ERMIS: Extracting Knowledge from Unstructured Big Data for Supporting Business Decision Making
Business managers support that decisions based on data analysis are better decisions. Nowadays, in the era of digital information, the accessible information sources are increasing rapidly, especially on the Internet. Also, the most critical information f
- PDF / 2,706,239 Bytes
- 12 Pages / 439.37 x 666.14 pts Page_size
- 41 Downloads / 175 Views
Industrial Systems Institute, ATHENA Research and Innovation Centre, Patras Science Park Building, Platani, 265 04 Rio, Patras, Greece {alexakos,nzervos}@isi.gr, [email protected], [email protected], [email protected], [email protected]
Abstract. Business managers support that decisions based on data analysis are better decisions. Nowadays, in the era of digital information, the accessible infor‐ mation sources are increasing rapidly, especially on the Internet. Also, the most critical information for business decisions is hidden in a large amount of unstruc‐ tured data. Thus, Big Data analytics has become the cornerstone of modern Busi‐ ness Analytics providing insights for accurate decision making. ERMIS (Exten‐ sible pRoduct Monitoring by Indexing Social sources) system is able to aggregate unstructured and semi-structured data from different sources, process them and extracting knowledge by semantically annotating only the useful information. ERMIS Knowledge Base that is created from this process is a tool for supporting business decision making about a product. Keywords: Big Data · Business Analytics · Ontologies · Data driven decision making · Knowledge extraction
1
Introduction
During the last decade the term Big Data has the pride of place in both academia and industry on the area of Business Analytics [1]. Big Data analytics methodologies and technologies permit the processing of large amount of data for providing accurate insights for a business [2]. Especially today, the century of digital information, busi‐ nesses can collect useful data from various sources such as Internet Sites, blogs, social media and IoT infrastructure, even from their own information systems. This not only means the need for processing of large volumes of data but also the necessity for the analysts to face the fact that these data volumes are increasing in high rates [3]. Since, the industry sector supports that decisions based on data analysis are better decisions, the utilization of Big Data analytics enables managers to conclude to decisions based on evidence rather than intuition [4]. Another immense challenge of Business Big Data analytics is the extraction of useful knowledge from the millions of unstructured data existing in various sources on the
© IFIP International Federation for Information Processing 2016 Published by Springer International Publishing Switzerland 2016. All Rights Reserved L. Iliadis and I. Maglogiannis (Eds.): AIAI 2016, IFIP AICT 475, pp. 611–622, 2016. DOI: 10.1007/978-3-319-44944-9_54
612
C. Alexakos et al.
Internet. This problem is entitled as Variety and it is one of the three major challenges in Big Data, called the three Vs of Big Data, alongside with Velocity and Volume [5]. Big data can be characterized in three types: (a) structured, (b) semi-structured and (c) unstructured. Structured data is provided in an already tagged and easily sorted format. Unstructured data is random thus it is difficult to be processed. Semi-structured data has separated data
Data Loading...
