The importance of data mining, user information behaviour and interaction audit for information literacy

Abstract

Purpose

This study aims to promote the need for advanced skills acquisition within the LIS and academic libraries. This study focuses on the importance of library management systems and the need for the graduates to be equipped with analytics skills. Combined with basic data, text mining and analytics, knowledge classification and information audit skills would benefit libraries and improve resource allocation. Agile institutional libraries in this big data era success hinge on the ability to perform depth analytics of both data and text to generate useful insight for information literacy training and information governance.

Design/methodology/approach

This paper adopted a living-lab methodology to use existing technology to conduct system analysis and LMS audit of an academic library of one of the highly ranked universities in the world. One of the benefits of this approach is the ability to apply technological innovation and tools to carry out research that is relevant to the context of LIS or other research fields such as management, education, humanities and social sciences. The techniques allow us to gain access to publicly available information because of system audits that were performed. The level of responsiveness of the online library was accessed, and basic information audits were conducted.

Findings

This study indicated skill gaps in the LIS training and the academic libraries in response to the fourth industrial technologies. This study argued that the role of skill acquisition and how it can foster data-driven library management operations. Hence, data mining, text mining and analytics are needed to probe into such massive, big data housed in the various libraries’ repositories. This study, however, indicated that without retraining of librarians or including this analytics programming in the LIS curriculum, the libraries would not be able to reap the benefits these techniques provided.

Research limitations/implications

This paper covered research within the general and academic libraries and the broader LIS fields. The same principle and concept is very important for both public and private libraries with substantial usage and patrons.

Practical implications

This paper indicated that librarianship training must fill the gaps within the LIS training. This can be done by including data mining, data analytics, text mining and processing in the curriculum. This skill will enable the news graduates to have skills to assist the library managers in making informed decisions based on user-generated content (UGC), LMS system audits and information audits. Thus, this paper provided practical insights and suggested solutions for academic libraries to improve the agility of information services.

Social implications

The academic librarian can improve institutional and LMS management through insights that are generated from the user. This study indicated that libraries' UGC could serve as robust insights into library management.

Originality/value

This paper argued that the librarian expertise transcends information literacy and knowledge classification and debated the interwoven of LMS and data analytics, text mining and analysis as a solution to improve efficient resources and training.

Citation

Ajibade, P. and Muchaonyerwa, N. (2023), "The importance of data mining, user information behaviour and interaction audit for information literacy", Library Hi Tech News, Vol. 40 No. 4, pp. 12-14. https://doi.org/10.1108/LHTN-09-2022-0109

Introduction

One of the important roles of the library and any librarian is to facilitate information discovery and often perform information literacy training. Because most academic libraries' data reside in the cloud and databases, data mining skills must be embedded in library training. Basic data mining, text mining, data analytics techniques as well as programming language skills are becoming prominent in the curricula of most library and information science (LIS) schools in the developed world and is a must for future librarians.

One of the critical roles of any librarian is to provide useful information to the library manager for an effective communication and decision-making purposes. The interaction audit of the users and their information behaviour is important to gather this information. These insights cannot be manually generated in a library that services over 20,000 users. It will require the ability to understand information behaviour and to provide data analysis and data analytics and text mining knowledge. The aid of a programming language will be essential because some of the data that we gather will be in the form of text and words; therefore, text mining and sentimental analysis are very vital.

Data mining literacy

Data literacy is the ability to use computational knowledge and soft skills to locate, aggregate and disseminate information. Using programs and software or cloud-based services is vital to distinguished librarians as knowledge workers. All information specialists need to be able to retrieve any data information or data set. The ability to use emerging technologies will only increase efficiency, turnaround time and the ability to increase the precision ratio of the information and the volume. Data mining in this knowledge society era is an additional ability to further information literacy within the library context (Berendt, 2012). Because it provides the ability to mine data and text through knowledge discovery (Olmer, 2008). The library has consistently adopted data mining (Katsurai and Joo, 2021). It was further discussed that this technology and skill is vital in developing knowledge products (Haravu and Neelameghan, 2003) and for knowledge organisations. Besides the use of data mining for knowledge discovery is mostly applied to bibliometric analysis within the LIS field (Jayasekara and Abu, 2018). It has been discussed that student online interaction can be understood by mining such information from various sources (He, 2013) to provide library management intelligence for decision-making through integrated mining exercises (Li and Wu, 2006). In most developing countries libraries with limited resources and financial capabilities to expend on large collection to devise means to minimise wasteful expenditure. Even the libraries in the developed world could limit the potential information overload by not curating items and resources that are not often consulted or used.

Information audits

The ability of librarians to understand the nature of problems either from the system audit perspective or information perspective enables them to be agile in daily operation. Information system audits is often conducted within the banking sector because of the sensitive nature of the information being processed (Wahyudi and Deswandi, 2016). Nevertheless, the same concept could be useful in the library and information service sector.

Just as blockchain technology provides immutability to records (Ahmad et al., 2021), information and system audits provide solid foundation for an agile information governance. And in the context of library, provincial library information systems have been documented through audits (Pradana et al., 2019). System audit and analysis would also help to understand system functions and how an individual interacts with the system. All these analyses could be done separately without any recourse to actual user data. The sentiment analysis could be done independently by using the data and text mining to perform the sentiment analysis.

Sentiment analysis

Recently, data mining is one of the techniques used to discover some of the information that exists within databases. One of the advantages of data mining is the potential to validate knowledge discovery (Gul et al., 2021). It provides a logical pattern and actionable intelligence for the library services to understand large set of textual data. One of the attributes of the online or virtual learning space have in common with the library management systems is the online environment in which these infrastructures are built to facilitate knowledge and information discovery. Text mining has recently been used to support massive online learning platforms (Dina et al., 2021). Because of large volume, libraries must be competent in big data collection, technologically enhanced searching and information retrieval.

Brahimi et al. (2016) note that data mining can help understand chat feeds (and Twitter feeds) polarity, the library can use the same techniques to understand users' sentiments. Text mining has been found to be useful in various contexts and fields of study. For example, the Python programming language was used to analyse social networks (Vidya Kumari and Kavitha, 2018), for opinion (Dhanalakshmi et al., 2016) or information mining (Gopal et al., 2011). The techniques have been used to study a social media platform (Li et al., 2020), just as the library could use it to study user behaviour. Depending on the use and context of research, mining student or faculty member opinion regarding the library information resources (Mathimagal and Jayalakshmi, 2021) could be the next innovative approach in achieving optimal resource usage and return on investment. The analysis of users, user interaction and content engagement because of user-generated content (UGC) could help the collection and acquisition department to strategically modify their strategies.

The library must also shift its thinking paradigm and begin to reason as a business entity that is looking to improve return on investment vis-à-vis content acquisition, engagement and utilisation. This strategy could also help the library understand any shift in training performance and user's perception (Noferesti and Shamsfard, 2016). It will enable the library to generate performance indicators (Saura et al., 2019) for the library services through the user UGC data (Pavaloaia et al., 2019). Irrespective of how valuable the application of sentiment analysis in the library context is, the ability to mine and contextualise a large set of data will be critical for an efficient information services. Information sensing (Ajibade, 2016) and big data analytics skill would enable the information librarian (Ajibade, 2016) to enhance their knowledge-sharing ability (Ajibade et al., 2019). Because most libraries now have social media accounts, performing sentiment analysis would be beneficial, just as Gupta et al. (2021) strategies to analyse TikTok user content. There are many techniques and programming languages that can mine quantitative and qualitative data. Python might be one of the beginner’s friendly choice. Surya Gunawan et al. (2020) used the program to perform public information analysis.

Mining public sentiments has been done before (Yan et al., 2020) and this can be used in the library context to find out what information users are mostly concerned with and the traffic and duration of time spent could help the library. This information could help them plan information training services or use such for collection development purposes. An information librarian could open a library training Twitter feed, collect real-time reaction and collate questions during training. The ability to quickly gather data and concerns could help the librarian resolve some of the participants' concerns. Albayrak and Gray-Roncal (2019) have explored this technique in another context to predict an event or monitor real-time data. This process will reduce the likelihood of unsatisfactory training sessions. It will minimise the need for a patron to seek further training or wait till the next training session to get some of the needed information.

Web interface audit and user interaction

Information system audits could provide useful information regarding problem that might be hard to understand if this audit is not conducted. From the root source (starting point), many branches (menu and sub menus) were analysed. The information architecture of the web interface was provided from the diagnosis. Initially, this type of analysis is usually conducted by the system librarian or library management system developer to understand the performance of the LMS or any online resources. One of the ideas behind this analysis was that diagnosing system and performing information audits should be routine task. Therefore, users’ interaction analysis and UGC feedback would make libraries much more efficient and tailor their information literacy training based on business intelligence from their patron’s data.

Having a fast, responsive, interactive LMS is crucial to encourage continual user access because of the low bandwidth and unavailability of mobile data networks in some remote areas. This graphical information is just one of the snippets of information and insights we have generated and the ability to create more would be necessary for libraries with limited resources.

Concluding remarks

The alignment of information technology with the library management depends on key factors. Firstly, the skills requisite to handle technical and technological requirements are vital. One of the suggested programming skills is Python as its syntax could be easily understood compared with other syntax for data or text mining. The training and retraining of the librarian to handle the system and information audits must be continuously carry out. This training enables librarians to perform robust information audits and business intelligence for decision-making that enhances optimal use of resources. The insight that in-depth data provides can help the library in designing responsive information literacy training. Furthermore, the analysis could help the library identify whether lack of engagement has anything to do with how the training is conducted, the format/types of resources or the user interface that is difficult to navigate.