Big data analysis system model using field data

Abstract

Today, thanks to the emergence of many methods of handling multi-dimensional data, enterprises have been able to analyze data above the TB size by utilizing BI (Business Intelligence) systems, and can apply the analyzed data to the decision-making process. However, small/medium-sized enterprises that lack IT experts or cannot make a sufficient investment in IT have difficulties catching up with this trend. In addition, the majority of enterprises have not introduced a Manufacturing Execution System (MES), and most of the enterprises that have managed onsite data have stored the data in the form of memos or Excel files, and thus cannot apply it to the decision making process. Thus, in this research, in order to strengthen the competitiveness of small/medium manufacturing enterprises, a Big Data analysis system model was developed to automatically collect, refine process and analyze the data used on the sites of manufacturing enterprises. Through ODE (Office Data Excavation) modules, this analysis system model automatically recognizes the pattern of documents for information (ex. Excel) recorded non-typically on sites, and then extracts, refines, and collects them in typical information. By providing the stored data through various charts by using D3.js, the open-source data visualization library, this model prepares the correlation between pieces of information and the multi-dimensional analysis base, and then effectively supports the decision making system. In addition, an economical Big Data analysis can be made by using the open-source Spago BI instead of the high-priced Big Data solution. © 2015Jae-Joong Kim, Baek-Min Seong, Min-Gyu Lee, Hyo-jungSohn and Jong-Bae Kim.