While the idea of technology as a driving force of history and society has been extensively studied in the history of the Industrial Revolutions, little attention has been paid to the social perception of those technologies. This thesis focuses on the discourses about technology observed in newspapers contemporary to the Second Industrial Revolution in Western countries. The goal is to highlight tensions between homogeneity and heterogeneity of information around the emerging technologies of that time, using computational methods. Topic Modelling models are applied to bring out the different discourses from the corpus, and the context in which technologies are discussed. To face the challenge of a multi-lingual dataset, Cross-Lingual Word Embeddings Models are used, allowing to compare topics across places and time. Finally, clusters of topics are computed using Network Analysis tools for community detection. General discussions, from the domestic to industrial use of technologies, common to the studied countries emerge from these analysis, as well as others topics specific to their societies.