Survey and modeling metadata schema relationship in agriculture domain for better Metadata Schema Service.
P. Lousangfa, Naiyana Sahavechaphan, Marut Buranarach, S. Vannrat, Asanee Kawtrakul, T. Nagatsuka, S. Ninomiya · 2008
Information is typically described by a set of element values based on either locally or globally pre-defined element set. Essentially, the application of globally pre-defined elements set, termed as standard metadata schemas, enables an efficient information sharing and interoperability across repositories. According to this, many organizations and committees (i.e. FAO and FDGC) have thus developed various standard metadata schemas which are distributedly available via the internet. The discovery for metadata schema of choice through various web sites, however, can be time-consuming and error prone, making (i) similar information across repositories to be defined using in-house schemas or even different standard ones; and (ii) organizations or committees to define redundant metadata schemas for similar information. These in turn hinder information sharing and interoperability. To address this, we have initiated Metadata Schema Service as the first portal for existing metadata schema available at different web sites. In this paper, we thus propose the classification and relationship model, namely CRM, as an underlying model of Metadata Schema Service. Particularly, CRM is a two-level model for classifying and relating metadata schemas. The upper level focuses on classification to facilitate the discovery for desired metadata schema, while the lower level defines relationship of metadata schema to support the development of new metadata schema. We have evaluated the applicability of CRM in agricultural and related domain as agricultural information sharing and interoperability is quite significant for agricultural countries. Specifically, we have (i) surveyed existing metadata schemas in agricultural and related domain; (ii) analyzed them in term of their classification and relationship; and (iii) instantiated CRM with the previous analysis result. In addition, the usefulness of CRM is shown with respect to metadata schema browsing and understanding.