A semantic matching engine for web service composition

Abstract

One of the main assets of service-orientation is composition, which consists of developing higher-level services by re-using well-known functionality provided by other services in a low-cost and rapid development process. However, considerable differences on structural, semantic and technical levels along with the growing number of available web services makes their discovery a significant challenging task. Therefore, services compatibility is an essential prerequisite to service composition. Measuring the similarity of services is an important and valuable task to get useful information about their compatibility. Similarity measure can be considered as an optimisation step before composing services since it enables to reduce the search time by functionally classifying similar services. This paper presents a practical approach to measure the similarity of web services. Both semantic and syntactic descriptions are integrated through specific techniques for computing similarity measures between services. Formal concept analysis (FCA) is then used to classify web services into concept lattices, and therefore generate a hierarchy of classes of similar web services. Service clustering is used to narrow down the search space and to enable rapid semantic matching of a service request against a large size pool of services. Following this step, a composition engine takes as inputs the set of similar services and the specification of the required service, and generates the candidate composition plans that realise the goal. To determine the composition plan, the composition is reduced to a planning problem.

Semantic Matching Efficiency of Supply and Demand Text on Cross‐Border E‐Commerce Online Technology Trading Platforms

“…At the same time, the traditional capsule network is improved, words that have nothing to do with text semantics are regarded as noise capsules, and smaller weights are assigned to reduce the impact on subsequent tasks [24]. The text semantic matching method first uses the pretrained GloVe model to map the two texts into a 300dimensional word vector matrix [25]. The word vector matrix is used as the input of the model, and weights are given by the attention mechanism module, and then, the results are input into the BiGRU network and the capsule network model, respectively.…”