A Review of: Adetayo, A. J. (2023). ChatGPT and librarians for reference consultations. Internet Reference Services Quarterly, 27(3), 131–147. https://doi.org/10.1080/10875301.2023.2203681 Objective – To investigate students’ use of ChatGPT and its potential advantages and disadvantages compared to reference librarians at a university library. Design – Survey research. Setting – A university library in Nigeria. Subjects – Students familiar with ChatGPT (n=54) who were enrolled in a library users’ education course. Methods – A survey was conducted in a sample of undergraduate students enrolled in a library users’ education course, who had previously used ChatGPT. Participants were asked questions based on six categories that reflected frequency of use, types of inquiries, frequency of reference consultations, desire to consult reference librarians despite the availability of ChatGPT, and potential advantages and disadvantages of ChatGPT compared to reference librarians. A 4-point Likert scale was used to measure the responses from often to never, strongly agree to strongly disagree, and rarely to frequently. Main Results – The sample of students who participated (n=54) were a diverse group whose age varied from below 20 (35.2%) to above 30 years (31.5%) and represented a variety of fields of study, such as engineering, business and social sciences, arts, law, sciences, basic and medical sciences. Regarding frequency of use, the author reported that 40.7% of participants occasionally used ChatGPT, and 26.1% and 16.7% used it frequently or very frequently, respectively. Of the five options that represented types of inquiries (religious, political, academic, entertainment, and work), academic and work-related inquiries were topics most often searched in ChatGPT. Participants indicated that they consulted reference librarians occasionally (40.8%), frequently (37%), or rarely (22.2%). Most students (87%) would continue to consult reference librarians despite the availability of ChatGPT. For questions that compared ChatGPT to reference librarians, four options were provided to describe potential advantages and four options were provided to describe potential disadvantages. Most students agreed or strongly agreed that ChatGPT is more user friendly (83.4%), that it includes a broad knowledge base (90.7%), is easily accessible (83.3%), and saves time by responding to questions quickly (98%) compared to reference librarians. Fewer than half of the students agreed or strongly agreed that ChatGPT’s knowledge base is not up to date (47.2%). Most agreed or strongly agreed that it cannot comprehend some questions (72.3%), that it cannot read emotions as a librarian would (74.1%), and that responses to questions may be incorrect (66.6%). The potential advantage with the strongest response score was that ChatGPT saves time by responding to questions quickly (mean 3.52). The potential disadvantage with the strongest response score was ChatGPT could not read emotion
{"title":"Students’ Perspective of the Advantages and Disadvantages of ChatGPT Compared to Reference Librarians","authors":"Mary-Kathleen Grams","doi":"10.18438/eblip30518","DOIUrl":"https://doi.org/10.18438/eblip30518","url":null,"abstract":"A Review of:\u0000Adetayo, A. J. (2023). ChatGPT and librarians for reference consultations. Internet Reference Services Quarterly, 27(3), 131–147. https://doi.org/10.1080/10875301.2023.2203681\u0000Objective – To investigate students’ use of ChatGPT and its potential advantages and disadvantages compared to reference librarians at a university library.\u0000Design – Survey research.\u0000Setting – A university library in Nigeria.\u0000Subjects – Students familiar with ChatGPT (n=54) who were enrolled in a library users’ education course.\u0000Methods – A survey was conducted in a sample of undergraduate students enrolled in a library users’ education course, who had previously used ChatGPT. Participants were asked questions based on six categories that reflected frequency of use, types of inquiries, frequency of reference consultations, desire to consult reference librarians despite the availability of ChatGPT, and potential advantages and disadvantages of ChatGPT compared to reference librarians. A 4-point Likert scale was used to measure the responses from often to never, strongly agree to strongly disagree, and rarely to frequently.\u0000Main Results – The sample of students who participated (n=54) were a diverse group whose age varied from below 20 (35.2%) to above 30 years (31.5%) and represented a variety of fields of study, such as engineering, business and social sciences, arts, law, sciences, basic and medical sciences. Regarding frequency of use, the author reported that 40.7% of participants occasionally used ChatGPT, and 26.1% and 16.7% used it frequently or very frequently, respectively. Of the five options that represented types of inquiries (religious, political, academic, entertainment, and work), academic and work-related inquiries were topics most often searched in ChatGPT. Participants indicated that they consulted reference librarians occasionally (40.8%), frequently (37%), or rarely (22.2%). Most students (87%) would continue to consult reference librarians despite the availability of ChatGPT. For questions that compared ChatGPT to reference librarians, four options were provided to describe potential advantages and four options were provided to describe potential disadvantages. Most students agreed or strongly agreed that ChatGPT is more user friendly (83.4%), that it includes a broad knowledge base (90.7%), is easily accessible (83.3%), and saves time by responding to questions quickly (98%) compared to reference librarians. Fewer than half of the students agreed or strongly agreed that ChatGPT’s knowledge base is not up to date (47.2%). Most agreed or strongly agreed that it cannot comprehend some questions (72.3%), that it cannot read emotions as a librarian would (74.1%), and that responses to questions may be incorrect (66.6%). The potential advantage with the strongest response score was that ChatGPT saves time by responding to questions quickly (mean 3.52). The potential disadvantage with the strongest response score was ChatGPT could not read emotion","PeriodicalId":45227,"journal":{"name":"Evidence Based Library and Information Practice","volume":null,"pages":null},"PeriodicalIF":0.8,"publicationDate":"2024-06-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141338172","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
A Review of: Subaveerapandiyan, A., Sunanthini, C., & Amees, M. (2023). A study on the knowledge and perception of artificial intelligence. IFLA Journal, 49(3), 503–513. https://doi.org/10.1177/03400352231180230 Objective – To assess the knowledge, perception, and skills of library and information science (LIS) professionals related to artificial intelligence (AI). Design – 45 statements were distributed to 469 LIS professionals via Google Forms to collect primary data. 245 participants responded to the structured questionnaire. Setting – University and college libraries in Zambia. Subjects – Zambian library and information science professionals.Methods – A descriptive approach was employed for the study. Data was gathered via a questionnaire. “The objective was to assess the statistical relationship between the knowledge, perception, and skills of LIS professionals (the independent variables) and AI (the dependent variable)” (Subaveerapandiyan et al., p. 506). The survey used a 5-point Likert scale with (1) strongly disagree being the lowest score and (5) strongly agree the highest. Means and standard deviations are included in data display tables. Thematic analysis was employed to analyze the data. SPSS was used for data analysis.Main Results – Survey results are presented in three tables. Table 1, “Awareness of AI among LIS professionals,” contains 21 statements related to AI use in various library environments and services, including reference (finding articles and citations, content summarization, detecting misinformation), circulation of library materials, security and surveillance, character recognition and document preservation, research data management, language translation, and others. The authors note that 44.1 percent of the respondents agreed that “AI is essential for the effectiveness and efficiency of library service delivery, enabling libraries to enhance and offer dynamic services for their users” (Subaveerapandiyan et al., 2023, p. 506). Table 2, “Perception of AI among LIS professionals,” contains 10 statements. Over 85 percent of respondents either strongly agreed or agreed that AI “makes library staff lazy” while 58.1 percent either strongly agreed or agreed that AI is a “threat to librarians’ employment” (Subaveerapandiyan et al., 2023, p. 506). The authors note that the “respondents also indicated barriers to the adoption of AI in libraries, such as the lack of LIS professionals’ skills and budgetary constraints” (Subaveerapandiyan et al., 2023, p. 506). Table 3 lists 13 competencies required by library professionals in the AI era. The majority of the respondents (an average of 65 percent) were in strong agreement that “electronic communication, hardware and software, Internet applications, computing and networking, cyber security and network management, data quality control, data curation, database management … are necessary competencies required by LIS professionals for them to be proficient in AI” (Subaveerapandiyan et
{"title":"A Study on the Knowledge and Perception of Artificial Intelligence","authors":"David M. Dettman","doi":"10.18438/eblip30436","DOIUrl":"https://doi.org/10.18438/eblip30436","url":null,"abstract":"A Review of:\u0000Subaveerapandiyan, A., Sunanthini, C., & Amees, M. (2023). A study on the knowledge and perception of artificial intelligence. IFLA Journal, 49(3), 503–513. https://doi.org/10.1177/03400352231180230\u0000Objective – To assess the knowledge, perception, and skills of library and information science (LIS) professionals related to artificial intelligence (AI).\u0000Design – 45 statements were distributed to 469 LIS professionals via Google Forms to collect primary data. 245 participants responded to the structured questionnaire.\u0000Setting – University and college libraries in Zambia.\u0000Subjects – Zambian library and information science professionals.Methods – A descriptive approach was employed for the study. Data was gathered via a questionnaire. “The objective was to assess the statistical relationship between the knowledge, perception, and skills of LIS professionals (the independent variables) and AI (the dependent variable)” (Subaveerapandiyan et al., p. 506). The survey used a 5-point Likert scale with (1) strongly disagree being the lowest score and (5) strongly agree the highest. Means and standard deviations are included in data display tables. Thematic analysis was employed to analyze the data. SPSS was used for data analysis.Main Results – Survey results are presented in three tables. Table 1, “Awareness of AI among LIS professionals,” contains 21 statements related to AI use in various library environments and services, including reference (finding articles and citations, content summarization, detecting misinformation), circulation of library materials, security and surveillance, character recognition and document preservation, research data management, language translation, and others. The authors note that 44.1 percent of the respondents agreed that “AI is essential for the effectiveness and efficiency of library service delivery, enabling libraries to enhance and offer dynamic services for their users” (Subaveerapandiyan et al., 2023, p. 506).\u0000Table 2, “Perception of AI among LIS professionals,” contains 10 statements. Over 85 percent of respondents either strongly agreed or agreed that AI “makes library staff lazy” while 58.1 percent either strongly agreed or agreed that AI is a “threat to librarians’ employment” (Subaveerapandiyan et al., 2023, p. 506). The authors note that the “respondents also indicated barriers to the adoption of AI in libraries, such as the lack of LIS professionals’ skills and budgetary constraints” (Subaveerapandiyan et al., 2023, p. 506).\u0000Table 3 lists 13 competencies required by library professionals in the AI era. The majority of the respondents (an average of 65 percent) were in strong agreement that “electronic communication, hardware and software, Internet applications, computing and networking, cyber security and network management, data quality control, data curation, database management … are necessary competencies required by LIS professionals for them to be proficient in AI” (Subaveerapandiyan et","PeriodicalId":45227,"journal":{"name":"Evidence Based Library and Information Practice","volume":null,"pages":null},"PeriodicalIF":0.8,"publicationDate":"2024-06-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141338404","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
A Review of: Wang, Y. (2022). Using machine learning and natural language processing to analyze library chat reference transcripts. Information Technology and Libraries, 41(3). https://doi.org/10.6017/ital.v41i3.14967 Objective – The study sought to develop a model to predict if library chat questions are reference or non-reference. Design – Supervised machine learning and natural language processing. Setting – College of New Jersey academic library. Subjects – 8,000 Springshare LibChat transactions collected from 2014 to 2021. Methods – The chat logs were downloaded into Excel, cleaned, and individual questions were labelled reference or non-reference by hand. Labelled data were preprocessed to remove nonmeaningful and stop words, and reformatted to lowercase. Data were then stemmed to group words with similar meaning. The feature of question length was then added and data were transformed from text to numeric for text vectorization. Data were then divided into training and testing sets. The Python packages Natural Language Toolkit (NLTK) and scikit-learn were used for analysis, building random forest and gradient boosting models which were evaluated via confusion matrix. Main Results – Both models performed very well in precision, recall and accuracy, with the random forest model having better overall results than the gradient boosting model, as well as a more efficient fit time, though slightly longer prediction time. Conclusion – High volume library chat services could benefit from utilizing machine learning to develop models that inform plugins or chat enhancements to filter chat queries quickly.
{"title":"Machine Learning Offers Opportunities to Advance Library Services","authors":"Samantha Kaplan","doi":"10.18438/eblip30527","DOIUrl":"https://doi.org/10.18438/eblip30527","url":null,"abstract":"A Review of:\u0000Wang, Y. (2022). Using machine learning and natural language processing to analyze library chat reference transcripts. Information Technology and Libraries, 41(3). https://doi.org/10.6017/ital.v41i3.14967\u0000Objective – The study sought to develop a model to predict if library chat questions are reference or non-reference.\u0000Design – Supervised machine learning and natural language processing.\u0000Setting – College of New Jersey academic library.\u0000Subjects – 8,000 Springshare LibChat transactions collected from 2014 to 2021.\u0000Methods – The chat logs were downloaded into Excel, cleaned, and individual questions were labelled reference or non-reference by hand. Labelled data were preprocessed to remove nonmeaningful and stop words, and reformatted to lowercase. Data were then stemmed to group words with similar meaning. The feature of question length was then added and data were transformed from text to numeric for text vectorization. Data were then divided into training and testing sets. The Python packages Natural Language Toolkit (NLTK) and scikit-learn were used for analysis, building random forest and gradient boosting models which were evaluated via confusion matrix.\u0000Main Results – Both models performed very well in precision, recall and accuracy, with the random forest model having better overall results than the gradient boosting model, as well as a more efficient fit time, though slightly longer prediction time.\u0000Conclusion – High volume library chat services could benefit from utilizing machine learning to develop models that inform plugins or chat enhancements to filter chat queries quickly.","PeriodicalId":45227,"journal":{"name":"Evidence Based Library and Information Practice","volume":null,"pages":null},"PeriodicalIF":0.8,"publicationDate":"2024-06-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141340981","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Call for Volunteers for EBLIP Journal: Peer Reviewers from Public Libraries","authors":"Editorial Team","doi":"10.18438/eblip30548","DOIUrl":"https://doi.org/10.18438/eblip30548","url":null,"abstract":"","PeriodicalId":45227,"journal":{"name":"Evidence Based Library and Information Practice","volume":null,"pages":null},"PeriodicalIF":0.8,"publicationDate":"2024-06-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141341847","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
A Review of: Moulaison-Sandy, H. (2023). What is a person? Emerging interpretations of AI authorship and attribution. Proceedings of the Association for Information Science & Technology, 60(1), 279–290. https://doi.org/10.1002/pra2.788 Objective – To examine how and which academic libraries are responding to emerging guidelines on citing ChatGPT in the American Psychological Association (APA) style through guidance published on the libraries’ websites. Design – Analysis of search results and webpage content. Setting – Websites of academic libraries in the United States. Subjects – Library webpages addressing how ChatGPT should be cited in APA format. Methods – Google search results for academic library webpages providing guidance on citing ChatGPT in APA format were retrieved on a weekly basis using the query “chatgpt apa citation site:.edu” over a six-week period that covered the weeks before and immediately after the APA issued official guidance for citing ChatGPT. The first three pages of relevant search results were coded in MAXQDA and analyzed to determine the type of institution, using the Carnegie Classification and membership in the Association of American Universities (AAU). As this was a period during which APA style recommendations for citing ChatGPT were shifting, the accuracy of the library webpage content was also assessed and tracked across the studied time period. Main Results – During the six-week period, the number of library webpages with guidance for citing ChatGPT in APA format increased. Although doctoral universities accounted for the largest number of webpages each week, baccalaureate colleges, baccalaureate/associate’s colleges, and associates’ colleges were also well-represented in the search results. Institutions belonging to the AAU were represented by a relatively small number throughout the study. Over half of the pages made some mention of APA’s recommendations being interim or evolving, though the exact number fluctuated throughout the period. Prior to the collection period, APA had revised its initial recommendations to cite ChatGPT as a webpage or as personal communication, but 40% to 60% of library webpages continued to offer this outdated guidance. Of the library webpages, 13% to 40% provided verbatim guidance from ChatGPT responses on how it should be cited. The final two weeks of the collection period occurred after April 7, 2023, when APA had published official recommendations for citing ChatGPT. In the week following this change, none of the webpages in the first three pages of results had been updated to fully capture the new recommendations. The study analyzed the nine webpages appearing in the first page of results for the second week after APA’s official recommendations were published, showing that three linked to the APA’s blog, zero provided further explanation on how to apply the recommendations, five included outdated guidance, and three gave guidance from ChatGPT’s responses to questions on how i
回顾：Moulaison-Sandy, H. (2023).什么是人？人工智能作者身份和归属的新兴解释。信息科学与技术协会论文集》，60(1), 279-290。https://doi.org/10.1002/pra2.788Objective - 通过图书馆网站上发布的指南，研究学术图书馆如何以及哪些学术图书馆正在响应新出现的美国心理学会（APA）风格的ChatGPT引用指南。 设计--对搜索结果和网页内容进行分析。背景--美国学术图书馆的网站。研究对象--图书馆网页，内容涉及如何以 APA 格式引用 ChatGPT。方法--在 APA 发布 ChatGPT 引用官方指南之前和紧随其后的六周内，使用 "chatgpt apa citation site:.edu "查询，每周检索学术图书馆网页的谷歌搜索结果。相关搜索结果的前三页在 MAXQDA 中进行了编码，并通过卡内基分类法和美国大学协会 (AAU) 会员资格进行分析，以确定机构类型。主要结果 - 在为期六周的时间里，图书馆网页中有关以 APA 格式引用 ChatGPT 的指导数量有所增加。虽然每周博士生大学的网页数量最多，但学士学位学院、学士/副学士学位学院和副学士学位学院在搜索结果中也占有很大比例。在整个研究过程中，属于 AAU 的院校数量相对较少。半数以上的网页都提到了 APA 的建议是临时性的或不断变化的，但具体数字在整个研究期 间有所波动。在收集期之前，APA 已经修订了其最初的建议，将 ChatGPT 引用为网页或个人通信，但 40% 至 60% 的图书馆网页仍然提供这种过时的指导。在这些图书馆网页中，有 13% 到 40% 提供了 ChatGPT 答复中关于如何引用的逐字指导。收集期的最后两周是在 2023 年 4 月 7 日之后，APA 发布了引用 ChatGPT 的官方建议。在这一变化之后的一周内，前三页结果中的网页都没有更新，无法完全捕捉到新的建议。研究分析了 APA 官方建议发布后第二周出现在搜索结果第一页的九个网页，结果显示，三个网页链接到了 APA 的博客，零个网页提供了如何应用建议的进一步解释，五个网页包含了过时的指导，三个网页从 ChatGPT 对如何引用问题的答复中提供了指导。结论--作者认为研究结果反映了三个相互关联的因素：一项新技术、图书馆员在大语言模型（LLM）方面的知识差距以及目前在作者身份方面的讨论方式，以及谷歌无法以优先考虑正确信息的方式对结果进行排序。为本科生提供服务的机构数量众多，这使作者得出结论：这类人群最需要有关引用 ChatGPT 的指导，而图书馆员的响应表明他们了解这一需求，即使指导本身并不准确。
{"title":"Academic Libraries’ Citation Guides to ChatGPT Show Mixed Levels of Accuracy and Currency","authors":"Abbey Lewis","doi":"10.18438/eblip30514","DOIUrl":"https://doi.org/10.18438/eblip30514","url":null,"abstract":"A Review of:\u0000Moulaison-Sandy, H. (2023). What is a person? Emerging interpretations of AI authorship and attribution. Proceedings of the Association for Information Science & Technology, 60(1), 279–290. https://doi.org/10.1002/pra2.788\u0000Objective – To examine how and which academic libraries are responding to emerging guidelines on citing ChatGPT in the American Psychological Association (APA) style through guidance published on the libraries’ websites. \u0000Design – Analysis of search results and webpage content.\u0000Setting – Websites of academic libraries in the United States.\u0000Subjects – Library webpages addressing how ChatGPT should be cited in APA format.\u0000Methods – Google search results for academic library webpages providing guidance on citing ChatGPT in APA format were retrieved on a weekly basis using the query “chatgpt apa citation site:.edu” over a six-week period that covered the weeks before and immediately after the APA issued official guidance for citing ChatGPT. The first three pages of relevant search results were coded in MAXQDA and analyzed to determine the type of institution, using the Carnegie Classification and membership in the Association of American Universities (AAU). As this was a period during which APA style recommendations for citing ChatGPT were shifting, the accuracy of the library webpage content was also assessed and tracked across the studied time period.\u0000Main Results – During the six-week period, the number of library webpages with guidance for citing ChatGPT in APA format increased. Although doctoral universities accounted for the largest number of webpages each week, baccalaureate colleges, baccalaureate/associate’s colleges, and associates’ colleges were also well-represented in the search results. Institutions belonging to the AAU were represented by a relatively small number throughout the study. Over half of the pages made some mention of APA’s recommendations being interim or evolving, though the exact number fluctuated throughout the period. Prior to the collection period, APA had revised its initial recommendations to cite ChatGPT as a webpage or as personal communication, but 40% to 60% of library webpages continued to offer this outdated guidance. Of the library webpages, 13% to 40% provided verbatim guidance from ChatGPT responses on how it should be cited. The final two weeks of the collection period occurred after April 7, 2023, when APA had published official recommendations for citing ChatGPT. In the week following this change, none of the webpages in the first three pages of results had been updated to fully capture the new recommendations. The study analyzed the nine webpages appearing in the first page of results for the second week after APA’s official recommendations were published, showing that three linked to the APA’s blog, zero provided further explanation on how to apply the recommendations, five included outdated guidance, and three gave guidance from ChatGPT’s responses to questions on how i","PeriodicalId":45227,"journal":{"name":"Evidence Based Library and Information Practice","volume":null,"pages":null},"PeriodicalIF":0.8,"publicationDate":"2024-06-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141344501","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
A Review of: Xiao, J., & Gao, W. (2020). Connecting the dots: reader ratings, bibliographic data, and machine-learning algorithms for monograph selection. The Serials Librarian, 78(1-4), 117-122. https://doi.org/10.1080/0361526X.2020.1707599 Objective – To illustrate how machine-learning book recommender systems can help librarians make collection development decisions. Design – Data analysis of publicly available book sales rankings and reader ratings. Setting – The internet. Subjects – 192 New York Times hardcover fiction best seller titles from 2018, and 1,367 Goodreads ratings posted in 2018. Methods – Data were collected using Application Programming Interfaces. The researchers retrieved weekly hardcover fiction best seller rankings published by the New York Times in 2018 in CSV file format. All 52 files, each containing bibliographic data for 15 hardcover fiction titles, were combined and duplicate titles removed, resulting in 192 unique best seller titles. The researchers retrieved reader ratings of the 192 best seller titles from Goodreads. The ratings were limited to those posted in 2018 by the top Goodreads reviewers. A Bayes estimator produced a list of the top ten highest rated New York Times best sellers. The researchers built the recommender system using Python and employed several content-based and collaborative filtering recommender techniques (e.g., cosine similarity, term frequency-inverse document frequency, and matrix factorization algorithms) to identify novels similar to the highest rated best sellers. Main Results – Each recommender technique generated a different list of novels. Conclusion – The main finding from this study is that recommender systems can simplify collection development for librarians and facilitate greater access to relevant library materials for users. Academic libraries can use the same recommender techniques employed in the study to identify titles similar to highly circulated monographs or frequently requested interlibrary loans. There are several limitations to using recommender systems in libraries, including privacy concerns when analyzing user behaviour data and potential biases in machine-learning algorithms.
{"title":"Machine-learning Recommender Systems Can Inform Collection Development Decisions","authors":"Kristy Hancock","doi":"10.18438/eblip30521","DOIUrl":"https://doi.org/10.18438/eblip30521","url":null,"abstract":"A Review of:\u0000Xiao, J., & Gao, W. (2020). Connecting the dots: reader ratings, bibliographic data, and machine-learning algorithms for monograph selection. The Serials Librarian, 78(1-4), 117-122. https://doi.org/10.1080/0361526X.2020.1707599\u0000Objective – To illustrate how machine-learning book recommender systems can help librarians make collection development decisions.\u0000Design – Data analysis of publicly available book sales rankings and reader ratings.\u0000Setting – The internet.\u0000Subjects – 192 New York Times hardcover fiction best seller titles from 2018, and 1,367 Goodreads ratings posted in 2018.\u0000Methods – Data were collected using Application Programming Interfaces. The researchers retrieved weekly hardcover fiction best seller rankings published by the New York Times in 2018 in CSV file format. All 52 files, each containing bibliographic data for 15 hardcover fiction titles, were combined and duplicate titles removed, resulting in 192 unique best seller titles. The researchers retrieved reader ratings of the 192 best seller titles from Goodreads. The ratings were limited to those posted in 2018 by the top Goodreads reviewers.\u0000A Bayes estimator produced a list of the top ten highest rated New York Times best sellers. The researchers built the recommender system using Python and employed several content-based and collaborative filtering recommender techniques (e.g., cosine similarity, term frequency-inverse document frequency, and matrix factorization algorithms) to identify novels similar to the highest rated best sellers.\u0000Main Results – Each recommender technique generated a different list of novels.\u0000Conclusion – The main finding from this study is that recommender systems can simplify collection development for librarians and facilitate greater access to relevant library materials for users. Academic libraries can use the same recommender techniques employed in the study to identify titles similar to highly circulated monographs or frequently requested interlibrary loans. There are several limitations to using recommender systems in libraries, including privacy concerns when analyzing user behaviour data and potential biases in machine-learning algorithms.","PeriodicalId":45227,"journal":{"name":"Evidence Based Library and Information Practice","volume":null,"pages":null},"PeriodicalIF":0.8,"publicationDate":"2024-06-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141343318","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Objective – This project sought to build upon a reader development tool, Many Roads to Wellbeing, developed by a health librarian in a mental health NHS Trust in Birmingham, England, by piloting reading group sessions in the main public library in the city using wellbeing-themed stories and poems. The aim was to establish whether a “wellbeing through reading” program can help reading group participants to experience key facets of wellbeing as defined by the Five Ways to Wellbeing. Methods – The program developers ran 15 monthly sessions at the Library of Birmingham. These were advertised using the Meetup social media tool to reach a wider client base than existing library users; members of the public who had self-prescribed to the group and were actively seeking wellbeing. A health librarian selected wellbeing-themed short stories and poems and facilitated read aloud sessions. The Library of Birmingham provided facilities and a member of staff to help support each session. Results – A total of 131 participants attended the 15 sessions that were hosted. There was a 95% response rate to the questionnaire survey. Of the respondents, 91% felt that sessions had helped them to engage with all of the Five Ways to Wellbeing. The three elements of Five Ways to Wellbeing that participants particularly engaged with were Connect (n=125), Take Notice (n=123), and Keep Learning (n=124). Conclusion – The reading program proved to be successful in helping participants to experience multiple dimensions of wellbeing. This project presents a new way of evaluating a bibliotherapy scheme for impact on wellbeing, as well as being an example of effective partnership working between the healthcare sector and a public library.
{"title":"“Wellbeing Through Reading”: The Impact of a Public Library and Healthcare Library Partnership Initiative in England","authors":"Anita Phul, Hélène Gorring, David Stokes","doi":"10.18438/eblip30475","DOIUrl":"https://doi.org/10.18438/eblip30475","url":null,"abstract":"Objective – This project sought to build upon a reader development tool, Many Roads to Wellbeing, developed by a health librarian in a mental health NHS Trust in Birmingham, England, by piloting reading group sessions in the main public library in the city using wellbeing-themed stories and poems. The aim was to establish whether a “wellbeing through reading” program can help reading group participants to experience key facets of wellbeing as defined by the Five Ways to Wellbeing. \u0000Methods – The program developers ran 15 monthly sessions at the Library of Birmingham. These were advertised using the Meetup social media tool to reach a wider client base than existing library users; members of the public who had self-prescribed to the group and were actively seeking wellbeing. A health librarian selected wellbeing-themed short stories and poems and facilitated read aloud sessions. The Library of Birmingham provided facilities and a member of staff to help support each session. \u0000Results – A total of 131 participants attended the 15 sessions that were hosted. There was a 95% response rate to the questionnaire survey. Of the respondents, 91% felt that sessions had helped them to engage with all of the Five Ways to Wellbeing. The three elements of Five Ways to Wellbeing that participants particularly engaged with were Connect (n=125), Take Notice (n=123), and Keep Learning (n=124). \u0000Conclusion – The reading program proved to be successful in helping participants to experience multiple dimensions of wellbeing. This project presents a new way of evaluating a bibliotherapy scheme for impact on wellbeing, as well as being an example of effective partnership working between the healthcare sector and a public library. ","PeriodicalId":45227,"journal":{"name":"Evidence Based Library and Information Practice","volume":null,"pages":null},"PeriodicalIF":0.8,"publicationDate":"2024-06-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141344183","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Evidence Summary Theme: All Things Virtual","authors":"Heather MacDonald","doi":"10.18438/eblip30483","DOIUrl":"https://doi.org/10.18438/eblip30483","url":null,"abstract":"","PeriodicalId":45227,"journal":{"name":"Evidence Based Library and Information Practice","volume":null,"pages":null},"PeriodicalIF":0.8,"publicationDate":"2023-12-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"138970724","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Objective - To compare Portland State University’s (PSU) local experience of using streaming media to national and international trends identified in a large qualitative study by Ithaka S+R. This comparison will help librarians better understand if the PSU Library is meeting the needs of faculty with its streaming media collection through a series of faculty interviews. Methods and Intervention - Two librarians from PSU participated in a large, collaborative, two-part study conducted by Ithaka S+R in 2022, with 23 other academic institutions in the United States, Canada, and Germany As part of this study, the authors conducted a series of interviews with faculty from PSU’s Social Work and Film Studies departments to gather qualitative data about their use, expectations, and priorities relating to streaming media in their teaching. Ithaka S+R provided guided interview questions, and librarians at PSU conducted interviews with departmental faculty. Local interview responses were compared to the interviews from the other 23 institutions. Results - PSU Library had a higher rate of faculty satisfaction than in the larger survey. Discussions raised concerns around accessibility of content, which was novel to PSU, and did not meaningfully emerge in the broader study. Local findings did line up with broader trends in the form of concerns about cost, discoverability, and lack of diverse content. Conclusions - The data collected by Ithaka S+R’s survey, which was the first part of their two-part study, is useful as it highlights the trends and attitudes of the greater academic library community. However, the second portion of the study’s guided interviews with campus faculty reinforced the importance of accessibility, the Library’s provision of resources, and the relationships between subject liaisons and departmental instructors. It emphasized that Portland State University’s Library has built a good foundation with faculty related to this area but has not been able to provide for every streaming instructional need. Reasons for this include limited acquisitions budgets, constraints of staff time, and market factors.
{"title":"Swimming Upstream in the Academic Library: Exploring Faculty Needs for Library Streaming Media Collections","authors":"Elsa Loftis, Carly Lamphere","doi":"10.18438/eblip30317","DOIUrl":"https://doi.org/10.18438/eblip30317","url":null,"abstract":"Objective - To compare Portland State University’s (PSU) local experience of using streaming media to national and international trends identified in a large qualitative study by Ithaka S+R. This comparison will help librarians better understand if the PSU Library is meeting the needs of faculty with its streaming media collection through a series of faculty interviews.\u0000Methods and Intervention - Two librarians from PSU participated in a large, collaborative, two-part study conducted by Ithaka S+R in 2022, with 23 other academic institutions in the United States, Canada, and Germany As part of this study, the authors conducted a series of interviews with faculty from PSU’s Social Work and Film Studies departments to gather qualitative data about their use, expectations, and priorities relating to streaming media in their teaching. Ithaka S+R provided guided interview questions, and librarians at PSU conducted interviews with departmental faculty. Local interview responses were compared to the interviews from the other 23 institutions.\u0000Results - PSU Library had a higher rate of faculty satisfaction than in the larger survey. Discussions raised concerns around accessibility of content, which was novel to PSU, and did not meaningfully emerge in the broader study. Local findings did line up with broader trends in the form of concerns about cost, discoverability, and lack of diverse content. \u0000Conclusions - The data collected by Ithaka S+R’s survey, which was the first part of their two-part study, is useful as it highlights the trends and attitudes of the greater academic library community. However, the second portion of the study’s guided interviews with campus faculty reinforced the importance of accessibility, the Library’s provision of resources, and the relationships between subject liaisons and departmental instructors. It emphasized that Portland State University’s Library has built a good foundation with faculty related to this area but has not been able to provide for every streaming instructional need. Reasons for this include limited acquisitions budgets, constraints of staff time, and market factors.","PeriodicalId":45227,"journal":{"name":"Evidence Based Library and Information Practice","volume":null,"pages":null},"PeriodicalIF":0.8,"publicationDate":"2023-12-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"138997461","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Objective – The goal of this study was to better understand to what extent digital repositories at academic libraries are active in promoting the collection of non-traditional research outputs. To achieve this goal, the researcher examined the digital repositories of universities in the United Kingdom who are signatories of the Declaration on Research Assessment (DORA), which recommends broadening the range of research outputs included in assessment exercises. Methods – The researcher developed a list of 77 universities in the UK who are signatories to DORA and have institutional repositories. Using this list, the researcher consulted the public websites of these institutions using a structured protocol and collected data to 1) characterize the types of outputs collected by research repositories at DORA-signatory institutions and their ability to provide measures of potential impact, and 2) assess whether university library websites promote repositories as a venue for hosting non-traditional research outputs. Finally, the researcher surveyed repository managers to understand the nature of their involvement with supporting the aims of DORA on their campuses. Results – The analysis found that almost all (96%) of the 77 repositories reviewed contained a variety of non-traditional research outputs, although the proportion of these outputs was small compared to traditional outputs. Of these 77 repositories, 82% featured usage metrics of some kind. Most (67%) of the same repositories, however, were not minting persistent identifiers for items. Of the universities in this sample, 53% also maintained a standalone data repository. Of these data repositories, 90% featured persistent identifiers, and all of them featured metrics of some kind. In a review of university library websites promoting the use of repositories, 47% of websites mentioned non-traditional research outputs. In response to survey questions, repository managers reported that the library and the unit responsible for the repository were involved in implementing DORA, and managers perceived it to be influential on their campus. Conclusion – Repositories in this sample are relatively well positioned to support the collection and promotion of non-traditional research outputs. However, despite this positioning, and repository managers’ belief that realizing the goals of DORA is important, most libraries in this sample do not appear to be actively collecting non-traditional outputs, although they are active in other areas to promote research assessment reform.
目的--本研究的目的是更好地了解学术图书馆的数字资源库在多大程度上积极促进了非传统研究成果的收集。为了实现这一目标，研究人员考察了英国签署了《研究评估宣言》（DORA）的大学的数字资源库，该宣言建议扩大评估工作中研究成果的范围。方法 - 研究人员编制了一份英国 77 所签署了 DORA 并拥有机构资源库的大学名单。利用这份名单，研究人员采用结构化协议咨询了这些机构的公共网站，并收集了以下数据：1）描述 DORA 签约机构的研究资料库所收集的产出类型及其提供潜在影响力衡量标准的能力；2）评估大学图书馆网站是否将资料库宣传为托管非传统研究产出的场所。最后，研究人员对资料库管理人员进行了调查，以了解他们参与支持其校园内 DORA 目标的性质。结果 - 分析发现，在所审查的 77 个资料库中，几乎所有资料库（96%）都包含各种非传统研究成果，尽管与传统成果相比，这些成果所占比例较小。在这 77 个资料库中，82% 的资料库采用了某种使用指标。不过，这些资源库中的大多数（67%）都没有为项目创建持久性标识符。在样本中，53% 的大学还拥有独立的数据存储库。在这些数据储存库中，90%都有持久性标识符，而且所有储存库都有某种指标。在对推广使用资源库的大学图书馆网站的审查中，47% 的网站提到了非传统研究成果。在回答调查问题时，资源库管理人员表示，图书馆和负责资源库的单位参与了 DORA 的实施，管理人员认为 DORA 在他们的校园中很有影响力。然而，尽管有这样的定位，而且资源库管理者认为实现 DORA 的目标非常重要，但样本中的大多数图书馆似乎并没有积极收集非传统成果，尽管他们在其他领域积极推动研究评估改革。
{"title":"Research Assessment Reform, Non-Traditional Research Outputs, and Digital Repositories: An Analysis of the Declaration on Research Assessment (DORA) Signatories in the United Kingdom","authors":"C. Hurrell","doi":"10.18438/eblip30407","DOIUrl":"https://doi.org/10.18438/eblip30407","url":null,"abstract":"Objective – The goal of this study was to better understand to what extent digital repositories at academic libraries are active in promoting the collection of non-traditional research outputs. To achieve this goal, the researcher examined the digital repositories of universities in the United Kingdom who are signatories of the Declaration on Research Assessment (DORA), which recommends broadening the range of research outputs included in assessment exercises.\u0000Methods – The researcher developed a list of 77 universities in the UK who are signatories to DORA and have institutional repositories. Using this list, the researcher consulted the public websites of these institutions using a structured protocol and collected data to 1) characterize the types of outputs collected by research repositories at DORA-signatory institutions and their ability to provide measures of potential impact, and 2) assess whether university library websites promote repositories as a venue for hosting non-traditional research outputs. Finally, the researcher surveyed repository managers to understand the nature of their involvement with supporting the aims of DORA on their campuses.\u0000Results – The analysis found that almost all (96%) of the 77 repositories reviewed contained a variety of non-traditional research outputs, although the proportion of these outputs was small compared to traditional outputs. Of these 77 repositories, 82% featured usage metrics of some kind. Most (67%) of the same repositories, however, were not minting persistent identifiers for items. Of the universities in this sample, 53% also maintained a standalone data repository. Of these data repositories, 90% featured persistent identifiers, and all of them featured metrics of some kind. In a review of university library websites promoting the use of repositories, 47% of websites mentioned non-traditional research outputs. In response to survey questions, repository managers reported that the library and the unit responsible for the repository were involved in implementing DORA, and managers perceived it to be influential on their campus.\u0000Conclusion – Repositories in this sample are relatively well positioned to support the collection and promotion of non-traditional research outputs. However, despite this positioning, and repository managers’ belief that realizing the goals of DORA is important, most libraries in this sample do not appear to be actively collecting non-traditional outputs, although they are active in other areas to promote research assessment reform.","PeriodicalId":45227,"journal":{"name":"Evidence Based Library and Information Practice","volume":null,"pages":null},"PeriodicalIF":0.8,"publicationDate":"2023-12-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"138997745","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：