Risk prevention perception method and system based on big data and multimodal intelligent algorithm

Abstract

The present invention provides a risk prevention perception method and system based on big data and multimodal intelligent algorithms, which extracts enterprise order information to obtain actual important customers and general customers; obtains multimodal data of general customers, evaluates general customers based on the multimodal data, and determines whether there are potential important customers; if there are actual important customers, reallocates the internal resources of the enterprise; if there are actual important customers or potential important customers, searches for potential customers based on the industries to which the actual important customers and potential important customers belong; by analyzing and evaluating the customers cooperating with the enterprise, actual important customers and potential important customers can be obtained, and when there are actual important customers, the amount of internal resources allocated to the actual important customers is reduced, and when there are actual important customers and potential important customers, searches for potential customers to seek cooperation, thereby reducing the dependence of the enterprise on the actual important customers and potential important customers and ensuring the normal operation of the enterprise.

Description

Technical Field

[0001] The present invention relates to the technical field of enterprise risk management, and in particular to a risk prevention perception method and system based on big data and multimodal intelligent algorithms. Background Art

[0002] During the production and operation process, enterprises will establish cooperative relationships with purchasers, and enterprises with excellent quality and good reputation will be favored by purchasers, thus establishing long-term cooperative relationships. Among these purchasers with whom cooperative relationships are established, there will be some purchasers whose purchase amount accounts for a large proportion of the enterprise. These purchasers will become the company's major customers. The enterprise will tend to focus more business and production on major customers, and gradually become dependent on major customers, causing major customer purchasers to occupy a dominant position in transactions, and price-cutting behavior is likely to occur in subsequent transactions. Since the company's dependence on major customer purchasers is strong, if the company does not accept the price-cutting behavior of major customers, the major customer purchasers will refuse to cooperate. Since the company's focus has already shifted to major customers, if the company loses the cooperation of major customer purchasers, the company will face operational risks. At present, in the operation process, the company has not analyzed whether the customers it cooperates with are major customers, nor will it take preventive measures against its dependence on major customers, resulting in certain operational risks for the company. Summary of the invention

[0003] In view of this, the present invention proposes a risk prevention perception method and system based on big data and multimodal intelligent algorithms, which can identify actual important customers and potential important customers, and take corresponding measures to avoid operational difficulties for the enterprise.

[0004] The technical solution of the present invention is achieved in this way:

[0005] The risk prevention perception method based on big data and multimodal intelligent algorithms includes the following steps:

[0006] Step S1, extracting enterprise order information, judging whether there are important customers according to the order information, and if so, outputting the important customers as actual important customers, and outputting the non-important customers as general customers;

[0007] Step S2: obtaining multimodal data of general customers, evaluating the general customers based on the multimodal data, and determining whether there are potential important customers;

[0008] Step S3: if there are actual important customers, reallocate the internal resources of the enterprise;

[0009] Step S4: If there are actual important customers or potential important customers, search for potential customers based on the industries to which the actual important customers and potential important customers belong.

[0010] Preferably, the specific steps of step S1 are:

[0011] Step S11, extracting enterprise order information, and obtaining all customer information and customer purchase volume from the order information;

[0012] Step S12: add up all purchase quantities to obtain the total purchase quantity, and add up the purchase quantities of the same customer to obtain the customer purchase quantity;

[0013] Step S13: Calculate the customer purchase ratio based on the customer purchase volume and the total purchase volume, and output customers whose customer purchase ratio is greater than the purchase threshold as actual important customers, and output other customers as general customers.

[0014] Preferably, the method for confirming the procurement threshold is: quantifying the production capacity, staff mobility and cash flow status of the enterprise, and adding up the quantified values ​​after assigning different weights.

[0015] Preferably, the specific steps of step S2 are:

[0016] Step S21: obtaining multimodal data published by general customers on network platforms and social media, wherein the multimodal data includes text, pictures and videos;

[0017] Step S22: Obtain the number of clicks, views, and forwardings of the multimodal data on the network;

[0018] Step S23: Calculate the increase in clicks, views and forwardings, and output general customers whose increase is greater than that of similar products as potential important customers.

[0019] Preferably, the multimodal data acquired in step S21 are texts, pictures and videos posted by general users on online platforms and social media for the purpose of promoting their own products.

[0020] Preferably, the specific steps of step S3 are:

[0021] Step S31: Acquire the internal resources of the enterprise, wherein the internal resources include human resources, material resources, financial resources and information resources;

[0022] Step S32: if there are actual important customers, obtain the resource ratio of the internal resources of the enterprise occupied by the actual important customers;

[0023] Step S33: Allocate the portion of the resource proportion that exceeds the resource threshold to general customers.

[0024] Preferably, the resource threshold of step S33 is set by the enterprise according to its own operating conditions.

[0025] Preferably, the specific steps of step S4 are:

[0026] Step S41, obtaining the industry to which the enterprise's customers belong according to the enterprise's order information;

[0027] Step S42: If there are actual important customers or potential important customers, the remaining customers in the industry to which the customers belong who have not cooperated with the enterprise are output as potential customers.

[0028] Preferably, it also includes:

[0029] Step S43: Obtain the transaction content between the enterprise and its actual important customers;

[0030] Step S44: Send the transaction content to the potential customer in the form of an invitation request.

[0031] The risk prevention perception system based on big data and multimodal intelligent algorithms includes:

[0032] The actual important customer classification unit is used to extract the enterprise order information and determine whether there are important customers based on the order information. If so, the important customers are output as actual important customers and the non-important customers are output as general customers.

[0033] A potential important customer classification unit is used to obtain multimodal data of general customers, evaluate general customers based on the multimodal data, and determine whether there are potential important customers;

[0034] Resource allocation unit, used to reallocate the internal resources of the enterprise if there are actual important customers;

[0035] A potential customer search unit is used to search for potential customers based on the industries to which the actual important customers and potential important customers belong if there are actual important customers or potential important customers;

[0036] The actual important customer division unit is data-connected with the potential important customer division unit, the resource allocation unit and the intended customer search unit respectively, and the potential important customer division unit is data-connected with the intended customer search unit.

[0037] Compared with the prior art, the present invention has the following beneficial effects:

[0038] The present invention provides a risk prevention perception method and system based on big data and multimodal intelligent algorithms, which obtains actual important customers and general customers according to the order information of the enterprise, evaluates the general customers to obtain potential important customers, and the actual important customers and potential important customers are customers with which the enterprise is highly dependent. When actual important customers exist, the internal resources of the enterprise are reallocated to avoid the internal resources of the enterprise being invested in actual important customers. When the actual important customers are uncooperative, the enterprise can still ensure that its internal resources are intact to serve other customers. When both actual important customers and potential important customers exist, potential customers can be searched, cooperation information can be sent to the potential customers, and the outlets of the enterprise's products can be increased, further reducing the dependence on actual important customers and potential important customers, thereby ensuring the normal operation of the enterprise. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only preferred embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0040] Figure 1 It is a flow chart of the risk prevention perception method based on big data and multimodal intelligent algorithm of the present invention;

[0041] Figure 2 This is a flow chart of step S1 of the risk prevention perception method based on big data and multimodal intelligent algorithm of the present invention;

[0042] Figure 3 This is a flow chart of step S2 of the risk prevention perception method based on big data and multimodal intelligent algorithm of the present invention;

[0043] Figure 4 This is a flow chart of step S3 of the risk prevention perception method based on big data and multimodal intelligent algorithm of the present invention;

[0044] Figure 5 This is a flow chart of step S4 of the risk prevention perception method based on big data and multimodal intelligent algorithm of the present invention;

[0045] Figure 6 This is a schematic diagram of the risk prevention perception system based on big data and multimodal intelligent algorithms of the present invention;

[0046] In the figure, 1 is the actual important customer classification unit, 2 is the potential important customer classification unit, 3 is the resource allocation unit, and 4 is the potential customer search unit. DETAILED DESCRIPTION

[0047] In order to better understand the technical content of the present invention, a specific embodiment is provided below, and the present invention is further described in conjunction with the accompanying drawings.

[0048] See also Figures 1 to 5 The risk prevention perception method based on big data and multimodal intelligent algorithm provided by the present invention comprises the following steps:

[0049] Step S1, extracting enterprise order information, judging whether there are important customers according to the order information, and if so, outputting the important customers as actual important customers, and outputting the non-important customers as general customers;

[0050] The specific steps are:

[0051] Step S11, extracting enterprise order information, and obtaining all customer information and customer purchase volume from the order information;

[0052] Step S12: add up all purchase quantities to obtain the total purchase quantity, and add up the purchase quantities of the same customer to obtain the customer purchase quantity;

[0053] Step S13: Calculate the customer purchase ratio based on the customer purchase volume and the total purchase volume, and output customers whose customer purchase ratio is greater than the purchase threshold as actual important customers, and output other customers as general customers.

[0054] Preferably, the method for confirming the procurement threshold is: quantifying the production capacity, staff mobility and cash flow status of the enterprise, and adding up the quantified values ​​after assigning different weights.

[0055] In the course of operation, an enterprise will establish cooperative relationships with several customers and provide products or parts according to the order information with the customers. With the improvement of customers' technical level and the increase of supply volume, the number of products supplied by the enterprise to customers is also increasing. The purchase volume occupied by some customers accounts for most of the enterprise, thus gradually becoming the enterprise's major customers. When the enterprise is too dependent on major customers, there will be certain operational risks. If the major customers cancel the cooperation, the enterprise's supply channels will be reduced, and the operation will be easily shut down. Therefore, in order to prevent operational risks, the present invention judges the customers of the enterprise, first obtains all customers cooperating with the enterprise from the order information of the enterprise, and then obtains the order information of all customers cooperating with the enterprise. The total purchase volume of the enterprise can be obtained based on the purchase volume of all customers. At the same time, for the same customer, there will be multiple orders. The customer purchase volume can be obtained by adding up the purchase volumes of different orders of the same customer. The customer purchase volume can be divided by the total purchase volume to get the customer purchase ratio. For the enterprise, a purchase threshold will be preset. When the customer purchase ratio is greater than the purchase threshold, it means that the purchase volume of the corresponding customer accounts for the majority of the enterprise, and the corresponding customer can be output as an actual important customer. Customers whose customer purchase ratio is less than the purchase threshold are general customers. For actual important customers, corresponding dependence reduction measures need to be taken.

[0056] As for the purchasing threshold, it is determined by the company's own situation. After quantifying the company's production capacity, personnel turnover and cash situation, they are assigned values ​​using a weighted assignment method, and the assigned data are added together to obtain the purchasing threshold. The stronger the company's production capacity, the less personnel turnover, and the more abundant the cash situation, the higher the purchasing threshold will be, and thus the higher the assessment standards for actual important customers will be.

[0057] Step S2: obtaining multimodal data of general customers, evaluating the general customers based on the multimodal data, and determining whether there are potential important customers;

[0058] The specific steps are:

[0059] Step S21, obtaining multimodal data published by general customers on network platforms and social media for their own product promotion, wherein the multimodal data includes text, pictures and videos;

[0060] Step S22: Obtain the number of clicks, views, and forwardings of the multimodal data on the network;

[0061] Step S23: Calculate the increase in clicks, views and forwardings, and output general customers whose increase is greater than that of similar products as potential important customers.

[0062] When determining whether actual important customers are obtained, general customers will also be obtained at the same time. Among these general customers, there will be some that can develop into actual important customers. Therefore, it is necessary to evaluate the general customers, mainly through some multimodal data released by general customers on online platforms and social media. These multimodal data include text, pictures and videos. General customers will promote products through multimodal data. If the effect of the promotion is good, it means that the public recognizes the general customers, and the shipment volume of general customers is likely to increase. When evaluating the effect of the promotion, it is necessary to obtain the clicks, views and forwarding volumes of these multimodal data, and calculate the increase in a certain period of time, and compare the increase with the increase of similar products. If the increase of general customers is greater than that of similar products, the corresponding general customers can be output as potential important customers. For potential important customers, since they have not actually developed into important customers, they do not occupy many internal corporate resources. Therefore, there is no need to reallocate resources for them, but to take other measures to reduce dependence.

[0063] Step S3: If there are actual important customers, reallocate the internal resources of the enterprise. The specific steps are:

[0064] Step S31: Acquire the internal resources of the enterprise, wherein the internal resources include human resources, material resources, financial resources and information resources;

[0065] Step S32: if there are actual important customers, obtain the resource ratio of the internal resources of the enterprise occupied by the actual important customers;

[0066] Step S33: Allocate the portion of the resource ratio that exceeds the resource threshold to general customers. The resource threshold is set by the enterprise according to its own operating conditions.

[0067] When there are actual important customers, the measure taken to reduce the dependence of the enterprise is to reallocate the internal resources of the enterprise. Before redistribution, the internal resources of the enterprise need to be acquired. The internal resources are composed of human resources, material resources, financial resources and information resources. When the enterprise cooperates with customers, it will allocate its internal resources to customers. When the internal resources occupied by the actual important customers are relatively large, if the actual important customers do not cooperate with the enterprise, the internal resources cannot be immediately circulated to other customers, which will lead to waste of internal resources, increase enterprise costs, and ultimately hinder the operation of the enterprise. Therefore, when there are actual important customers among the customers of the enterprise, the resource proportion of the internal resources occupied by the actual important customers is obtained, and the part of the resource proportion that exceeds the resource threshold is allocated to other general customers. The resource threshold needs to be set by the enterprise according to its own operating conditions. When the resource proportion of the actual important customers is greater than the resource threshold, it means that the actual important customers occupy too many resources and need to be properly redistributed to reduce dependence on the actual important enterprises.

[0068] Step S4: If there are actual important customers or potential important customers, search for potential customers based on the industries to which the actual important customers and potential important customers belong. The specific steps are as follows:

[0069] Step S41, obtaining the industry to which the enterprise's customers belong according to the enterprise's order information;

[0070] Step S42: If there are actual important customers or potential important customers, output the remaining customers in the industry to which the customer belongs who have not cooperated with the enterprise as potential customers;

[0071] Step S43: Obtain the transaction content between the enterprise and its actual important customers;

[0072] Step S44: Send the transaction content to the potential customer in the form of an invitation request.

[0073] For actual important customers, in addition to reallocating resources, a search for potential customers will also be conducted. At the same time, when there are potential important customers among the company's customers, since they are not actual important customers, there is no need to reallocate resources. Instead, a search for potential customers will be conducted. Among them, it is necessary to obtain the industry to which the customers cooperating with the company belong, and output the remaining customers in the same industry on the market who have not cooperated with the company as potential customers. By increasing the number of customers cooperating with the company, the dependence on important customers can be further reduced. After obtaining potential customers, the transaction content and invitation request between the company and the actual important customers will be sent to the potential customers. The transaction content with the actual important customers can be used to demonstrate the company's own strength and sincerity, thereby improving the cooperation achievement rate with potential customers.

[0074] See also Figure 6The risk prevention perception system based on big data and multimodal intelligent algorithms includes:

[0075] The actual important customer classification unit 1 is used to extract the enterprise order information, and judge whether there is an important customer according to the order information. If there is, the important customer is output as the actual important customer, and the non-important customer is output as a general customer;

[0076] The potential important customer classification unit 2 is used to obtain multimodal data of general customers, evaluate the general customers based on the multimodal data, and determine whether there are potential important customers;

[0077] Resource allocation unit 3, used to reallocate the internal resources of the enterprise if there are actual important customers;

[0078] The potential customer search unit 4 is used to search for potential customers based on the industries to which the actual important customers and potential important customers belong if there are actual important customers or potential important customers;

[0079] The actual important customer division unit 1 is data-connected with the potential important customer division unit 2 , the resource allocation unit 3 and the intended customer search unit 4 , respectively. The potential important customer division unit 2 is data-connected with the intended customer search unit 4 .

[0080] The actual important customer classification unit 1 can classify actual important customers, and the potential important customer classification unit 2 can classify potential important customers. The resource allocation unit 3 reallocates the resource proportion of actual important customers. When there are actual important customers and potential important customers, the potential customer search unit 4 can search for potential customers and provide them to the enterprise for the enterprise to choose for cooperation, thereby reducing the enterprise's dependence on important customers and ensuring the normal operation of the enterprise.

[0081] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A risk prevention perception method based on big data and multimodal intelligent algorithms, characterized in that: The following steps are involved: Step S1, extracting enterprise order information, judging whether there are important customers according to the order information, and if so, outputting the important customers as actual important customers, and outputting the non-important customers as general customers; Step S2: obtaining multimodal data of general customers, evaluating the general customers based on the multimodal data, and determining whether there are potential important customers; Step S3: if there are actual important customers, reallocate the internal resources of the enterprise; Step S4: If there are actual important customers or potential important customers, search for potential customers based on the industries to which the actual important customers and potential important customers belong; The specific steps of step S3 are: Step S31: Acquire the internal resources of the enterprise, wherein the internal resources include human resources, material resources, financial resources and information resources; Step S32: if there are actual important customers, obtain the resource ratio of the internal resources of the enterprise occupied by the actual important customers; Step S33: allocating the portion of the resource proportion that exceeds the resource threshold to general customers; The specific steps of step S1 are: Step S11, extracting enterprise order information, and obtaining all customer information and customer purchase volume from the order information; Step S12: add up all purchase quantities to obtain the total purchase quantity, and add up the purchase quantities of the same customer to obtain the customer purchase quantity; Step S13: Calculate the customer purchase ratio according to the customer purchase volume and the total purchase volume, output the customer whose customer purchase ratio is greater than the purchase threshold as the actual important customer, and output the other customers as general customers; The specific steps of step S2 are: Step S21: obtaining multimodal data published by general customers on network platforms and social media, wherein the multimodal data includes text, pictures and videos; Step S22: Obtain the number of clicks, views, and forwardings of the multimodal data on the network; Step S23, calculating the increase in clicks, views and forwarding, and outputting general customers whose increase is greater than that of similar products as potential important customers; The resource threshold of step S33 is set by the enterprise according to its own operating conditions.

2. The risk prevention perception method based on big data and multimodal intelligent algorithm according to claim 1 is characterized in that: The method for confirming the procurement threshold is: quantifying the production capacity, personnel mobility and cash flow status of the enterprise, and adding up the quantified values ​​after assigning different weights.

3. The risk prevention perception method based on big data and multimodal intelligent algorithm according to claim 1 is characterized in that: The multimodal data acquired in step S21 are texts, pictures and videos posted by general users on online platforms and social media for the purpose of promoting their own products.

4. The risk prevention perception method based on big data and multimodal intelligent algorithm according to claim 1 is characterized in that: The specific steps of step S4 are: Step S41, obtaining the industry to which the enterprise's customers belong according to the enterprise's order information; Step S42: If there are actual important customers or potential important customers, the remaining customers in the industry to which the customers belong who have not cooperated with the enterprise are output as potential customers.

5. The risk prevention perception method based on big data and multimodal intelligent algorithm according to claim 4 is characterized in that: Also includes: Step S43: Obtain the transaction content between the enterprise and its actual important customers; Step S44: Send the transaction content to the potential customer in the form of an invitation request.

6. A system using the risk prevention perception method based on big data and multimodal intelligent algorithm as described in any one of claims 1 to 5, characterized in that: include: The actual important customer classification unit is used to extract the enterprise order information and determine whether there are important customers based on the order information. If so, the important customers are output as actual important customers and the non-important customers are output as general customers. A potential important customer classification unit is used to obtain multimodal data of general customers, evaluate general customers based on the multimodal data, and determine whether there are potential important customers; Resource allocation unit, used to reallocate the internal resources of the enterprise if there are actual important customers; A potential customer search unit is used to search for potential customers based on the industries to which the actual important customers and potential important customers belong if there are actual important customers or potential important customers; The actual important customer division unit is data-connected with the potential important customer division unit, the resource allocation unit and the intended customer search unit respectively, and the potential important customer division unit is data-connected with the intended customer search unit.

Images