Big Data Analytics - Literature review Example

Introduction The term “big data” is normally used as a marketing concept refers to data sets whose size is further than the potential of normally used enterprise tools to gather, manage and organize, and process within an acceptable elapsed time. In fact, the size of these huge data sets is believed to be a continually growing target. Additionally, the size of big data is presently ranging from a few dozen terabytes to a number of petabytes of data in a single data set (Josyula, Orr, & Page, 2011, p. 89). This paper presents a detailed analysis of big data analytics. This paper will discuss various aspects associated with big data analytics. This paper will discuss some of the uses and issues of big data analytics. “Big Data” as its name indicates is a collection of huge amounts of formless and meaningless data which are generated by high-quality and heavy software applications belonging to a varied group of software applications such as social networks, a wide variety of scientific computing applications, medical information systems, e-government applications, and many more. The research has shown that data that is used and processed by these different software applications share some common attributes. Some of these common characteristics can include large-scale data (which defines the distribution and size of data stores), scalability issues (it define the functionalities and features software applications processing across-the-board, huge data repositories such as big data), ensuring and maintaining advanced Extraction-Transformation-Loading (ETL) processing on low-level, unstructured and meaningless data to some extent meaningful information; designing and implementing straightforward and understandable analytics over big data stores with the purpose of attaining intelligence and extracting valuable facts and information from them. Additionally, in the past few years, analytics over big data stores has caught the attention of researchers and organizations. In addition, the research has shown various application areas where these analytics can play a significant role. In this scenario, scientific computing is believed to be one of the most important application areas for the reason than in this domain academic researchers and scientific create huge amounts of data every day in the results of their experiments and tests (for instance consider fields such as astronomy, high-energy physics, biomedicine, biology and many others). On the other hand, extracting valuable information and knowledge for different useful tasks on the basis of these huge, comprehensive data stores seems to be impracticable for common database management systems and other similar analysis tools (Cuzzocrea, Song, & Davis, 2011; Lopez, 2012). Figure 1Big Data Process In this scenario, figure1 demonstrates the process of big data analytics. First of all data is collected from different sources. As discussed above these sources vary from social networks to different information systems and web applications. Hence, the size of this data is so huge that it is difficult to measure. In this scenario, understanding and using this data for useful tasks is almost impossible. Therefore, there is a need for a framework that could help users understand and make effective use of this data. For this purpose, there are a number of frameworks and users can select a suitable framework according to their needs and requirements. After selecting a framework, this framework is applied to data and some coding is done. After that the users can obtain results that they can use to drive decisions and perform the desired operations (Fisher, DeLine, Czerwinski, & Drucker, 2012; Lopez, 2012). Though, the term “big data” is used in different ways in different disciplines. However, in their paper (Chaudhuri, 2012) define some common characteristics of the big data idea as they have to do with analytics: Investigating unstructured data and text to determine if these sources can be used to get useful information for particular purpose (depending on the field for which this data is being collected). Minimizing the time difference between data collection and performing operations on these data to make is effective, which is sometimes known as near real-time business analytics. Making use of these analytics to drive business decisions further than the capabilities provided by the traditional business intelligence (BI) stack. Searching for and implementing economical, extremely scalable analytics frameworks. Without a doubt, with the big data a wide variety of organizations get a wonderful opportunity to control a number of sources of data (comprising both unstructured and structured forms of data). Additionally, web data are based on exclusive characteristics for the reason that its scalability, its numerical redundancy and accessibility of user responses and feedback (through click information and query logs) have made the collection and extraction of easy to use data (such as entities) from web data particularly attractive. In fact, a wide variety of tools and techniques for entity extraction have already been productively utilized for determining sources to different product labels, people, or locations mentioned on websites. For instance, in the enterprise applications voice of the customer are used to discover opinions, sentimental knowledge and useful trends against a particular set of products and services available on a web site and blogs. In this scenario, these kinds of applications allow business organizations to develop products in the light of the results of these applications. These results are derived from these huge amounts of data. There is another useful instance of a functional service that can include product data conflation derived from web information, for instance identifying frequent synonyms of products from the click information and web query log. Without a doubt, the identification of these high class services, providing a huge group of resulting data assets, and offering easy to use tools and applications for information extraction collectively can help develop a platform that is capable of effectively utilizing the social media and web data for a wide variety of applications (Chaudhuri, 2012; Kumar, Niu, & Re, 2013). In addition, one of the most important advantages of big data is it allows the business organizations to identify and discover deep insights in data that can be effectively used for decision making. In this scenario, machine learning is believed to be one of the most useful technologies that can be used to open these insights. In fact, machine learning has already been used for a long time in a wide variety of areas (for instance internet search, fraud detection and marketing and advertising) (Chaudhuri, 2012; Kumar, Niu, & Re, 2013). At the present, in order to make effective use of these huge amounts of data, business organizations and users should possess adequate knowledge of data querying and implementing latest techniques to explore data. However, they face some difficulties and challenges while competently searching for deep insights in data. Some of the challenges that they face in this process can include various questions such as, how to recognize applicable sets of data without difficulty from a large number of data sources, what and how to make use of data cleanup tools and techniques for instance estimated links between different data sources, deciding the techniques to select samples and results of a query gradually, and how to get effective demonstration? Additionally, the development and effective use of these applications requires effective systems skills and various algorithmic issues and problems in each of the above mentioned challenges (Chaudhuri, 2012; Kumar, Niu, & Re, 2013). In the same way, (Labrinidis & Jagadish, 2012; Rabl, Sadoghi, Jacobsen, omezVillamor, MuntesMulero, & Mankovskii, 2012) discuss a number of challenges that users and organizations can face while using big data analytics. According to their viewpoint in view of the fact that big data is based on a collection of data from various sources and some sources such as sensor networks, can create astounding amounts of unstructured data. However, it can be purified and condensed by orders of scale. In this scenario, one biggest challenge for data scientists is to establish these purification and compression processes in such a way that they do not miss valuable facts. In addition, mining of these huge amounts of data can require dirt free, integrated, resourcefully available and reliable data, declarative mining and query interfaces, effective mining programming, and big data processing setting. In this scenario, one of the serious problems is the lack of harmonization among various database systems used to maintain the data and offer SQL querying (Labrinidis & Jagadish, 2012; Rabl, Sadoghi, Jacobsen, omezVillamor, MuntesMulero, & Mankovskii, 2012). At the present, there are a wide variety of effective and efficient tools that can used to deal with big data related problems and issues. For instance, (Cheng, Qin, & Rusu, 2012) discuss a tool known as Glade, which presents an easy to use interface to perform a wide variety of investigative actions and a dedicated runtime for aggregation. In this scenario, the interface consists of a standard UDA interface at the same time as the run-time looks like the relational aggregate operator. In addition, this tool provides users with an extended capability of a database system with dedicated aggregation operators (Cheng, Qin, & Rusu, 2012). Some other tools comprise IBM's Jeopardy-winning Watson system, Google's Knowledge Graph, Apple's Siri, and the systems recommended by Amazon and Netflix (Kumar, Niu, & Re, 2013; Manadhata, 2012). Without a doubt, big data analytics seem to be an up-and-coming type of knowledge work, which offers a large number of capabilities and opportunities for both academic and business perspectives. In fact, it is also useful for those who do not want to collect huge amounts of data; the importance of big data analytics cannot be unobserved. In addition, it is believed to be a significant new path towards learning about how people use computing. In fact, the organizations are already making use of big data in the form of product usage data collected from tens of thousands to millions of customers. It is expected that in the future the use of big data will further grow for the betterment and improvements of different disciplines (Fisher, DeLine, Czerwinski, & Drucker, 2012; Lopez, 2012). In conclusion, this paper has presented a detailed analysis of a latest emerging IT concept known as big data analytics. It is an admitted fact that organizations heavily depend on huge amounts of data for deriving useful facts in order to improve their business performance. In this scenario, these data are collected from a variety of sources such as social networks and other information systems. Though, in this process they can face a number of serious problems. However, there are a number of tools that can be used to effectively deal with these problems. The research has shown that in the future the use of big data analytics will further grow and expand in other areas of life. References Chaudhuri, S. (2012). What next?: a half-dozen data management research goals for big data and the cloud. PODS '12 Proceedings of the 31st symposium on Principles of Database Systems (pp. 1-4). Scottsdale, Arizona, USA: ACM. Cheng, Y., Qin, C., & Rusu, F. (2012). GLADE: big data analytics made easy. SIGMOD '12 Proceedings of the 2012 ACM SIGMOD International Conference on Management of Data (pp. 697-700). Scottsdale, Arizona, USA: ACM. Cuzzocrea, A., Song, I.-Y., & Davis, K. C. (2011). Analytics over large-scale multidimensional data: the big data revolution! DOLAP '11 Proceedings of the ACM 14th international workshop on Data Warehousing and OLAP (pp. 101-104). Glasgow, Scotland, UK: ACM. Fisher, D., DeLine, R., Czerwinski, M., & Drucker, S. (2012). Interactions with big data analytics. Interactions, Volume 19 Issue 3, pp. 50-59. Josyula, V., Orr, M., & Page, G. (2011). Cloud Computing: Automating the Virtualized Data Center (1st ed.). New York: Cisco Press. Kumar, A., Niu, F., & Re, C. (2013). Hazy: Making it Easier to Build and Maintain Big-data Analytics. Queue - Web Development, Volume 11 Issue 1, p.30. Labrinidis, A., & Jagadish, H. V. (2012). Challenges and opportunities with big data. Proceedings of the VLDB (pp. 2032-2033). Istanbul, Turkey: ACM. Lopez, X. (2012). Big data and advanced spatial analytics. COM.Geo '12 Proceedings of the 3rd International Conference on Computing for Geospatial Research and Applications (p. 5). New York: ACM. Manadhata, P. K. (2012). Big data for security: challenges, opportunities, and examples. BADGERS '12 Proceedings of the 2012 ACM Workshop on Building analysis datasets and gathering experience returns for security (pp. 3-4). Raleigh, North Carolina, USA: ACM. Rabl, T., Sadoghi, M., Jacobsen, H., omezVillamor, S. G., MuntesMulero, V., & Mankovskii, S. (2012). Solving big data challenges for enterprise application performance management. Proceedings of the VLDB Endowment, Volume 5 Issue 12, pp. 1724-1735. Read More