The present invention discloses a
Web service recommending method based on association rules. Although association rules are introduced into an existing
Web service recommending method and execution efficiency and
service quality of a
Web service composition are improved, only recommendation of strongly associated services can be generated and a part of services that are uncommon but have high association capabilities are ignored due to a low usage rate. According to the present invention, the association
rule mining is improved, i.e. the support degree for the association rules is modified. The improved association rule recommending method guarantees association strength between all atomic services in the
service composition not to be reduced due to the increase of the length of the
service composition and is improved in stability and accuracy compared with a conventional recommending method. According to the Web service recommending method based on the association rules, which is disclosed by the present invention, historical data of the
service composition is analyzed and an implicit internal relationship between individual Web services is mined, so that a potential demand of a user can be found and a
system selects the optimal service composition to recommend to the user. The Web service recommending method based on the association rules, which is provided by the present invention, provides a convenient and efficient service recommending method to users and has a wide application prospect.