Translating user preferences into fuzzy rules for the automatic selection of services

Abstract

Abstract-This article proposes an approach for including user preferences and quality of service characteristics in the selection process of services. Our approach consists of a Domain ontology for the service description vocabulary, and a trader that facilitates user-preference-based service selection, combining imperfect service matching and ranking algorithms. The novelty of our approach lies in the fact that we automatically generate fuzzy rules starting from individual user preferences and use them in a fuzzy inference process that ranks the service candidates. We present our experiments to evaluate our approach using a prototype implementation of a service broker.

Combining boolean consistent fuzzy logic and ahp illustrated on the web service selection problem

“…In the approach presented in Ref. 29 fuzzy rules are automatically generated from the decision maker's preferences and used in a fuzzy inference process that ranks the web services. The premises of the fuzzy rules specify possible combinations of values for the attributes, and the corresponding conclusion specifies the degree of acceptability for that web service.…”

Qos Based Selection and Composition of Web Services-a Fuzzy Approach

“…Furthermore, different web service applications with different QoS requirements as in (Ran, 2003;Sora et al, 2009), will compete for network and system resources such as bandwidth and processing time. Nevertheless, an enhanced QoS for a web service will bring competitive advantage for service provider.…”

Mapping a fuzzy logic approach for QoS-aware service selection on current web service standards

“…In order to complete its information about the semantics of QoS parameters found in the description of service candidates, it interacts with the Domain Ontology Service to obtain additional semantic information. The ranking is done the FuzzyRanking subsystem (FR), which implement the approach based on automatically generating fuzzy rules starting from individual user preferences and using them in a fuzzy inference process ranking the web service candidate, as described in our previous work [5].…”

“…In our previous works [5], [6], we have developed a novel fuzzy logic approach for the specification, selection and ranking of services according to individual QoS preferences. Our approach uses fuzzy inference for ranking the candidates, but based on sets of automatically generated fuzzy rules for each set of individual preferences.…”

“…Fuzzy rules have a big expressive power, and the fact that they are generated automatically makes this approach user-friendly. As we have shown in [5], the advantage of using fuzzy inference with generated rules for service ranking is that it is more flexible and has a bigger and more controllable expressive power than the classical fuzzy multicriteria decision making (FMCDM), especially in the case when several properties with different importance degrees are considered.…”