Nathanael Pribady
•
Teachers College, Columbia University
•
np2903@tc.columbia.edu
This study utilizes machine learning techniques, including natural language processing and sentiment analysis, to dissect and interpret the evolution of educational philosophy across 60 key Western philosophical texts. By employing methods like Latent Dirichlet Allocation (LDA), knowledge inference, concept mapping, and sentiment analysis, we investigate how philosophers from different historical periods have conceptualized the role, methods, and outcomes of education. Key findings include the identification of dominant themes in various eras, such as ethics and human nature in the Ancient period, shifting to science, language, and socio-economic considerations in the Modern era.
History of Philosophy Dataset containing over 300,000 sentences from more than 60 philosophical texts, spanning ten major schools of thought: Plato, Aristotle, Rationalism, Empiricism, German Idealism, Communism, Capitalism, Phenomenology, Continental Philosophy, and Analytic Philosophy.
We utilized a combination of text analysis, topic modeling, and sentiment analysis to identify dominant themes and emotional tones in philosophical works across different eras. We employed association rule mining to explore relationships between philosophical schools and constructed knowledge graphs and concept maps to visualize semantic interconnections and thematic evolution.
Figure 1: Distribution of topics across Ancient and Modern philosophical eras