Mashup technology enables developers to create new applications more readily by combining existing services. As its popularity grows, research on service recommendation for mashup creation has gained increasing attention. Existing recommendation methods have the following limitations: either they are susceptible to data sparsity problems, or they exhibit over-smoothing when aggregating high-order neighbors, resulting in similar and non-specific node feature representations, or they only focus on bipartite graphs and neglect the rich heterogeneous information in the mashup-service ecosystem. To address these issues, we propose a service recommendation method for mashup creation based onlightheterogeneous hypergraphcontrastivelearning (LHGCL). This method first constructs a heterogeneous hypergraph by combining mashup information, service information, the mashup-service interaction data, and their related attribute information. Then, it designs a light hypergraph neural network to capture the high-order relationships between mashups and services. Next, it applies contrastive learning to enhance the representations of mashups and services. Finally, it utilizes the enhanced feature vectors of mashups and services to predict mashup preferences for services. Comprehensive experiments conducted on the real-world ProgrammableWeb dataset demonstrate the superiority of the proposed method and the effectiveness of its key modules.

Third-party APIs (Application Programming Interfaces) are widely used in modern software development nowadays. Inspired by traditional recommender systems, recommending appropriate third-party APIs to developers has attracted a lot of research interest. Existing methods mainly focus on applying techniques such as Matrix Factorization (MF), Factorization Machine (FM), graph neural network (GNN) and hypergraph neural network (HGNN) to solve the recommendation problem. However, some limitations have not been well explored in existing methods: 1) MF and FM based API recommendation methods have difficulties in capturing the high-order interactions between users and APIs and are subject to noisy features. 2) GNN based methods can only be applied to simple graph structures, and suffer from the over-smoothing problem when aggregating high-order neighbor information. 3) HGNN based methods are focused on homogeneous hypergraphs and do not take the extra node attributes into consideration. To tackle the limitations, this paper proposes a third-party API recommendation method based on Heterogeneous Hypergraph Attention Network (HHAN). This method first constructs a heterogeneous hypergraph by exploiting the user-API interaction data and extra API attribute information. It then aggregates the neighbor information on the heterogeneous hypergraph to capture the high-order relationships between APIs and users. Finally, a node - and hyperedge-specific attention mechanism is designed to distinguish the importance of different types of neighbors. Extensive experiments on a real-world dataset crawled from ProgrammableWeb.com demonstrate the effectiveness of the proposed method.

Third-party APIs have been widely used to develop various applications.As the number of third-party APIs grows, it becomes increasingly challenging to quickly find suitable APIs that meet users' requirements.Inspired by recommender systems, API recommendation methods have been proposed to address this issue.However, previous API recommendation methods are insufficient in utilising the high-order interactions between users and APIs, and thus have limited performance.Based on the model of lightweight graph convolutional neural network, this paper proposes an effective API recommendation method by exploiting both low-order and high-order interactions between users and APIs.It first learns the embedding of users and APIs from the user-API interaction graph, and then adopts a weighted summation operator to aggregate the embeddings learned from different propagation layers for API recommendation.Extensive experiments are conducted on a real dataset with 160,309 API users and 21,031 Web APIs, and the results show that our method has significantly better precision and recall than other state-of-the-art methods.