Trust Behavior Based Recommender System for Software Usage

Trust plays an important role in software usage. It helps users overcome perceptions of uncertainty and risk and engages in trust behaviors. In this position paper, we propose using the trust behavior as a main characteristic to provide software recommendations. We explore a model of trust behavior that contains three root constructs: using behavior, reflection behavior and correlation behavior. The behavior information can be automatically collected through software usage observation. We further design an algorithm that provides recommendations on software based on the correlation of trust behaviors in order to overcome the challenges of existing recommender systems.