Complex systems analysis identifies Italian political and geopolitical shifts
Headline and summary by local AI
A new method using complex systems and statistical physics identifies major historical turning points in Italian media by analyzing 600,000 news articles.
The approach examines a 15-year archive of the Italian newspaper La Repubblica published between 1985 and 2000. By applying Natural Language Processing to the text, the researchers tracked lexical and semantic shifts to map how discourse changed during a period of intense political and social transition in Italy.
The analysis detected specific shifts such as the transition from the First Republic to the Second Republic and the impact of the Gulf War and Kosovo War. These transitions appeared as distinct changepoints in word salience and semantic similarity. The findings show that periods of war correspond to lower entropy in the semantic space, which indicates a more focused and less diverse media discourse.
This research demonstrates how computational linguistics can provide quantitative insights into historical dynamics without needing prior labels for the data. It offers a framework for studying how society and media evolve by treating news archives as complex systems.
Selected and summarized by local AI How we work
