Big Data and Big Data Analytics

Big Data and Big Data Analytics is important when the rapidly growing data volumes make companies to face new challenges and forces to think to choose options. The hot topic of Big Data provides currently not only encourages curiosity to know more, but also in practice creates confusion within some companies: With the given gripping reports on the use of Big Data is not Could Big Data method also: Under what circumstances is big data analytics is to recommend ?

Big Data, Big Data Analytics and Data Warehouse

Big Data for the IT decision makers brings the essential question : Can the benefits that Big Data promises, even with the existing technologies, in simpler means – like using a data warehouse can solve the problems.

Data warehouse systems contain data sets collected periodically from the transaction systems, filtered and aggregated for analysis. A data warehouse is often filled in daily a huge load of operations with these data and therefore has a latency period. The Pivot is the most widely used analytical tool in the field of data warehousing. Often Pivot tables represent the characteristics of quantitative variables according to qualitative criteria in tabular format – they are broken down by products and by sales regions. The use of other statistical methods – Keyword Data Mining – is in the Data Warehouse possible but not as common as the analysis with the pivot tables. The main reason is the aggregated, coarse-grained data that statistically hide the rewarding properties and exclude based on criterion. This is another important reason is the complexity of the data mining process.
 

Big Data and Big Data Analytics

The definition and criteria of Big Data are different depending on the source. But there is one point all usually agree : For Big Data, the data sets are no longer tagged as efficient to be managed by the conventional means. This does not include aggregated values ??obtained in real-time data obtained for example from transaction systems, scientific experiments, simulations or sensors.

In addition, the term Big Data Analytics has been established. The latter include analytical methods to gain insights from large data sets. The special methods from statistics, marketing and information technology are merged to advantage in big data analytics. Big Data systems use the pattern recognition to identify trends and patterns in real time and to discover previously unknown or suspected relationships between individual parameters. Systems that can be checked for anomalies, for example, can be used to identify potential credit card fraud in real time. Thus, thousands of credit card transactions per second can be checked immediately. Unlike traditional data warehouses, they are real-time data warehouses. By aggregating data, the data is coarse-grained, as with data warehouses only limited statistical analysis are possible. The result : the detection of trends, patterns and relationships in data warehouses is rather crude and limited to longer latencies.

The fine granularity and low latency of data resources of big data are important prerequisites for segmentation in real time, such as when making online purchases. Data warehouses also offer the possibility for the formation of segments. However, these are coarse-grained and have a longer latency than the segments of big data systems.