The rise of the concept of big data seems to be just yesterday, but thanks to this era of rapid development, we can already see many mature big data application tools. In a very short time, we can accurately locate and analyze in the vast ocean of data, and get the results we want. Of course, the progress of these technologies is not driven by banks. Large retailers, online shopping malls and technology companies of various categories are the leaders of big data. However, after their exploration, big data has also opened a door for banks. The door to precision marketing. In the long run, if banks can make full use of the advantages of big data, they can make significant progress in market segmentation, customer service, customer research, product development, product testing, etc., and to some extent completely change the way banks serve customers. , methods and channels for selling products. Of course, the premise of all this is that banks can find the right methods and tools to enter the big data era. For banks, using correct quantitative models and analysis methods to meet their current business needs is the key to rationally utilizing big data to achieve more economic returns. The experience of other industries has proved that big data is good, but if the data cannot be effectively screened and used correctly, it will only end up losing the wife and losing the army. In particular, banking is a relatively special and sensitive industry. It is unrealistic to completely carry out the so-called big data revolution at the global level. The correct approach is to start with small specific businesses and key nodes, so as to be able to be integrated into the bank's existing management structure and external regulatory mechanisms. Accept the method and gradually incorporate big data into the bank's operating system. For example, currently the banking industry is generally worried about two things: retaining customers and meeting customer expectations. For these two problems, sentiment analysis and behavior prediction under the big data mechanism can play an unexpected role. Analyzing Customer Sentiment Traditional methods of collecting and investigating customer opinions are often based on one group. Through surveys and research on some groups of customers, banks can obtain all aspects of their customers. With the advancement of the times, this method is no longer effective in obtaining the latest trends in customers' financial consumption and discovering customers' hidden needs. The most fatal point is that this method of collecting customer information and data often takes longer and costs more, but the final results are often unable to respond to changes in customer needs in real time. The so-called sentiment analysis refers to the collection of a customer's comments and activities on online platforms including social networks, including not only his own part, but also other friends he has recently associated with. The data thus obtained is processed through a set of scientific The designed calculation and analysis system can determine the recent emotional trend of a specific customer, provide help in predicting customer actions, and help banks designate specific response measures. Here, "emotion" does not simply represent the customer's emotional changes, but also includes the customer's attitude, stance, emotional tendencies, etc. This was something that was extremely difficult to grasp in previous investigation and analysis tools, but in this era of self-media, such information is scattered on the Internet and is extremely easy to obtain and analyze. Moreover, the methods for capturing and analyzing this data have become quite mature. From technical men who stay at home to serious academics, everyone is launching such tools. The bank only needs to choose a relatively stable technology supplier, and provide real-time feedback and integration of the results into its own system. It can determine the customer's reaction to the bank's products, services, pricing or policy adjustments at the first time, and take appropriate measures. Respond appropriately. If the customer's reaction is favorable to the bank, the bank can intervene in time to guide the customer's emotions to achieve better service and sales; if the customer expresses unfavorable emotions towards the bank, the bank can also detect it in time and actively deal with it. Further enhance customer service experience. Here are a few examples of customer statements that banks must pay attention to in a timely manner: "XXX Bank is indeed very useful for small and micro businesses, but it lacks suitable same-day account delivery service!" "XX Bank's online account balance check The functions are indeed well designed, but some details of customer service really need to be improved.” From an ordinary person’s perspective, these are just two simple expressions of customer opinions.

When you know the customer's emotional changes and their possible purchasing needs, as long as you can deliver what the customer needs in a timely manner in a suitable way, the customer will naturally be happy to accept it. It is important to harness the power of big data in the right way. One negative effect that big data may bring is the invasion of customer privacy. The Target department store mentioned above is an example. After this incident, Target adjusted the way it sent out promotional ads: when it discovered that a customer might be pregnant, Target would still send her a brochure containing products needed by pregnant women. , but through visual layout, cross-arrangement of other categories of products, etc., accurate product recommendations can be achieved without arousing customers' resentment of being "peeped". Ultimately, with the help of big data, Target's sales grew from US$44 billion to US$67 billion between 2002 and 2010. It is worth mentioning that big data applications can also help banks achieve effective risk control. Some foreign financial institutions have already used big data to help with risk control in financial product transactions, credit card consumption, etc. Especially in products such as credit cards and unsecured loans, through models established by big data, banks can accurately know a customer's life and consumption situation, and then choose whether to issue a card/loan to him or not. Increase the limit and delay the repayment period. Once a customer behaves abnormally, the bank will be aware of it in the shortest possible time and take corresponding measures to prevent risk cases from occurring. In short, although it is not perfect yet, big data has an unlimited future.