GraphEvol: A Graph Evolution Technique for Web Service Composition

Abstract

Web service composition can be thought of as the combination of reusable functionality modules available over the network to create applications that accomplish more complex tasks. Evolutionary Computation (EC) techniques have been applied with success to the problem of fully automated Web service composition, which is when candidate services are selected at the same time that the best configuration in which to connect those candidates is determined. Genetic Programming (GP) is the EC technique traditionally employed in fully automated Web service composition, with solutions encoded as trees instead of their natural Directed Acyclic Graph (DAG) form. The problem with a tree representation is that it complicates the enforcement of dependencies between service nodes, which is much easier accomplish in a DAG. This motivates the proposal of GraphEvol, an evolutionary technique that uses DAGs directly to represent and evolve Web service composition solutions. GraphEvol is analogous to GP, but it implements the mutation and crossover operators differently. Experiments were carried out comparing GraphEvol with GP for a series of composition tasks, with results showing that GraphEvol solutions either match or surpass the quality of those obtained using GP, at the same time relying on a more intuitive representation.

Robustness Estimation and Optimisation for Semantic Web Service Composition With Stochastic Service Failures

“…This field has been widely studied in the past few years, focusing on developing effective and efficient EC techniques to find composite services with optimised QoS. To achieve such a goal, researchers have been working on developing new and effective representations of composite services for EC techniques, such as DAG-based [24,17,25,26], tree-based [4,7,27,28,29], tree-like based [7], and permutation-based representations [5,23,8]. The majority of these works also propose different representation-dependent genetic operators to explore large searching spaces, while some of them [8,17] propose new sampling techniques for breeding new promising solutions from the learned distributions (i.e., Node or Edge Histogram Matrix) of historical solutions.…”

“…In the presentation of this paper, we mainly focus on two constructors, sequence • and parallel , similar to most automated service composition works [8,25,28,27,51], where a composite service is often represented in the form of a directed acyclic graph (DAG, denoted as G). In a DAG, nodes represent web services (also called component services) and edges represent robust causal links.…”

Evolutionary Multitasking for Semantic Web Service Composition

“…This representation allows better scalability by eliminating the replicas of subtrees based on the tree-based representations in [8], [9], [11]. Apart from evolving trees, graph evolution techniques are employed to evolve graph-based composite solutions, such as GraphEvol [19]. On the other hand, Particle Swarm Optimization, Genetic Algorithm, and Estimation of Distribution Algorithm have been employed to solve the single-tasking service composition problem [2], [4], [6], [10].…”

EDA-based approach to comprehensive quality-aware automated semantic web service composition

“…In the presentation of this paper, we mainly focus on two constructors, sequence • and parallel , similar as in most automated service composition works [5], [8], [10], [11], [32], [33], where service composition solutions are represented as a Directed Acyclic Graph (DAG). We can easily calculate QoS of a composite service that is represented as a DAG [10] according to Table 1.…”

“…Service composition solutions are commonly represented as Directed Acyclic Graphs (DAGs) [5], [8], [10], [11], [32], [33]. Let G = (V, E) be a DAG-based composite solution from Start to End, where nodes correspond to the services and edges correspond to the robust causal links.…”