The 10 top misconceptions about Big Data

This article is intended to deal with this and a few other misconceptions surrounding Big Data and thus encourage Big Data projects and bring more clarity to the subject.

1. only Social Media–Data are Big Data:

Even if the general discussion about Big Data Social Media-While the most commonly cited data sources are probably the Internet and other channels, there are many other sources as well.

For example:

• Bio/medical Data from test series
• Data of technical networks
• Data of transmitters, such as RFID tags
• Geodata to flood-prone regions
• Machine or sensor data
• Traffic data
• Weather data
• Audio data from call centers
• Image data from video surveillance.

2. only in the area of CRM are there business cases:

The focus on CRM (Customer Relationship Management) as the business Case for Big Data analytics is closely related to the focus on Data from Social Media-channels.

However, the Big Data-supported Analysis of customer relationships is not of equal benefit and added value for every company.

Which business cases actually come into question for a company depends on the company's industry, the industry focus, the company's business model, the corporate strategy, the IT strategy, the BI-Strategy etc.

3. Big Data is only of added value for trading companies:

Also in the context of CRM as the business Case stands the Acceptancethat Big Data is only of added value for retail or media companies.

Already the reference to the many Big Data sources refutes these Acceptance. For example, pharmaceutical companies could provide medical Data from test series as a Big Data source for information on the mode of action of drug combinations.

Insurance companies could Geodata on wind and weather conditions as a Big Data source for assessing risks to be insured.

Manufacturing companies could use sensor data to proactively maintain their production equipment.

So there are plenty of potential business cases for non-trading companies as well.

4. Big Data has added value only for large companies:

In the current discussion about business cases for Big Data arises in part because of the companies Amazon mentioned as Big Data users, Apple, Facebook or Google the impression that the Analysis from Big Data is only of use to very large companies and, because of the resources required, can only be implemented by them.

The size of a company is neither a decisive factor for value-added business cases nor a guarantee for the successful implementation of Big Data projects.

Smaller companies, in particular, are very well placed to develop innovative business cases around Big Data stand.

And the resources that are supposedly required can now also be deployed via SaaS (Software as a Service) and IaaS (Infrastructure as a Service) models. So: no obstacles for smaller companies per se.

5. Big Data is an IT issue:

Big Data projects should be initiated by the business departments of a company. These know the Business processes and can demonstrate the added value of the Analysis from Big Data assess accordingly.

In this regard, it is advisable to look for the business Case to also make use of "fresh" suggestions from outside the company. In any case, the business cases should fit the company's business model.

Big Data is therefore primarily a specialist-driven issue and not a technological end in itself.

However, it would be just as big a mistake to conclude that IT should be "left out" of Big Data projects or, at best, kept informed.

The Analysis from Big Data is made possible by suitable tools. Here, IT supports the business departments in the selection of technology and tools.

The early involvement of IT and close cooperation between the business units and IT can avoid a costly proliferation of tools and are therefore essential.

6. Big Data requires IT investments in any case:

Big Data does not always require investments in new software and infrastructure. To Analysis from Big Data For example, the technology that has been established over many years Data Mining come into play.

Data Mining can serve as the technology of choice for business cases around text analytics, predictive analytics, and preventive analytics.

At least Data Mining-basic functionalities are already provided today with some BI-platforms offered "out-of-the-box".

If bottlenecks arise on the company's internal infrastructure due to the volumes of data to be analyzed, the IaaS models already mentioned offer themselves as a measure.

An exact Analysis Nevertheless, it is imperative to identify potential resource requirements, including those in the technical area.

7. there is a lack of innovative visualization techniques:

The Visualization unknown contexts and facts for an ever larger as well as heterogeneous target group is indeed a challenge.

In the meantime, however, there are promising approaches, such as tree maps, to provide context and facts in Big Data to visualize.

This is clearly illustrated by the "needle in a haystack": Imagine we are standing in front of a haystack and are supposed to look for the much sought-after needle - hopeless!

How helpful would it be if this "foreign object" from the haystack's point of view were marked in some way, and ideally if we were still made aware of this marking?

Also a combined Visualization can be one approach. Why not, for example, think about a 3-DVisualization of a car the presentation of warranty and goodwill costs of individual components or component groups?

Interrelationships are thus much more clearly recognizable and can be analyzed more interactively than in a column report grouped by components.

8. the mass alone makes Data Big Data:

Big Data do not become "Big" data simply because of their quantity.

Usually "the three Vs" Volume, Variety and Velocity are used for the definition of Big Data consulted.

Here, the first V, "Volume," confirms the obvious association of very large volumes of the most detailed Data - but is only one feature of several.

9. unstructured alone Data are Big Data:

The second V, "Variety," in addition to "Volume," characterizes the diversity of the Data and comprises variously structured, semi-structured and unstructured Data.

10. real-time processing Data are Big Data:

The third V, "Velocity", characterizes the frequency and duration of the data flow. In this respect, real time data are to be processed Data one of the possible characteristics of data delivery intervals for Big Data.

Conversely, however, not all real time delivered data are necessarily Data Big Data.

The following are a few examples:

For the operational reporting of order transactions, all changes from the transactional source system are sent immediately after the change to a BI-system supplied.

Although this results in a continuous Data stream - the supplied Data are clearly structured and the reporting only takes these into account. Data => no Big DataAnalysis.

The previous data delivery is supplemented with "O-Ton" comments from the customer during the order process in the call center.

These comments are unstructured and thus different to analyze, but are intended to be included in a more comprehensive Analysis of customer data, order processes and customer satisfaction.

In this scenario, both structured and unstructured Data for the Analysis is used, also the Data flow not continuous or regular.

At the end of the day, only a company-specific view can determine the added value of using Big Data decide

A serious examination of the topic is recommended in any case.