Warning: fopen(/home/virtual/kcse/journal/upload/ip_log/ip_log_2024-04.txt): failed to open stream: Permission denied in /home/virtual/lib/view_data.php on line 88 Warning: fwrite() expects parameter 1 to be resource, boolean given in /home/virtual/lib/view_data.php on line 89 Impact and perceived value of the revolutionary advent of artificial intelligence in research and publishing among researchers: a survey-based descriptive study
Skip Navigation
Skip to contents

Science Editing : Science Editing

OPEN ACCESS
SEARCH
Search

Articles

Page Path
HOME > Sci Ed > Volume 10(1); 2023 > Article
Original Article
Impact and perceived value of the revolutionary advent of artificial intelligence in research and publishing among researchers: a survey-based descriptive study
Riya Thomasorcid, Uttkarsha Bhosaleorcid, Kriti Shuklaorcid, Anupama Kapadiaorcid
Science Editing 2023;10(1):27-34.
DOI: https://doi.org/10.6087/kcse.294
Published online: February 16, 2023

Enago Academy, Enago, Mumbai, India

Correspondence to Riya Thomas academy@enago.com
• Received: January 21, 2023   • Accepted: February 3, 2023

Copyright © 2023 Korean Council of Science Editors

This is an open access article distributed under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

  • 3,335 Views
  • 336 Download
  • 1 Web of Science
  • 1 Crossref
  • 1 Scopus
  • Purpose
    This study was conducted to understand the perceptions and awareness of artificial intelligence (AI) in the academic publishing landscape.
  • Methods
    We conducted a global survey entitled “Role and impact of AI on the future of academic publishing” to understand the impact of the AI wave in the scholarly publishing domain. This English-language survey was open to all researchers, authors, editors, publishers, and other stakeholders in the scholarly community. Conducted between August and October 2021, the survey received responses from around 212 universities across 54 countries.
  • Results
    Out of 365 respondents, about 93% belonged to the age groups of 18–34 and 35–54 years. While 50% of the respondents selected plagiarism detection as the most widely known AI-based application, image recognition (42%), data analytics (40%), and language enhancement (39%) were some other known applications of AI. The respondents also expressed the opinion that the academic publishing landscape will significantly benefit from AI. However, the major challenges restraining the large-scale adoption of AI, as expressed by 93% of the respondents, were limited knowledge and expertise, as well as difficulties in integrating AI-based solutions into existing IT infrastructure.
  • Conclusion
    The survey responses reflected the necessity of AI in research and publishing. This study suggests possible ways to support a smooth transition. This can be best achieved by educating and creating awareness to ease possible fears and hesitation, and to actualize the promising benefits of AI.
Background/rationale
Artificial intelligence (AI) and machine learning have transformed several industries since their advent, facilitating easier and quicker automation of numerous processes. Likewise, AI-based technologies are being developed and implemented in the academic publishing industry to assist authors, editors, publishers, and stakeholders from allied industries. The deployment of AI in academic publishing has helped to tackle issues related to peer review, searching pertinent literature using scholarly databases, detecting plagiarism, identifying data fabrication, automated text analysis, content translation, content personalization, search engine optimization analysis, chatbots, and much more [1]. However, the views of authors and researchers who are at the epicenter of the publishing system are often not known. Although there are growing concerns over the potential misuse of AI in research and publishing, the drivers of the system must express their views to support and bring about changes that authors want to see in the publishing landscape [2]. Thus, it is imperative to examine and understand the perceived value and impact of AI on the future of academic publishing. Understanding authors’ and researchers’ viewpoints could provide stakeholders of science with an unambiguous idea of how to adopt AI to make the research and publishing system more efficient than it already is. Based on the data we gathered with this global survey, this report presents some thought-provoking trends that we identified.
Objectives
This study aimed to gain a better understanding of the perceptions and awareness of AI among researchers and other stakeholders of the scholarly community. Predominantly, we focused on identifying the adoption rate and popularity of AI-based tools amongst editors, publishers, and authors. Furthermore, the study also aimed to identify the perceived benefits of AI and concerns related to the use of AI in research and academic publishing.
Ethics statement
This study was based on a survey about journal publishing, the items of which included no sensitive personal information. No Institutional Review Board approval was required. The participants agreed to participate voluntarily in the survey.
Study design
This was a survey-based descriptive study.
Setting
We launched a global survey titled “Role and impact of AI on the future of academic publishing” [3] that was distributed online to collect the viewpoints of stakeholders in the scholarly community. The survey was conducted in the English language between August 27, 2021 and October 3, 2021, and was open to researchers, authors, journal editors, publishers, and other scholarly stakeholders through the SurveyMonkey tool.
Content validity test
To evaluate the content validity, a Likert scale was used to measure subject matter experts’ satisfaction regarding the clarity and coverage of the items of the questionnaire and to understand whether the individual items seemed relevant for determining the knowledge, perceptions, and attitudes of the respondents for the study. It was tested by four subject matter experts, two of whom are active professionals in editing and publication support services and the other two work in the development of natural language processing and AI-assisted tools. The subject matter experts agreed with the validity of these items for the survey on the role and impact of AI in the academic landscape. Table 1 shows that the questionnaire achieved high content validity according to the subject matter experts.
Participants and variables
After sending the survey questionnaire to email addresses listed in the Enago database and to internet users using social networks (Facebook and LinkedIn), 365 responses were collected. There was no exclusion criterion. Hence, the total target number could not be estimated. All items of the survey questionnaire were variables.
Data source/measurement
The survey covered several topics, such as researchers’ awareness of researchers about the applications of AI in scholarly publishing, whether AI has revolutionized the publishing domain, and the expected prospects of AI for advancing academic research and publishing. The survey questionnaire is available in Suppl. 1.
Bias and study size
There was no bias in selecting participants. Sample size estimation could not be performed due to the nature of this survey-based descriptive study. All responses were included in the analysis.
Statistical methods
Descriptive statistics were applied to observe some trends in the overall perceptions of AI in the scholarly world, which are presented in this article.
Characteristics of participants
The survey garnered 365 responses from researchers, authors, journal editors, publishers, and other stakeholders in the scholarly publishing industry at 212 universities across 54 countries. The collated data sample represented viewpoints from several countries and diverse academic roles. Data on sex, age, geographical distribution, the size of organizations/institutions, and respondents’ job titles are summarized in Table 2.
Awareness about applications of AI in academic publishing
Answers to the two questions related to the respondents’ knowledge about AI app ing importance in most industries, it is not surprising that about 86% of the respondents (Fig. 1) in this survey had a sufficient working knowledge of AI or at least a basic understanding of AI and its concepts. Around 13% had heard these concepts but did not understand them very well. A small portion (1%) of respondents chose other responses, and a few of them also specified that although they had a the oretical understanding of the concept, they had not yet applied it in a working environment. Predictably, plagiarism detection was the most widely known application, which was recognized by about 50% of respondents (Fig. 2). Image recognition (42%) was another fairly widely known application of AI, followed by data analytics (40%), language enhancement (39%), text analysis (34%), text summarization (33%), and metadata creation (28%). Bots that write manuscripts are a relatively new application that only a few participants (6%) were aware of. Other responses included automated reasoning and logic.
How has AI revolutionized academic publishing?
The answers to question 3, “What are the benefits of implementing AI in research and publishing?,” are presented in Fig. 3. Among the perceived benefits of AI, automation of repetitive tasks (57%), reduction of overall cost and time (52%), and improved quality of the output (50%) were the most prominent responses. The responses to question 4, “What do you anticipate to be the primary obstacle in implementing AI?,” are given in Fig. 4. The lack of competency in understanding AI (42%), difficulties in integrating AI-based solutions into existing IT infrastructure (41%), and the lack of technical expertise and specialized equipment/software (38%) were stated as the major challenges. Consequently, respondents suggested that academic institutions or publishers must invest in additional training (39%) or rely heavily on AI-trained staff (35%) to implement and use AI-based tools. Furthermore, the cost of implementation (upfront investment) and maintenance, uncertain return on investment, lack of standards, and other legal and compliance issues were also identified as challenges faced by the respondents. Moreover, 35% of respondents said that the primary reason for not adopting AI was the fact that their organizational culture had not yet recognized the need for it. Overall, there did not seem to be a single dominant reason for the limited use of AI; instead, multiple concerns contributed to this factor. Another obstacle specified was the instability of internet connections. The next question (question 5) in this section aimed to understand the areas where AI would require improvement: “Which problems in academic publishing will be difficult to solve using AI?” The identification of predatory publishers (57%), the problem of data fabrication or fake data (44%), and review bias (46%) were reported to be the most pressing challenges, followed by the problem of inaccurate translations (32%).
Prospects of AI
This section of the survey focused on the expectations/requirements of respondents to use AI tools more effectively. In this section, we wanted to determine whether any specific AI assistance was required by the academic community. The responses to question 6, “What kind of AI assistance or access do you need in your current role?,” are shown in Fig. 5. Most responses (>40%) suggested the need for AI tools that could help them with global demographic analysis, perform automated text analysis, and monitor for copyright infringement. A significant number of participants also proposed developing AI-powered tools that could perform predictive analysis (35%) and manage royalties (25%). Some participants also mentioned that they would need assistance for translation and understanding machine learning and the scope of AI in the academic publishing domain. The answers to question 7, “Do you or your department need expert advice on how you can use AI to facilitate your publication journey?,” are presented in Fig. 6. The most response (77%) from our survey participants was that they needed expert advice on how to successfully and effectively implement AI to ease their publication journey. However, about 18% mentioned that they might require AI assistance in the near future.
Interpretation
Considering demographics, women are still underrepresented in fields such as computing, information technology, engineering, mathematics, and physics. According to the data released in the World Economic Forum’s Global Gender Gap Report 2021, globally, only 32% of AI professionals are female [4]. This report validates the gender disparity seen among our survey participants. Furthermore, the respondents primarily belonged to the age groups of 18–34 and 35–54 years, which presently constitute the largest consumer groups for AI. Majority of the respondents were observed to be academics from research institutions and organizations. Hence, the results provide a fair understanding of the usage and impact of AI tools at present. Meeting authors’ and researchers’ needs and expectations would present a strong case for adopting and improving AI in the academic publishing landscape.
The majority of the respondents identified in the survey were graduates, postgraduates, or researchers. Their digital experience will help push the boundaries of AI. This survey reveals that AI-powered plagiarism detection tools are widely used and provide a hassle-free solution for academics and professionals in the publishing industry. Accurate comprehension and dissemination of scientific literature are crucial aspects of academic writing and publishing. Sifting through the millions of available documents, such as review articles, research papers, or patents, to extract key information relevant to one’s research is a challenge. Software that assists in image recognition, language enhancement, and the creation of summaries and metadata is also among the widely known applications of AI in academic publishing. Respondents were also aware of tools that help in performing data analytics tasks such as automatic tagging, identification of entities, and the identification of metadata such as title and author. AI-powered bots are now assisting in the composition of the first draft of manuscripts, thereby revolutionizing scientific writing. In recent years, various support services that use AI technology for manuscript writing have become increasingly available [5]. AI-based applications are being developed to assist authors and publishers in performing activities with minimal human intervention and greater efficiency. For example, significant efforts have been made to automate parts of the peer review process, and recent years have seen the application of AI in assisting the peer review process. The ever-increasing growth in manuscript submissions has necessitated peer review automation, and improving the automation level of peer review has gained increasing attention over the past few years [6,7]. Knowledge and integration of AI applications into online publishing platforms will help create highly advanced and focused tools.
Some of the key benefits of AI tools identified by the respondents are relatively straightforward processes such as finding potential peer reviewers, scanning articles suitable for manuscript submission, and identifying language or grammatical errors. These tools enable researchers and publishers to rapidly complete routine or mundane tasks with the aid of a machine. It is important to focus on research that not only makes AI more capable but also maximizes its societal benefit. Evidence from the survey suggests that the lack of AI skills and difficulties in the application of AI solutions to existing infrastructure are the most common hurdles. To overcome the expertise issue, researchers and publishers could consider collaborating with external research organizations (for acquiring AI training, skills, and technology). The lack of awareness of the potential benefits and applications of AI also appears to be a significant barrier to the large-scale adoption of AI. Research organizations can achieve the most significant performance improvements when humans and machines work together. Thus, realizing this immense potential of AI, research organizations must take the initiative to recruit skilled training staff and expertise to achieve the desired outcomes.
AI-assisted tools are redefining the scholarly landscape. Academicians must be aware of what is unfolding before them and what preceded them. Making data-driven decisions is the need of the hour. With text analysis, global demographic analysis, and predictive analysis, researchers can use AI tools to convert unstructured text into meaningful data. Most of our survey participants agreed that AI has the potential to augment the academic publishing process. However, many institutions and organizations are still in the nascent stage in this regard. They are unaware of how to implement AI solutions in their workflow and where to begin. We think that there is a pressing need for focused webinars, conferences, and workshops to support and facilitate researchers’ understanding and usage of AI tools and algorithms in practice. These sessions must also provide insights into the impact of AI technologies on academic institutes and research organizations, the implications AI will have for the publishing processes, and the ways in which its performance can be enhanced.
Limitations and suggestions for further studies
While the survey was globally accessible, we found that Australia/Oceania had the lowest participation (3%). In contrast, Africa showed greater participation (9%), reflecting the fact that it is joining the global AI revolution [8], followed by Europe (8%) and South America (4%). As a next step, we plan to conduct additional surveys focusing on the identification of malpractice with AI, its integration with research and publishing demands, and awareness of trends in AI to derive comprehensive worldwide conclusions. As the AI landscape changes, there is also merit in assessing the detrimental and disruptive role AI can play in research integrity, science communication, and scholarly publishing.
Conclusion
The results of this survey provide interesting insights into how AI-supported innovations are perceived and used by key stakeholders in the academic ecosystem, such as publishers, editors, reviewers, authors, and many more. Academic stakeholders are already experimenting with AI tools to improve the current workflow and efficiency. Most of the survey respondents have claimed that a major limitation to implementing AI is the lack of knowledge, trained in-house staff, and resources. The ever-increasing demand for quality publications in the race to“publish or perish” will undeniably increase the adoption of AI in academic publishing. As technology improves, not only will it assist in increasing efficiency and reducing costs in the current research ecosystem, but it might also transform the world of research completely. This study outlines the challenges currently faced by the consumers of AI-based tools in academia and suggests possible ways to support a smooth transition. This can be best achieved by educating and creating awareness to ease the possible fears and hesitation, and to actualize the promising benefits of AI.

Conflict of Interest

No potential conflict of interest relevant to this article was reported.

Funding

The authors received no financial support for this article.

Supplementary material is available from the Harvard Dataverse at https://doi.org/10.7910/DVN/DTHJDG.
Suppl. 1.
Survey form of the questionnairs.
kcse-294-Supplementary-1.pdf
Fig. 1.
Understanding of the concepts of artificial intelligence, machine learning, internet of things, clustering, and other related topics.
kcse-294f1.jpg
Fig. 2.
Different artificial intelligence tools used by respondents.
kcse-294f2.jpg
Fig. 3.
Benefits of implementing artificial intelligence in research and publishing.
kcse-294f3.jpg
Fig. 4.
Key concerns associated with artificial intelligence (AI).
kcse-294f4.jpg
Fig. 5.
Artificial intelligence assistance required by respondents.
kcse-294f5.jpg
Fig. 6.
Expert advice on how to use artificial intelligence to facilitate the publication journey.
kcse-294f6.jpg
Table 1.
Content validity for the study of role and impact of artificial intelligence on academic publishing
No. Criteria Needs improvement (%) Fair (%) Good (%) Very good (%) Excellent (%)
1 Clarity of items in the questionnaire: The vocabulary level, language structure, and conceptual level of the questions meet the level of respondents. The questionnaire directions and items are written clearly and are easy to understand. 0 0 0 50 50
2 Organization and presentation of items: The items are organized and presented in a logical and sequential manner. 0 0 25 0 75
3 Adequateness of items: The questions are designed to determine the knowledge, perceptions, and attitudes of the respondents for the particular study. The number of questions is representative enough of the concept defined for the particular study. 0 0 0 75 25
4 Attainment of purpose: The instrument as a whole is relevant and could answer the basic purpose for which it is designed. 0 0 0 75 25
5 Objectivity: No aspect of the questionnaire suggests bias (such as gender stereotyping, etc.) on the part of the study. The items on the instrument can elicit responses that are stable, definite, consistent, and not conflicting. 0 0 0 50 50
Table 2.
Demographic findings of participants (n=365)
Category No. (%)
Sex
 Male 204 (56)
 Female 150 (41)
 Others 7 (2)
 Preferred not to disclose 4 (1)
Age group (yr)
 18–34 178 (49)
 35–54 161 (44)
 55–74 22 (6)
 ≥ 75 4 (1)
Geographical distribution
 North America 190 (52)
 Asia 88 (24)
 Africa 33 (9)
 Europe 29 (8)
 South America 14 (4)
 Australia/Oceania 11 (3)
Size of organization/institution
 Small (2–100) 142 (39)
 Medium (100–500) 69 (19)
 Large (> 500) 128 (35)
 Individual 26 (7)
Job title
 Postgraduate student 106 (29)
 Doctoral student 80 (22)
 Established researcher (having > 5 publications) 59 (16)
 Graduate student 55 (15)
 Journal editor 22 (6)
 Postdoctoral fellow 18 (5)
 Publisher 7 (2)
 Others 18 (5)

Figure & Data

References

    Citations

    Citations to this article as recorded by  
    • The impact of generative AI tools on researchers and research: Implications for academia in higher education
      Abdulrahman M. Al-Zahrani
      Innovations in Education and Teaching International.2023; : 1.     CrossRef

    Figure
    Related articles

    Science Editing : Science Editing