The Use of Artificial Intelligence in Visualizing Historical and Cultural Objects on Social Media:

A Sentiment Analysis of Public Reactions to AI Generated Images of Indonesian Heritage

Abstract

Visual reconstructions generated by Artificial Intelligence (AI) are swiftly gaining popularity in both academic circles and entertainment industries. Without credible sources, disseminating AI-generated visual depictions of historical figures and ancient sites on social media may perpetuate misinformation and cognitive bias. AI-generated reconstructions have been developed within the scientific practice of History, Anthropology, and Archaeology. However, there are limited studies regarding the utilisation of AI to visualise historical objects, in terms of public reaction to the AI-generated visual. By conducting sentiment analysis of comments on AI Nusantara’s TikTok platform, the study intends to provide an overview of public response to the visual design of historical figures and sites generated by AI technology. The public response to AI is important because many researchers are using social media as a source of data to forecast and explain human behaviour in real life. Related to the use of AI technology that penetrates various fields of life, this research in general can be useful for reading public responses and tastes so that they can project the use of AI in the future. Sentiment analysis methodology was employed to analyse comments posted under uploaded AI-generated images of historical figures and sites. The sentiment classification process uses the Naïve Bayes classifier algorithm in the RapidMiner software. This study finds that public sentiment regarding the utilisation of AI is categorised into three domains: accuracy, technology, and common knowledge. The categorisation is based on scientific opinion and local knowledge opinion. Accuracy relates to the similarity of AI-generated visuals to public expectations. Technology refers to the public's curiosity about the tools that are used to create the visualisation. Common knowledge refers to public perception and custom. This study also illustrates how AI can facilitate the creation of creative characters which are accurate by integrating academic references.