Increase revenue by billions of USD / year by processing large data

Knowing how to handle big data (big data) including unstructured data from sensors, videos, social networking sites ., organizations and businesses can increase billions of dollars in revenue and improve operational efficiency .

Potential value

Talking to the Vietnamese press on the morning of April 5, 2012, Mr. Kaleem Chaudhry, Director of Enterprise Technology Division of Oracle Asia-Pacific Region, said big data (unstructured data) is being collected. attract the attention of executives (CEO) in the world by the ability to add value to organizations and businesses. Specifically, thanks to the application of processing technology for big data, the US health care system has increased the value of 300 billion USD; service providers increased their annual revenue to 100 billion USD, while the US retail industry increased net profit by over 60% per year .

Increase revenue by billions of USD / year by processing large data Picture 1
The concept and values ​​of big data have been introduced to the Vietnamese press by Oracle
In the morning of April 5, 2012, in Hanoi.

To 'show' the big data concept, Mr. Kaleem gave 3 examples.

The first evidence he raised was in a large retail store, suddenly seeing natural revenue decline, if only looking at the traditional data warehouse (structured data like word, excel), only can capture the status of ongoing transactions, which products sell better, . but it is difficult to know why the revenue is declining. However, through the collection of information related to the store is expressed through the stories of the social network's residents, such as Facebook, Twitter . (eg hot products of the store If the size is suitable for potential customers, or the product has not been offered for sale at a place where there are many people waiting .), the store can be compiled into specific reports, through which can There is an answer to the decline in revenue.

Next, Mr. Kaleem pointed out that in a telecommunications enterprise with hundreds of thousands of different switches, the technology of processing big data will exploit data from file logs, in-depth analysis to capture information of customers (for example, problems with telecommunications equipment, or fault coverage, .). Thereby, telecommunications enterprises can set appropriate ways of behaving, such as timely offering preferential packages to users .

Third, at a national security agency, big data assisted the criminal investigation department in retrieving criminal information, such as data from sensors at ticket control stations (record number of crime car plates running through), or financial transactions via electronic banking, or phone calls, .

'Mining' with an integrated device

Big data is characterized by large capacity, diverse formats, so if you want to handle big data, special processing technology is required.

According to Kaleem Chaudhry, Oracle is currently the only company with technology to handle big data. Along with 2 other products, Exadata (technology for processing structured data from many different sources) and Exalytics (specializing in analyzing support to produce reports in friendly and easy-to-use formats), Bigdata contributes Help organizations and businesses improve their information assets performance to make good and effective management decisions.

The outstanding advantage in Oracle's big data processing technology is the integration of many software into one device box, including Oracle Linux, Oracle NoSQL, Open-source R distribution, Cloudera's Distribution software (including Apache Hadoop). ), Cloudera Manager. Thanks to this integration, customers will not have to struggle to buy discrete products and calculate how to build and connect these discrete products into an integrated product that is effective.

The condition to be able to use Oracle's large data processing technology is an organization, the enterprise must have very strong hardware configuration, with specific parameters including: 216 CPU core hardware, 864 GB RAM, 648TB disk capacity; 40Gb / s InfiniBand, rack, node interconnection; 10 Gb / s Ethernet, data center connection.