This paper explores the evolution of user interaction in Facebook's social network, focusing on the activity network, which represents actual interactions rather than just friendships. The study reveals that links in the activity network tend to come and go rapidly over time, with the strength of ties generally decreasing as the social network link ages. For example, only 30% of Facebook user pairs interact consistently from one month to the next. Despite the rapid changes in individual links, many graph-theoretic properties of the activity network remain stable over time. The research collected data on friendship relationships and interactions for a large subset of the Facebook New Orleans network, examining over 60,000 users and over 800,000 logged interactions over two years. The data shows that interactions between user pairs vary widely, with some pairs interacting infrequently and others frequently. The study found that interactions between infrequently interacting user pairs are often triggered by site mechanisms, such as Facebook's birthday reminder feature. In contrast, frequently interacting user pairs show a more stable interaction pattern, with activity levels decreasing over time. The paper also examines the evolution of the activity network's structure over time. It finds that while the activity network changes rapidly, many of its structural properties, such as average node degree, clustering coefficient, and average path length, remain relatively stable. This suggests that despite the dynamic nature of individual links, the overall structure of the activity network remains consistent over time. The study highlights the importance of considering the dynamic nature of user interactions in social networks. It shows that while the activity network is highly dynamic, it also exhibits stable structural properties, which can be useful for understanding and analyzing social networks. The findings have implications for the design of social network systems, emphasizing the need to consider both the dynamic nature of individual links and the stability of overall network properties.This paper explores the evolution of user interaction in Facebook's social network, focusing on the activity network, which represents actual interactions rather than just friendships. The study reveals that links in the activity network tend to come and go rapidly over time, with the strength of ties generally decreasing as the social network link ages. For example, only 30% of Facebook user pairs interact consistently from one month to the next. Despite the rapid changes in individual links, many graph-theoretic properties of the activity network remain stable over time. The research collected data on friendship relationships and interactions for a large subset of the Facebook New Orleans network, examining over 60,000 users and over 800,000 logged interactions over two years. The data shows that interactions between user pairs vary widely, with some pairs interacting infrequently and others frequently. The study found that interactions between infrequently interacting user pairs are often triggered by site mechanisms, such as Facebook's birthday reminder feature. In contrast, frequently interacting user pairs show a more stable interaction pattern, with activity levels decreasing over time. The paper also examines the evolution of the activity network's structure over time. It finds that while the activity network changes rapidly, many of its structural properties, such as average node degree, clustering coefficient, and average path length, remain relatively stable. This suggests that despite the dynamic nature of individual links, the overall structure of the activity network remains consistent over time. The study highlights the importance of considering the dynamic nature of user interactions in social networks. It shows that while the activity network is highly dynamic, it also exhibits stable structural properties, which can be useful for understanding and analyzing social networks. The findings have implications for the design of social network systems, emphasizing the need to consider both the dynamic nature of individual links and the stability of overall network properties.