AVALIAÇÃO DE SIMILARIDADE ENTRE CONCEITOS REPRESENTADOS PELA XBRL

Abstract

XBRL is an XML-based language consolidated as the standard for publishing and exchange of financial information on the Web. XBRL is extensible, feature that expanded its area of operation and increased the use by the financial community. XBRL concepts are the basis for the financial information. Due to the expanded use of XBRL and the increase in the creation of new financial concepts, it becomes relevant the application of the Data Quality Management (DQM) discipline in this context. Similarity evaluation is an important process in data management and serves as support for some of its core activities. The classification of XBRL concepts that are not identical according to the similarity between them is useful in the study of concepts (grouping), the integration of concepts (duplicate detection) in version control (change detection) in information retrieval (order results) and other applications. The objective of this work is to propose a process to evaluate similarity between concepts of XBRL. Therefore, a literature review on techniques for similarity evaluation and a survey on the characteristics of the concepts represented by XBRL were performed; a process for similarity evaluation between concepts represented by XBRL, based on techniques from the literature, was proposed and a case study was made, which applied the proposed process in a developing taxonomy.