Deep Learning Framework for Online Interactive Service Recommendation in Iterative Mashup Development

Recent years have witnessed the rapid development of service-oriented computing technologies. The boom of Web services increases the selection burden of software developers in developing service-based systems (such as mashups). How to recommend suitable follow-up component services to develop new mashups has become a fundamental problem in service-oriented software engineering. Most of the existing service recommendation approaches are designed for mashup development in the single-round recommendation scenario. It is hard for them to update recommendation results in time according to developers' requirements and behaviors (e.g., instant service selection). To address this issue, we propose a deep-learning-based interactive service recommendation framework named DLISR, which aims to capture the interactions among the target mashup, selected services, and the next service to recommend. Moreover, an attention mechanism is employed in DLISR to weigh selected services when recommending the next service. We also design two separate models for learning interactions from the perspectives of content information and historical invocation information, respectively, as well as a hybrid model called HISR. Experiments on a real-world dataset indicate that HISR outperforms several state-of-the-art service recommendation methods in the online interactive scenario for developing new mashups iteratively.