Google Bard Generated Literature Review: Metaverse

Abstract

Literature review articles aim to present studies in a field to researchers in a regular, systematic and meaningful way. It is often a very difficult process to reach all the studies in the field to be reviewed or to identify and evaluate the important ones. With this study, the use of Artificial Intelligence robots and moreover generative artificial intelligence in literature review processes is evaluated. Google Bard was used to detect artificial intelligence's ability to generate a literature review article. First, Bard was asked questions to write about some of the Metaverse-related topics in this article. Some of the texts in the study were generated entirely by the answers given to the questions asked by Bard. In addition, 10 articles on Metaverse published in the last three years (2021, 2022 and 2023) were collected by searching Google Scholar with the word "Metaverse". Afterwards, these studies were interpreted by Bard. Bard was told to paraphrase the summary parts of the related studies and the produced texts were shared in the study. All produced texts were checked through ithenticate and the results were evaluated. In addition, the texts were evaluated semantically. Additionally, a comparison with the capabilities of OpenAI ChatGPT is given. The results are promising; However, it was observed that the plagiarism matching rate of the paraphrased texts was higher when compared to the answers given to the questions. This article is an experiment to show that the collection and expression of knowledge can be accelerated with the help of artificial intelligence. It is considered that the relevant tools will be used more and more effectively in academic literature in the future.