Adjusting Matching Algorithm to Adapt to Dynamic Subscriptions in Content-Based Publish/Subscribe Systems

Abstract

Content-based publish/subscribe systems enable on-demand event distribution based on users' interests. In dynamic environments, such as social networks and stock exchanges, the subscriptions that express users' interests update frequently, which changes the subscriptions' matchability which is defined as the matching probability of subscriptions with events. In the presence of dynamic subscriptions, it is challenging to maintain the performance stability of matching algorithms as the subscriptions' matchability is an important factor that impacts the performance of matching algorithms. So far, this issue has not been well addressed in the literature. In this paper, we design a matching algorithm that has the ability to adjust its behavior to adapt to dynamic subscriptions, aiming at stabilizing the performance of matching algorithms. To achieve this objective, a lightweight adjustment mechanism is proposed and adopted on a selected test bench, which gives rise to Maema, a matchability adaptive event matching algorithm. The effectiveness of Maema is extensively evaluated through a series of experiments using both synthetic and real-world data. Experiment results show that Maema not only possesses the beneficial adaptability, but also performs more efficiently.