Sentiment Analysis of Primary Historical Sources

Sentiment analysis is a research area that has experienced considerable growth over the last decade. The scope of sentiment analysis is to extract feelings, opinions, and emotions from text. This paper employs state-of-the-art multilingual sentiment analysis algorithms in the field of primary historical sources to assist historians in discovering new insights. In particular, we use Sentic APIs, a neurosymbolic AI toolkit, to process the travel accounts of Ibn Battuta (in Arabic), Marco Polo (in Italian), and Ma Huan (in Chinese), available in the Engineering Historical Memory web application. Results show that sentiment analysis can be a powerful tool for historians and researchers to extract valuable insights from historical resources, shedding light on the emotions, opinions, and societal changes that shaped the past.