Connecting information between Technology and Culture: A case study on the information spread in the LinkedIn network of CultTech

Abstract

This thesis presents the development of a freely accessible webcrawler designed to infer social connections between individuals using minimal initial information and publicly available web content. Its proposed algorithm explores the efficiency of co-occurrence patterns, sentence proximity, language consistency, and other features. The program aims to identify genuine social ties from unstructured online data. The research addresses the significant challenge of reconstructing social networks without relying on extensive private or pre-existing datasets, emphasising the methodological hurdles such as name ambiguity, content heterogeneity, and validation limitations. While preliminary results indicate potential, the work also highlights the conservative nature of the approach and current limitations in accuracy and applicability, particularly concerning diverse populations and online content reliability. Overall, this tool offers a modest avenue for exploratory social network analysis within low-resource, ethical, and academic contexts.