Digital Scholar Bytes: Evaluating Generative AI Research Platforms

The emergence of ChatGPT and other AI technologies has significantly transformed the landscape of library research, introducing complexities alongside innovative solutions. In our February 9th workshop, A Review of Generative AI Tools for Research, we explored the following platforms that can enhance your research mastery: Perplexity, Elicit, Consensus, Lateral, ScholarAI, Keenious, OpenAlex, Semantic Scholar, Research Rabbit, and Scite. These tools are engineered to streamline your research workflow, making the process more efficient and effective.

 

Advantages

These research platforms can help the researcher find relevant papers without having to worry about a perfect keyword match or using Boolean operators. They offer summaries of research output from the topic specific to the question, and extract key information from the papers for further perusal and evaluation. These platforms can discover hidden relationships among papers since they examine the source more deeply than select fields in a library database. Last, they enhance your research workflow with such benefits as research and discovery, note-taking and organization, summarizing papers, performing meta-analyses, and writing assistance.

Limitations

At this stage, these platforms do have limitations. There are free and subscription-based models and you do get more with a subscription. Depending on your research demands, you may only need the free version. Research Rabbit, OpenAlex, and Semantic Scholar are free to use. Second, while the documents retrieved are sourced predominately from Semantic Scholar, hallucinations can creep into the process so you should always check your results. Last, there are gaps in what is not covered–subscription journals and paywalls–so using these platforms with your library databases will ensure that your are not missing relevant material.

Assistance

If you would like assistance with using these products, contact Kevin Gunn, Coordinator of Digital Scholarship, or Charles Gallagher, Research and Instruction Librarian.

Platforms Reviewed

Scite

Perplexity

Elicit

Consensus

Lateral

ScholarAI

Keenious

OpenAlex

Research Rabbit

Semantic Scholar

 

Digital Scholar Bytes: My Kind of Data!

Love Data Week 2024, scheduled for February 12-16, presents a unique opportunity for scholars, students, librarians, and data enthusiasts to delve into the multifaceted world of data. This year’s theme, “My Kind of Data,” underscores the personal connection we all have with data and its diverse applications. Understanding and managing data becomes not just useful but crucial for academic and professional growth. The international, yearly event is sponsored by the ICPSR (Inter-university Consortium for Political and Social Research).

The Importance of Data Literacy

With data misinterpretation and misinformation rampant, Love Data Week promotes data literacy and sound research data management strategies. The week-long program offers webinars and workshops aimed at enhancing understanding and skills in handling data. These events are not only urgent but essential in building a community that values and accurately interprets data. See the list of events that ICPSR is hosting.

Unique Celebrations: Adopt a Dataset

One aspect of Love Data Week is the “Adopt a Dataset” program. Much like adopting a pet, participants can choose a dataset from a diverse list, ranging from health studies to social surveys, and learn about the data it contains. This novel approach not only fosters a deeper understanding of specific datasets but also encourages participants to share their findings, thus contributing to a broader discourse on data utilization. To adopt a dataset:

Join the Global Conversation

Love Data Week is more than a series of events; it’s a global conversation about the role of data in our lives. By participating, you’re not only gaining valuable knowledge and skills but also joining a community of like-minded individuals who understand the power of data. Whether you’re attending a webinar, adopting a dataset, or simply engaging in the social media conversation using #LoveData24, you’re contributing to a movement that values and celebrates data in all its forms.

In conclusion, Love Data Week 2024 offers a platform for learning, sharing, and celebrating the diverse world of data. It’s an opportunity to enhance your data literacy, engage in creative activities, and contribute to a global discussion on the importance of data in shaping our future.

Catholic University Libraries

If you are interested in learning more, see our Digital Scholarship guide.

Additional Resources

U.S. Government (data.gov)

Google Dataset Search

The Open Data Handbook

Love Data Week- National Library of Medicine

Research & Instruction: American Fiction, 1774-1920

Embark on a literary journey through the epochs of American history with American Fiction, 1774–1920. This resource encompasses more than 17,800 works of prose fiction written by Americans from the political beginnings of the United States through World War I, including thousands never before available online. This landmark digital collection is based on authoritative bibliographies including thousands never before available online.

The researcher can explore the development of American literature through novels, short stories, romances, fictitious biographies, travel accounts, and sketches. Revealing the socioeconomic, political, and religious tenor of America through this time period allows students and researchers “to answer key questions about history, society, identity, psychology, race, gender, and culture.”

The collection comes from three major sources. First, nearly all the works found in Lyle H. Wright’s American Fiction: A Contribution Toward a Bibliography. This three-volume set consists of: American Fiction, 1774–1850; American Fiction, 1851–1875; and American Fiction, 1876–1900. The second major source is the titles published from 1901 to 1910 which were gathered from major American fiction collections across the United States. Third, titles published between 1911 and 1920 come from the William Charvat Collection of American Fiction at The Ohio State University Libraries.

Wright’s criteria was to include only first editions (or the earliest copy of a work) of novels and short story collections. He excluded such items as annuals, anthologies, and collections of anecdotes. The researcher should note that although Wright sought to include only fictitious works, some slave narratives as Solomon Northup’s Twelve Years a Slave (1853), Jermain Wesley Loguen’s The Rev. J. W. Loguen, as a Slave and as a Freeman, a Narrative of Real Life (1859), and Harriet Jacobs’s Incidents in the Life of a Slave Girl (1861), are include in the database.

Search Options

The researcher can search by the entire document, keyword, document title, author, publisher, place of publication, front matter, image caption, illustrator, and Gale document number. In addition to straight search inquiries, there are two search tools that allow the researcher to explore the collection. First, the Topic Finder allows the researcher to find new topics or keywords and discover new connections found in the top results. Second, the Term Frequency tool can explore the coverage of search terms graphed over time and illuminate trends in coverage of multiple topics.

The platform allows the researcher to see the search term present in a number of works including how often and exactly where in a particular work (see image). A great feature of the platform is the OCR confidence reading. The researcher can see the original text and how accurately it has been read by optical character recognition. A ‘dirty’ OCR document may need to be manually reviewed before further searching and text analysis can be performed.

Teaching with American Fiction, 1774-1920

One of the nice options of American Fiction, 1774-1920 is the ability to perform a deep analysis of a particular author. James Fenimore Cooper was the first major American author to garner an international audience extolling American frontier life. His early works–The Pioneers, The Last of the Mohicans, and The Prairie (published between 1823 and 1827)–delve into the hardships of the early pioneers in New York State and the encroachment on Native American life.  One can use the Topic Finder and Term Frequency research tools to explore the texts. Next, using a particular topic or term, one could contrast these early works with his later works–The Pathfinder (1840) and The Deerslayer (1841)–to see not only how Native life changed but also Cooper’s evolving social commentary.  For a deeper analysis of the texts, these works could be collected into a dataset for further analysis in the Gale Digital Scholar Lab. The image below represents a topic modeling search on Cooper’s works.

Text Analysis

Since American Fiction, 1774-1920 is part of the Gale Digital Scholar Lab, the researcher can do a textual analysis of the entire database or build a subset of documents. Once logged in (or having created an account) in the Lab, select the ‘Build’ option (top right hand corner of the screen). Select ‘view all limiters in Advanced Search’ option. A list of available databases is given (select American Fiction, 1774-1920). While the search interface for both the Gale Primary Sources and the Gale Digital Scholar Lab are the same, the Lab allows the researcher to customize one’s own unique dataset by pulling material from Gale Primary Sources or by importing external documents from other full text databases. In addition to building individualized datasets, the Lab has six text analysis tools for examining the content further:  Ngrams, Parts of Speech, Document Clustering, Named Entity Recognition, Topic Modeling, and Sentiment Analysis.

 

Digital Scholar Bytes: Data Privacy Week

Data privacy is the safeguarding of individuals’ personal information and the preservation of their right to control the collection, processing, storage, and sharing of their data. Numerous challenges surround this crucial concept: data breaches, identity theft, lack of consent, data profiling, inadequate security measures (e.g. weak encryption), misuse of collected data, insufficient regulatory frameworks and enforcement mechanisms, social engineering attacks, location tracking, and the integration of emerging technologies such as biometrics and artificial intelligence without proper safeguards further compound the challenges, emphasizing the need for robust measures to address these multifaceted issues.

The first step in combating these threats is user education and data literacy. Data Privacy Week (January 21-27, 2024) is a public awareness program created by the National Cybersecurity Alliance to meet this need. The Alliance is a non-profit organization whose mission is to ‘advocate for the safe use of all technology and educate everyone on how best to protect ourselves, our families, and our organizations from cybercrime.’ The Alliance creates partnerships between governments and corporations in answering this need.  This week, the Alliance is offering daily conversations called ‘Talking Data‘ with such topics as ‘Giving Data Brokers the Slip,’ ‘Protect Our Kids’ Privacy,’ and ‘Privacy on the AI Frontiers.’ Next week, the Alliance will be offering an Introduction to Privacy webinar on January 29th for entry-level privacy professionals and non-privacy professionals. Register here.

Helpful Resources

Note:

Catholic University of America is hosting a all day conference on Generative AI and National Security on January 31st. Register.

 



			

Digital Scholar Bytes: Celebrating World GIS Day: Unveiling the Power of Spatial Intelligence

GIS Day

In the realm of modern technology, where data is the new currency, Geographic Information Systems (GIS) stand as the unsung heroes orchestrating a symphony of spatial intelligence. November 15 marks World GIS Day, a celebration that shines a spotlight on the transformative impact of GIS in our interconnected world.

Originally conceived by Jack Dangermond, the founder and president of Esri, GIS Day is for people to explore GIS technology and to demonstrate how geospatial technology affects everyone. Real real-world applications of GIS are making a difference in business, government and society. At its essence, GIS are more than just maps and coordinates; it’s a dynamic system that integrates data with geographic locations, unraveling patterns, relationships, and insights that might otherwise remain hidden.

“GIS Day is a wonderful opportunity for professionals from around the world to get together to share their amazing work,” said Dangermond. “These individuals are using GIS technology to take a geographic approach in addressing unprecedented challenges, from COVID-19 to climate change. Organizations across the globe are taking part in grassroots events that help celebrate geospatial science and how it impacts the real world for good. Our users should be proud of their achievements in the field, which GIS Day is meant to showcase.”

Here are four examples of how GIS are improving our society.

Better Traffic Management

GIS are the architects of smart cities. It optimizes traffic flow, designs efficient public transportation systems, and aids in disaster preparedness. By visualizing data in a geographic context, city planners can make informed decisions that enhance the quality of urban life. This can include considerations for green spaces, zoning regulations, and infrastructure development.

Disease Mapping and Epidemiology

GIS enables the mapping of disease outbreaks, helping health professionals visualize the spread of infections. It assists in identifying high-risk areas, facilitating targeted interventions and resource allocation. The first case of using geography and spatial analysis to track disease was by Charles Picquet in 1832 in Paris, during the cholera outbreak.

Disaster Response and Management

When disasters strike, time is of the essence. GIS aid in rapid response and effective management of natural disasters. By overlaying real-time data with geographic information, emergency responders can streamline their efforts, saving lives and minimizing damage.

AI-powered GIS

The integration of AI with GIS revolutionizes spatial analysis and decision-making. AI enhances GIS capabilities through autonomous agents. These agents automate data interpretation, enable predictive modeling, and extract meaningful insights from vast geospatial datasets.

 

In conclusion, World GIS Day is not just a celebration of a technology, it is a celebration of the profound impact GIS has on our understanding of the world. From unraveling the mysteries of our environment to empowering individuals with the tools to create change, GIS are a testament to the transformative power of spatial intelligence. Let’s celebrate the achievements of GIS and look forward to the innovations that will shape our spatial future.

Further Reading

National Geographic Society, and ESRI. 2023. National Geographic and ESRI: Exploring with GIS. A documentary mini-series.
ESRI. What is GIS? 2023. ESRI.
SpatialPost. 2023. The Evolution Of GIS: From Mapping To Spatial Intelligence. SpatialPost. November 12.
Zhenlong Li & Huan Ning. 2023. “Autonomous GIS: the next-generation AI-powered GIS,” International Journal of Digital Earth, 16:2, 4668-4686.

Digital Scholar Bytes: World Digital Preservation Day: Safeguarding Our Digital Heritage

Happy World Digital Preservation Day 2023!

On November 2nd every year, the global digital preservation community comes together to celebrate World Digital Preservation Day. This annual event serves as a reminder of the critical importance of preserving our digital heritage and the challenges that come with it. This year’s theme: “Digital Preservation: A Concerted Effort.”

The History of World Digital Preservation Day

World Digital Preservation Day (WDPD) is a relatively new addition to the calendar of events in the digital preservation field. It was first established in 2010 by the Digital Preservation Coalition (DPC), a not-for-profit membership organization that works to secure our digital legacy. Professionals from a wide range of fields, including libraries, archives, museums, and research institutions participate.

The primary goal of WDPD is to create global awareness about the challenges of preserving digital content and to inspire actions that safeguard our digital heritage for future generations. This event provides a platform for experts and enthusiasts to share knowledge, experiences, and best practices in the field of digital preservation.

The Main Issues in Digital Preservation

Preservation of Digital Materials: The rapid pace of technological advancement results in the constant obsolescence of hardware and software. This makes it challenging to preserve and access digital materials, including documents, images, and multimedia content. WDPD highlights the need for active preservation strategies to ensure long-term access to digital assets.

Data Loss and Digital Obsolescence: Digital data is fragile and susceptible to loss. The risk of losing valuable information due to factors like hardware failure, media decay, and format obsolescence, is paramount. The importance of migration and emulation to combat obsolescence issues cannot be understated.

Digital Privacy and Ethical Considerations: Preserving digital content often involves handling sensitive information. Ethical considerations surrounding digital preservation, such as privacy concerns, consent, and data protection regulations like General Data Protection Regulation (GDPR), are complicated.

Access to Digital Heritage: One issue that WDPD addresses is making sure that digital heritage is accessible to future generations. Standards and best practices that support long-term access to digital content and research data is part of the discussion.

Research and Innovation: Preserving digital heritage effectively fosters research and innovation. WDPD underscores the potential for data reuse, computational analysis, and the development of new technologies in digital preservation.

The Future of Digital Preservation

As the digital world continues to evolve, the challenges of preserving digital heritage become more complex. The exponential growth of digital content, the rise of big data, and the development of new technologies all impact the way we approach digital preservation.

The future of digital preservation will be shaped by such technologies as blockchain, artificial intelligence, and decentralized storage offer promising solutions for long-term data preservation. These innovations will be crucial in addressing the challenges of data loss and obsolescence.

Interdisciplinary collaboration between diverse fields, such as computer science, data science, and information management, will lead to innovative preservation strategies that accommodate the evolving digital landscape. International collaboration in digital preservation will become increasingly essential. Cross-border partnerships and standards will help ensure that digital heritage remains accessible worldwide.

Digital Stewardship educational initiatives focusing on digital stewardship and preservation will become more prevalent. These programs will equip professionals with the knowledge and skills needed to manage and preserve digital content effectively. In 2010, the Library of Congress created the Digital Preservation OutReach and Education Network program for such professionals. WRLC sponsored the program in 2016 and three CUA Librarians attended.

World Digital Preservation Day is a vital event in the digital preservation community, serving as a reminder of the importance of safeguarding our digital heritage. It addresses critical issues related to digital preservation, inspires collaboration, and offers a glimpse into the future of this ever-evolving field.

Further reading

Adams, Caralee. 2022. In an Ever-Expanding Library, Using Decentralized Storage to Keep Your Materials Safe. Internet Archives Blog. March 11.

Digital Preservation Coalition. https://www.dpconline.org/

Digital Preservation Handbook. https://www.dpconline.org/handbook

Digital Preservation Coalition. The Global ‘Bit List’ of Endangered Digital Species. https://www.dpconline.org/digipres/champion-digital-preservation/bit-list

Jaillant, Lise, and Arran Rees. 2023. Applying AI to digital archives: trust, collaboration and shared professional ethics. Digital Scholarship in the Humanities 38 (2): 571-585.

Oldham, Krista M. 2020. Partners for Preservation: Advancing Digital Preservation through Cross-Community Collaboration. The American Archivist 83 (2): 487–490.

News and Events: WRLC, a Valuable Resource

 

For over 35 years, the Washington Research Library Consortium has taken decisive action to empower students and faculty in our member institutions. Through our collaborative partnership, WRLC consistently achieves substantive outcomes by:

  • Hosting a preservation-quality space for print materials
  • Actively promoting college affordability through open access books, journals, and textbooks
  • Ensuring a secure environment for digital content
  • Strengthening partner library staff through centralized training, support, and professional development initiatives.

The WRLC community consists of:

  • WRLC Staff: 18
  • Member Institutions’ Library Staff: 550+
  • Students Served: 110,000+
  • Faculty Supported: 9,000+

Digital Resources:

  • JSTOR E-books Provided: 88,000
  • Cost per Use: $0.60
  • DeGruyter E-books and Journals: 17,000

 

Collaborative investments from our partners within WRLC empower us to collectively achieve ambitious objectives on a scale that would be cost-prohibitive when pursued independently. Our shared vision underscores the principle that each financial contribution to WRLC not only yields substantial returns but also fortifies our community of trust. This interdependent relationship serves as a cornerstone for the advancement of pedagogy, scholarship, and research, profoundly benefiting our academic ecosystem.

 

 

 

Gale Digital Scholar Lab: Understanding Sentiment Analysis

Gale Digital Scholar Lab has six analysis tools built into it with Sentiment Analysis being one. Sentiment Analysis is the method of revealing trends or patterns in positive, negative, or neutral sentiment within a collected set of documents.  “Sentiment Analysis uses the AFINN Lexicon to assign a “sentiment score” to each document within a researcher’s content set, then provides two interactive visualizations of this data. In this webinar, Senior Digital Humanities Specialist Dr. Sarah Ketchley will show how to run the Sentiment Analysis tool, explain the recent upgrade to this tool, and discuss how Sentiment Analysis can be used in Digital Humanities projects to answer research questions.”

Date: Monday October 30, 3:00-4:00pm

Register here: https://cengage.zoom.us/webinar/register/8916952637130/WN_mvfHxuYaTna7cNrSwH97DQ

Instructor: Dr. Sarah Ketchley. Digital Humanist Scholar, Gale

Gale Literature Resource Center New Enhancements

Gale’s Literature Resource Center has added new enhancements and content that will help with your research needs. This webinar will be “a tour of the all-new browse functionality, 700+ portal pages focused on works of literature and literary topics, new primary and historical document sets, plus access to an additional 100+ original works of literature. The enhancements are aimed to empower students working with primary sources and to provide instructors with easy-to-find materials that inspire learning.”

When: Wednesday, October 25, 1:00 PM – 1:45 PM

Register here: https://cengage.zoom.us/webinar/register/2316920231155/WN_FvoHj-q4Seig3dm2qGYY1A

If you have general questions about Gale Literature Resource Center, please contact Kevin Gunn, liaison librarian to the Departments of English and Modern Languages and Literatures.

Peer Review Models and Misconceptions

As part of Peer Review Week (September 25-29, 2023), it is advantageous for us to explore the changing nature of peer review and the models that have sprung up over the years. For example, as the editor-in-chief for a Taylor & Francis academic journal, College & Undergraduate Libraries, I regularly return manuscripts back to the author as they do not seem to understand that anonymity is required in peer review (the journal uses double-blind peer review), the traditional standard for most academic journals. Double-blind peer review is where the author does not know the peer reviewers (and vice versa) and the peer reviewers do not know each other. This working in isolation allows the reviewer to offer an honest assessment without the taint of bias towards the author or his affiliation.

The Peer Review Process

The image from Wiley gives an overview of the peer review process. When the author submits their manuscript, the editor determines whether it is a good fit for the journal. This will include examining the content, relevancy, methodology, and timeliness of the manuscript. If the manuscript is rejected, the editor will explain why. If the editor returns the manuscript and asks for initial revisions, he is interested in sending the manuscript to reviewers. Once the revised manuscript is returned to the editor, he identifies relevant reviewers. The reviewers will usually have four weeks to review the manuscript. The comments can vary in depth and length, depending on the reviewer. In the end, the reviewer decides to accept ‘as is,’ accept with revisions, or reject outright. The editor will make the final call. Most manuscripts are ‘accept with revisions.’ This means the editor asks the author to make certain changes suggested by the reviewer. The purpose of this process is to strengthen the intellectual content of the manuscript.

 

Types of Peer Review

I mentioned double-blind peer review as the traditional standard of peer review but there are many others. Different peer-review models are given below and some publishers practice multiple models.

  • Open Review. Open review makes parts of the peer review process public, either before or after publication. Examples: ScienceOpen and PLOS.
  • Journal-independent peer review. This model allows authors to have their manuscripts reviewed before they submit to a specific journal. Example: PeerRef.
  • Co-review. Collaborate on reviewing a manuscript with another reviewer. IOP Publishing is currently using this process for their entire journal portfolio.
  • AI-assisted review. AI is being slowly rolled in a number of ways (although doubts still remain). (Kousha and Thelwall 2023).
  • Post-publication peer review. Once an article is published and available in an open repository, other scholars can come and review it. (Enago Academy).
  • Preprint publication. Manuscripts are shared with other scholars before initiating the peer review process. (ArXiv and eLife).

 

Common Misconceptions regarding Peer Review

In working with authors, I find that many have misconceptions about what constitutes a peer review. Here are some of the common ones.

Misconception: Peer Review is a One-Size-Fits-All Process

As seen above, peer review approaches vary widely, and there’s no one-size-fits-all method. One manuscript may need a different process of evaluation than another. Furthermore, different journals and disciplines can have varying models of peer review to suit their objectives and the nature of the research. For example, you may encounter single-blind review, where the reviewers remain anonymous, or double-blind review, where both reviewers and authors are anonymous to each other. Each review process protects the author and reviewers from potential bias. Howver, in some cases, double-blind review enhances anonymity, but may not be feasible, especially in cases where the author’s identity is obvious. On the flip side, some journals practice open review, where the identities of both reviewers and authors are disclosed, fostering transparency. This approach encourages constructive criticism while holding reviewers and authors accountable.

Misconception: Peer Reviewers are Infallible Experts

Peer reviewers are indeed experts in their fields, but they are not infallible. We all make errors in judgment, oversight, or interpretation, just like any other professionals. Furthermore, biases can inadvertently influence reviewers. Unconscious biases related to gender, ethnicity, or institutional affiliations can impact evaluations. To address this, many journals now focus on diversity in their reviewer pool. For my journal, the publisher has encouraged me to solicit reviewers from countries such as China and India. Editors play a crucial role in ensuring the reliability of peer review. They can act as intermediaries between authors and reviewers, assessing reviews for fairness and rigor. In some cases, they may seek additional opinions or clarification when conflicting reviews arise.

Misconception: Peer Review is a Fast Process

Peer review can be a time-consuming process. From the submission of a manuscript to its final publication, several months may pass, or even longer. This extended timeline can be frustrating for authors eager to share their findings. Factors including journal policies, reviewer availability, rounds of review, editorial process, and scope and complexity of the manuscript can impact the duration of the process. Preprint servers can help authors disseminate their work before the official peer review process and post-publication review can help in discussing research findings.

Misconception: Reviewers Primarily Look for Groundbreaking Findings

At my journal, we look for groundbreaking findings but we also accept manuscripts that contribute to an existent body of knowledge. The peer review process evaluates research based on several criteria, including validity, clarity, and methodology, not just the novelty of findings. Reviewers assess whether the research is conducted rigorously and whether the evidence supports the conclusions. Reviewers will assess the appropriateness of research methods, statistical analyses, and data collection techniques as well.

Misconception: Published Papers are the Gold Standard

While publication in a peer-reviewed journal is a significant achievement, it is not the only standard of research quality. Not all high-quality research gets published. The peer review process is selective, and even well-conducted studies may face rejection for various reasons, including limited journal space, a perception of limited novelty, or specific journal scope. Conversely, not all published work is of high quality. Mistakes, errors, or even ethical lapses can sometimes find their way into published papers. The peer review process helps to identify and mitigate these issues but is not infallible. Last, research impact, citations, and real-world applications also play a role in assessing the quality and influence of a paper. Some groundbreaking research may not gain immediate attention but becomes influential over time.

Enjoy the rest of Peer Review Week!

 

Further reading

Kousha, K. and Thelwall M. 2023. Artificial intelligence to support publishing and peer review: A summary and review. Learned Publishing. 2023.

Leung T.I., de Azevedo Cardoso T., Mavragani A., Eysenbach G. 2023. Best Practices for Using AI Tools as an Author, Peer Reviewer, or Editor. Journal of Medical Internet Research 2023.

Willis, Michael. 2022. How Can Technology Aid Research Integrity? Wiley blog. November 22.

Wulf, Karin et al. 2023. Ask the Chefs: What is the Single Most Pressing Issue for the Future of Peer Review? Scholarly Kitchen, September 22, 2023.