Evaluating AI Text Detection Tools for Distinguishing Human-Written from AI-Generated Abstracts in Persian-Language Journals of Library and Information Science

This study examines the proportion of women as first authors in major library and information science (LIS) journals over the years 1981–2020. Author name and year data were collected for 10 LIS journals – five that are associated more with library topics and five with information science topics – and analysed using the genderize.io tool. Both general trends over time and comparisons of information science versus library science journals are presented. The findings indicate significant growth in the proportion of women authors among the LIS journals, but primarily concentrated only among the library science journals, with information science journals falling well behind. Representation of women authors (~60%) still lags well below the overall representation of women in librarianship (~80%). These findings suggest that there is still considerable growth needed to decrease the gender gap among authorship in top LIS journals.

Presently, journals are considered as the most important tools of information science and knowledge growth throughout the world. Due to the increase in the number of scientific journals, the selection, evaluation and determination of the authenticity of these resources by the authorized organizations has attached more significance to them. Thus, the present study aims at investigating the compliance of Iranian library and information science journals with basic standards of journal evaluation through Thomson Reuters’ viewpoint. This is an applied research, which has been conducted through comparative analysis. It evaluates the authenticity of scientific journals through four basic standards of Thomson Reuters, namely on-time publishing of journals, observing international publishing laws, full-text in English, and peer review. The population of the research includes all the active scientific Iranian journals in the field of library and information sciences (12 journals). Results showed that the mean of correlation ratio between the studied journals and standards was 75%. On-time publishing and full-text in English were observed only in 33% and 58% of the studied journals respectively. However, observing international laws of publishing and peer review are in optimal status. Studies are needed in order to find out the compliance of the other countries’ library science journals with international standards, (specially developing ones) to help those countries to identify the existing gaps which will assist them to present their researches in the international level through being indexed in authentic databases. Obviously more research is needed in this area, as Thomson Reuters has published standards other than the basic standards.

Asian and Pacific Region Authorship Characteristics in Leading Library and Information Science Journals

Authorship characteristics from the Asian and Pacific region in the top twenty journals in library and information science are studied. Data was collected searching the Institute for Scientific Information (ISI) Web of Science databases. Major findings of this study are: there are a total of 1,317 articles for the period 1967 to 2005; the most productive countries are, in rank order, Australia, China, South Korea, Taiwan, Singapore, Japan, New Zealand, Malaysia, Thailand and Philippines; and 77.6 percent of authors in the top twenty library and information science journals contributed a single article. Among the library science journals about 50 percent were written by multiple authors, while 73.1 percent of articles in the information science journals were written collaboratively. The most productive individual authors in the region are reported. The strongest collaboration within the region took place between Australia and China; China and Singapore; Australia and New Zealand.

Although there are journals whose published papers are cited evenly, there may be journals whose impact factors (IF) are distorted by a few popular papers. This study measures the citation inequality in papers published in 44 international journals in the field of library and information science using the Gini Coefficient and analyzes whether there are differences in citation distribution depending on the research areas of Library Science (LS), Information Science (IS), and Scientometrics (SM). Moreover, it explores whether the citation inequality index was related to the journal’s IF and whether the IF was distorted by a few popular papers. Analysis revealed that first, the inequality in the number of citations of papers included in library and information science (LIS) journals was found to be g = 0.62. Moreover, if the IF is high, the Gini coefficient is low, therefore the number of citations becomes even (r = −0.367, p < 0.05). Additionally, it was confirmed that equality increased as the non-citation rate decreased. Second, there was a significant difference in the Gini coefficients of papers included in journals in the LS, IS, and SM fields (p < 0.05). The number of citations of papers included in SM journals was the most even (g = 0.5), whereas LS journals had the most uneven citation distribution (g = 0.7). Third, even in LIS journals, we found cases in which journal IF inflation occurred because of a few highly cited papers. This phenomenon has been observed in journals including COVID-19 related topics or bibliometric methodologies that attracted attention.

This study employs the method of direct citation to analyze and compare the interdisciplinary characteristics of the two disciplines of library science and information science during the period of 1978---2007. Based on the research generated by five library science journals and five information science journals, library science researchers tend to cite publications from library and information science (LIS), education, business/management, sociology, and psychology, while researchers of information science tend to cite more publications from LIS, general science, computer science, technology, and medicine. This means that the disciplines with larger contributions to library science are almost entirely different from those contributing to information science. In addition, researchers of library science frequently cite publications from LIS; the rate is as high as 65.61%, which is much higher than the rate for information science, 49.50%. However, a decreasing trend in the percentage of LIS in library science indicates that library science researchers tend to cite more publications from non-LIS disciplines. A rising trend in the proportion of references to education sources is reported for library science articles, while a rising trend in the proportion of references to computer science sources has been found for information science articles. In addition, this study applies an interdisciplinary indicator, Brillouin's Index, to measurement of the degree of interdisciplinarity. The results confirm that the trend toward interdisciplinarity in both information science and library science has risen over the years, although the degree of interdisciplinarity in information science is higher than that in library science.

Citation analysis of all the journal articles published in the journal of Library and Information Science (JOLIS) from 2004-2009 is carried out. 72 articles were published in the journal during five (5) years covered. Highest number of (14) articles were published in 2007 and 2008.A total of 998 references were generated by the journals, indicating that 13.7 average citation per articles. The result shows that journals were the most cited materials as it accounted for 37.14% of the total citations, followed by books with 33.14%. The individual articles that had the highest citation had 44 references and was published in 2008. The findings further show that 62 library and information science (LIS) journals cited produced 172 citations. African Journal of Library, Archives and Information science led the ten (10) most cited library and information science journals (LIS) with 40 citations in the journal. 15(24.19%) of the (LIS) journals were published in Nigeria, 45(72.58%) were from outside African continent, while only 2 (3.22%) other journals were from Africa. The majority (38.2%) of materials cited was published in 1995 and beyond, authorship pattern shows that (79.85%) of the materials cited was written by single authors while only 8.8% of the total citations were Internet resources.

Altmetrics indicators are useful for assessing the impact of research and have been increasingly used alongside traditional citations in recent years. Mendeley provides readership statistics which give an early indicator of the impact of research outputs. This study aimed to investigate how Mendeley early readership indicator was associated with later citations across nine selected library and information science (LIS) journals. This study examined bibliographic data of 9 LIS journals extracted from the Scopus database over a 17 months period from June 2019 to November 2020. Data were extracted using Webometric Analyst. Spearman’s rank correlation coefficient was used to characterize the relationship between these two variables. The number of readers per paper, and of each journal are described using mean, standard deviation, frequency, and geometric mean. Readership growth was increased in all selected journals, but citation growth was unstable in most of the journals. Early readership statistics positively correlated with early citation analysis in all journals except the Journal of Educational Media and Library Science, which had a weaker positive correlation. The correlation between early readership and later citation numbers varied, with some journals being moderately positive and some weakly positive.

This study aimed to examine the publication trends of Scimago Journal &amp; Country Rank (SJR)-indexed library and information science (LIS) journals by examining publication share growth by country and region. For this study, we used LIS journals indexed in the SJR in 2000, 2010, and 2020. The results showed that the most frequent publishers of SJR-indexed LIS journals are large commercial publishers. The top three publishers since 2000 were Taylor &amp; Francis, Emerald, and Springer Nature, despite their publication share among SJR-indexed journals declining since this period. The top three countries in journal publishing were the United States, the United Kingdom, and the Netherlands. As the large commercial publishers have established themselves in these countries, the publication shares of these countries have also declined slightly because of the slowing growth rate of their publishers. Similarly, publication shares in both Northern America and Western Europe have declined slightly since 2000. However, the large commercial publishers, based mainly in the United Kingdom and the United States, are likely to take the lead in publishing SJR-indexed international LIS journals over the next decade. The results suggest that publishers from non-Western countries will need to publish significantly more international LIS journals indexed in the major databases to achieve a significant publication share. In conclusion, more innovative ways to support journals published in non-Western countries are needed in order to meet the essential selection criteria of the Scopus and Web of Science journal indexes.

This stated preference study approached the issue on sub-categorization of the information science---library science (IS---LS) journals listed in the Journal Citation Report (JCR) 2011. To investigate this, 243 active authors/editors publishing in this field were requested to indicate their preferred category to 83 journal titles listed in JCR 2011 from four options: information science (IS), library science (LS), information systems (ISys) and do not know/undecided. Based on the popularity count, respondents assigned 39 titles to LS, 23 titles to IS and 21 titles to ISys. Twenty-five titles received high "do-not-know" counts--these are titles in non-English languages, information management and publishing sub-fields. Only one title in LS was grouped in the highest quartile by impact factor, compared to 8 titles in IS and 11 in ISys. This indicates that LS journals are hardly represented among the top 25 % of the impact factor distribution of JCR's ranked IS---LS journals. Respondents show concern about the "fit" of information systems journals in the IS---LS category.

Retraction of scholarly publications ensures that unqualified knowledge is purged from the scientific community. However, there appears to be little understanding about how this is practiced among library and information science (LIS) journals. Hence, this study investigated the correction and retraction practices of LIS journals. Journals included in the Web of Science’s information science and library science subject category were selected for the study and the characteristics of the articles corrected or retracted in those journals between 1996 and 2016 were examined. Findings show that there were 517 corrections and five retractions in LIS journals during the period. Most of the corrections made to articles in LIS journals were minor while the reasons for article retraction included plagiarism, duplication, irreproducible results and methodological errors. Our findings also reveal that on average it took about 587 days for an article to be retracted while some of the retracted articles continued to be cited after retraction. The study concluded that the average number of errors per correction was lower than what had been observed in medical journals while some of the retracted articles continued to receive positive post-retraction citations. It also recommended the inclusion of a check on the validity of literature cited by authors at the review stage as part of the quality control mechanism by publishers of LIS journals.

A Review of:&#x0D; Cirasella, J., &amp; Bowdoin, S. (2013). Just roll with it? Rolling volumes vs. discrete issues in open access library and information science journals. Journal of Librarianship and Scholarly Communication, 1(4). http://dx.doi.org/10.7710/2162-3309.1086&#x0D; &#x0D; Abstract &#x0D; &#x0D; Objective – To understand the prevalence of, motivations for, and satisfaction with using a rolling-volume publishing model, as opposed to publishing discrete issues, across open access academic journals in library and information science.&#x0D; &#x0D; Design – A 12 question survey questionnaire.&#x0D; &#x0D; Setting – English-language, open access library and information science (LIS) journals published in the United States of America.&#x0D; &#x0D; Subjects – A total of 21 open access LIS journals identified via the Directory of Open Access Journals that were actively publishing, and that also met the authors’ standard of scholarliness, which they established by identifying a journal’s peer-review process or other evidence of rigorous review. Based on responses, 12 journals published using discrete issues, while 9 published as rolling volumes or as rolling volumes with some discrete issues.&#x0D; &#x0D; Methods – In late 2011, the study’s authors invited lead editors or primary journal contacts to complete the survey. Survey participants were asked to identify whether their journal published in discrete issues, rolling volumes, or rolling volumes with occasional discrete issues, with the latter two categories combined as one for ease of results analysis. Survey logic split respondents into two groups, either discrete-issue or rolling-volume. Respondents in both categories were posed similar sets of questions, with the key difference being that the questions directed at each category accounted for the publication model the journals themselves identified as using. Editors from both groups were asked about the reasons for using the publication model they identified for their journal: within the survey tool, authors provided 16 potential reasons for using a discrete-issue model, and 13 potential reasons for using a rolling-volume model. Respondents from both groups were asked to mark all reasons that applied for their respective journals. The survey also included questions about whether the journal had ever used the alternate publishing model, the editor’s satisfaction with their current model, and the likelihood of the journal switching to the alternate publishing model in the foreseeable future. &#x0D; &#x0D; Main Results – The authors collected complete responses from 21 of the original 29 journals invited to participate in the study, a response rate of 72%. For the 12 journals that identified as using discrete issues, ease of production workflow (91.7%), clear production deadlines (75.0%), and journal publicity and promotion (75.0%) were the three most common reasons for using a discrete-issue model. For the nine journals using rolling volumes, improved production workflow (77.8%), decreased dependence on production deadlines (77.8%), and increased speed of research dissemination (66.7%) were the three most common reasons cited for using a rolling-volume model. &#x0D; &#x0D; Findings show that overall satisfaction with a journal’s particular publication model was a common factor regardless of publishing model in use, though only the rolling-volume editors unanimously reported being very satisfied with their model. This high satisfaction rate is reflected in editors’ positions that they were very unlikely to switch away from the rolling-volume method. While a majority of editors of discrete-issue journals also reported being very satisfied or somewhat satisfied with their current model, the mixed responses to whether they would contemplate switching to the alternate model suggests that awareness of the benefits of rolling-volume publishing is increasing. &#x0D; &#x0D; Conclusion – Researchers discovered a greater incidence of rolling-volume model journals with open access LIS journals than anticipated, suggesting that this is an area where additional research is necessary. The relative newness of the rolling-volume model may be a contributing factor to the high satisfaction rate among editors of journals using this model, as journal editors are likely to be more deliberate in selecting this model over the traditional discrete-issue publishing model. Workflow and production practices were identified as key characteristics for selecting a publishing model regardless of the model selected, and therefore this is another area in need of further investigation.

The purpose of this paper was to identify the existing criteria of various conventional databases that are considered in evaluating journals in the scholarship of library & information science (LIS) journals and accordingly developed a tool box to evaluate some selected journals of LIS in South Asian Association of Regional Cooperation (SAARC) countries. Journals have been evaluated by applying 30 criteria based on the current measures used by Thompson Reuter, SCOPUS, SciELO, LISA, LISTA, etc. The result indicates that although there are a considerable number of journals being published in India since long time, only a few journals are qualitatively strong. The review policy as mentioned in documentation, subject coverage can be considered as their strength, the geographic non-diversity of members in Editorial Board, contributors are their weakness. Most of journals are indexed in LISA and LISTA, however no journals are yet to include in JCR. Overall, to cope-up with international standard journals need to consider their publication policy thoroughly. The findings of the study seems to be useful for (a) academics – to know the list of journals which adhere to the quality requirements of LIS discipline; (b) librarians – to know the core LIS journals of SAARC countries in LIS discipline for their clients; (c) policy makers – to measure the weightage of publication, while evaluating performance for career and promotion, to evaluate individual research performance, while releasing grants for academic projects. DESIDOC Journal of Library & Information Technology , 2014, 34(2), pp. 152-161. http://dx.doi.org/10.14429/djlit.34.4968

Several studies discussed the characteristics of authors who published in Library and Information Science journals. Although none focused specifically on Arab librarians as authors, the current study attempts to reveal the scholarly contributions to library literature by Arab librarians. The study describes and analyses the journal research publications in Library and Information Science journals by professional librarians from 1981 to 2010. Single-author articles are found to be highly followed by two and three authored articles. The average degree of collaboration between authors in Library and Information Science journals is 9.64% (only 19 journal articles written by at least two or three authors). Finally, this study provides recommendations to Arab librarians to encourage them to be engaged in research in the Library and Information Science discipline.

A study of Thomson‐Scientific ISI ranked Library and Information Science (LIS) journals (n = 52) is reported. The study examined the stances of publishers as expressed in the Copyright Transfer Agreements (CTAs) of the journals toward self‐archiving, the practice of depositing digital copies of one's works in an Open Archives Initiative (OAI)‐compliant open access repository. Sixty‐two percent (32) do not make their CTAs available on the open Web; 38% (20) do. Of the 38% that do make CTAs available, two are open access journals. Of the 62% that do not have a publicly available CTA, 40% are silent about self‐archiving. Even among the 20 journal CTAs publicly available there is a high level of ambiguity. Closer examination augmented by publisher policy documents on copyright, self‐archiving, and instructions to authors reveals that only five, 10% of the ISI‐ranked LIS journals in the study, actually prohibit self‐archiving by publisher rule. Copyright is a moving target, but publishers appear to be acknowledging that copyright and open access can co‐exist in scholarly journal publishing. The ambivalence of LIS journal publishers provides unique opportunities to members of the community. Authors can self‐archive in open access archives. A society‐led, global scholarly communication consortium can engage in the strategic building of the LIS information commons. Aggregating OAI‐compliant archives and developing disciplinary‐specific library services for an LIS commons has the potential to increase the field's research impact and visibility. It may also ameliorate its own scholarly communication and publishing systems and serve as a model for others.