Performance of Metadata Structure of Digital Identifier Systems: Comparative Evaluation

The main solution to the problems of persistency and uniqueness in identification of digital objects in a web environment is provided by using digital identifiers instead of URL. The main basis of this solution is resolution mechanism that is used in digital identifier systems. Resolution is the use of indirect names instead of URLs; what worked for
the DNS (Domain Name System) in stabilizing internet hostnames should work for digital object references. Considering that this mechanism is known as the technical backbone of the digital identifier systems, these systems are trying different ways to implement this mechanism with high technical and content quality. Accordingly, performance evaluation of the resolution structure in digital identifier systems is considered in this study To achieve this goal, two-step process was designed and implemented on the basis of illustrative evaluation method. In the first step, the basic framework of metadata structure component evaluation was designed with Delphi method. Designed basic framework in this step contains 9 performance indicators: conceptual and technical persistency of resolution, interoperability, reliability, speed and performance, scalability, multiple resolution, Internet functionality and no need to install. In the second step, according to the designed basic framework, an illustrative comparison of 6 well-known worldwide digital identifier systems was carried out, using TOPSIS method. Results of this study revealed that DOI, Handle and UCI identifier systems show high performance in their metadata structure, while PURL and ARK systems respectively have mediocre and low efficiency in this component and URN Systems do not have any performance in this component due to the lack of implemented resolution structures. Analysis of the results also indicated that the main reasons for high performance systems are using of Web proxy as a user interface, increasing the speed and performance using solutions such as replication and cache and implementation of multiple resolution mechanisms. The results of this study can be used by researchers of this field and administrators and users of these systems.