The paper presents a system for customized metadata facility adapted to expose domain specific metadata along with language attributes. Quality metadata is crucial for efficient information retrieval. Use of standard and common vocabulary leads to normalization of differences that arise due geographic, cultural, domain specific environments even amongst digital libraries with same or similar collection scopes. The present work demonstrates implementation of language attributes (which is missing in the official version of DSpace) for metadata elements. As a case study we demonstrate the AGRIS (Agricultural Information System) application profile customized for metadata in DSpace repositories. It serves as a full text content management platform for the agricultural community and especially the languagewise disparate and geographically distributed AGRIS network through enriched domain metadata in standard format for the description, management and better metadata sharing.