Document Type



Internet services in distributive organization are normally built on an open network environment. In the environment internet service provisioning cannot be expected executing in a single close organization [1]. Identity management in cross-organization becomes an issue for handling Internet service and distributive business process.

The “identity” in cross-organization web service is defined as global identity rather than private identity from client. Global identity searching table that registers all related service organization is the normal way we used to [2]. Through global identity searching table the target service organization can be looked up directly. For some business program, however, global identity is not necessary registered in specific service organization, e.g. IMSI (International Mobile Subscriber Identification) registration in NP (Number Portability) Service [3], for instance. In NP Service each IMSI can be re-assigned to different mobile service provider if the IMSI apply the re-assignment.

In the example about IMSI in NP service mentioned above, if there is an internet service will be executed according to identity management with IMSI, then it will be a challenge to find the organization for available IMSI in very short response time. To dynamically re-assign a IMSI in different mobile service provider, the traditional global identity searching table will not be practical due to frequently changing the registration of IMSI in different service provider.

To give an intelligent searching model for cross-organization global identity management is a better way than a static global identity searching table management in Web service. In this article the XCS (eXtended Classifier System) classifier system [4] will be proposed as the kernel system. With the characteristics in machine learning and rules management, the XCS-based intelligent searching model can help to predict where the web service can find the global identity in the open and cross-organization environment.