AI in Academia: How is Big Data Changing the Landscape?

Undoubtedly, COVID-19 significantly catalyzed a technological shift at universities worldwide. Today, academia and education are undergoing one of their most substantial transformations. Millions of students, tutors, and professors have had to adapt to online classes to continue their research and studies.

Nevertheless, research suggests technological transformation was already underway in academia before the global pandemic. In recent years, Artificial Intelligence (AI) and Machine Learning (ML) have profoundly impacted how we learn and study. It has also played a crucial role in how universities meet the demand for data scientists and innovative scientific research.

In this article, we will discuss how AI's evolution is shaping the future of academia and the growing demand for expertise in this fast-growing technology.

How will AI change academia?

AI and ML technologies have evolved dramatically over the past decade while the field of their application continues to expand. As these technologies advance and become more sophisticated, experts are confident they will revolutionize academia.

For example, AI has enabled overburdened professors to reduce their workload by automating tasks such as data entry and grading, freeing them up to focus on more important things. For students, this technology can customize the learning experience to their needs.

Throughout the globe, universities, and scientific institutions are recruiting data-savvy individuals seeking to tackle academic, business, and societal challenges with the support of AI technologies. Elsevier, a leader in scientific information, research, and analytics, has been working with the global research community to understand better how the evolution of AI will affect the world of academic research. In their Research Futures Report 2.0, they noted some key findings regarding AI:

• Content shared on preprint servers, which are online repositories sharing data from scholarly articles awaiting peer review, are now accepted by 67% of researchers as a good source of communication, an increase of 43% before the pandemic.
• Attitudes toward using AI have grown more favorable, with 16% of researchers using AI in their work.
• Acceptance of AI in peer-reviewed work is gaining; now, 21% of researchers say they would read AI-assisted articles, a five percent increase since 2019.
• However, 58% of researchers said they would not be willing to read AI-generated articles.

The new methods of scientific collaboration and impact assessments brought about by AI are transforming the world of academia as we know it. With these visions beginning to come to fruition, AI will continue to play a crucial role in scientific research. Two studies illustrate the value of AI-enabled technologies for science.

MORE: How virtual learning tools enable research

Powering AI initiatives in academia

Amidst global scrutiny of news, particularly its authenticity, a 2018 Massachusetts Institute of Technology (MIT) paper found that applying ML tools to help identify and detect fake news should still be further improved. Indeed, the MIT study resulted in the development of a device that can detect language patterns that characterize real and fake news and differentiate between the two with the support of ML and natural language processing (NLP) techniques.

This example showcases the progressive approach the academic community displays concerning AI and ML technology and provides a best-practice example of how these technologies can tackle societal issues.

AI-enabled processes have further allowed researchers to quantify and validate gender stereotypes and sexist language in literature. An unsupervised ML study, presented at the 2019 Association for Computational Linguistics meeting, concluded that negative terms in literature and language disproportionately identify female characters or subjects. In a collaborative effort, researchers from different institutions and private entities used AI and ML-based technologies to emphasize literature and language's impact on protecting gender inequalities.

The benefits AI and ML technologies provide in these cases are just two examples. AI also benefits the scientific community by fighting plagiarism and identifying inaccurate results and statistics.

AI expertise on the move

The increasing demand for AI and ML concepts has led to a hard-fought competition between academic institutions and the private sector to nurture and invest in junior AI researchers. Unsurprisingly, the potential for careers in the AI and ML fields is good, with the Bureau of Statistics predicting a 13 percent increase in AI-related occupations between 2016 and 2026.

According to a 2019 Worcester Polytechnic Institute study, the percentage of ads seeking expertise in AI, ML, and data mining has almost doubled in the five years prior to the study—that number has likely only continued to grow. The scarcity of AI researchers across all academic career levels has also benefited those involved in this industry. Higher salaries, improved resource allocation, and enhanced data sets—both in academia and the private sector—have substantially transformed the industry.

MORE: How to improve academic research for students and educators

Cooperation not confrontation

Fostering collaborative efforts is critical for developing an inclusive approach toward AI and ML technologies and their application in academia and beyond. In early 2019, Dutch Prime Minister Mark Rutte inaugurated Elsevier's TechHub, which supports efforts to streamline the development of AI-enabled technologies between the government, academia, and the private sector. Linked to Elsevier's TechHub, the Amsterdam-based Innovation Center for Artificial Intelligence is an additional example of how cooperation between the private sector and academic institutions can enhance the application of AI and ML technologies.

Undoubtedly, AI will substantially influence the world of science in the years and decades to come. It will make academia more accessible while simultaneously more cohesive, reducing current limitations researchers often face.