Perceptions and use of big data analytics for information management among librarians in selected university libraries in Kwara state,Nigeria

Introduction

Mikalef et al. (2020) describe information management as a life cycle information and its generation, acquisition, appraisal, storage, utilization, control, access and archiving are all regarded as information governance. Apparently, the management of information includes decision-making positions (structural practices), decision-making duties (procedural practices), and personal growth and responsibility (relational practices). Tan et al. (2014) corroborated that the operational practices of information management focused on data movement, retention, cost allocation, and data analysis revolve around the information management. Whereas, structural practices identify important decision-makers and their roles and it facilitates effective communication, create formal connections between technical and business staff, and encourage knowledge exchange, instruction, and strategic planning.

Information management performs an essential role in libraries, and library needs to have the ability to preserve, organize, and make resources accessible to their users. Libraries need to maintain their collections with effective procedures and systems that work well with the wealth of information available today. Information assets include textual documents, data, images, recordings, and information kept in paper, electronic, and other media types (Arua and Ukwuaba, 2016). University libraries are turning information management into a learning engine by assessing user queries, resource consumption, and social media interactions using machine learning and natural language processing. Hence, the lifetime of information, include its generation, collection, storage, quality control, sharing, retrieval, and preservation are often covered by information management techniques and activities (Cox and Pinfield, 2014; Whyte and Tedds, 2011). This information lifecycle can be implemented or promoted using different technologies, which may include emerging big data analytics.

Big data analytics has the potential to revolutionize university libraries through dynamic collection, curation, personalized recommendations, research trend, and prediction by utilizing state-of-the-art tools and methodologies. Garcia and Wang (2013) demonstrated that 5 Vs model characterize big data which are large volume, tremendous velocity, great variety, low veracity, and high value. Sonawane (2018) argued that big data technologies describe a new generation of technologies and architecture and it is designed to economically extract value from very large volumes of a wide variety of data by enabling high velocity, capture, discovery and analysis of data. According to Zakir et al. (2015), big data analytics is the use of advanced computer techniques to examine large, complex datasets, uncover hidden patterns, and extract actionable insight. The authors showed that Hadoop, HDFS, NoSQL, MapReduce, MongoDB, Cassandra, PIG, HIVE, and HBASE are only a few of the essential technologies that make up big data analytics. Big data is essential for academic institutions, governments, and scientific research due to its benefits. Libraries leverage big-data techniques to improved their services, found inefficiencies in information management, anticipate needs, optimize resource allocation, and demonstrate the value of library services (Chen et al., 2015).

Hodonu-Wusu et al. (2020) posited that big data analytics is transforming library information management and this is done by examining large datasets and finding patterns, trends, and insights. However, this process aids with resource allocation, collection development, and user behaviour analysis, allowing libraries to personalize services, enhance user experiences, and make informed decisions. Qualified librarians and other information professionals need to fulfill their responsibilities in terms of gathering, maintaining, and promoting national information (Arua and Ukwuaba, 2016). Meanwhile, there is need for library resources to ensuring global information resources are effectively, appropriately, and quickly made available to various societal segments. In this age, the importance of globalization of information resources cannot be overemphasized. By this, information resources can be democratized in order to entrench their accessibility and usage.

Kitchin and McArdle (2016) demonstrated that big data analytics assists companies in finding new business prospects by using the insights generated by big data analytics. Business may make incremental advancements to their goods or services by detecting client feedback and using real-time operational monitoring. McAfee et al. (2012) show that there are improvements in organizations that use big data analytic to adjust their offers appropriately. Hence, big data analytics is used to detect new possibilities and risks of information management in the library. In this study, however, the focus is to examine the possibilities regarding the use of big data analytics for information management. Meanwhile, there are several sectors that have big data analytics. For instance, Internet service providers, predictive maintenance, and banks are some examples of businesses that use transaction data to deliver customized services (Kitchin, 2021).

Chiang et al. (2018) disclosed that big data analytics functions in information management to gathering data of guaranteeing its reliability and quality of information. The authors stated that the use of analytic is pertinent in analysis, trend detection, and legal obligations depend on the insight given by big data applications. Shabbir and Gardezi (2020) concluded that big data analytics is seen as a new paradigm for knowledge assets and it serves as “fourth paradigm of science”. This big data analytic is seen by high-achieving companies, libraries and industries; and it is pertinent as an essential way of differentiation and serves a means of expansion. Nevertheless, there is little or no connection between librarians’ perceptions and use of big data analytics to drive information management in university library, especially in the Nigerian context. In the Nigerian clime, anecdotal observation shows that big data analytics is not such that has gained general acceptability, especially for information management. Also, limited or no study has been conducted in this area, especially in the Nigerian context. Therefore, this study seeks to examine perceptions and use of big data analytics for information management among librarians in selected university libraries in Kwara State.

Review of related literature

Different scholars have explored the idea of big data analytics for the provision of library services. The following reviews under the different sub-headings provide evidence of related evidence from past studies in the current area of study:

Methodology

Cross-sectional research design was used for this study. This design is favoured because it allows data to be gathered across different geographical area in an effort to simultaneously characterize a phenomenon as it is. The population of this study comprises librarians working in selected university libraries in Kwara State, Nigeria with a minimum of bachelor’s degree in Library and Information Science. They are considered as professional librarians, hence their consideration to the study’s target group to enhance the credibility of the information/data collected. Forty-eight (48) professional librarians make up the population for this study (See Table 1). The target population is naturally small since the study focused on specialised professionals in unique institutional settings. Hence, the study adopted the total enumeration sampling technique.

OPEN IN VIEWER

Table 1.

Population of the study.

S/N	Institutions	Population
1	University of ilorin library, ilorin	23
2	Kwara state university library, malete	16
3	Al-hikmah university library, ilorin	5
4	Ojaja university library, ilorin	4
Total		48

As mentioned, due to the small population, total enumeration sampling technique was adopted, which result in a sample size of 58 that represent the entire population. Questionnaire was used for data collection. The questionnaire is divided into six (6) parts, which consist of the respondents’ demographic information and responses to items on the research questions. The Section A elicits the demographic details of the respondents, which include gender, year of experience, and educational qualification. The Section B of the research questions focuses on the perceptions of big data analytics for information management, which has five (5) items with five-point Likert scale of Strongly Agreed = 5 to Strongly Disagreed = 1. The Section C focuses on the extent of using big data analytics for information management, which has six (6) items with four-point Likert scale of High Extent = 4 to No Extent = 1. Section D focuses on the purpose of using big data analytics (BDAs) for information management, which has six (7) items with five-point Likert scale of Strongly Agreed = 5 to Strongly Disagreed = 1. Section E focuses on the challenges associated with the usage of big data analytic for information management, which has five (7) items with five-point Likert scale of Strongly Agreed = 5 to Strongly Disagreed = 1.

The questionnaire was validated by two (2) scholars of Library and Information Science and one (1) scholar of Computer Science. Three (3) copies of the questionnaire were given to the experts, and their comments were implemented in the questionnaire before administration on the respondents. Meanwhile, the choice of the scholar from Computer Science is informed by the fact that the study is related to the field of Computer Science since it focuses on big data analytics. For the test of the questionnaire’s reliability, 12 copies of the questionnaire were administered on respondents at Ladoke Akintola University of Technology, Ogbomoso, Oyo State, Nigeria. Split-half method was used to test the internal consistency and the Cronbach alpha (α) is reported as seen in Table 2. It can be seen in Table 2 that each section of the questionnaire has over 0.60 Cronbach alpha (α), with an overall average of 0.78.

OPEN IN VIEWER

Table 2.

Reliability of the instrument.

S/N	Headings	Cronbach alpha
1	Big data analytics for information management	0.83
2	Extent of using big data analytics for information management	0.81
3	Purpose of using big data analytics for information management	0.69
4	Challenges associated with the usage of big data analytic for information management	0.78
Average total		0.78

A Letter of Introduction was requested and collected from the Director for Centre of Research and Development, Kwara State University, Malete, Nigeria. This letter was shown to the heads of the different university libraries where data was collected to seek their permission to gather data for the study. The letters were approved, stamped, and signed by the different university libraries’ heads. The data was collected by the researchers physically with the help of two (2) research assistants. Each research assistant collected data at the University of Ilorin Library, Ilorin, and Ojaja University, Eyenkorin, Kwara State, Nigeria. The research assistants were trained on how to go about the data collection before giving them the letter of introduction. The purpose of the study was explained to the respondents before administering copies of the questionnaire on them.

Descriptive statistics, which include frequency count, simple percentage, mean, and standard deviation were used to analyse the collected data from the respondents. However, data collected was cleaned, coded and analysed using the IBM-Statistical Product and Service Solutions (SPSS) version 26. Upon request by the student researcher, the Ethical Approval Letter was obtained from the Director Centre of Research and Development, Kwara State University, Malete. The Reference No. of the Letter is KWASU/CR&D/REA/2024/0065. Informed consent of the respondents was sought and gotten before the administration of the questionnaire on the respondents. Participation in the study was voluntary.

Results

The total sample of the study is forty-eight (48) copies of the questionnaire was administered on the respondents, however only forty-four (44) copies were returned completely filled and found usable for the study. This indicates 91.67% return rate, which is considered appropriate for the study. The analyses of the collected copies of the questionnaire are presented in Table 3:

OPEN IN VIEWER

Table 3.

Respondents’ demographic information.

Items	Frequency	Percentage %
Gender
Male	26	59.1
Female	18	40.9
Total	44	100%
Years of experience
1–5 years	20	45.5
6–10 years	15	34.1
11–15 years	4	9.1
16 years and above	5	11.4
Total	44	100%
Educational qualification
Bachelor’s	20	45.5
Master’s	20	45.5
PhD	4	9.1%
Total	44	100%

Table 3 shows the demographic analysis of the respondents, which disclosed some notable patterns across various categories. Gender distribution indicates a significant majority of male respondents (59.1%) compared to female respondents (40.9%).The Table also shows the experience levels of the respondents: individuals with 1-5 years of experience constitute the largest group (44.5%), while those with 11-15 years represent the smallest cohort (9.1%). Regarding educational background, Bachelor’s degree holders and Master degree have the highest rate (45.5%), with a notably lower representation of respondents with PhD (9.1%).

Table 4 indicates that the respondents agreed that perceptions of big data analytics for information management include BDA can be used to plan, organize and structure information (x̄ = 4.50), BDA impact the quality and accuracy of information management (x̄ = 4.20), BDA identify areas for improvement for information resources management (x̄ = 4.25), and BDA can be used to process, control, evaluate, and report information(x̄ = 4.41). Meanwhile, it can be observed from the table that more than average of the respondents agreed that BDA has the capability to process information in university libraries (86.4%).

OPEN IN VIEWER

Table 4.

Perceptions of big data analytics for information management.

Statements	SA f(%)	A f(%)	N f(%)	D f(%)	SD f(%)	Mean (x̄)	SD
BDA can be used to plan, organize and structure information	22 (50.0%)	22 (50%)	0 (0.0)	0 (0.0)	0 (0.0)	4.50	0.50
BDA impact the quality and accuracy of information management	35 (79.5%)	9 (20.5%)	0 (0.0)	0 (0.0)	0 (0.0)	4.20	0.41
BDAs identify areas for improvement for information resources management	19 (43.2%)	21 (47.7%)	0 (0.0)	4 (9.1%)	0 (0.0)	4.25	0.87
BDA has the capability to process information in university libraries	8 (18.2%)	30 (68.2%)	0 (0.0)	4 (9.1%)	2 (4.5%)	3.87	0.97
BDA can be used to process, control, evaluate, and report information	18 (40.9%)	26 (59.1%)	0 (0.0)	0 (0.0)	0 (0.0)	4.41	0.98

Table 5 shows the extent of using big data analytic for information management. The Table indicates that less than half of the respondents were lowly used QlikView to process the framework of large data sets (15.9%), and Splunk platform for machine data analysis for information management (34.1%). The Table also shows that more than half of the respondents moderately used Tableau for data visualization in information management (68.1%), moderately used Apache Spark for data analysis engine for information management (56.2%), moderately used SAS Visual Analytic for data exploration and predictive analytics for information management (52.6%), and moderately used Apache Hadoop to gain insights from structured and unstructured data in information management (67.4%).

OPEN IN VIEWER

Table 5.

Extent of using big data analytics for information management.

Statements	HE F (%)	ME F (%)	LE F (%)	NE F (%)	Mean (x̄)	SD
I Use qlikview to process the framework of large data sets	16 (36.4%)	12 (27.3%)	7 (15.9%)	9 (20.5%)	2.80	1.15
I Use tableau for data visualization in information management	5 (11.4%)	25 (56.8%)	6 (13.6%)	8 (18.2%)	2.61	0.92
I Use splunk platform for machine data analysis for information management	9 (20.5%)	17 (38.6%)	15 (34.1%)	3 (6.8%)	2.72	0.87
I Use Apache spark for data analysis engine for information management	4 (9.1%)	23 (52.3%)	12 (27.3%)	5 (11.4%)	2.60	0.81
I Use SAS visual analytic for data exploration and predictive analytics for information management	14 (31.8%)	17 (38.6%)	8 (18.2%)	5 (11.4%)	2.91	0.98
I Use Apache hadoop to gain insights from structured and unstructured data in information management	6 (13.6%)	27 (61.4%)	5 (11.4%)	6 (13.6%)	2.75	0.86

Table 6 indicates that most of the respondents agreed that the purpose of using big data analytics for information management include BDAs are used to analyze user data (x̄ = 4.15), BDAs are used to provide data analysis for research (x̄ = 4.11) and BDAs are used to drive decision making and strategic planning (x̄ = 4.20). However, it can be observed from the Table that more than half of the respondents agreed that BDAs are used to streamline information process (76.4%), used to demonstrate users-centered information (72.6%), used to optimize resource allocation (77.4%), and used for borrowing management (53.1%).

OPEN IN VIEWER

Table 6.

Purpose of using big data analytics for information management.

Statements	SA F (%)	A F (%)	N F (%)	D F (%)	SD F (%)	Mean (x̄)	SD
I Use BDAs to analyze user data	20 (45.5%)	17 (38.6%)	3 (6.8%)	2 (4.5%)	2 (4.5%)	4.15	1.05
I Use BDAs to provide data analysis for research	14 (31.8%)	25 (56.8%)	3 (6.8%)	0 (0.0)	2 (4.5%)	4.11	0.89
I Use BDAs to drive decision making and strategic planning	20 (45.5)	19 (43.2%)	1 (2.3%)	2 (4.5%)	2 (4.5%)	4.20	1.02
I Use BDAs to streamline information process	6 (13.6%)	31 (70.5%)	4 (9.1%)	2 (6.8%)	0 (0.0)	3.90	0.70
I Use BDAs to demonstrate users-centered information	9 (20.5%)	28 (63.6%)	3 (6.8%)	4 (9.1%)	0 (0.0)	3.95	0.80
I Use BDAs to optimize resource allocation	7 (15.9%)	31 (70.5%)	2 (4.5%)	2 (4.5%)	2 (4.5%)	3.88	0.89
I Use BDAs for borrowing management	10 (22.7%)	19 (43.2%)	6 (13.6%)	6 (13.6%)	3 (6.8%)	3.61	1.18

Table 7 illustrates that the respondents agreed that challenges associated with the usage of big data analytic include data quality issues (x̄ = 4.45), complex data (text document, videos and audio files) (x̄ = 4.52), epileptic power supply (x̄ = 4.27), and diversity data types (x̄ = 4.0).

OPEN IN VIEWER

Table 7.

Challenges associated with the usage of big data analytic for information management.

Items	SA F (%)	A F (%)	N F (%)	D F (%)	SD F (%)	Mean (x̄)	SD
Scalability limitations	10 (22.7%)	26 (59.1%)	1 (2.3%)	4 (9.1%)	3 (6.8%)	3.81	1.10
Data quality issues	22 (50%)	21 (47.7%)	1 (2.3%)	0 (0.0)	0 (0.0)	4.45	0.62
Complex data (text doc., videos and audio files)	26 (59.1%)	17 (38.6%)	1 (2.3%)	0 (0.0)	0 (0.0)	4.52	0.73
Epileptic power supply	16 (36.4%)	26 (59.1%)	2 (4.5%)	0 (0.0)	0 (0.0)	4.27	0.69
Data privacy	13 (29.5%)	9 (20.5%)	0 (0.0)	21 (47.7)	1 (2.3%)	3.27	1.38
Limited technical expertise	4 (9.1%)	18 (40.9%)	1 (2.3%)	21 (47.7%)	0 (0.0)	3.11	1.12
Diversity data types	7 (15.9%)	32 (72.7%)	3 (6.8%)	2 (4.5%)	0 (0.0)	4.0	0.64

Discussion

The findings showed that perceptions of big data analytics for information management among librarians in selected academic libraries in Kwara State include big data analytic can be used to plan, organize and structure information, big data analytic is used to impact the quality and accuracy of information management, big data analytic is used to identify areas for improvement for information resource management and big data analytic is used to process, control, evaluate, and report information. The finding is different from that of Kamupunga and Chunting (2019), which indicated that librarian perceived big data analytics as expensive technologies for information management which made up of high-velocity, high-variety, and high-volume information that is used creatively to improve understanding and decision-making in universities libraries. Meanwhile, the finding is similar to that of Müller et al. (2016), which demonstrated that librarians perceived big data analytic in information management in the areas of organising and structuring information. Also, it was shown that big data analytic is used to impact the quality and accuracy of information management.

The finding shows the extent of using big data analytic for information management. The finding indicates that less than half of the librarians lowly used QlikView and Splunk to process the framework of large data sets and machine data analysis for information management. This implies that there is a need for university libraries may to focus on providing more awareness and workshop on the use of these big data analytics tools in information management. With this, information service provision can be entrenched and service delivery would be enhanced. The finding shows that more than half of the librarians moderately used Tableau for data visualisation, moderately used Apache Spark for data analysis engine, moderately used SAS Visual Analytic for data exploration and predictive analytics, and moderately used Apache Hadoop to gain insights from structured and unstructured data. Meanwhile, the finding of the current study is different from previous studies (Mikalef et al., 2020; Müller et al., 2016), which revealed that SAS Visual Analytic and QlikView is extensively explored for predictive analytics for information management. The finding aligns with Obitade (2019), which shows that healthcare information management consider the use of big data analytics like Apache Spark for data analytics engine for information management.

The result revealed that the purpose of using big data analytics for information management include to analyze user data for research and drive decision making and strategic planning. The finding is similar to Khan and Vorley (2017), which demonstrated that big data analytics in knowledge management (KM) is use to provide data analysis for research, analyse user data and used to drive decision making and strategic planning. This indicates that big data analytics are important for information processing in the cycle of information management. This means that librarians in university libraries can provide research support service to other academic staff of universities using their expertise of big data analytics tools like Tableau, Apache Hadoop, and so on. Meanwhile, the finding of the current study is dissimilar to previous studies (Elgendy and Elragal, 2014; Zakir et al., 2015), which established those big data analytics may be used practically for data visualization and knowledge management (KM) enhancement by enhancing the insights for a variety of business operations, including supply chain and marketing management.

The finding indicated that challenges associated with the usage of big data analytic include data quality issues, complex data (text document, videos, and audio files), epileptic power supply, and diversity data types. The finding differs from Bansal et al. (2018), which illustrated that inaccurate users’ requirements, system failures, mobile payments pose security risks, and substantial financial investment for data management, analytics tools, and infrastructure serves as the pitfalls mitigating the use of big data analytics for information management. The finding is similar to Golub and Hansson (2017), which showed that extensive and varied datasets, unstable power supply and data quality issues were the challenges associated with the use of big data analytics. The finding is different from Chen et al. (2015), which demonstrated that intricate nature of privacy, unavoidable data leaks and assurances of confidentiality were challenges associated with the usage of big data analytic.

Implications

Theoretically, the findings of this study contribute to discourse on the role of big data analytics in information management within academic libraries. Contrary to previous evidence by Kamupunga and Chunting (2019), which highlighted the perception of big data analytics as primarily high-cost and complex, this study reveals a more functional perspective, emphasizing its practical applications in organizing, structuring, and improving the quality of information. This divergence suggests that the theoretical understanding of big data analytics in information management may be evolving, with a shift towards recognising its value in enhancing operational efficiency rather than focusing solely on its complexity and cost. Additionally, the study aligns with Müller et al. (2016) in identifying the importance of big data analytics in improving information accuracy and quality, thereby reinforcing theories that emphasize the role of big data in enhancing decision-making processes in academic libraries.

Practically, the study provides valuable insights for library professionals in information management. The moderate use of tools like Tableau, Apache Spark, and SAS Visual Analytics suggests that while there is awareness and adoption of big data analytics tools, there is still room for more extensive usage and integration. The study highlights the need for further training and resources to enable librarians to fully harness these technologies for information management. Moreover, the challenges identified, such as epileptic power supply and complex data types, suggest that libraries in Kwara State, Nigeria and university libraries from similar developing countries may need to prioritize infrastructural improvements and adopt strategies to manage data quality issues effectively. By addressing these practical challenges, academic libraries can enhance their information management processes, leading to better service delivery and more accurate, data-driven decisions.

From a societal perspective, the study’s findings underscore the growing acceptance and integration of big data analytics in information management in academic libraries. By highlighting the practical uses of big data analytics—such as data visualization, predictive analytics, and strategic decision-making—this study suggests that librarians and academic institutions are increasingly leveraging these technologies to improve information services. This shift has broader societal implications, as it indicates a move towards more data-driven decision-making in university libraries, which could lead to more informed and efficient management of information resources. Furthermore, the challenges identified, such as data quality issues and diverse data types, point to the need for continued investment in infrastructure and training to fully realize the potential benefits of big data analytics in academic libraries.

Conclusion

The study concludes that librarians perceived big data analytic for information management in university libraries to plan, organize and structure information The study establishes that librarians did not extensively use big data analytic for information management such as Tableau for data visualization in information management, Apache Spark for data analysis engine for information management, SAS Visual Analytic for data exploration and Apache Hadoop to gain insights from structured and unstructured data in information management. The study also recognises that librarians used big data analytic to analyze user data, to provide data analysis for research and for strategic planning. Lastly, the study highlighted that data quality issues, complex data (text document, videos, and audio files), epileptic power supply, and diversity data types were challenges associated with the use of big data analytics.

Recommendations

Based on the results of the study, the followings are the suggested recommended:

(i) University libraries should implement regular training programs and workshops to enhance librarians’ understanding and skills in big data analytics like QlikView and Splunk, making them more confident and proficient in using these tools for information management.