Clustering Techniques Within Service Sector

Clustering Techniques Within Service Sector

İbrahim Yazici (İstanbul Technical University, Turkey), Ömer Faruk Beyca (İstanbul Technical University, Turkey) and Selim Zaim (İstanbul Technical University, Turkey)
Copyright: © 2017 |Pages: 14
DOI: 10.4018/978-1-5225-2148-8.ch005
OnDemand PDF Download:
No Current Special Offers


Due to big data availability in markets recently, processing and making predictions with data have been becoming more difficult, and this difficulty has been affecting management decisions. As a result, competitiveness for companies are related to analyze and utilize big data in order to achieve company targets. Transforming big data into business advantage has become a vital management tool across all industries. There are many data mining techniques that are being applied to plenty of problems. One of the frequently utilized data mining technique is clustering method. Clustering techniques aim to group a set of objects in clusters that more similar objects are in the same cluster. Main utilization aim of clustering techniques is segmenting or clustering or grouping objects. Clustering techniques and their utilization within service sector by aim of clustering technique and their methodologies are presented. Energy, social media and bank sectors are found that the mostly user of clustering techniques within service sector for segmenting customers based on searched papers.
Chapter Preview


Big data refers to large and complex datasets, whether they are structured or unstructured, growing in size and complexity. Big data term is often used instead of analytics term. Even the aim of both of these terms is to extract useful information from data, then, to acquire business advantages from the useful information, three distinctions are available among them. These distinctions are named as the three Vs (McAfee & Brynjolfsson (2012));

  • Volume: Collections of much more information from transactions, records, tables and files. Big data are defined with terabytes even with petabytes.

  • Velocity: Data streams or delivery from any kind of machine. Storing and processing data whether in real time or not is growing.

  • Variety: Availability of data stream sources such as clickstreams, geo-spatial feedbacks, RFID, logs, clickstreams. And some audios and videos that are hard to be categorized (Russom (2011)).

The three Vs are illustrated in Figure 1. The figure illustrates the components of the three Vs of big data.

Figure 1.

The three Vs in big data (Russom (2011))


In addition to the three Vs, two dimensions are regarded for big data: variability and complexity. Variability means periodic or non-periodic peaks of data stream creating more unstructured data, hence, difficulty of manageability of the data. Complexity means that the availability of difficulty in data cleansing, linking, transforming across systems due to the fact that data come from various resources In today’s world, not only storing and having data but also processing and transforming them into the business advantage is important for the companies. Big data management may provide efficient customer solutions, and reductions in cost, time and improvements in decision making.

Measurability and manageability is related terms in any industry. To manage a system better, the measurability of the processes or components in the system is a vital issue for companies McAfeeBrynjolfsson et al. (2012). Advances in data acquiring and storage technology provide huge amount of data to companies, and enable them to analyze and process the acquired data to manage the system more efficiently. In 2011 alone over 1.8 zettabytes (1 zettabyte = 1021 bytes) of information created IDC (2011). Today, over 300 users generate 500 millions of tweets every day, 200 billions of tweets every year, and 1.7 billion of Facebook users responds to 4 million of posts every minute (InternetLiveStats (2016)). Information has become profuse, and been increasing geometrically every minute. Social media providers such as Twitter, Facebook stores huge amount of daily data to the extent of 10 TB. Acquired much more information from sensors or machine-to-machine data, or about customers such as demographic and transaction history information are valuable assets for companies to design their management strategies by storing and analyzing them. Growing data in the worldwide creates burden for companies to analyze data, and, then, make decision with obtained results to reach the specified targets. To process data and transform them into business advantage, data mining tools are utilized for companies to reach the manageability by succeeding in measurability as mentioned before. However, conventional data mining techniques are inadequate to understand and analyze big data. New technologies in big data enable us to extract useful information from abundant data which can be used in various areas such as healthcare, energy, social media, finance and banking etc.

Complete Chapter List

Search this Book: