Process Innovation

Big Data Analytics

Big Data Analytics is the use of a large collection of data gathered and collected from inside and outside the company. Making use of such datasets is generally a very complex thing to do and using traditional processing applications may not be enough. This gap in the traditional processing applications has actually stimulated the burgeoning and growth of multiple companies, interested in capitalizing on Big Data Analytics.

There are several definitions of Big Data Analytics. This can create complexity, given the presence of complex linkages and hierarchies among all data (Troester 2012). Academic literature does not agree on one unique definition of Big Data Analytics. Three different perspectives (Hu et al. 2014) are possible:

  • • According to the attributive definition, “Big Data technologies describe a new generation of technologies and architectures, designed to economically extract value from very large volumes of a wide variety of data, by enabling high-velocity capture, discovery, and/or analysis” (Carter 2011).
  • • Based on the comparative definition, instead, “Big Data are datasets whose size is beyond the ability of typical database software tools to capture, store, manage, and analyze” (Manyika et al. 2011).
  • • The architectural definition cites Big Data as projects “where the data volume, acquisition velocity, or data representation limits the ability to perform effective analysis using traditional relational approaches or requires the use of significant horizontal scaling for efficient processing.” (NIST)[1]

Big Data Analytics provides opportunities in existing environments. It also creates new opportunities for financial institutions’ stakeholders. These opportunities were not possible by dealing with structured content in traditional ways. Big Data Analytics has three characteristics—the so- called 3 Vs:

  • • Volume: The quantity of data should be relatively large. The word “relative” refers to the organization: a small organization might consider as Big Data Analytics a relatively lower volume of data with respect to large organizations. Big Data Analytics refers to the large—and exponentially growing—amount of data flooding in and out of every financial services company and that have been internally generated. Examples of these can be found in a variety of sources, including:
    • - the structured granular call detail records (CDR) in a call center;
    • - the detailed sensor data from telematics devices, such as personal computers (PCs)s, mobile, ATM, Point of Sale (POS), and so on;
    • - external information, including open data, marketing research, and other behavioral data;
    • - unstructured data from social media, reports of different types, and so on.
  • • Velocity: Financial institutions must be able to process, access, analyze, and report huge volumes of information as quickly as possible in order to make timely decisions, especially in the operational environment. Financial institutions also need to (Bhargava 2014)
  • - reduce latency to optimize transparency, cross-selling, and upselling in the different channels;
  • - provide quick enterprise Intranet documents search to study the impact of different events and decisions;
  • - decrease the business delivery time for reports in a data warehousing environment. There is the need of resources and solutions for fast processing of the data, in such a way that they cannot “age” too much:
  • - clickstreams and ad impressions capture user behavior at millions of events per second;
  • - machine-to-machine processes exchange data between billions of devices; and
  • - infrastructure and sensors generate massive log data in real time.
  • • Variety: The majority of organization’s data (estimated on average around 85%) is unstructured. This means that further elaborations are necessary in order to analyze data that do not flow into the organization in a constant manner; peak loads may occur with daily, seasonal, or event-triggered frequencies. Furthermore, different sources may require different architectures and technologies for the analysis (audio, text, video, and so on). Data can come from disparate sources beyond the usually structured environment of data processing. It would include mobile, online, agent-generated, social media, text, audio, video, log files, and more. Big Data Analytics is not just numbers, data, and strings. Big Data Analytics is also documents, geospatial data, three-dimensional data, audio, photos and videos, and unstructured text, including log files and social media. The processing of such variety of information is not easy. Traditional database systems address smaller volumes of structured data, fewer updates with a predictable, consistent data structure. In general, it is possible to classify Big Data Analytics as:
  • - Structured: Most traditional data sources are structured.
  • - Semi-structured: Many sources of Big Data Analytics are semi-structured.
  • - Unstructured set of data: such as video data and audio data.

The analysis of unstructured data types is a challenge. Unstructured data differ from structured data in that their format, which varies widely. They cannot be stored in traditional relational databases without significant effort at data transformation. Sources of unstructured data, such as email, word documents, pdfs, geospatial data, and so on are becoming a

relevant source of Big Data Analytics, and also for financial institutions. There are three other Vs that are important to consider:

  • • There should be a concern about the “veracity” of data. It refers to the messiness or trustworthiness of the data. With many forms of Big Data, quality and accuracy are less controllable. The quality, dependability, reliability, and consistency of data are critical issues for financial institutions looking to extract from data meaningful information to support their decision-making processes. The consequences are different. The impact of veracity on Big Data Analytics is much wider than on small data. In some cases, such as in voice-to-text conversions or social network conversations, data quality can result in meaningful information. This is true especially if financial institutions are trying to analyze macro-level phenomena, such as in sentiment analysis.
  • • “Vulnerability” is also important. Due to the variety of Big Data Analytics, ensuring data privacy for unstructured data might be a challenge.
  • • Last but most important, “value” refers to the ability to turn the data into value. Value for the customer is the most important of the Big Data Analytics’ characteristics. If the customer finds value in the relationship with a financial services company, the value should accrue also to the company. Big Data Analytics use should add value for the customers and the organization. Financial institutions that adopt customer-centric approaches can get valuable insights from data analysis. It is important that financial institutions make a case for any attempt to collect and leverage data. It is easy to fall a victim to the latest fashion and launch Big Data Analytics initiatives without a clear understanding of its business value. In order for financial institutions to derive true value from Big Data Analytics, they must enable innovations in products, processes, organizations, and business models.

Scholars and practitioners have identified the main challenges of Big

Data Analytics governance in (Cavanillas et al. 2016):

  • • Analysis
  • • Capture
  • • Data curation and quality
  • • Querying
  • • Data security and privacy
  • • Search
  • • Sharing
  • • Storage
  • • Transfer
  • • Visualization

This chapter focuses on the extraction of value from data: large amounts of structured and unstructured data contain a variety of useful information that managers can use in pursuing their objectives in a more efficient, effective, and economical way.

In the survey “Big Data in Big Companies” (2013), Tom Davenport interviewed managers from more than 50 businesses in an effort to understand the ways through which companies create value. According to Davenport, Big Data Analytics allow significant cost reductions not by simply bringing cost advantages. It also helps in identifying new paths and ways for doing business. Implementing Big Data Analytics implies better decision-making processes, with reference to both time and quality; decision-makers have the opportunity to analyze new sources of data in a faster way. That could end up in the discovery of completely “uncharted oceans”, as new markets, products, or services. Big Data Analytics can help also in cross-selling and risk management (see Fig. 4.5).

  • [1], Accessed 06 August 2016.
< Prev   CONTENTS   Source   Next >