Augmenting Labeled Probabilistic Topic Model for Web Service Classification

Abstract

Web service classification has become an urgent demand on service-oriented applications. Most existing classification algorithms mainly rely on the original service descriptions. That leads to low classification accuracy, since it cannot fully reflect the semantic feature specific to a service category. To solve the issue, this article proposes a novel approach for web service classification, including service topic feature extraction, service functionality augmentation, and service classification model learning. The characteristic is that the original service descriptions can be semantically augmented, which is fed to deriving a service classifier via labeled probabilistic topic model. A benefit from this approach is that it can be applied to an online service management platform, where it assists service providers to facilitate the registration process. Extensive experiments have been conducted on a large-scale real-world data set crawled from ProgrammableWeb. The results demonstrate that it outperforms state-of-the-art methods in terms of service classification accuracy and convergence speed.