10 large data cases that the CIO must know

Before you listen to Gartner's analyst Doug Laney in 55 minutes about 55 Big data applications, you may have doubts about whether the big data is landing. Laney's speech, like the complete works of Shakespeare, may be "less entertaining and more informative" (perhaps for technicians). This presentation is a comprehensive explanation of the 3v features of large data: produced (type), velocity (generating speed), and volume (scale). The inventor of the term describes large data in this way – dating back to 2001 years.

These 55 examples are not used to bluff, Laney's intention is to illustrate the practical application of large data, listeners should think how to make large data in their own company and promote business development. "There may be some examples that don't come from your current industry, but you need to think about how to do it," he said. "Laney said.

Here are 10 typical cases:

1. Macy's real-time pricing mechanism. According to demand and inventory, the company based on the SAS system for up to 73 million kinds of goods for real-time price adjustment.

2. TIPP24 AG's betting and forecasting platform for the European gaming industry. The company uses Kxen software to analyze billions of of transactions and customer characteristics, and then through the predictive model for specific users to carry out dynamic marketing activities. The move reduced the 90% forecast model build time. SAP is trying to buy kxen. "SAP wants to use the acquisition to reverse its long-standing weakness in predictive analysis. Laney Analysis.

3. Wal-Mart search. The retail oligopoly has designed the latest search engine Polaris for its website Walmart.com, using semantic data for text analysis, machine learning, and synonym mining. According to Wal-Mart, the use of semantic search technology has increased the completion rate of online shopping by 10% to 15%. "For Wal-Mart, that means billions of dollars." "Laney said.

4. Video analysis of the fast food industry (Laney did not name the company). The company uses video to analyze the length of waiting queues, and then automatically changes what the electronic menu displays. If the queue is longer, displays food that can be supplied quickly, and if the queue is shorter, displays those foods that have a higher profit but are prepared for a relatively long time.

5. Morton Steak Shop Brand awareness. Morton began his social show when a customer made a joke by tweeting to the Chicago steak chain to take it to Newark Airport in New York, where he would arrive after a day of work. First of all, the analysis of Twitter data, found that the customer is our regular customers, is also a popular Twitter. According to the customer's previous orders, speculated on the flight, and sent a waiter in a tuxedo to provide dinner for the client. It may sound too bizarre, but you have to look at yourself: "Do I have the ability to do that?" "Laney said.

6. PredPol Inc. PredPol, working with police in Los Angeles and Santa Cruz and a group of researchers, predicts the probability of a crime based on variants of the Earthquake prediction algorithm and crime data, which can be as accurate as 500 square feet. In Los Angeles, where the algorithm was used, the distribution of theft and violent crime fell by 33% and 21%.

7. Tesco PLC (special easy to purchase) and operation efficiency. The supermarket chain collects data from 7 million refrigerators in its data warehouse. Through the analysis of these data, more comprehensive monitoring and proactive maintenance to reduce overall energy consumption.

8. Anglo Express (American Express, AMEX) and business intelligence. In the past, Amex can only realize the hindsight report and lag prediction. "Traditional bi has not been able to meet the needs of business development. "Laney thinks. As a result, Amex began building models that were truly capable of predicting loyalty, based on historical transaction data, with 115 variables for analysis and prediction. The company says it has been able to identify 24% of its customers who will be lost in the next four months.

9. Express Scripts LC Co. Products manufacturing. The company found that those who needed to take the medicine were often the most likely to forget the drug. As a result, they developed a new product: A ringing medicine lid and automatic phone calls to remind patients to take medicine on time.

Infinity Property & Casualty Corp. 's Dark Data (dark). Laney's definition of dark data is that data collected for a single goal is often archived and idle, and its true value is not fully tapped. In certain cases, this data can be used for other purposes. The company used the cumulative claims Division report to analyze fraud cases, and through the algorithm to recover the 12 million dollar subrogation amount.

(Responsible editor: The good of the Legacy)