Big Data – Part I – Transforming Information Architecture

Big Data is another theme that frequents technical forums, social networks, computer training courses and is part of the portfolio of many technology companies. With this, of course it is on the priorities of corporate managers as a present or future investment. Much has already evolved in the context of Big Data, but will we examine the extent of what this innovation can bring? Are we prepared to act upon the benefits of Big Data, translated into facts for businesses and individuals? Does our public administration, our consultants, and professionals view Big Data in a comprehensive way?

In fact, we have some means to understand Big Data, or even, different approaches to its study. All must be seen so that we can measure its actual impact on the present and the future. There is a path that would be the technological approach, and that is where many begin, due to a software designer or system architect. This can be done as an evolution of understanding of Business Intelligence and Data Warehouse software; or as an evolution of the study of data management, databases, SQL and numerous extensions; or as administration of large-volume systems of unstructured data such as Hadoop; or finally via the study of MapReduce technology and distributed systems.

They would be paths through technology, usually traced by those who come from foundation software for Big Data and who seeks to acquire information from this new world from their basic knowledge.

Another path is usually traced by Telecom or microelectronics engineers, by statistics, and by the new strand of Data Scientists from diverse backgrounds and even a business analyst with an interest in innovation and a certain technical bias. In this case, they will probably be interested and begin their study with more or less recent techniques, but today components of the Big Data world as Predictive Models, Pattern Recognition and Classification, Regression, Simulation, Genetic Algorithms… These terms are more familiar to and statisticians who choose to merge their knowledge with innovative techniques. This would be the case then for engineers who want to evolve techniques of signal recognition and streaming, or analysis of social networks; information analysts who investigate neural networks, crowdsourcing, and self-learning.

Anyway, there is a good set of techniques that today participate in the study of Big Data will need to be supported by systems or software, or by technology. Thus, these techniques will give these professionals the means to operationalize their innovation demanded by some research, by the company, or by the evolution of their knowledge.

A third way is done by technology managers or companies, by journalists, by innovation analysts or consultants, and even by sociologists, economists or anthropologists interested in evolution that is more human-related. In this case, a more business-oriented approach, behavior, and the study of socioeconomic tendencies are involved.

Big Data presents transformational impacts on all these items. This interest may be due to a need to know the whole (as we are doing now), by researchers or journalists, as well as innovators in specific areas such as health, public administration, retail, manufacturing, etc.

There is a vast and interesting field on this path. You can examine the advantages to a business, public service, research or academic function—on how to have a range of data previously never acquired, ready for use, worked in an intelligent way and crossed with many other rich sources, also handled. Thus, someone in the healthcare area can work in the research area with richer data, since Big Data will bring information previously unavailable and now captured from various sources, either from anonymous statistics of health wearables, or consumption of certain drugs in social segments, by age, etc., or from social media with behavioral data, anonymous results of tests, treatments, medical records, etc.

         

As they say, an augmented reality for various activities. As intended in BI, offer data for analysis and decision around health with much more information, but only achieved with the resources implemented with Big Data.

We can think of several applications, for example, the public administrator can do an improved management according to the capture of all this information. Sources of social media, geospatial and localization, intersection with use of private services and analysis of feeling (subjective part of texts) will be made available and treated for the best decision of the public manager.

With objectives of better service delivery, winning competitors and gaining more profit, providing more information and innovating in your area from Big Data is a path taken by those who are looking at Big Data for a differentiated solution to an existing problem.

Not only do the new basic technologies for Big Data allow its advent, another connected world as we have seen in other chapters. From the billions of connected mobile devices, a lot of people provide valuable information on Facebook, Twitter, LinkedIn or Google Plus, a generation of knowledge whose maintainer is the network itself, mashup services that gather information previously unthought, an extensive and unprecedented menu.

We still add computing power, cloud computing, and a new vision of features as services provide the ingredient for Big Data to prevail.

In any of these ways, we realize that deploying Big Data in a company is no easy task. Existence of talents, a business process solution orientation, a well-aligned vision of the team of managers and clear business goals are needed above all else. Bringing Big Data into a company is not the same as a Data Warehouse, and it needs to structure itself for this new demand and to stay competitive.

So, if you are involved in creating a center of excellence in Big Data in your company, you want to do better over an existing area, or even be responsible for creating initial studies for a future project, have in your agenda some key points:

• Have a very focus on creating value through the Big Data differential. On the one hand, evaluate the most advantageous points for your organization with a focus on business processes. Let’s say that the moment your company cuts operating costs by process improvement, investing in new marketing channels may be the priority. Whether or not, improving the sales force’s business process and logistics is a priority. Being aligned with organizational goals and knowing the processes that support them is very important for projects that will make a difference. After that, understand what potential data sources can be worked on in your Big Data architecture. They can be external data from social networks and crowdsourcing applications of the market to give quality to your marketing process. It can be a geo-locator application connected to customer service to leverage the business process. It can be a treatment of the operational and collection of M2M information in an analytical context. Anyway, do not want to embrace everything, but choose a combination of the priorities of your company with what would be feasible or better—a quick return project (quick-win). Since a Big Data project tends to always have good complexity (especially the first one), it needs to be well planned.

• Gather talent in a company excellence group. Big Data can and should have participants from other departments, a staff that is analytically focused and open to innovation, as well as supporting operational areas such as IT infrastructure. But it is important to have permanent Champions in the team. You do not need to start with a large group, but having someone focused on data intelligence, analytical models is important. Today there are many experts known as data scientists in this regard. A person who can be a focal point to the external and internal sources that will feed Big Data, usually with a bias of the traditional managed data will assist in the processing and storage part of semi-treated or untreated data. Get used to giving a communication plan, when it comes to something new, with a certain time of return on investment, this will protect you in alignment with the goals outlined.

• Raise the tone of Big Data in the best sense. In line with company goals and creating a return expectation, link your goals to your company strategy. And even change the strategy when you understand you can. Working with Big Data is having new insights and you will be one of the first to be able to communicate them. Get used to thinking not as an IT professional, marketing, operations, but a business person. Of course, this will bring good fruit, but if you support Big Data with something that it can really provide you, have a differentiated strategy for your company in the market.

• In addition to business processes, create a data-driven Big Data process. You need to understand which of the data you are working on are private and whose business you need to protect. Which of these are confidential. What is public? What will need to be complemented with information providers, through integrated external applications or services? Which information partners do you need to support? What are the sources of all this information? What retention, storage plan, security policy guide data? What is the contingency of provision in case of external sources of partners? How much do you really need to store, or cannot, let for a period in a stream with Cloud Computing? Lots of options and details… you need to let your information flow well clear and documented, engage the organization around a common goal.

(to be continued)

.                       

[wp_ad_camp_1]