Cloud-Based Data Classification Framework for Cultural Heritage Conservation

Abstract

Nineveh is one of ancient cities of Iraq, ruled by various civilizations throughout the ages. As a result, Nineveh retains different types of tangible and intangible cultural heritage with different values that show its importance. This research contributes to the digital archiving process of the culture heritage of Nineveh by suggesting a cloud-based framework that classifies the text data obtained from various heterogonous data sources according to the type, values, civilization, and time to which they belong. We used four classical machine learning algorithms to train the classifier, such as Multinomial Naive Bayes, Support Vector Machines, Random Forest, and K Nearest Neighbors. We then chose the classifier with the highest accuracy to classify the obtained texts automatically. The finding showed that the K-Nearest Neighbors classifier is the best classifier to be adopted in the classification process.