Artificial Intelligence (AI) and the literature review process: Synthesis
Although there is potential for AI tools such as GPT to create summaries from abstracts which would help in the synthesis of the research, "there were errors that suggest the technology is not yet ready for such a task" (Qureshi, 2023: 2).
There is one AI tool, Scite.ai, that has been applied in the literature to the evaluation process.
AI tools for critical evaluation
SciteAI
SciteAI, designed by Nicholson et al. (2021), categorizes citations according to whether articles are contrasting, supporting or mentioning an original article. Scite.ai is very keen for you to take a 7 day free trial but will provide you with a certain amount of free information that you can use for critical evaluation.
Using SciteAI
Enter the article title of a paper into Scite.ai. This will give you details of the number of publications citing this work. It will also tell you whether there is any supporting or contrasting evidence. This will work best if the paper is a few years old.
In this example, the search was for a 2019 paper:
Enter "Meta‐analysis of 16 studies of the association of alcohol with colorectal cancer" into the search box.
There are 78 more recent papers that have cited the 2019 original. 47 mention the paper but there are three which provide contrasting evidence and one which provides supportive evidence. These papers will give you the opportunity to provide critical evaluation of the content.
Evidence
You need to cite papers to provide supporting evidence for your arguments. In addition, you need to provide a critical assessment and evaluation of the papers you have cited. Recent articles have used Scite.ai to examine the reasons why authors are citing the paper (citation sentiment), whether they are positive (supporting) negative (contradicting) or neutral (Mentioning) (Bakker et al., 2023; Lund and Shamsi, 2023; Sebo and Shamsi, 2023).
Please be aware that Scite does include preprints which have yet to be peer-reviewed and open archives. It is also based on using natural language processing which may not always capture correctly nuances of meaning or translations, introducing language and cultural biases. However, Scite.ai is a "valuable tool for efficiently and effectively analyzing the types of citations made in various disciplines" (Lund and Shamsi, 2023: 4896).