Traditional recommendation systems in software engineering (RSSE) analyze artifacts stored in large repositories to create relevant recommendations. More recently, researchers have started exploring interaction data as a new source of information—moving closer to the creation and usage of the artifacts rather than just looking at the outcome. In software engineering, interaction data refers to the data that captures and describes the interactions of developers with artifacts using tools. For instance, the interactions might be the edits or selections that affect specific source code entities or webpages (artifacts) using an integrated development environment or a web browser (tools). Interaction data allows to better investigate developers’ behaviors, their intentions, their information needs, and problems encountered, providing new possibilities for precise recommendations. While various recommendation systems that use interaction data have been proposed, there is a variety in the data being collected, the way the data is collected, and how the data is being processed and used. In this chapter, we survey and summarize the major approaches for RSSEs that create recommendations based on interaction data. Along with this, we propose a conceptual framework for collecting and processing interaction data for the purpose of recommendation.