2.1 Background

The conceptual lines addressed in this article correspond to characteristics of Big Data and Open Government Data (Figure 1), and how evaluating and assuring their quality opens a path to generating value according to their use and appropriation.

Figure 1 Characteristics of Big Data and Open Government Data. 

For ^{Cai and Zhu (2015)}, Big Data addresses unique characteristics, such as volume, speed, accuracy, value, and variety, so quality assurance involves a series of challenges such as facilitating the discovery and promoting the use of data, as a process that enables its value and strengthens the trust of its origin to facilitate its understanding and discovery; and surmounting shortcomings that hinder the analysis process, such as inaccurate, obsolete, wrong, or incomplete data (i.e., not suitable for use).

The United Nations Economic Commission for Europe, UNECE (2014), defined the general principles for evaluating the quality of Big Data, which are suitability for use, flexibility, ability to extend its context, and application to various situations. Finally, it highlights the relationship between the effort required to ensure its quality versus the value obtained.

To evaluate the quality of use of Big Data, we take as a reference the "3 C" model, exposed by ^{Williams and Tang (2020)}, which addresses: a) Contextual Consistency, which takes into account the capacity of the data set to be useful in the domain of interest independent of its format; b) Temporal Consistency, which focuses on the time interval where the data can be used; and c) Operational Consistency, which represents the extent to which the data can be processed for analysis.

In the past, the consumers of the data corresponded to the same group of producers, thus allowing a high level of control to guarantee its quality. However, with the era of Big Data, ^{Cai and Zhu (2015)}, noted that the users of the data do not necessarily correspond to the producers of the data. This new reality increases the complexity of quality measurement since the data set itself is involved, as well as the technology required for its management and its spatial and operational problems, as ^{Merino et al. (2016)} explained.

^{Mahecha et al. (2018)}, identified that public institutions produce and manage data very quickly, which is published in predetermined formats in a structured way to facilitate access to the public and its reuse in different contexts; such data is called Open Government Data (OGD). While Open Linked Data was defined by Maestre et al. (2019), as a set of best practices for the publication and use of open and structured data on the Web that allows the exploration of new relationships between data and the development of new applications for the cities.

The objective of OGD projects and initiatives is to facilitate cooperation among public administration, politics, industry, and citizens, strengthening transparency, democracy, participation, and collaborative work (^{Kucera, 2015}). According to ^{Attard et al. (2015)}, the quality of OGDs is interdisciplinary, multidimensional, and transversal in its life cycle, since the following occurs: i) The data are created to ensure their completeness ; ii) The data are selected according to their objectivity, timeliness, and accuracy; iii) The data are harmonized to ensure accessibility, ease of use, and processing and iv) The data are published so that they are available to the public.

Data availability promotes reuse and discovery and facilitates the exploration and exploitation of data, processes through which the knowledge obtained from the information that correlates the data under study is built, innovated, connected, and managed.

In turn, Ministerio de Tecnologías de la Información y las Comunicaciones (MinTIC, 2019a), in the "Guide for the Use and Exploitation of Open Data in Colombia," specified how the principles of OGD chart the path to generating public value in the following ways:Promoting transparency and social control, helping people to take a more active role in society, facilitating product innovation and the creation of new business models, improving government management and efficiency of services, allowing planning and forecasting of future scenarios and strengthening the exploitation of the potential use of data.

^{Ciancarini et al. (2016)}, defined open data as a subdomain of Big Data, and data quality problems are exacerbated in Big Data; for this reason, it becomes fundamental to focus efforts on evaluating and guaranteeing the quality of this representative subdomain.

About the above, the quality of the OGDs must be given as a process before the opening of the data to provide value in the creation and generation of new business models in the business sphere and, in turn, to facilitate innovation in the public sector, based on ideas and best practices from the private sector. Therefore, the development of data opening strategies generates value in discovering trends and identifying new perspectives of a problem to obtain patterns that support innovation.

2.2 Method

The methodology used in this study, corresponds to a literature review based on the work of ^{Maestre-Gongora and Colmenares-Quintero (2018)}. Based on the object of the study, the following questions are proposed to be resolved: Q1. How is data quality addressed in the life cycles of Big Data and OGD? and Q2. What types of values are generated from Big Data and OGD?

The search was done in the following sources: IEEE Xplore digital library, Enlace Springer, and Scopus, using the following keywords: Open Government Data Quality, Big Data Quality, Open Government Data Quality Methodology, and Big Data Quality Methodology.

In the selection of documents, the title, keywords and summary were verified, so that these documents would make a decisive contribution to the research and thus avoid irrelevant references. The following filters were applied to the pre-selected documents: year, type of publication: articles, theses, normative documents, technical documents and reports from multilateral organizations.

Subsequently, we identified how each of the selected articles addressed: relationship with the object of study: Emphasis - relationship with the object of study: (Big Data, Government Open Data), Sector of application (Public, business, academic), Type of Value (Public, Political, Social, Commercial, Economic or Technical, other), Affecting Factors (Benefits, limitations, challenges and risks), Quality framework (model, policies, principles, process, guide, life cycle, best practices, methodology, among others). Then, the type document (journal and conference articles), and the range of publication dates, 2010 to 2021, were reviewed.

Therefore, the final selection included a total of 32 documents, which contribute to the results and discussion of the research questions. Finally, 32 documents were selected, as shown in Table 1.

Documents classification. The 32 relevant documents selected were classified according to the following criteria: country, language, type of document (articles, theses, normative documents, technical documents and reports from multilateral organizations), subject matter (OGD or Big Data).

3.1 Context of research in OGD and Big Data

Of the documents selected, 47% were generated in the Americas, corresponding to Argentina, Colombia, the United States, and Mexico. Within the remaining 53%, there are publications from Germany, China, Spain, Greece, Holland, England, Ireland, Italy, Malaysia, Pakistan, Poland, the Czech Republic, and Singapore. Colombian publications accounted for 25% of the total number of documents and were taken as a contextual basis. On the other hand, 16% of the records correspond to publications made by multilateral organizations, such as the Inter-American Development Bank (IDB), the UNECE, and the American Society for Quality (ASQ).

It is evident that a large part of the literature selected for this research is in English, 64%, because English-speaking countries have explored the study's object more; the remaining 36% is in Spanish. The authors of topics of study have dedicated considerable time in their research to the area of Emerging Information Technologies, 55% of the selected literature is focused on open government data, 19% shows the relationship of Big Data with open government data, 7% discusses the relationship of Big Data with open data, and 19% concentrates only on Big Data.

3.2 Relationship between data quality addressed in the life cycle of Big Data and OGD

Open Government Data life cycle

In the "Guide for the use and exploitation of Open Data in Colombia," the MinTIC (2019a), describes the open data life cycle, which consists of four stages: a) Establish the opening plan, b) Structure and publish the data, c) Communicate and promote the use of the data (disseminate, link stakeholders, consolidate, and position), and d) Monitor quality and use (measure quality, measure use, and measure the impact).

The quality evaluation is established in the documentation and structuring stage, which is done by applying and measuring parameters and indicators. These indicators must correlate use with quality, demonstrating that it is only possible to generate value from the data being used. They are only used if these data meet quality standards. Figure 2 presents a parallel between the life cycle phases of Big Data and OGD. The quality assurance component is immersed in evaluating the data for Big Data and documenting and structuring the data for OGDs. This component enables the next phase that corresponds to the exploitation and use of government data, through which the value immersed in the data is obtained.

Figure 2 Big Data and OGD life cycle relationship. 

OGD is emerging as a recent trend that encourages collaboration and citizen participation, and enables reusing data to pursue public and private innovation. Consumers of these data require a solid foundation of trust that activates the use of public information, which can only be acquired after the implementation of a data quality assurance process that strengthens and intensifies the inherent value of the data and activates its potential in different contexts.

3.3 Types of values generated from Big Data and OGD

In 60% of the documents reviewed, the topic of "Value" is addressed, in some of them, from the public, economic, social, commercial, and, to a lesser extent, political and technical approach. Research covering the concept of public value at the social, economic, and commercial levels is the most relevant literature. Political value highlights governments’ and citizens' responsibilities to contribute to mitigating corruption collaboratively. Technical value focuses on the technology and the platform's characteristics and requirements that support the data.

Later, when analyzing the studies selected by the type of public, social, commercial, and economic value generated by guaranteeing the quality of Big Data, it was found that the public, social, and economic value has been simultaneously established in ^{Kalampokis et al. (2011)} and Consejo Nacional de Política Económica y Social de la República de Colombia (2018); while the commercial value has been analyzed independently (^{Merino et al., 2016}; ^{Wahyudi et al., 2018}; ^{Wieczorkowski, 2019}; ^{Williams & Tang, 2020}). Table 3 shows the types of value generated and the sources that refer to them.

In the literature reviewed related to OGD, the type of public, social, commercial, and economic value is also identified, where more than 80% of the authors have established public value as the most identified and affirmed by using OGD, while the least referenced is commercial value.

It is evident that the only author who considers in his study the four types of value is ^{Attard et al. (2016)}, and he names two additional types of value (political and technical value), while in the document "Quality Requirements for Open Data" produced by MinTIC (2019b), the public value is discriminated according to its political, strategic, and ideological impacts and the value of legitimacy and respect.

In many studies, the public, social, and economic values are established simultaneously, as shown in Table 4 (^{Bracken & Greenway, 2019}; Consejo Nacional de Política Económica y Social de la República de Colombia, 2018; Cumbre Global de la Alianza para el Gobierno Abierto, 2015; Ministerio de Tecnologías de la Información y las Comunicaciones de Colombia, 2016a; 2019a; 2019b).

Some of the public value-oriented data perspectives highlight the public value in reuse, collaboration, and innovation. Regarding the public value in reuse, OGD may be of interest to the business sector and the general public. For the data to be reused, it is necessary to guarantee its quality level to facilitate the connection between the real-world objects and the description that the data provides of these objects. ^{Kalampokis et al. (2011)}, defined that the challenges for achieving this ideal situation vary between the lack of available information and the large amount of data produced by the public sector.

A considerable percentage of data is shared by several entities and may expose different results from the same object or real-world problem, preventing the creation of new value-added services and products. Linking other actors in the data ecosystem, promotes the reuse of available information and facilitates feedback on detected inconsistencies, thus ensuring access and generating value from the use of data (Ministerio de Tecnologías de la Información y las Comunicaciones de Colombia, 2019a).

Alternatively, citizen participation is a useful mechanism to build knowledge collaboratively. OGDs promote the active intervention and empowerment of citizens and facilitate transparency through accountability and the nation's growth as a whole, as indicated by (^{Zuiderwijk & Janssen, 2015}). ^{Attard et al. (2015)}, presented that OGDs are related to decision and policymaking rather than sporadic participation in election processes. The value obtained from collaboration transversally acts in the process of continuous improvement of the government's services. MinTIC (2016a), in the Open Data Guide, declared that this contributes to transparency, social control by stakeholders, and predictive measurement of policies' impact at the social, cultural, and economic levels.

Additionally, ^{Yang and Kankanhalli (2013)}, reported that innovation has had an evolution, going from a closed concept, in which ideas and the flow of knowledge occur within an organization, to an open concept, where development ideas have an external origin and are welcome to generate new products and services.

The concept of OGD relates more explicitly to the theme of value, which can be seen because the authors address several types of value in their publications simultaneously. In contrast, big data addresses them individually. The value of OGD is based on ensuring its quality and transformation into relevant public information to serve citizens by extracting the maximum value from the data, which are built through a process of reuse in fields of innovation and creation of services. Public value is the most prevalent value in the OGD literature, followed by economic value, while commercial value seems to have more impact on Big Data.