Cataloguing to Facilitate Big Data Analytics

Abstract

“Big Data” is the popular term used to denote the collection of large data sets possessed by multiple systems. The inherent characteristics of this Big Data are the difficulty in processing due to sheer scale and accessibility of data and also un manageability through a traditional Database Management System. The size of this data set is ever increasing with increasing pace and addition of multi-exabytes per day. Apart from these, big data normally comprise of heterogeneous dataset, both structured and unstructured and also containing diverse data and file formats. It is very difficult to locate and retrieve the relevant information in real time from the universe of big data. Librarians, coming out of the walled library, can be expected to contribute in this task with their expertise in information organization and management. In this paper various challenges to the big data are identified and to address the challenges mechanisms for creating big data catalogue have been identified. Various mechanisms are discussed and compared and it is proposed to use the technique of library classification and cataloguing to catalogue the datasets in the big data thereby facilitating the information retrieval in the universe of big data.