Abstract:Cloud application programming interface (API) is a software intermediary that enables data exchange, business logic or functionality delivery between applications, infrastructures and IoT devices for supporting service oriented architecture. Currently, the number of cloud API in the Web is increasing and the number of cloud API with similar functionality is very large. Since quality of service (QoS) can well differentiate the performance of similar cloud API, QoS prediction has become the critical base for fast, personalized and high-quality cloud API selection and recommendation. However, contexts are used indirectly through context aware neighbors in most existing researches and the combined feature interactions are treated equally in previous factorization machine based methods, which together hinder the accurate QoS prediction. To address the above concerns, we first conduct data analysis on real-world QoS datasets and provide conclusive evidence to verify the necessity of incorporating contextual information and differentiating feature interaction. Then, we propose a context-aware QoS prediction approach via automatic feature interaction named CAFI. Contextual information of both user and cloud API sides are directly fed into CAFI through feature embedding. Moreover, the importance weights of feature interactions are learned by a generalized regularized dual averaging optimizer, so as to reward effective feature interaction and penalize noise feature interaction automatically. Lastly, final QoS prediction is obtained in an ensemble way by balancing the results of linear regression, automatic feature interaction and non-linear interaction. Extensive experiments on two public real-world QoS datasets demonstrate that CAFI can significantly improve QoS prediction accuracy. And the proposed CAFI approach is promising to encourage service providers provide high quality APIs and enable developer get desired cloud APIs quickly, thereby promoting the development of API economy. • We discover a phenomenon that there is a positive correlation between contextual information and cloud API QoS data. • We observe a fact that the combined feature interactions contributes differently to QoS prediction. • We propose a unified context-aware QoS prediction approach via automatic feature interaction. • We demonstrate the effectiveness of CAFI by incorporating contexts and differentiating feature interaction.

Context and Auto-Interaction Are All You Need: Towards Context Embedding Based QoS Prediction Via Automatic Feature Interaction for High Quality Cloud API Delivery

Contexts Enhance Accuracy: on Modeling Context Aware Deep Factorization Machine for Web API QoS Prediction.

Modeling Feature Interactions for Context-Aware QoS Prediction of IoT Services

A QoS Guarantee Framework for Cloud Services Based on Bayesian Prediction

Time-Aware and Sparsity-Tolerant QoS Prediction Based on Collaborative Filtering

QoS Prediction of Web Services Based on Two-Phase K-Means Clustering

Location-aware Collaborative Filtering and Feature Interaction Learning for QoS Prediction

Recurrent Neural Network Based Collaborative Filtering for QoS Prediction in IoV

Context-Aware QoS Prediction with Neural Collaborative Filtering for Internet-of-Things Services.

Context-aware QoS Prediction for Web Service Recommendation and Selection.

Personalized Context-Aware QoS Prediction for Web Services Based on Collaborative Filtering.

AERQP: adaptive embedding representation-based QoS prediction for web service recommendation

Personalized Services Recommendation Based on Context-Aware QoS Prediction

Context-Aware and Adaptive QoS Prediction for Mobile Edge Computing Services

Personalized QoS Prediction of Cloud Services via Learning Neighborhood-based Model

QoS Prediction in Intelligent Edge Computing Based on Feature Learning

CCeACF: Content and Complementarity Enhanced Attentional Collaborative Filtering for Cloud API Recommendation

FEDGE: an Interference-Aware QoS Prediction Framework for Black-Box Scenario in IaaS Clouds with Domain Generalization

Personalized Qos Prediction For Web Services Via Collaborative Filtering

MF-Informer for long-term QoS prediction in edge-cloud collaboration environments

Alleviating Data Sparsity in Web Service QoS Prediction by Capturing Region Context Influence.