Impediments to and readiness for Linked Data application in libraries: Pakistani information professionals’ perspective

Abstract

Findings revealed that there is little evidence that exists of the implementation of LD technologies by libraries in developing countries. This study aims to examine the willingness of information professionals along with their perceived barriers to the implementation of LD technologies in Pakistani libraries. Data was collected from Pakistani librarians in an online questionnaire survey. There is broad interest and willingness to implement Linked Data technologies in Pakistani libraries while perceptions of the presence of barriers remain. Participants are willing to explore and learn more about LD technologies as well as advocate and promote the adoption of these technologies. They also want to attend events about LD technologies and their applications in libraries. Though they identified the existence of different potential barriers that have to do with themselves (e.g. lack of awareness of basic LD concepts), their institutions/libraries (e.g. they lack best practices regarding applications of LD in libraries), and the technology itself (e.g. new, complex, costly). These dynamics have implications for LD applications in libraries, especially those in developing countries. The current study is a valuable addition to the literature as no study has been conducted regarding the willingness to adopt LD in libraries and barriers to its implementation. Findings regarding barriers to implementation would helpful for policymakers and IT experts consider these challenges and work to minimize these challenges for the implementation of LD in Libraries.

Keywords

Barriers-Linked Data interest and willingness-Linked Data librarians metadata RDF semantic web

Introduction

Linked Data (LD) is a concept that is still evolving and information environments such as libraries, especially those in developed countries, are slowly adopting its associated technologies. However, although it has been on the horizon and widely utilized in most other sectors for over a decade, libraries in developing countries are yet to consider adopting it. Linked Data enable the publishing and sharing of structured data on the World Wide Web through standards such as the Resource Description Framework (RDF). They also make interconnecting data from diverse sets of sources, including external sources, possible (Bizer et al., 2009). The publishing and interlinking of data online created a global data space, a Web of data, or a Semantic Web. The use of LD to link online resources and data also increases the potential value and usefulness of published data. All these are a culmination, over the last couple of decades, of the evolution of the World Wide Web from a hypertext-based Web of documents mainly built for human users to LD which is a Web of Data meant to be, for the most part, machine-readable.

In other words, LD is the logical next step in the development of the World Wide Web. Because, in addition to making persistent identification of LD possible, it allowed individual data elements to be represented, identified, interlinked, and machine-processable & accessible. LD also seems to have redefined the WWW in other ways. First, it made the creation, management, and sharing of large datasets online possible. The interlinking of individual data elements and datasets and making their relationships processable and understandable by both machines and humans facilitate their visibility, accessibility, and reuse.

All the characteristics and benefits of LD are an ideal fit for information environments such as libraries and other cultural heritage institutions. Because, not only do these institutions have, as part of their collections, data and information resources that are rare, historical, and valuable, some of the resources are often rarely used. In developing countries, information resources and collections are less likely to be available in a digital format, hence not widely accessible. Although many other factors, including IT infrastructure and financial and social capital in libraries, may affect the implementation of LD technologies in libraries (Warraich and Rorissa, 2018). The current study focused on the application of LD technologies in information environments in the context of a developing country. The findings of the study may be generalizable to developing countries with similar characteristics.

The main functions and responsibilities of information environments such as libraries are selection, identification, acquisition, organization, management, retrieval, dissemination, and preservation of information resources to provide information services to potential users. As such, they have traditionally been considered the main facilitators of access to information in all its forms. However, with the proliferation of personal and mobile devices, more and more users search the Web first, at least initially, to meet their information needs. Unless information environments invest in, adopt, and implement technologies such as LD that will allow them to widely publish data on their resources online, thereby creating an online presence or representation of resources in their collections through LD records, they would expect to serve their potential users only partially.

By their very nature, LD technologies have been designed to not only interlink and widely share data online but also enable information and data service providers to make their collections widely available and accessible as well as increase their reach and use. Publishing LD records will also open the possibility of more access to collections and resources, in all kinds of information environments, by the so-called “Google” or “Internet” generation and could draw more of them to the information environment. LD has shown to be an effective tool for increasing the visibility of resources and collections held by information environments such as libraries (Debattista et al., 2018), especially when users utilize search engines for their information seeking (Ali and Warraich, 2018). Students at all levels tend to overwhelmingly use search engines for their educational and research needs (Halavais, 2013). However, without any use of LD technologies, a small proportion of libraries’ digital collections shows up in search engine results.

LD technologies may provide solutions to these challenges and help libraries direct traffic to their wealth of information resources (Winer, 2014). There are signs that some libraries are publishing their catalogs and authority files using the LD format and principles (Hidalgo-Delgado et al., 2019). Investigations into the implementation of LD in libraries have also been conducted (Alemu et al., 2012; Ali and Warraich, 2018; Cagnazzo, 2017; Monyela, 2022; Weidner et al., 2019) and frameworks for publishing LD have been proposed (Hidalgo-Delgado et al., 2019).

While limited online accessibility and availability of library collections, resources, and catalogs are some of the common issues for any information environment, they are more acute for those in developing countries like Pakistan. The country, with a population of about 200 million (sixth in the world), has five provinces, 177 universities & degree-awarding institutions, and a 55% literacy rate. All universities and degree-awarding institutions in Pakistan have central and seminar libraries that provide information services to members of the broader university communities and specific entities such as departments, colleges & schools, institutes, and centers. As part of its support for higher learning institutions, over the last 10 years, the government of Pakistan had invested in information and communication technologies (ICTs) to allow the institutions and others to develop and provide e-learning, e-library, and e-government services, thereby creating a digital culture that is necessary to transition the country to an information society. The increased adoption and use of ICTs come with expectations, by both the government and users, of academic institutions, their libraries, and information professionals to create digital information environments with new and innovative ways of making their collections and resources widely available as well as providing efficient and effective information services. We believe that LD has the potential to be an enabling technology that will help higher learning institutions in Pakistan and their libraries meet those expectations.

Previous studies on LD in libraries and information centers explored challenges to and motivating factors for its implementation (Wang and Yang, 2018; Warraich and Rorissa, 2020); LD and digital libraries (Hidalgo-Delgado et al., 2019; Raza et al., 2019); implementation of open source LD systems (Weidner et al., 2019); technical, ethical, legal and other issues in adoption of LD in libraries (Saleem et al., 2018); LD problems for digital libraries (Debattista et al., 2018); librarians’ knowledge and view about LD and its allied tools, librarians experience about LD and its challenges, and benefits for libraries, archives, and museums (McKenna et al., 2018). Although these studies reported different challenges to and perceptions of library and information professionals regarding the implementation of LD technologies in libraries and information centers, some of them provided examples of best practices. Besides, only a few of the studies were empirical. Most of them were limited in scope and were conducted in developed countries. Empirical studies of librarians’ level of readiness and barriers to implementing LD technologies in Pakistani libraries were also missing from previous literature.

The current study is an attempt to examine the willingness of information professionals along with their perceived barriers to the implementation of LD technologies in Pakistani libraries. An added reason for the study is the fact that LD is one of the little understood phenomena, especially concerning its adoption and implementation by libraries in general and Pakistani libraries in particular. The main aim of the study is to explore information professionals’ willingness and level of interest to adopt and implement LD technologies in Pakistani libraries. It also tries to present any perceived impediments to LD applications in Pakistani libraries. Therefore, the current study explored Pakistani LIS professionals’ willingness and impediments to implementing LD technologies in academic libraries.

The current study provided interest and willingness to adopt Linked Data in libraries such as the interest in attending events related to Linked Data standards, willingness to explore ways to incorporate Linked Data standards in bibliographic records, and interest in exploring the creation of bibliographic records for incorporating in the library. The study also provides barriers to implementing LD technology in libraries. These barriers included a general lack of awareness of LD among information professionals, lack of practice of LD in libraries, Linked Data technology being too new and unproven, and the high cost of LD technology for implementation.

Background

Linked data applications and libraries

According to Bizer et al. (2009), although the World Wide Web is responsible for a radical shift in how documents and knowledge are shared and accessed online, the Web’s ability to link data, in addition to documents, was limited until recently. However, the increased adoption of Linked Data (LD) principles and technologies in the last decade transformed the web from a global information space where only documents could be interlinked into a space where a web of data with meaningful relationships among related objects could be created. Due to the novelty of LD, the literature establishes a relationship between the elements of the Semantic Web and library catalogs (Berners-Lee and Swick, 2006; Coyle, 2012; Gracy et al., 2013; Greenberg et al., 2009; Hannemann and Kett, 2010; Skevakis et al., 2014; Winer, 2014).

While these studies periodically provide an overview of LD in libraries, others have discussed the application of LD technologies in different disciplines. However, this study focuses only on the relevant literature in the context of libraries. Some of the LD initiatives in libraries are direct results of efforts by the W3C’s Library LD Incubator Group (Isaac et al., 2011; Wang and Yang, 2018). Other LD initiatives include those by the Library of Congress; the German National Library; the Swedish National Library; the Hungarian National Library; and the European Digital Library (Isaac et al., 2011). It is no surprise that these and other libraries are leading such efforts. Because, libraries, generally, offer a natural setting for implementing LD projects (Singer, 2009). They are institutions that often collect, curate, manage, provide access to, preserve, and disseminate data and information to end-users (Miller and Westfall, 2011). Salo (2010) acknowledged the potential of LD in libraries in the context of service and stated that “Linked data deserves special mention here, not so much for its technical details as for the mindset of building data and metadata with the express intent of easy sharing and remixing.” On the other hand, Stevenson (2012) advises lowering expectations for LD’s potential to transform information access and sharing but, rather, shifting attention to building a viable LD business model that could lower the cost of MARC-based cooperative cataloging and increase its quality.

Because libraries create and process MARC-based bibliographic records that do not usually lend themselves to conversion to LD-compliant RDF format. It is assumed that, currently, library data do not allow easy and automatic conversion due to non-standard URI tags for some information objects (Greenberg and Robertson, 2002). In other words, some challenges remain when it comes to the application of LD technologies and standards in some information environments such as libraries. According to Gonzales (2014), “the benefits of LD to libraries and their users are potentially great, but so are the many challenges to its implementation” (p. 10). This point is significant for the current work because it focuses on information professionals’ level of interest and perceived barriers to explore LD technologies for compatible bibliographic data.

More than any work on LD and its applications in libraries, the work by LaPolla (2013) served as a foundation and catalyst for our work, mainly because the population under study was library catalogers in the United States. LaPolla (2013) recognized the potential that Semantic Web technologies hold to improve the library catalog. He also argues that the eventual access to information resources can, partly, be realized if librarians have adequate levels of understanding of the key concepts and positive attitudes toward LD applications and standards. LaPolla (2013) also believes that scarce financial resources and lack of best practices concerning Semantic Web technologies such as LD will be among potential barriers to implementing LD in libraries.

Ali and Warraich (2018) compiled a detailed characterization of potential challenges to implementing LD technologies in libraries and information centers, through a systematic review of the literature. Some of the challenges identified include those that are technical and ontological; others are related to LD tools, existing data sets, lack of guidance, and non-availability of open licenses to implement LD technologies. McKenna et al. (2018) have also identified major challenges related to LD tools, despite prior knowledge of LD concepts by the majority of their study’s participants. Saleem et al. (2018) also reported technical, legal, and ethical issues and challenges when it comes to the implementation of LD in libraries. Rasmussen Pennington and Cagnazzo (2018), based on a survey of Scottish information professionals, concluded that participants lacked understanding of and awareness about LD as well as expertise and technical skills, while their libraries lack financial resources and time.

The main motivation for the current study is the dearth of relevant literature on the nature of applications of LD technologies in libraries in general and those in developing countries in particular. There is a disparity between libraries in developed and developing countries in terms of their automation level, level of skill of their staff, levels of the budget available to them as well as cultural barriers they encounter to discharge their responsibilities (Ramzan, 2004). Even with these challenges, there appears to be a positive outlook on the role and potential of LD technologies in libraries of all types. However, there are few studies in general and fewer empirical studies that attempted to assess the level of efforts by and interest of information professionals to implement LD technologies in libraries in developing nations such as Pakistan and the potential barriers they anticipate. Consequently, we attempted to address the following two research questions:

What are the levels of interest in and willingness to the application of Linked Data technologies in Pakistani libraries?

What are the perceptions of information professionals in Pakistan concerning barriers to Linked Data applications in their libraries?

Methods

The current study is part of a broader project about the nature of the application of Linked Data (LD) technologies in academic libraries in Pakistan. Specifically, its focus is on the examination of information professionals in Pakistani libraries concerning their willingness and level of interest in the application of LD as well as their perceptions of any potential barriers to LD application in their libraries. Data collection from information professionals working in higher education institutions in Pakistan was conducted through an online questionnaire.

The questionnaire was developed through a thorough review of the relevant literature as well as inputs from experts who participated in a sorting activity to ensure the content validity of the items in the survey instrument. Multiple items, using a seven-level Likert Scale (1 = Strongly Disagree, 2 = Disagree, 3 = Slightly Disagree, 4 = Neutral (Neither Disagree nor Agree), 5 = Slightly Agree, 6 = Agree, 7 = Strongly Agree), were utilized to measure several constructs, including the two main ones (interest and willingness; and barriers to the application of LD technologies)

A sorting task

Before the questionnaire was distributed to potential participants, a sorting task was designed and deployed online where six expert judges participated in the sorting task. The sorting task aimed to ensure the reliability and validity of measurement scales. The consistent sorting of an item into a construct is an indication of convergent validity. A sorting task is considered one of the most common ways to assess construct validity including discriminant and convergent validity. It is also used to see if constructs have adequate domain coverage through analysis of items in the questionnaires (Moore and Benbasat, 1991). In this activity, a measure of inter-judge agreement for pairs of five (5) judges was used (sorting data for the sixth expert judge had to be excluded due to low agreement values). All expert judges were initially contacted through email that described the purpose of the study. When they agreed to contribute to the sorting task, they were sent a link for OptimalSort with clear instructions about the sorting task. Although the range of Cohen’s Kappa (Cohen, 1960) values (for measuring inter-rater reliability between variables) and the raw agreement scores were between 0.58 & 0.88 and 0.47 & 0.85, respectively, the average values of those two measures were 0.76 and 0.69. These were considered acceptable (Jarvenpaa, 1989; Todd and Benbasat, 1989; Vessey, 1984) leading to the conclusion that the scale development process formed a questionnaire with construct validity.

Population and data collection

In the absence of a readily available sampling frame, the selection of a random sample was not possible. Therefore, as a faculty member of a LIS school in Pakistan and someone familiar with the network of library and information professionals in the country, the first author identified and recruited potential participants All 177 higher learning institutions and universities to have at least one library or information center with librarians and/or information professionals. Potential participants were contacted through listservs, social media, email, and the telephone and by sharing the URL to the questionnaire, when possible. Respondents were purposively selected. Following a few reminders and efforts to encourage librarians to participate in the survey, within 4 weeks, 136 were distributed questionnaires, of which 86 fully completed ones, were received. We noticed that because LD is a relatively new and not well-known concept and technology, especially when it comes to its applications in libraries, a few of the potential participants might have hesitated to complete the questionnaire.

Findings and discussion

Demographic information

Of the 136 distributed questionnaires, 100 were received. A data cleaning procedure revealed that 14 of them were partially completed while 86 were completed fully and used for subsequent analyses. In Pakistan, information professionals progress through the following ranks (in order from the lowest to the highest rank): “assistant librarian,” “librarian,” “senior librarian,” “deputy librarian,” and “chief librarian.” A little over half of the respondents or 45 (52.33%) indicated “librarian” as their designation or job title. “Librarian” is the most common job title among Pakistani information professionals. Thirteen (15.12%), six (6.98), one (1.16%), and eight (9.30%) of the respondents were “assistant librarian,” “senior librarian,” “deputy librarian,” and “chief librarian,” respectively (See Figure 1). The remaining 13 respondents (15.12%) picked “other” as their job title. The various titles that these 13 had included metadata librarian, assistant director, archive officer, regional library coordinator, and other academic designations, including lecturer, assistant professor, and associate professor.

Figure 1.

Respondents’ designation or job title.

In terms of the respondents’ highest level of education, two-thirds or 58 (67.44%) of them hold a Masters’ degree while 25 (29.7%) of them hold an MPhil degree (which is a research degree that one earns after about 18 years of education in a 30-credit program). The remaining three (3.49%) respondents are Ph.D. holders.

Pakistan is the sixth most populous country with five provinces and federal capital. Sixty (69.77%) of the respondents were from Punjab, the province with more than half of the country’s population and higher education institutions. Ten respondents (11.63%) were from the federal capital, Islamabad, followed by six (6.98%) from Khyber Pakhtunkhwa. Only 10 respondents were from the other three provinces: Sindh—5 (5.81%), Balochistan—3 (3.49%), and Gilgit-Baltistan—2 (2.33%).

In terms of their primary responsibilities, because there is a shortage of library professionals at university libraries in Pakistan, a professional often assumes more than one role/duty to run their small or medium-sized library effectively (Warraich, 2011). More than half of our survey participants, 48 (55.81%), indicated their primary role/duty includes user services while 36 (41.86%) were responsible for acquisitions. Thirty-four (39.53%) and 31(36.05%) respondents worked in cataloging services and information technology & user instruction services. The remaining 25 (29.07%), 23 (26.74%), and 17 (19.77%) respondents had circulation services, collection development, and special collections, respectively, as their primary responsibilities.

Respondents’ interest and willingness to adopt Linked Data technologies in libraries

On a seven-point Likert scale (1 = Strongly Disagree to 7 = Strongly Agree), respondents were asked to rate the level of their interest and willingness to adopt LD technologies in their libraries through five statements. An option, “Don’t Know/Not sure/Not Applicable,” was available to those who are not able to rate a particular item/statement. The computed value of Cronbach’s (1951) Alpha for the five statements used to measure interest and willingness to adopt LD is 0.886, an indication that the items that make up the construct’s measure have relatively high internal consistency or reliability. Furthermore, we performed factor analysis to know which item belongs to their corresponding factor. For this purpose, we conducted a Principal Component Analysis (PCA) using the Kaiser–Guttman criterion (Guttman, 1954; Kaiser, 1960, 1970) where components with eigenvalues greater than one are retained. Results of the PCA (see Table 1) showed that the five items form a single composite variable that explains about 70% of the variation in the set of five items, hence our conclusion that all five items/statements measure a single construct, “Interest and Willingness to Adopt Linked Data.” PCA results also depict (Table 5) that five items of a single composite variable (barriers to implement Linked Data technology) explained about 67% variance and eigenvalues of all the constructs were greater than one. Hence, we concluded that all the items were measured to construct “barriers to implement Linked Data technology”

Table 1.

Total variance explained through principal component analysis—interest and willingness to adopt Linked Data.

Component	Initial Eigenvalues			Extraction sums of squared loadings
Component	Total	% of variance	Cumulative %	Total	% of variance	Cumulative %
1	3.468	69.351	69.351	3.468	69.351	69.351
2	0.629	12.572	81.923
3	0.420	8.404	90.328
4	0.311	6.225	96.553
5	0.172	3.447	100.000

Results of our analysis (see Table 2) show that information professionals in Pakistan are very much interested in and willing to adopt LD technologies in their libraries. All five items/statements received a mean rating of at least 5.19, well above the neutral value of 4. The item with the highest mean rating (5.59 on a seven-point scale) is their willingness to ‘attend events where the topic of creation of bibliographic records that aim to incorporate LD standards is discussed, although the same item also had the highest Standard Deviation (1.508) of all the five statements. This is an indication that, while there is no broader consensus, respondents are at least amenable to exploring more about LD by, among other things, educating themselves on the applications of LD through attending events where relevant topics are discussed.

Table 2.

Respondents’ interest and willingness to adopt Linked Data in libraries (N = 86).

Statements	Mean	S. D.
Interested in attending events where the topic of creation of bibliographic records that aim to incorporate Linked Data standards is discussed	5.59	1.508
Willing to explore ways to incorporate Linked Data standards in bibliographic records management at my institution/library	5.33	1.359
Interested in exploring the creation of bibliographic records that aim to incorporate Linked Data standards at my institution/library	5.30	1.275
Willing to promote incorporating Linked Data standards in bibliographic records management at my institution/library	5.29	1.396
Interested in advocating for creation of bibliographic records that aim to incorporate Linked Data standards at my institution/library	5.19	1.355

Scale: 1 = Strongly Disagree, 2 = Disagree, 3 = Slightly Disagree, 4 = Neutral (Neither Disagree nor Agree), 5 = Slightly Agree, 6 = Agree, 7 = Strongly Agree, Don’t Know/Not sure/Not Applicable.

Among the indicators or items of levels of interest and willingness to adopt LD technologies in libraries, the statement “willing to explore ways to incorporate Linked Data standards in bibliographic records management” had the second highest mean rating (Mean = 5.33, SD = 1.359). This is another sign that information professionals in Pakistan are on track to being ready to adopt LD technologies and standards in their libraries by not only showing interest in learning more about the technologies, exploring further, and advocating for the technologies but also being willing to promote the adoption of these technologies

For instance, the third highest rated statement, “interested in exploring the creation of bibliographic records that aim to incorporate Linked Data standards at my institution/library” (mean rating value of 5.30, SD (1.275) shows that Pakistani information professionals have moved beyond simply being interested in the technologies to actively exploring ways to incorporate them into their libraries’ operations. We can also safely say that the respondents would also like to familiarize themselves with the technologies and standards through the exploration to promote them and advocate for their adoption to convince others, such as library management and other stakeholders.

The two groups of respondents, in terms of their level of education, are master’s and MPhil/Ph.D. degree holders. To ascertain that the two groups are similar in the rating of their interests and willingness to adopt LD technologies in their libraries, total or summated scale scores were determined by adding responses (Grable et al., 2015) to the five items used to measure interest and willingness. Table 3 presents the descriptive statistics (mean and standard deviation scores) for the summated scale for both groups of respondents.

Table 3.

Descriptive statistics (of the summated scale) of information professionals’ interest & willingness to adopt Linked Data in libraries (N = 86).

Level of education	N	Mean	S. D.	Std. error mean
Master’s degree	58	26.5455	6.20046	0.81416
MPhil & PhD	28	27.0357	4.66255	0.88114

It is evident from Table 3 that both groups, Masters and MPhil/Ph.D. degree holders, have similar mean rating scores on the five items used to measure their interest and willingness to adopt LD technologies in their libraries. To further test the assumption that the two groups have little difference in terms of their mean rating scores, an independent samples t-test was performed and the assumption of homogeneity of variance was tested and satisfied via Levene’s (1960) test (F (84) = 0.703, p = 0.404) (see Table 4). The independent samples t-test also confirmed that there was no statistically significant difference between the two groups concerning their interest and willingness to adopt LD technologies in Pakistani libraries (t (84) = −0.370, p = 0.712 (two-tailed)).

Table 4.

Independent samples t-test between Masters and MPhil/Ph.D. degree holders for their interest and willingness.

	Levene’s test for equality of variances		t-test for equality of means
	F	Sig.	T	Df	Sig. (two-tailed)	Mean difference	Std. error difference	95% confidence interval of the difference
	F	Sig.	T	Df	Sig. (two-tailed)	Mean difference	Std. error difference	Lower
Equal variances assumed	0.703	0.404	−0.307	84	0.712	−0.49025	1.32346	−3.12210
Equal variances not assumed			−0.409	68.97	0.684	−0.49025	1.19969	−2.88360

Therefore, it is evident and safe to conclude that our respondents’ level of education had no bearing on the ratings of their interest and willingness to adopt LD technologies in their libraries. That is, the mean ratings by those with Masters degrees were not significantly different from the mean ratings by those with MPhil/Ph.D. degrees.

Barriers to implementing Linked Data technologies in Pakistani libraries

Once again, five items/statements were used to measure Pakistani information professionals’ perceived barriers to implementing LD technologies in their libraries. Respondents were asked to rate each item/statement on a seven-level Likert Scale ranging from Strongly Disagree (1) to Strongly Agree (7). Cronbach’s (1951) Alpha value for the five measurement items/statements is 0.875. This shows that the items/statements have high internal consistency or reliability. Also, similar to the procedures used for the “Interest and Willingness to Adopt Linked Data” construct above, before proceeding with subsequent data analysis, a Principal Component Analysis was conducted to assess whether redundancy exists among the measurement items/statements for the “Barriers to implement LD technologies in Pakistani libraries” construct, using the Kaiser–Guttman criterion (Guttman, 1954; Kaiser, 1960, 1970). Because only one component was extracted from the five items/statements (See Table 5), it was assumed that all five items/statements are measures of the construct.

Table 5.

Total variance explained through principal component analysis—Barriers to implementing Linked Data technologies in Pakistani libraries.

Component	Initial Eigenvalues			Extraction sums of squared loadings
Component	Total	% of variance	Cumulative %	Total	% of variance	Cumulative %
1	3.335	66.704	66.704	3.335	66.704	66.704
2	0.706	14.126	80.830
3	0.395	7.908	88.738
4	0.305	6.091	94.829
5	0.259	5.171	100.000

Mean ratings, by participants of our survey, of all the five items/statements used to measure “Barriers to implement LD technologies in Pakistani libraries” were above 4.97, meaning that they either “Slightly Agree” or “Agree” or “Strongly Agree (Table 6).” Among the five items/statements, those that received the two highest mean ratings have to do with a lack of awareness, among information professionals, of basic LD concepts and a lack of best practices regarding applications of LD in libraries. These two seem to be intertwined, in that, they are both about the shortcomings of information professionals and/or their institutions, rather than the technologies. They are also critical to the adoption of LD technologies, or any other technology, in libraries. Our findings are consistent with previous research by Rasmussen Pennington and Cagnazzo (2019). They also found a lack of awareness about LD among information professionals at Scottish and European national libraries. The next three items/statements have to do with barriers that are directly associated with LD technologies, rather than information professionals and/or their institutions. These items/statements are about the fact that the technology is either new or complex or costly. They are also indicators of apprehension by participants of our survey and their institutions to adopt LD technologies. Our findings align, for the most part, with results obtained through previous studies by others (e.g. LaPolla, 2013), although LaPolla (2013) found cost as a major barrier to the implementation of the technologies in libraries. Another study also provided that the complexity of LD technology is a challenge for the adoption of LD in libraries. Researchers opined that the MARC format is a challenge as it is not suitable for replacement, Disparity in the terminology of library-related and web-related standards, and complexity of LD technology standards such as RDF, SPARQL, etc. (Alemu et al., 2012). The findings regarding the barriers to the adoption of LD in libraries are also aligned with the findings of a systematic review. Researchers reported challenges related to financial and human resources (Ali and Warraich, 2018).

Table 6.

Descriptive statistics—barriers to implementing Linked Data technologies in Pakistani libraries.

Statements	Mean	S. D.
There is a general lack of awareness of the basic concepts of Linked Data among library and information professionals	5.92	1.321
There is lack of best practices with respect to Linked Data and their application in libraries	5.53	1.403
Linked Data technology is too new and unproven in libraries	5.26	1.489
The technology is too complex to implement a Linked Data compliant catalog in libraries	5.01	1.515
Implementing Linked Data technology would be too costly for libraries in terms of financial and human resources	4.97	1.456

Scale: 1 = Strongly Disagree, 2 = Disagree, 3 = Slightly Disagree, 4 = Neutral (Neither Disagree nor Agree), 5 = Slightly Agree, 6 = Agree, 7 = Strongly Agree, Don’t Know/Not sure/Not Applicable.

Once again, the mean ratings of the two groups of respondents, those holding Masters’ and MPhil/Ph.D. degrees, concerning barriers to implementing LD technologies in Pakistani libraries were compared to see if there are any discernable differences. Table 7 shows that the two groups of respondents rated the five items/statements similarly. In addition, an independent samples t-test (t (84) = −0.549, p = 0.585 (two-tailed)) and homogeneity of variance test through Levene’s (1960) test (F(84) = 0.489, p = 0.486) (see Table 8) confirmed that there was no statistically significant difference between the two groups concerning their mean ratings. That is, both groups have similar opinions when it comes to their perceived barriers to implementing LD technologies in Pakistani libraries.

Table 7.

Descriptive statistics (of the summated scale) of barriers to implement Linked Data technologies in Pakistani libraries (N = 86).

Level of education	N	Mean	S. D.	Std. error mean
Master’s degree	58	26.4483	6.24969	0.82063
MPhil & PhD	28	27.1928	5.06741	0.95765

Table 8.

Independent samples t-test between Masters and MPhil/Ph.D. degree holders to their perceptions of barriers to Linked Data application in Pakistani libraries.

	Levene’s test for equality of variances		t-test for equality of means
	F	Sig.	T	Df	Sig. (two-tailed)	Mean difference	Std. error difference	95% confidence interval of the difference
	F	Sig.	T	Df	Sig. (two-tailed)	Mean difference	Std. error difference	Lower
Equal variances assumed	0.489	0.486	−.549	84	0.585	−.74448	1.35670	−3.44242
Equal variances not assumed			−.590	64.689	0.557	−.74448	1.26116	−3.26341

Summary, conclusions, and recommendations

To address an apparent gap in the literature concerning the nature of the application of LD technologies by libraries in developing countries, we surveyed information professionals in Pakistan. Using a systematically developed questionnaire (including a literature review and inputs from subject experts through a sorting task), data were collected on their level of interest and willingness to implement LD technologies in Pakistani libraries as well as their perceived barriers to LD applications. Five items/statements constituted the scale used to measure each of the two constructs, where each item/statement was rated on a seven-level Likert Scale (1 = Strongly Disagree, 2 = Disagree, 3 = Slightly Disagree, 4 = Neutral (Neither Disagree nor Agree), 5 = Slightly Agree, 6 = Agree, 7 = Strongly Agree). Two questions about the perceived willingness and barriers to implementing LD technologies in academic libraries guided the study.

The systematically developed questionnaire was distributed to all potential participants - information professionals in Pakistani libraries. Evaluation of the 100 responses received revealed that only 86 of them were complete responses and these were subsequently analyzed. In terms of the internal consistency or reliability of the two sets of items/statements, high Cronbach’s (1951) Alpha values (0.886 for the level of interest and willingness and 0.875 for barriers to LD application in their libraries) indicated that both sets of items/statements had high internal consistency or reliability. Also, Principal Component Analyses (PCA) showed that a single component or composite variable explained a large portion of the variation in each set of items/statements (Tables 1 and 5).

Analysis of our survey data revealed that there is a reasonably high level of interest and willingness to implement LD technologies in Pakistani libraries, although perceptions of the presence of barriers to LD application in Pakistani libraries remain. While participants showed interest in and willingness to explore and learn more about LD technologies as well as advocating for and willing to promote the adoption of the technologies, barriers that have to do with themselves (e.g. lack of awareness of basic LD concepts), their institutions/libraries (e.g. they lack best practices regarding applications of LD in libraries), and the technology itself (e.g. new, complex, costly). These barriers and challenges can be minimized as a previous study (Ali and Warraich, 2018) provided that the professional associations in developing countries should motivate professionals to initiate LD technology in libraries by organizing conferences and workshops. The area of LD technology should be focused on research activities in developing countries. International organizations such as IFLA, and OCLC should devise guidelines for the implementation of LD technology in libraries and deal with the diverse nature of LD technology challenges. Results obtained were consistent across the two groups of participants, those with Master’s and MPhil/Ph.D. degrees, as evidenced by results of the independent samples t-tests that showed no statistically significant differences.

The findings of the current study clearly show that information professionals in Pakistan are open and ready, to learn about and implement LD technologies in their libraries, despite potential impediments. The findings also point to, among other things, the need to organize training workshops and conference presentations/panels or discussions that focus on not only the benefits of implementing LD technologies in libraries but also to share any best practices and initiatives that others could benefit from. In addition, library and information science (LIS) schools or programs and/or professional organizations, perhaps through a collaborative effort, need to either introduce new courses on LD technologies and their applications in libraries or add modules to existing and relevant courses. Higher learning and cultural heritage institutions that offer library and information services need to encourage their information professionals, and provide them with the necessary level of support, to pursue professional development opportunities. They also need to invest in the implementation of LD technologies in libraries given the benefits of these technologies, especially to end-users.

LD is an emerging technology that has the potential to revolutionize information organization, management, sharing, access, and dissemination. At the same time, they are yet to be fully explored broadly and even less explored in the context of libraries. Therefore, it takes more than a study such as ours to uncover the full picture of the nature of their adoption and implementation in Pakistani libraries.

Our study and its findings relied on a survey of information professionals who work in academic libraries in Pakistan. What is more, two constructs were considered to answer two research questions, among many relevant constructs and research questions. As such, it could be considered a useful baseline study that future researchers could use as a starting point and/or for comparison purposes rather than one that presents a holistic picture of information professionals’ perception of the application of LD technologies in libraries. Therefore, a more comprehensive future study with a broader scope is recommended to both confirm the findings of the current study and to provide a fuller picture of the adoption and implementation of LD technologies in libraries around the globe.

The findings of the study may be generalizable to developing countries with similar characteristics. As we know that the findings of the study from one sample can be generalized for a population of similar characters. In the context of the current study, the libraries in developing countries share the same kind of characteristics such as IT infrastructure, LIS professionals’ skills, etc. and we may assume that they have to face the same kind of challenges to implement emerging technologies in libraries including LD. The findings of the study regarding the interest and willingness of LIS professionals to adopt Linked Data in libraries also showed their keen interest and willingness to do it. This finding is also transferable for developing country professionals to adopt LD technologies in their libraries.

Limitations

In the absence of a readily available sampling frame, the selection of a random sample was not possible. Respondents were purposively selected who are well versed in technology and library automation. The response rate is low as LD is a relatively new and not well-known concept and technology, especially when it comes to its applications in libraries. Few of the potential participants might have hesitated to complete the questionnaire. Moreover, the authors only took academic libraries as the population because academic libraries, and specifically university libraries have better IT infrastructure, automation, and digital/electronic services in Pakistan. Public libraries in Pakistan are in poor condition due to the absence of professional staff, outdated collection, and insufficient budgets, (Warraich, 2013). Public libraries were not included in the study as the public library system is not well developed in Pakistan.

Footnotes

Appendix

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The first author conducted this research at the State University of New York (SUNY) at Albany, NY-USA as Fulbright Visiting Scholar Fulbright funded this research project.

ORCID iD

Nosheen Fatima Warraich

Author biographies

Nosheen Fatima Warraich has been serving as professor and Director at the Institute of Information Management, University of the Punjab, Lahore, Pakistan. She is the recipient of a Fulbright post-doc fellowship in 2015 and the ASIS&T new leader award in 2017. Her research gets published in prestigious journals and conferences in the field.

Abebe Rorissa is the director of the School of Information Sciences. Previously, he served as Associate Dean for Faculty Development and Associate Professor in the College of Emergency Preparedness, Homeland Security, and Cybersecurity at the State University of New York (SUNY) at Albany.

References

Alemu

Stevens

Ross

, et al. (2012) Linked data for libraries: Benefits of a conceptual shift from library-specific record structures to RDF-based data models. New Library World 113(11/12): 549–570.

Ali

Warraich

(2018) Linked Data initiatives in libraries and information centres: A systematic review. The Electronic Library 36(5): 925–937.

Berners-Lee

Swick

(2006) Semantic Web Development. Cambridge, MA: Massachusetts Institute of Technology (MIT)/World Wide Web Consortium (W3C). Available at: https://www.researchgate.net/publication/235026219_Semantic_Web_Development (accessed 15 December 2021).

Bizer

Heath

Berners-Lee

(2009) Linked Data - the story so far. International Journal on Semantic Web and Information Systems 5(3): 1–22.

Cagnazzo

(2017) Linked Data: Implementation, use, and perceptions across European National Libraries (Doctoral Dissertation). Glasgow, UK: University of Strathclyde.

Cohen

(1960) A coefficient of agreement for nominal scales. Educational and Psychological Measurement 20(1): 37–46.

Coyle

(2012) Chapter 2: Semantic web and Linked Data. Library Technology Reports 48(4): 10–14.

Cronbach

(1951) Coefficient alpha and the internal structure of tests. Psychometrika 16: 297–334.

Debattista

McKenna

Brennan

(2018) Understanding information professionals: A survey on the quality of Linked Data sources for digital libraries. In: Panetto

Debruyne

Proper

, et al. (eds) On the Move to Meaningful Internet Systems. OTM 2018 Conferences. OTM 2018. Lecture Notes in Computer Science, vol. 11230. Cham: Springer, pp.537–545.

10.

Gonzales

(2014) Linking libraries to the web: Linked Data and the future of the bibliographic record. Information Technology and Libraries 33(4): 10–22.

11.

Grable

Heo

Rabbani

(2015) Financial anxiety, physiological arousal, and planning intention. Journal of Financial Therapy 5(2): 2.

12.

Gracy

Zeng

Skirvin

(2013) Exploring methods to improve access to music resources by aligning library data with Linked Data: A report of methodologies and preliminary findings. Journal of the American Society for Information Science and Technology 64(10): 2078–2099.

13.

Greenberg

Robertson

(2002) Semantic web construction: an inquiry of authors’ views on collaborative metadata generation. In Proceedings of the 2002 international conference on Dublin Core and metadata applications: metadata for e-communities; supporting diversity and convergence, Florence, Italy (pp. 45–52). Firenze University Press. Available at: http://dcpapers.dublincore.org/pubs/article/view/693/689 (assessed 15 September 2022).

14.

Greenberg

White

Carrier

, et al. (2009) A metadata best practice for a scientific data repository. Journal of Library Metadata 9(3-4): 194–212.

15.

Guttman

(1954) Some necessary conditions for common-factor analysis. Psychometrika 19: 149–161.

16.

Halavais

(2013) Search Engine Society. Hoboken, NJ: Wiley.

17.

Hannemann

Kett

(2010) Linked data for libraries. In: World library and information congress 2010: 76th IFLA general conference and assembly, Gothenberg. Available at: https://www.ifla.org/past-wlic/2010/149-hannemann-en.pdf (assessed 10 October 2022).

18.

Hidalgo-Delgado

Mariño-Molerio

, et al. (2019) A Linked Data-based semantic interoperability framework for digital libraries. Revista Cubana de Ciencias Informáticas 13(1): 14–30.

19.

Isaac

Waites

Young

, et al. (2011) Library Linked Data incubator group: Datasets, value vocabularies, and metadata element sets. W3C Incubator Group Report. Available at: https://www.w3.org/2005/Incubator/lld/XGR-lld-vocabdataset-20111025/ (assessed 24 July 2022).

20.

Jarvenpaa

(1989) The effect of task demands and graphical format on information processing strategies. Management Science 35(3): 285–303.

21.

Kaiser

(1960) The application of electronic computers to factor analysis. Educational and Psychological Measurement 20: 141–151.

22.

Kaiser

(1970) A second generation little jiffy. Psychometrika 35: 401–415.

23.

LaPolla

(2013) Perceptions of librarians regarding semantic web and Linked Data technologies. Journal of Library Metadata 13(2-3): 114–140.

24.

Levene

(1960) A robust test for equality of variances. In: Olkin

Ghurye

Hoeffding

, et al. (eds) Contributions to Probability and Statistics: Essays in Honour of Harold Hotelling. Stanford, California: Stanford University Press, pp.278–292.

25.

McKenna

Debruyne

O’Sullivan

(2018) Understanding the position of information professionals with regards to Linked Data: A survey of libraries, archives, and museums. In: Proceedings of the 18th ACM/IEEE on joint conference on digital libraries (JCDL’18), pp.7−16. New York: ACM.

26.

Miller

Westfall

(2011) Linked Data and libraries. The Serials Librarian 60(1-4): 17–22.

27.

Monyela

(2022) Knowledge organisation in academic libraries: The Linked Data approach. In: Masenya

(ed.) Innovative Technologies for Enhancing Knowledge Access in Academic Libraries. Hershey, PA: IGI Global, pp.71–88.

28.

Moore

Benbasat

(1991) Development of an instrument to measure the perceptions of adopting an information technology innovation. Information Systems Research 2: 192–222.

29.

Pennington

Cagnazzo

(2018, July) Relationship status: Libraries and linked data in Europe. In: Challenges and Opportunities for Knowledge Organization in the Digital Age. Würzburg: Ergon-Verlag, pp.662–665.

30.

Ramzan

(2004) Effects of IT utilization and knowledge on librarians’ IT attitudes. The Electronic Library 22(5): 440–447.

31.

Rasmussen Pennington

Cagnazzo

(2019) Connecting the silos: Implementations and perceptions of Linked Data across European libraries. Journal of Documentation 75(3): 643–666.

32.

Raza

Mahmood

Warraich

(2019) Application of Linked Data technologies in digital libraries: A review of literature. Library Hi Tech News 36(3): 9–12.

33.

Saleem

QUA

Butt

Warraich

(2018) Applications of Linked Data technologies in libraries: Technical and ethical considerations. In: 2018 international conference on information management and processing (ICIMP), pp.11–15. New York: IEEE.

34.

Salo

(2010) Retooling libraries for the data challenge. Ariadne (Issue 64). Available at: http://www.ariadne.ac.uk/issue64/salo/ (assessed 30 June 2022).

35.

Singer

(2009) Linked library data now! Journal of Electronic Resources Librarianship 21(2): 114–126.

36.

Skevakis

Makris

Kalokyri

, et al. (2014) Metadata management, interoperability, and Linked Data publishing support for natural history museums. International Journal on Digital Libraries 14(3-4): 127–140.

37.

Stevenson

(2012) Linking lives: Creating an end-user interface using Linked Data. Information Standards Quarterly 24: 14–23.

38.

Todd

Benbasat

(1989) An Experimental Investigation of the Impact of Computer-Based Decision Aids on the Process of Preferential Choice. Kingston, Ontario, Canada: School of Business, Queen’s University.

39.

Vessey

(1984) An investigation of the psychological processes underlying the debugging of computer programs (Unpublished PhD Dissertation). Department of Commerce, University of Queensland, Australia.

40.

Wang

Yang

(2018) Linked Data technologies and what libraries have accomplished so far. International Journal of Librarianship 3(1): 3–20.

41.

Warraich

(2011) Human resource management in university libraries of the Punjab. Pakistan Journal of Library and Information Science; Lahore 12(1): R1–R2.

42.

Warraich

(2013) The future of quiescent public library system: a developing country perspective. Available at: http://library.ifla.org/id/eprint/273 (assessed 10 November 2022).

43.

Warraich

Rorissa

(2018) Adoption of Linked Data technologies among university librarians in Pakistan: Challenges and prospects. Malaysian Journal of Library & Information Science 23(3): 1–13.

44.

Warraich

Rorissa

(2020) Application of Linked Data technologies in libraries: Pakistani information professionals’ attitudes and perceptions. The Electronic Library 38(5/6): 1035–1051.

45.

Weidner

Washington

Liu

(2019) Planting cedar: An open source Linked Data vocabulary manager at the University of Houston libraries. Journal of Library Metadata 19(1-2): 53–68.

46.

Winer

(2014) Judaica Europeana: An infrastructure for aggregating Jewish content. Judaica Librarianship 18: 88–115.