Data Mining Ancient Script Image Data Using Convolutional Neural Networks

Abstract

The recent surge in ancient scripts has resulted in huge image libraries of ancient texts. Data mining of the collected images enables the study of the evolution of these ancient scripts. In particular, the origin of the Indus Valley script is highly debated. We use convolutional neural networks to test which Phoenician alphabet letters and Brahmi symbols are closest to the Indus Valley script symbols. Surprisingly, our analysis shows that overall the Phoenician alphabet is much closer than the Brahmi script to the Indus Valley script symbols. CCS CONCEPTS • Computing methodologies → Machine learning; Machine learning approaches;

The design and implementation of AIDA

“…The AIDA system has a potential to be a widely used resource for many scholars in the humanities in the fields of classics, history and linguistics. As a future work, we hope to extend the system with other ancient languages, such as Sumerian [5,18], Elamite [6], and the Indus Valley Script [3,20]. As our database grows, we also investigate the possibility of using ElasticSearch [9] to make queries more efficient.…”

Convolutional Neural Networks Analysis Reveals Three Possible Sources of Bronze Age Writings between Greece and India

“…Arora et al [36] compared neural networks and support vector machines using the Devanagari script, which is a Brahmi descendant. However, these earlier works did not apply CNNs to generate various script classification dendrograms, as in the current paper, and in the preliminary conference papers of the authors [37,38]. The authors also applied non-neural network-based techniques to identify allographs within the Indus Valley script [39].…”

Data mining ancient scripts to investigate their relationships and origins

“…Earlier work of the authors shows the similarity between the Indus Valley script and other scripts using CNNs [6,7]. However, the use of neural networks to generate script families is a new domain.…”