus Exclusion: Urban Project Development in Latin America and Africa. Sustainability 9: 2038. [Google Scholar] [CrossRef]
  31. Kowalczyk, Martin, and Peter Buxman. 2015. An ambidextrous perspective on business intelligence and analytics support in decision processes: Insights from a multiple case study. Decision Support Systems 80: 1–13. [Google Scholar] [CrossRef]
  32. Kricke, Meike, and Stefan Neubert. 2017. Inclusive Education as a Democratic Challenge—Ambivalences of Communities in Contexts of Power. Education Science 7: 12. [Google Scholar] [CrossRef]
  33. Kundo, Hare Krisna. 2018. Micro politics of social safety net programmes: The case of the food-for-work programme in Bangladesh. Development Policy Review 36: O815–O830. [Google Scholar] [CrossRef]
  34. Li, Yan, Manoj A. Thomas, and Kweku-Muata Osei-Bryson. 2016. A snail shell process model for knowledge discovery via data analytics. Decision Support Systems 91: 1–12. [Google Scholar] [CrossRef]
  35. Liang, Xuedong, Dongyang Si, and Xinli Zhang. 2017. Regional Sustainable Development Analysis Based on Information Entropy—Sichuan Province as an Example. International Journal of Environmental Research and Public Health 14: 1219. [Google Scholar] [CrossRef]
  36. Liu, Ying, Yanliu Lin, Na Fu, Stan Geertman, and Frank van Oort. 2018. Towards inclusive and sustainable transformation in shenzhen: Urban redevelopment, displacement patterns of migrants and policy implications. Journal of Cleaner Production 173: 24–38. [Google Scholar] [CrossRef]
  37. Martin-Sanchez, F., and V. Aguiar-Pulido. 2017. Analytics and Decision Support Systems in Global Health Informatics. Global Health Informatics 2017: 195–217. [Google Scholar]
  38. Mohanty, Aparajita, Shashikala Gurpur, and Chaitra R. Beerannavar. 2014. Rethinking Inclusive Development: A Human Rights Critique of South Asia. Procedia—Social and Behavioral Sciences 157: 128–36. [Google Scholar] [CrossRef]
  39. Mutchler, Jan E., Yang Li, and Ping Xu. 2018. How strong is the social security safety net? using the elder index to assess gaps in economic security. Journal of Aging and Social Policy 31: 1–15. [Google Scholar] [CrossRef]
  40. Noga, Jacqueline, and Gregor Wollbring. 2013. An Analysis of the United Nations Conference on Sustainable Development (Rio +20) Discourse Using an Ability Expectation Lens. Sustainability 5: 3615–39. [Google Scholar] [CrossRef]
  41. Nwagboso, Chris. 2012. Security Challenges and Economy of the Nigerian State (2007–2011). American International Journal of Contemporary Research 2: 244–58. [Google Scholar]
  42. Oberlin, Kathleen C., and Oren Pizmony-Levy. 2016. Beyond Access: Social Safety Nets and the Use of Free Health Care Services. Sociological Perspectives 59: 640–67. [Google Scholar] [CrossRef]
  43. Okewu, Emmanuel, Sanjay Misra, Rytis Maskeliūnas, Robertas Damaševičius, and Luis Fernandez-Sanz. 2017. Optimizing Green Computing Awareness for Environmental Sustainability and Economic Security as a Stochastic Optimization Problem. Sustainability 9: 1857. [Google Scholar] [CrossRef]
  44. Okewu, Emmanuel, Sanjay Misra, Luis Fernandez-Sanz, Rytis Maskeliūnas, and Robertas Damaševičius. 2018. An e-environment system for socio-economic sustainability and national security. Problemy Ekorozwoju 13: 121–32. [Google Scholar]
  45. Okewu, Emmanuel, Sanjay Misra, Luis Fernandez-Sanz, Foluso Ayeni, Victor Mbarika, and Robertas Damasevičius. 2019. Deep Neural Networks for Curbing Climate Change-Induced FarmersHerdsmen Clashes in a Sustainable Social Inclusion Initiative. Problemy Ekorozwoju—Problems of Sustainable Development 14: 143–55. [Google Scholar]
  46. Patel, Leila. 2018. Social Protection in Africa: Beyond Safety Nets? The Journal of Sociology & Social Welfare 45: 4. [Google Scholar]
  47. Peterman, Amber, Anastasia (Naomi) Neijhoft, Sarah Cook, and Tia M. Palermo. 2017. Understanding the linkages between social safety nets and childhood violence: A review of the evidence from low-and middle-income countries. Health Policy and Planning 32: 1049–71. [Google Scholar] [CrossRef]
  48. Pitrėnaitė-Žilėnienė, Birute, and Fabrizio Torresi. 2014. Integrated Approach to a Resilient City: Associating Social, Environmental and Infrastructure Resilience in its Whole. European Journal of Interdisciplinary Studies, Bucharest Economic Academy 6: 1. [Google Scholar]
  49. Porter, Catherine, and Radhika Goyal. 2016. Social protection for all ages? Impacts of Ethiopia’s Productive Safety Net Program on child nutrition. Social Science and Medicine 159: 92–99. [Google Scholar] [CrossRef]
  50. Pradhan, Mohammad Abdul Hannan, and Sabiha Afrin. 2015. A Review of Social Safety Nets Programs for Women in Bangladesh: Issue and Challenges. Advances in Economics and Business 3: 149–56. [Google Scholar]
  51. Raheem, Ibrahim D., Kazeem O. Isah, and Abdulfatai A. Adedeji. 2018. Inclusive growth, human capital development and natural resource rent in SSA. Economic Change and Restructuring 51: 29–48. [Google Scholar] [CrossRef]
  52. Rahman, Mohammad Mahbubur. 2014. Estimating the Average Treatment Effect of Social Safety Net Programmes in Bangladesh. Journal of Development Studies 50: 1550–69. [Google Scholar] [CrossRef]
  53. Sherman, Jennifer. 2013. Surviving the Great Recession: Growing Need and the Stigmatized Safety Net. Social Problems 60: 409–32. [Google Scholar]
  54. Silva, Bhagya Nathali, Murad Khan, and Kijun Han. 2018. Towards sustainable smart cities: A review of trends, architectures, components, and open challenges in smart cities. Sustainable Cities and Society 38: 697–713. [Google Scholar] [CrossRef]
  55. Song, Jin Yu, and Hui Kang Kim. 2010. A Study of Decision Tree in Detecting Intrusions. Journal of the Korean Data Analysis Society 12: 983–96. [Google Scholar]
  56. Spooner, Joseph. 2017. Seeking shelter in personal insolvency law: Recession, eviction, and bankruptcy’s social safety net. Journal of Law and Society 44: 374–405. [Google Scholar] [CrossRef]
  57. Spross, Jeff. 2015. America’s social safety net is way too skimpy—And horribly designed. The Week, May 8. [Google Scholar]
  58. Stranz, Hugo, Patrik Karlsson, and Stefan Wiklund. 2017. The wide-meshed safety net. decision-making on social assistance eligibility in Sweden. European Journal of Social Work 20: 711–23. [Google Scholar] [CrossRef]
  59. Tom, Edet J., and Amadu J. Attai. 2014. The Legislature and National Development: The Nigerian Experience. Global Journal of Arts Humanities and Social Sciences 2: 63–78. [Google Scholar]
  60. Tovar, Jorge, and Miguel Urrutia. 2017. The Impact of Social Safety Net Programs on Household Savings in Colombia. Developing Economies 55: 23–37. [Google Scholar] [CrossRef] [Green Version]
  61. Turner, Carl W., James R. Lewis, and Jakob Nielsen. 2006. Determining usability test sample size. In International Encyclopedia of Ergonomics and Human Factors. Edited by Waldemar Karwowski. Boca Raton: CRC Press, pp. 3084–88. [Google Scholar]
  62. Umukoro, Nathaniel. 2013. Poverty and Social Protection in Nigeria. Journal of Developing Societies 29: 305–22. [Google Scholar] [CrossRef]
  63. United Nations. 2017. United Nations Development Programme. Inclusive Development. [Google Scholar]
  64. Wogu, Ikedinachi Ayodele, Sanjay Misra, Patrick Assibong, Adewole Adewumi, Robertas Damasevicius, and Rytis Maskeliunas. 2018. A Critical Review of the Politics of Artificial Intelligent Machines, Alienation and the Existential Risk Threat to America’s Labour Force. In Computational Science and Its Applications—ICCSA 2018. Basel: Springer International Publishing, pp. 217–32. [Google Scholar] [CrossRef]
  65. Wogu, Ikedinachi Ayodele P., Sanjay Misra, Patrick A. Assibong, Esther Fadeke Olu-Owolabi, Rytis Maskeliūnas, and Robertas Damasevicius. 2019. Artificial intelligence, smart classrooms and online education in the 21st century: Implications for human development. Journal of Cases on Information Technology 21: 66–79. [Google Scholar] [CrossRef]
  66. Wu, Caesar, Rajkumar Buyya, and Kotagiri Ramamohanarao. 2016. Big Data Analytics = Machine Learning + Cloud Computing. Big Data 2016: 3–38. [Google Scholar]
  67. Wu, Wenqing, Kexin Yu, Chien-Chi Chu, Jie Zhou, Hong Xu, and Sang-Bing Tsai. 2018. Diffusion of Corporate Philanthropy in Social and Political Network Environments: Evidence from China. Sustainability 10: 1897. [Google Scholar] [CrossRef]
  68. Zavadskas, Kazimieras Edmundas, Jurgita Antucheviciene, and Prasenjit Chatterjee. 2019. Multiple-Criteria Decision-Making (MCDM) Techniques for Business Processes Information Management. Information 10: 4. [Google Scholar] [CrossRef]
Figure 1. The component-based layered architecture of the Social Safety Net Expert System (SSNES).
Figure 1. The component-based layered architecture of the Social Safety Net Expert System (SSNES).
Admsci 09 00055 g001
Figure 2. The Social Safety Net Expert System (SSNES).
Figure 2. The Social Safety Net Expert System (SSNES).
Admsci 09 00055 g002
Figure 3. Decision tree for determining a genuine social safety net program (SSNP) beneficiary: here, “Aged = 0” means Age is below 65, “Aged = 1” means Age is 65 and above, “Married = 0” means widow/widower, “Married = 1” means married, “Employed = 0” means unemployed, “Employed = 1” means employed, and the end leaves show the probability of genuine beneficiary status.
Figure 3. Decision tree for determining a genuine social safety net program (SSNP) beneficiary: here, “Aged = 0” means Age is below 65, “Aged = 1” means Age is 65 and above, “Married = 0” means widow/widower, “Married = 1” means married, “Employed = 0” means unemployed, “Employed = 1” means employed, and the end leaves show the probability of genuine beneficiary status.
Admsci 09 00055 g003
Table 1. Sample ongoing social investment programs in Nigeria.
Table 1. Sample ongoing social investment programs in Nigeria.
Welfare SchemeTarget AudienceDetails
The N-Power Volunteer CorpUnemployed youths (15–24 years of age)Empowering unemployed youths, including graduates and nongraduates, with skills to engage in productive ventures such as teaching, farming, and community services. Monthly stipend paid.
Conditional Cash TransfersPoorest of the poor (elderly, disabled, widows, etc.)Monthly stipend paid to the elderly, disabled, widows, etc. considered to be the poorest of the poor. Community-based approach is used to identify them through traditional and religious leaders.
Government Enterprise and Empowerment Program (GEEMP)Small-holding farmers, cooperatives, market women, enterprising youths, and other small-scale businesses, particularly in rural areasSoft loans ranging from 10,000 to 100,000 Nigerian naira given as start-up support to enhance entrepreneurial activities among the poor (farmers, artisans, traders, etc.).
Home-Grown School Feeding ProgramStudents in public primary schoolsA minimum of one meal served per day to students in public primary schools.
Rapid Response Initiative (RRI)Poor with health challengesA free medical service for urban and rural poor with health challenges with a particular focus on surgery operations.
Table 2. A sample of poverty-related advocacy data.
Table 2. A sample of poverty-related advocacy data.
CaseAdvocacy StrategyAttention-CatchingMessage RetentionClass: Impact_Rating
1.OutdoorYesHighEffective
2.OutdoorYesLowNot Effective
3.OutdoorYesFairEffective
4.OutdoorNoHighNot Effective
5.OutdoorNoFairNot Effective
6.OutdoorNoLowNot Effective
7.IndoorYesHighEffective
8.IndoorYesFairEffective
9.IndoorYesLowNot Effective
10.IndoorNoHighNot Effective
11.IndoorNoFairNot Effective
12.IndoorNoLowNot Effective
13.MixedYesHighEffective
14.MixedYesLowNot Effective
15.MixedNoHighNot Effective
16.MixedNoFairNot Effective
17.MixedNoLowNot Effective
Table 3. A sample of the data of vulnerable people.
Table 3. A sample of the data of vulnerable people.
PersonMarital StatusPhysical StateAgeEmployment StatusClass: Genuine_Beneficiary
1.WidowNot DisabledBelow 65UnemployedYes
2.WidowNot DisabledBelow 65EmployedNo
3.WidowerDisabled65 and AboveUnemployedYes
4.WidowerNot DisabledBelow 65UnemployedNo
5.MarriedNot Disabled65 and AboveEmployedYes
6.MarriedDisabled65 and AboveUnemployedYes
7.MarriedDisabledBelow 65UnemployedYes
8.WidowDisabledBelow 65UnemployedYes
9.WidowNot Disabled65 and AboveUnemployedYes
10.WidowerDisabledBelow 65EmployedNo
11.WidowerNot Disabled65 and AboveEmployedYes
Table 4. Information gain of each attribute.
Table 4. Information gain of each attribute.
AttributeInformation Gain
Age0.548
Marital Status0.315
Physical State0.186
Employment Status0.105

Share and Cite

MDPI and ACS Style

Okewu, E.; Misra, S.; Okewu, J.; Damaševičius, R.; Maskeliūnas, R. An Intelligent Advisory System to Support Managerial Decisions for A Social Safety Net. Adm. Sci. 2019, 9, 55. https://doi.org/10.3390/admsci9030055

AMA Style

Okewu E, Misra S, Okewu J, Damaševičius R, Maskeliūnas R. An Intelligent Advisory System to Support Managerial Decisions for A Social Safety Net. Administrative Sciences. 2019; 9(3):55. https://doi.org/10.3390/admsci9030055

Chicago/Turabian Style

Okewu, Emmanuel, Sanjay Misra, Jonathan Okewu, Robertas Damaševičius, and Rytis Maskeliūnas. 2019. "An Intelligent Advisory System to Support Managerial Decisions for A Social Safety Net" Administrative Sciences 9, no. 3: 55. https://doi.org/10.3390/admsci9030055

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop