Adrien Guille's PhD work focuses on information diffusion in online social networks. The aim is to understand how information spreads and to develop tools for analyzing this process. The work presents three main contributions: (i) a survey of developments in the field, (ii) T-BaSIC, a graph-based model for predicting information diffusion, and (iii) SONDY, an open-source platform for analyzing social network dynamics. T-BaSIC is a time-based model that accounts for the temporal aspect of diffusion, allowing for more accurate predictions. SONDY provides functionalities for detecting topics, analyzing network structures, and comparing different techniques for social data mining. The work addresses challenges in information diffusion, including identifying popular topics, understanding diffusion paths, and identifying influential spreaders. It also highlights the importance of considering both network and temporal factors in modeling diffusion. The study shows that T-BaSIC outperforms existing models in predicting the temporal dynamics of information diffusion, although it underestimates the volume of diffusion. The research contributes to the understanding of information diffusion in online social networks and provides tools for analyzing and predicting this phenomenon.Adrien Guille's PhD work focuses on information diffusion in online social networks. The aim is to understand how information spreads and to develop tools for analyzing this process. The work presents three main contributions: (i) a survey of developments in the field, (ii) T-BaSIC, a graph-based model for predicting information diffusion, and (iii) SONDY, an open-source platform for analyzing social network dynamics. T-BaSIC is a time-based model that accounts for the temporal aspect of diffusion, allowing for more accurate predictions. SONDY provides functionalities for detecting topics, analyzing network structures, and comparing different techniques for social data mining. The work addresses challenges in information diffusion, including identifying popular topics, understanding diffusion paths, and identifying influential spreaders. It also highlights the importance of considering both network and temporal factors in modeling diffusion. The study shows that T-BaSIC outperforms existing models in predicting the temporal dynamics of information diffusion, although it underestimates the volume of diffusion. The research contributes to the understanding of information diffusion in online social networks and provides tools for analyzing and predicting this phenomenon.