第二种深入计算我们称之为数据分析处理 -- 有些人称之为商业情报处理，即从大量信息中提炼出有用情报，并分析出以前得不到或看不见的各种关系和发展趋势。

　　Banks are looking at spending patterns and other demographic data to see which customers are more profitable over the long haul. Health-care companies are analyzing millions of patient records to find hidden indicators of disease.

　　银行在分析消费模式和其它统计数据，以弄清哪些客户长期的收益状况较好。医疗企业在分析成百万的病例以从中发现隐藏的疾病迹象。

　　These tools are also helping slash the staggering cost of insurance fraud in the health-care industry, which is a hundred-billion-dollar problem in the United States alone. Insurance companies can now spot every billion practices. One company in the United States has saved 38 million dollars, having invested only 400 thousand in this technology.

　　这些工具还有助于削减 health-care 产业中惊人的保险欺骗损失，仅在美国这个损失就高达上千亿美元。保险公司现在能够发现任何欺诈行为。有一家美国公司仅在这种技术上投资 40 万美元就减少了 3800 万美元的损失。

　　In one instance they found a doctor, who was sending it a bill once a week for a procedure that particular - usually was done once or twice in a life time. At some times the patterns and relationships that are uncovered are truly baffling. One retail chain discovered the following correlation -- for whatever reason, new fathers buy disposable baby diapers and beer on the same shopping trip. This led to many, many thoughtful ideas not at least which was they never discount diapers and beer on the same day.

　　有一次他们发现一个医生每周都要送来一份医疗检查帐单，而其中的一些项目一生也只会做一次或两次。有时发现的一些情况和关系着实令人难以理解。有一家零售连锁店发现了如下的相关联系：不管出于什么原因，新爸爸们总是在一次购物中同时购买一次性婴儿尿片和啤酒。这一点引发了很多很多很有思想的想法，例如他们从不同时对尿片和啤酒打折。