In the era of big data, the solution of bank BI application is discussed

Big data is hailed as a new driving force for development and creation in the 21st century, and BI (Business Intelligence) has become the most popular data application solution at the moment. According to the data, among the five industries with the highest investment in big data IT in China, the Internet is the highest, followed by telecommunications, finance, government and medical care. In the financial industry, banks are allocated the first place, followed by securities and insurance.

How to effectively apply new information technologies such as big data and cloud computing to create value and wealth and create the future is a huge opportunity and challenge for us.

The following is a detailed and comprehensive introduction to the application of bank big data.

1. Big data financial application scenarios

From the technical characteristics of big data and the application exploration of banks in recent years, the application of big data in bank business intelligence is mainly reflected in the following aspects:

In the era of big data, the solution of bank BI application is discussed

1. Customer management and analysis

Customer management applications include customer profiling, customer segmentation, and customer pricing.

In the era of big data, the solution of bank BI application is discussed

(1) Client portraits, including personal client portraits and corporate client portraits. Personal customer profiles include demographic characteristics, spending power data, interest data, risk appetite, financial data, etc.; corporate customer profiles include production, circulation, operation, finance, sales, customer data, and upstream and downstream data of related industrial chains.

Customer portraits are the foundation of customer analysis. To do a good job in customer portraits, banks should not only consider the data collected by the bank's own business, but also consider integrating more external data to expand their understanding of customers. Such as social media, Weibo, e-commerce website and other data.

(2) Customer segmentation: Customer segmentation methods can be divided into groups through various dimensions, such as entering into segmentation through customer behavior analysis. Customer segmentation is the basis of product development and marketing. Differentiated services are possible through customer segmentation, so that the products and services provided are more directly directed to specific service groups. With the basis of customer segmentation, banks can avoid a homogeneous sales strategy.

In the era of big data, the solution of bank BI application is discussed

2. Precision Marketing

Precision marketing is refined marketing based on customer management, including real-time marketing, cross-marketing, event marketing, personalized marketing, potential customer mining, and advertising.

(1) Real-time marketing: Marketing is carried out according to the real-time status of customers. If the customer happens to visit the outlet to deposit a large amount of current account, the lobby manager will recommend relevant financial products to the customer based on the captured information.

(2) Event Marketing: Marketing according to a recent event of the customer. For example, consider life-changing events (job change, marital status change, home purchase, etc.) as marketing opportunities.

(3)交叉营销:不同业务或产品的交叉推荐。如某银行借记卡发卡量为1000多万张,而贷记卡发卡量为30万多张,借记卡发卡量是贷记卡的300倍,可通过数据挖掘,分析借记卡的使用情况,特别是商场刷卡消费、网上付款较多、金额较大客户,向他们宣传贷记卡的透支、积分及活动优势,以提高贷记卡发卡量。

(4)个性化营销:银行可以根据客户喜好进行服务或者进行银行产品的个性化推荐。例如,可根据客户的年龄、资产规模、理财偏好等,对客户群进行精准定位,分析出其潜在金融服务需求,进而有针对性的营销推广。

3、风险管理

数据挖掘在银行风险管理方面也是用途很广,包括风险评估、反歁诈和反洗钱等。

(1)中小企业贷款风险评估。银行可通过企业的生产、流通、销售、财务等相关信息结合大数据挖掘方法进行贷款风险分析,量化企业的信用额度,以便更有效的开展中小企业贷款。

(2)实时欺诈交易识别和反洗钱分析。银行可以利用持卡人基本信息、卡基本信息、交易历史、客户历史行为模式、正在发生行为模式(如转账)等,结合智能规则引擎(如从一个不经常出现的国家为一个特有用户转账或从一个不熟悉的位置进行在线交易)进行实时的交易反欺诈分析。

4、业务、服务创新

银行可将客户行为转化为信息流,并从中分析客户的个性特征和风险偏好,更深层次地理解客户的习惯,智能化分析和预测客户需求,从而进行产品创新和服务优化。如银行目前对大数据进行初步分析,通过对还款数据挖掘比较区分优质客户,根据客户还款数额的差别,提供差异化的金融产品和服务方式。

5、精细化管理

精细化管理,其指导思想就是以数据为依据进行管理,包括成本核算、资本管理、绩效考核等方面,具体应用如资源配置、舆情分析等。

(1) 市场和渠道分析优化。通过数据分析,可以了解营业网点渠道的资源配置情况,高柜、低柜窗口是否饱和,自助和人工窗口比例是否恰当,应该如何进行资源优化调 整;通过大数据,银行可以监控不同市场推广渠道尤其是渠道推广的质量,从而进行合作渠道的调整和优化。同时,也可以分析哪些渠道更适合推广哪类银行产品及服务,从而进行渠道推广策略的优化。

(2) 舆情分析。银行可以通过爬虫技术,抓取社区、论坛和微博上关于银行以及银行产品和服务的相关信息,并通过自然语言处理技术进行正负面判断,掌握银行、银行产品及客户服务方面的负面信息,及时发现和处理问题;对于正面信息,可以加以总结并继续强化。同时,银行也可以抓取同行业的银行正负面信息,及时了解同行优势,以作为自身业务优化的借鉴。

6、历史数据归档管理

历史数据归档是大数据的最基本的应用,解决了传统数据平台在处理P级以上数据容量能力不足问题。历史数据管理也是各银行技术部门困扰而又未能很好实现的问题。目前很多银行借大数据技术的应用趋式,利用关系型数据库与大数据技术的互补模式,偿试利用大数据技术,以历史数据归档和查询应用为突破口,既满足历史数据归档、客户查询、司法查询以及审计之需求,又培养一批大数据技术人才。

7、征信服务

征信是为个人或企业建立信用档案,采集、客观记录其信用信息,并对外提供信用信息服务的一种活动。按业务模式可分为企业征信和个人佂信,按服务对象可分为信贷征信、商业征信、雇佣征信等。

在互联网金融发展如火如荼之际,基于大数据技术的互联网征信应运而生。对金融业,征信完善了对风险的识别、判断、评估和管理,有利于加快授信过程,分级定价,降低优质借款人借贷成本,大幅提高信贷效率,如蚂蚁小贷,放款时间基本在3分钟以内,金额从几千至几万。

8、IT治理

以数据为中心的信息化建设把IT治理摆在更加突出的位置,如何保证计算机系统的正常运营,成为保证银行业务正常运行的关键。通过大数据技术,可以从运营监控和日志分析中发挥巨大作用。

(1)运营全景视图:通过对系统的动态实时采集,建立可视化的运营管理视图,进行系统跟踪和调度。

(2)日志分析:通过收集操作系统、网络系统运行日志分析,预测未未来可发生的情况;特别是对应用系统操作日志的分析,有助于了解操作人员对系统功能的使用情况,以助于应用系统的优化完善。

二、大数据时代BI特征

大数据时代下的商业智能与传统BI相比有许多特点,具体分析如下:

In the era of big data, the solution of bank BI application is discussed

1、结构化数据及非结构化数据的处理

能够分析处理的数据更多样化,除了包括传统BI的结构化数据化外,还包括传统BI不能处理的非结构化数据,如图像、文字、语音等。这正是体现了大数据数据处理的多样性特性(Variety)。

2、分布式数据库

传统BI采用的是集中式数据库,数据规模一般只能达到PB级,而大数据时代的BI,采用了分布式数据库,数据规模从PB到TB级,能够实现海量数据处理。或者采用饭不是计算将结果数据六赞存在HIVE中通过帆软FineBI一类的大数据BI工具展现,这体现的大数据的巨大数据量处理特性(Volume)。

3、分布式-计算向数据靠拢

传统BI采用集中式数据库,数据向计算靠拢,应用中主要以离线计算、批量处理为主,而大数据BI,采用分布式云计算技术,实现了计算向数据靠拢,支持实时及离线的计算模式,能够支持在线计算。这种快速的计算、实时特点体现了大数据的快速特性(Velocity)。

4、智能决策和解读数据应用

传统的商业智能技术,只能实现“使用算法”,进行“数据统计”,通过“报表展示”,达到“看数据”目的,而大时代的商业智能,对海量数据,深度应用算法,达到找到数据的相关性,通过自动化分析,解读数据,最后进行“智能决策”,因此体现了更强大的商业价值,这正是大数据4V特征Value的具体体现。

三、大数据的应用实施探讨

从当前的银行应用来看,大数据的应用实施通常包括三个方面:基础架构建设、数据接入和应用主题挖掘。

1、基础架构建设

基础架构是支撑大数据应用的基础。基础架构建设通常包括底层组件开发、基础服务组件开发和应用服务组件开发。基础架构通常采用商业化或开源HADOOP组件来集成;基础服务组件是大数据平台的管理和服务的重要组成部份;应用服务组件通常根据应用需要进行开发和布署。例如之前文章《Hadoop技术在商业智能BI中的应用》中讲到帆软大数据BI基于Hadoop架构做的银行业的数据分析。

In the era of big data, the solution of bank BI application is discussed

2、数据接入

数据是我们分析的资源,没有足够大、足够全面的数据就没谈不上大数据分析。因此我们收集的数据不仅包括内部数据、还包括外部数据,不仅包括结构化数据,还包括非结构化数据;不仅包括静态数据,还要包括动态、快速变化数据的采集。

3、应用主题挖掘

Data mining is our ultimate goal. Therefore, we need to have big data thinking first, in order to provide fine management for our management, provide services for decision support, and serve our business economy, in order to have more business innovation and services. Innovation enables precise marketing of products and services to our customers. The data has no size, but the pattern and position of using the data are divided into sizes. Faced with the most popular application scenarios, it can be implemented from top to bottom through top-level design. For example, the "Gui Nong Cloud" recently planned by Guizhou Rural Credit Co., Ltd. is a "high-level" project, including one big data platform, one portal, and six business platforms; The pain point is the starting point. Faced with the actual problems of capital, technology and external data acquisition, we can't wait and miss the good times, and we can't be big and comprehensive, such as the loss of bank customers, where are the high-quality customers in the agency salary payment, and how to achieve borrowing. Cross-marketing between debit cards and credit cards, etc., on this basis, explore, establish an analysis model, discover the correlation of data, and put it into production applications after successful verification.

Guess you like

Origin http://43.154.161.224:23101/article/api/json?id=326316389&siteId=291194637