How can large data be used in the financial field and create value?

Here is the answer to the data from Daniel Wang hold:

First, what you call the Big data software is not very clear to people. At least in the industry less mention of large data software this argument, it can be said that large data technology architecture, may also say data mining software. But I understand that your large data software should be a series of work and systems around large data analysis and application.

Second, the financial data you mentioned, this involves a wide range of my limited contact to see a lot, such as the fund's sales data, customer holdings and transaction data, customer contact data, customer site browsing data, such as the bank involved in the account data, customer basic information data; For example, insurance companies have customers to buy insurance data and so on. Generally speaking, can be divided into the following major categories: Customer basic attributes data, customer product purchase data, customer transaction behavior data, customer preference data ... What you can do analysis needs to see what data you can get, if you can integrate the industry's third party data, you can do more digging. For example, if you are a fund company that can get users on the Web browsing behavior data, you can determine whether the user has recently paid attention to related products, there is no concern about competitors products.

Third, to talk about large data specific in the financial industry application areas, according to my limited contact, the following areas need large data to play a greater role.

1, the user credit: This is actually the earliest application of data mining one of the fields, the domestic data mining is basically based on the need for credit classification mining algorithms and development. It is an important direction to judge the user's credit risk based on the large data. In particular, many credit evaluation systems are now dependent on foreign assessment agencies, and if you can build a credit evaluation mechanism based on large data (see what data you can get), this will have a market.

2, Trading Risk control: This is different from the user credit. The original data mining can realize the user static credit evaluation, based on the large data flow processing ability can realize the user dynamic evaluation, that is, transaction risk judgment. For example, when you find that the same account carries out credit card transactions in different regions at almost the same time, the risk of the transaction arises at this time. Customers ' credit cards may be stolen and fraudulent transactions may exist.

3, the present forecast: The current internet finance is a very big specific is to break the original liquidity and profitability can not be characterized. And now a lot of "baby" can be both, in addition to innovation-related, in the technical level if you can achieve large data on the product support, will do more efficient. Specifically, the "baby" needs to meet the needs of users every day, which requires a strong reserve liquidity, less reserves, there will be a run on the stock, and the funds can not be fully utilized, can not generate more revenue. Therefore, we need to build a forecast model to realize the effective budget and management of capital demand.

4. Marketing monitoring and evaluation: this is an area that is easily overlooked because it involves specific tactical work. After most people are concerned about the final effect of marketing results, such as engaged in a customer marketing products, see the final transformation of how much, but in fact there are many links may affect the user's transformation. such as contact, such as attraction, such as the lag of consumption and so on. These need to rely on large data based on more customer-accurate answers.

5, the loss of warning: If you can get the data can be insight into users throughout the use of related products, you can insight into the potential loss of user risk and whereabouts. For example, you may find that a higher-quality customer has recently been inactive for a period of time, which can be risky, but is it a recent one without a deal? Or is there someone else he loves? This need depends on large data for insight. Users may be paying attention to or already buying competitor products during this time, which can provide greater marketing management value.

......

In fact, there are many, not much to say, no field more analytical value, depending on the core of your concerns.

Forth, on the direction of development: I think a trend has actually been achieved is the integration of financial and third-party industry resources, such as financial and internet enterprises, financial and large data holders of resource integration. No matter what industry's resource integration, an undeniable fact is the innovation of traditional financial instruments, so think about what the current financial products have problems, may be able to find innovative direction. For example, the benefits of the baby and the loss of both in fact is to challenge the traditional financial both the lack of both. At least for ordinary investors. In addition, innovation will be reflected in the specific marketing aspects, such as relying on a variety of effective contact to achieve marketing, relying on social media to achieve marketing innovation. I do not know that you have recently been concerned about no, there are some innovations in practice. For example, Taikang life based on micro-letter to achieve the love of the spread and marketing, that is, the user initiates the care of activities, sent to friends Circle, friends can send 1 yuan care, this 1 yuan is actually to help launch users to buy insurance products. You go to the micro-letter to see.