The chapter discusses the development and implementation of Fab, a hybrid content-based and collaborative recommendation system designed to address the limitations of both traditional content-based and collaborative filtering methods. Fab combines the strengths of both approaches to provide more accurate and personalized recommendations. The system addresses scaling issues common to web services, such as an increasing number of users and documents, and automatically identifies emerging communities of interest, enhancing group awareness and communication. Content-based recommendation in Fab involves comparing the content of web pages with user profiles, using features extracted from text to recommend similar items. Collaborative recommendation, on the other hand, identifies users with similar tastes and recommends items they have liked. Fab's hybrid approach leverages both methods, allowing it to recommend items that are both similar to a user's past preferences and relevant to their current interests. The Fab system is implemented as a distributed architecture with three main components: collection agents, selection agents, and a central router. Collection agents gather web pages relevant to specific topics, while selection agents deliver these pages to users based on their profiles. The central router routes pages to the appropriate users. The system dynamically adapts by weeding out unsuccessful collection agents and duplicating successful ones, ensuring efficient and effective recommendations. Experiments with a small number of users show that Fab's profile construction methods are effective, and the system outperforms other benchmarks in terms of user satisfaction. Future work will focus on scaling up the system to handle a larger number of users and further investigating the dynamic processes involved in the hybrid recommendation system.The chapter discusses the development and implementation of Fab, a hybrid content-based and collaborative recommendation system designed to address the limitations of both traditional content-based and collaborative filtering methods. Fab combines the strengths of both approaches to provide more accurate and personalized recommendations. The system addresses scaling issues common to web services, such as an increasing number of users and documents, and automatically identifies emerging communities of interest, enhancing group awareness and communication. Content-based recommendation in Fab involves comparing the content of web pages with user profiles, using features extracted from text to recommend similar items. Collaborative recommendation, on the other hand, identifies users with similar tastes and recommends items they have liked. Fab's hybrid approach leverages both methods, allowing it to recommend items that are both similar to a user's past preferences and relevant to their current interests. The Fab system is implemented as a distributed architecture with three main components: collection agents, selection agents, and a central router. Collection agents gather web pages relevant to specific topics, while selection agents deliver these pages to users based on their profiles. The central router routes pages to the appropriate users. The system dynamically adapts by weeding out unsuccessful collection agents and duplicating successful ones, ensuring efficient and effective recommendations. Experiments with a small number of users show that Fab's profile construction methods are effective, and the system outperforms other benchmarks in terms of user satisfaction. Future work will focus on scaling up the system to handle a larger number of users and further investigating the dynamic processes involved in the hybrid recommendation system.