A short message channel intelligent matching system based on multi-dimension weight distribution

Abstract

The application discloses a kind of short message channel intelligent matching system based on multidimensional weight distribution, specifically related to short message channel intelligent matching technical field, to solve the problem of channel matching weight mechanism rigidification, cost control extensive, insufficient sending scheduling stability in prior art, including channel hierarchical screening, dimension weight matching, cost accounting control, four big modules of dispatch execution self-healing;Through customer industry, short message type and so on attribute completes channel preliminary screening, locks available channel pool, in combination with customer subscription grade sets differentiated weight, based on cost, quality and so on multidimensional index calculates channel performance coefficient, selects optimal primary and backup channel;While constructing dynamic premium logic realizes fine pricing, quantifies channel abnormal degree by sliding window, executes gradient shunting and automatic switching, the application can realize channel accurate matching, gross profit effective guarantee and sending stability improvement, adapts to enterprise differentiation service demand.

Description

Technical Field

[0001] This invention relates to the field of intelligent matching technology for SMS channels, and more specifically, to an intelligent matching system for SMS channels based on multi-dimensional weight allocation. Background Technology

[0002] As a core communication channel for enterprises to reach end users, industry SMS is widely used in various scenarios such as identity verification, transaction notification, marketing promotion, government emergency response, and public service announcement. Its delivery rate, timeliness, compliance, and cost controllability directly determine the business operation effect and end user experience of enterprises.

[0003] However, existing technologies for matching and scheduling industry SMS channels still have many technical shortcomings and cannot adapt to industry development and regulatory requirements. These shortcomings are specifically reflected in the following aspects: The rigid weighting mechanism of channel matching makes it impossible to achieve coordinated optimization of multi-dimensional indicators and adapt to the differentiated needs of customers. Existing solutions mostly adopt fixed priority polling and single-dimensional priority scheduling logic, which can only achieve the optimal single objective among cost, quality, timeliness and load, and cannot take into account the coordinated balance of multi-dimensional indicators. At the same time, the weight allocation adopts a fixed mode and cannot be dynamically adjusted according to the differentiated needs of customers' contract levels. It cannot meet the priority guarantee requirements of member customers for sending quality and timeliness, nor can it adapt to the core demand of low-cost customers for cost control. The accuracy of channel matching is insufficient and it is impossible to select the primary and backup channels with the best overall performance. The company suffers from insufficient cost accounting and gross profit control capabilities, and lacks a refined differentiated pricing mechanism. Existing cost accounting technologies mostly adopt a static unit price calculation model, failing to calculate the weighted average cost based on the transmission volume allocation ratio of the primary and backup channels. This results in a disconnect between cost accounting and channel matching. Furthermore, the premium level is simply set based on customer level, without dynamic adjustments based on customer cooperation duration, historical cooperation contributions, and payment performance. This makes it impossible to achieve differentiated pricing and to conduct pre-verification and guarantee of the minimum gross profit margin for different customer levels. This easily leads to uncontrolled profit margins and a mismatch between pricing and customer value, making it difficult to achieve a balance between corporate profits and customer service.

[0004] The aforementioned shortcomings not only lead to unstable delivery quality and difficulty in controlling operating costs in the industry's SMS services, but also make it impossible to adapt to increasingly stringent regulatory requirements and the differentiated service needs of enterprise customers, thus hindering the standardization and high-quality development of the SMS service industry. To address this, an intelligent matching system for SMS channels based on multi-dimensional weight allocation has been launched. Summary of the Invention

[0005] To overcome the above-mentioned deficiencies of the prior art, embodiments of the present invention provide an intelligent matching system for SMS channels based on multi-dimensional weight allocation.

[0006] To achieve the above objectives, the present invention provides the following technical solution: A smart SMS channel matching system based on multi-dimensional weight allocation includes: The channel hierarchical filtering module is used to obtain SMS sending requests initiated by customers, combine customer attributes with preset filtering rules, conduct a preliminary screening of all available channel processes, and lock the available channel pool. The dimension weight matching module is used to output the channel performance coefficient based on multi-dimensional indicators and weight allocation for all channels in the available channel pool, and select the primary and backup transmission channels according to the channel performance coefficient. The multi-dimensional indicators include cost indicators, quality indicators, timeliness indicators and load balancing indicators. The cost accounting and control module is used to output a quotation plan based on the unit SMS price of the selected primary and backup channels, combined with customer cooperation data, after comprehensive processing. The quotation plan includes the quotation unit price charged to the customer, the gross profit margin, and the gross profit verification results. The scheduling and execution self-healing module is used to execute SMS sending tasks according to the selected primary and backup channels after the customer determines the unit price in the quotation plan. It monitors the operating status of the primary channel in real time, identifies the degree of abnormality, makes a judgment, and selectively performs automatic channel switching.

[0007] Specifically, customer attributes include the customer's industry affiliation, the type of SMS sent, and the complaint rate of the requested channel; Each of the fully available channels is pre-marked with the supported customer industry affiliation and SMS type; The system can search all available channels and simultaneously support channels based on the current customer's industry affiliation and SMS sending type. In the search results, channels with complaint rates lower than the customer's required complaint rate are retained as an available channel pool.

[0008] Specifically, the filtering logic for the primary and backup transmission channels; For all channels in the available channel pool, cost indicators, quality indicators, timeliness indicators, and load balancing indicators are extracted and combined with the weight set corresponding to the customer's contract level. After comprehensive processing, the channel performance coefficient is obtained. The customer contract level includes member customers, standard customers, and low-cost demand customers. A weight set is set for each different customer contract level. For all channels in the available channel pool, sort the channels from high to low according to their performance coefficients. Starting from the left, select the channel ranked 1st as the primary sending channel for this sending request, and select the channels ranked 2nd-3rd as backup sending channels.

[0009] Specifically, the calculation logic for multi-dimensional indicators; Obtain the pre-marked unit SMS price of the channel as a cost indicator; Extract the total number of SMS messages sent within a set time zone before the current time point; count the number of SMS messages successfully delivered to end users within the total number of SMS messages sent, and calculate its proportion in the total number of messages sent, as the channel arrival rate; count the number of user complaints within the total number of SMS messages sent, and calculate its proportion in the total number of messages sent, as the channel complaint rate; combine the channel arrival rate and channel complaint rate with preset arrival weights and complaint weights for comprehensive processing to obtain quality indicators; The average time taken to send SMS messages within a set time zone is used as a timeliness indicator. The maximum concurrent capacity of the preset channel is calculated, the actual transmission volume of the channel within a unit of time before the current time point is counted, and the ratio between the actual transmission volume and the maximum concurrent capacity is calculated as a load balancing indicator.

[0010] Specifically, the logic behind obtaining the unit price of the quote received by the customer; The system analyzes the requested SMS volume in the customer's SMS sending request, combines it with the preset sending volume allocation ratio and the unit SMS price per channel, and uses a formula... Output the average unit cost of the current customer's SMS sending request. ; in The unit SMS rate for the main channel, The preset allocation of sending volume for the main channel is obtained by multiplying the requested SMS sending volume by the main channel execution ratio in the preset sending volume allocation ratio; and The unit SMS price for the two backup channels. and The backup sending volume allocated to the two backup channels is obtained by multiplying the requested SMS sending volume by the backup channel execution ratio in the preset sending volume allocation ratio; the quotation unit price applied to the current customer's SMS sending request is obtained by multiplying the average unit cost by the premium factor.

[0011] Specifically, the logic for obtaining the premium coefficient; The premium coefficient is determined by combining customer contract level, historical cooperation records, and cooperation duration from customer cooperation data. Analyze the customer's contract level and set a set of benchmark coefficients for member customers, standard customers, and low-cost demand customers respectively; Acquire the duration of customer cooperation and convert it into cumulative cooperation months; then, utilize the mapping rules between cumulative cooperation months and duration value coefficient to convert it into duration value coefficient. Extract the current customer's contribution percentage, revenue contribution percentage, and timely payment percentage from the customer's historical cooperation records and process them comprehensively to obtain the current customer's cooperation performance index; then, use the mapping rules between the cooperation performance index and the customer incentive coefficient to convert it into the customer incentive coefficient. Premium coefficient = Benchmark coefficient × Duration value coefficient × Customer incentive coefficient.

[0012] Specifically, the gross profit margin is obtained by subtracting the average unit cost from the quoted unit price and then dividing by the quoted unit price. Preset the minimum gross profit margin requirements corresponding to different customer contract levels; The determined gross profit margin is compared with the preset minimum gross profit margin corresponding to the customer's contract level. If it is higher than the minimum gross profit margin, the verification result is determined to meet the gross profit requirement; otherwise, the verification result is determined to not meet the gross profit requirement.

[0013] Specifically, the logic for identifying the degree of abnormality in the main channel's operating status; Monitor the degree of anomalies in the main channel by sending real-time status and channel concurrency ratio; The real-time status of the sent data includes the window send success rate, the window send arrival rate, and the window response timeout rate. The abnormal state coefficient of the main channel is output by comprehensively processing the window sending success rate, window sending arrival rate, window response timeout rate, and concurrency ratio.

[0014] Specifically, the calculation logic for sending real-time status and the proportion of channel concurrency; The number of SMS messages that successfully received receipts in the main channel within the current sliding window is counted, and their proportion in the total number of SMS messages submitted for sending in the main channel within the current sliding window is calculated to obtain the window sending success rate. The number of SMS messages successfully delivered to end users in the current sliding window is counted, and its proportion in the total number of SMS messages submitted for sending in the current sliding window is calculated to obtain the window delivery rate. Within the sliding window, count the number of times the response time of a single request in the main channel exceeds the preset timeout threshold, calculate its proportion in the total number of submissions within the window, and obtain the window response timeout rate; Within the sliding window, the average actual concurrency per unit time of the main channel is statistically analyzed, and its proportion in the preset maximum concurrency capacity of the main channel is calculated to obtain the concurrency proportion.

[0015] Specifically, the detailed process of automated channel switching; If the abnormal state coefficient of the main channel in the current sliding window is higher than the preset abnormal threshold coefficient, the pre-edited diversion strategy will be executed in the next sliding window. The task volume of the next sliding window main channel is evenly distributed to the first and second backup channels for execution, based on a preset proportion of the task volume to be sent.

[0016] The technical effects and advantages of this invention are as follows: (1) To achieve refined cost control and differentiated pricing, effectively guarantee the company's operating profit. By combining the distribution ratio of the main and backup channels to calculate the average unit cost, the traditional static unit price accounting model is replaced, allowing cost accounting and channel matching to be deeply linked. At the same time, a multi-dimensional dynamic premium coefficient model is constructed. Based on the customer's contract level, a benchmark coefficient is set, and then dynamically adjusted in combination with the cooperation duration, historical contribution, and payment performance. This achieves differentiated pricing. Furthermore, the gross profit margin can be pre-verified based on the customer level to ensure that the actual gross profit is not lower than the minimum requirement. This solves the problem of uncontrolled profit margin and mismatch between pricing and customer value in the existing technology, and achieves a balance between corporate profit and customer service. (2) Significantly improve the accuracy of channel matching, achieve precise adaptation to the differentiated needs of customers, lock the available channel pool through customer attributes; then set a differentiated weight set through customer contract level, take into account the four dimensions of cost, quality, timeliness and load balancing to calculate the channel performance coefficient, and accurately select the optimal primary and backup channels. This not only solves the defects of fixed weight and single-dimensional scheduling in the existing technology, but also achieves multi-objective collaborative optimization, and can meet the quality and timeliness needs of member customers and the cost control needs of low-cost customers respectively, while avoiding compliance risks from the source of the channel. (3) Construct an automated anomaly self-healing scheduling system to significantly improve the stability and robustness of SMS sending. Based on the sliding window, monitor the four core indicators of the main channel in real time: sending success rate, arrival rate, response timeout rate, and concurrency ratio. Quantify the abnormal state coefficient and accurately identify the degree of abnormality. Execute a gradient diversion strategy according to the difference in the degree of abnormality to realize the dynamic diversion of task volume. After the anomaly is recovered, it can be smoothly switched back to the main channel, effectively ensuring the continuity of sending tasks. Attached Figure Description

[0017] Figure 1 This is a flowchart of an intelligent matching system for SMS channels based on multi-dimensional weight allocation, as described in this invention. Detailed Implementation

[0018] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0019] like Figure 1As shown, a module for an intelligent SMS channel matching system based on multi-dimensional weight allocation is as follows: The channel hierarchical filtering module is used to obtain SMS sending requests initiated by customers, combine customer attributes with preset filtering rules, conduct a preliminary screening of all available channel processes, and lock the available channel pool. Before initial screening of the channels, it is necessary to conduct comprehensive pre-compliance verification and risk control on all SMS sending requests initiated by customers, and block illegal requests: The sensitive word database is divided into three levels: Level 1 sensitive words (fraud-related, politically related, illegal and irregular words), Level 2 sensitive words (marketing-related restricted words), and Level 3 sensitive words (industry-specific restricted words). Analysis process: The NLP (Natural Language Processing) algorithm is used to segment the SMS text content and perform a full match verification with a pre-set hierarchical sensitive word library; if a level 1 sensitive word is matched, the sending request is directly blocked and a violation log is recorded; if a level 2 / 3 sensitive word is matched, the content is manually reviewed, the content risk level is marked, and the information is synchronized to the subsequent channel matching module to filter compliant channels that match the content risk level.

[0020] Specifically: Customer attributes include the customer's industry affiliation, the type of SMS sent, and the complaint rate of the requested channel. Our clients' industries include, but are not limited to, finance, government affairs, healthcare, education, e-commerce, and logistics. The types of text messages sent include, but are not limited to, verification codes, transaction notifications, marketing promotions, government emergency messages, and public service announcements. Each of the fully available channels is pre-marked with the supported customer industry affiliation and SMS type; The system can search all available channels and simultaneously support the current customer's industry affiliation and SMS sending type. In the search results, channels with a complaint rate lower than the customer's required complaint rate are retained as an available channel pool. Additional filtering rules: Prioritize provincial network channels that perfectly match the customer's mobile phone number's location; if the number of provincial network channels is insufficient, supplement with nationwide channels that have cross-provincial sending permissions for the corresponding region; finally, use a three-network channel with full regional coverage as a backup; the operator affiliation of the end user's mobile phone number being sent by the customer (China Mobile / China Unicom / China Telecom / Virtual Operator) requires that the channel must have the corresponding operator's legal sending permissions; for example, if the requirement is only to send China Mobile mobile phone numbers, channels without mobile operator sending qualifications will be directly eliminated; The specific coverage area requested by the customer (precise coverage within the province / full coverage across provinces / unified coverage across all three networks nationwide) directly determines the type of channel to be initially screened. For example, if the customer requires unified coverage across all three networks nationwide, channels with full coverage across all three networks will be screened first, and then supplemented with full network channels; if the customer requires exclusive coverage within the province, only provincial network channels for the corresponding province will be screened.

[0021] The dimension weight matching module is used to output the channel performance coefficient based on multi-dimensional indicators and weight allocation for all channels in the available channel pool, and select the primary and backup transmission channels according to the channel performance coefficient. The multi-dimensional indicators include cost indicators, quality indicators, timeliness indicators and load balancing indicators. Specifically: Obtain the pre-marked unit SMS price of the channel as a cost indicator; Extract the total number of SMS messages sent by a channel within a set time zone before the current time from a pre-built database; the database can be used to store historical SMS sending records for different channels; The number of SMS messages successfully delivered to end users is counted from the total number of SMS messages sent, and its proportion in the total number of SMS messages sent is calculated as the channel arrival rate. The number of user complaints is counted in the total number of SMS messages sent, and the percentage of complaints in the total number of SMS messages sent is calculated as the channel complaint rate. The channel arrival rate and channel complaint rate are combined with preset arrival weights and complaint weights to obtain quality indicators. Quality Indicators ,in and These represent the channel arrival rate and the channel complaint rate, respectively. and This indicates the arrival weight and complaint weight.

[0022] The average time taken to send SMS messages within a set time zone is obtained; that is, the arithmetic mean of the time taken for each SMS message sent within the set time zone is taken, and the time taken is counted from the time the request is submitted until the end user successfully receives the message; this serves as a timeliness indicator.

[0023] The maximum concurrent capacity of the preset channel; that is, the maximum concurrent capacity per unit time; the actual transmission volume of the channel in the unit time before the current time point, and the ratio between the actual transmission volume and the maximum concurrent capacity, as a load balancing indicator; For all channels in the available channel pool, cost indicators, quality indicators, timeliness indicators, and load balancing indicators are extracted and combined with the weight set corresponding to the customer's contract level. After comprehensive processing, the channel performance coefficient is obtained. Customer contract levels include membership customers, standard customers, and customers with low-cost needs; Different customer contract levels are assigned a set of weights; each weight set contains the weight coefficients corresponding to cost indicators, quality indicators, timeliness indicators, and load balancing indicators, and the sum of these coefficients is one. Additional notes: Standards for setting the weight size within the weight set; For member customers: Prioritize delivery quality and timeliness, and adjust the weights as follows: Quality indicator weight > Timeliness indicator weight > Load balancing indicator weight > Cost indicator weight, with the sum remaining at 1; Standard customers: use the default base weighting coefficient; For customers with low-cost requirements: Prioritize cost control and adjust the weights as follows: Cost indicator weight > Quality indicator weight > Load balancing indicator weight > Timeliness indicator weight, with the sum remaining at 1.

[0024] Channel performance coefficient ,in , as well as This refers to cost metrics, timeliness metrics, and load balancing metrics. , , , These represent the benchmarks for quality indicators, cost indicators, timeliness indicators, and load balancing indicators for each channel, respectively. Each benchmark is obtained by taking the arithmetic mean of the cost, quality, timeliness, and load balancing indicators for each channel. The weight coefficients of each indicator within the weight set set based on the customer's contract level.

[0025] For all channels in the available channel pool, sort the channels from high to low according to their performance coefficients. Starting from the left, select the channel ranked 1st as the primary sending channel for this sending request, and select the channels ranked 2nd-3rd as backup sending channels.

[0026] The cost accounting and control module is used to output a quotation plan based on the unit SMS price of the selected primary and backup channels, combined with customer cooperation data, after comprehensive processing. The quotation plan includes the quotation unit price charged to the customer, the gross profit margin, and the gross profit verification results. To safeguard corporate profit margins and achieve refined cost control; Specifically: The system analyzes the requested SMS volume in the customer's SMS sending request, combines it with the preset sending volume allocation ratio and the unit SMS price per channel, and uses a formula... Output the average unit cost of the current customer's SMS sending request. ; in The unit SMS rate for the main channel, The preset allocation of sending volume for the main channel is obtained by multiplying the requested SMS sending volume by the main channel execution ratio in the preset sending volume allocation ratio; and The unit SMS price for the two backup channels. and The fallback transmission capacity allocated to the two backup channels; It is obtained by multiplying the requested SMS sending volume by the backup channel execution ratio in the preset sending volume allocation ratio; The sum of the execution ratio of the main channel and the execution ratio of the two sets of backup channels is one, which can be allocated according to 6:2:2 or 8:1:1 to ensure the priority of the main channel allocation. The technical personnel will make adjustments later.

[0027] All allocated sending volumes are rounded to ensure that the total allocated volume for each channel is equal to the requested SMS sending volume.

[0028] Based on the average unit cost, and combined with the customer's contract level, historical cooperation records, and cooperation duration in the customer cooperation data, a premium coefficient is determined to obtain the quotation unit price based on the current customer's SMS sending request, which is obtained by multiplying the average unit cost by the premium coefficient. Analyze the customer's contract level and set a set of benchmark coefficients for member customers, standard customers, and low-cost demand customers respectively; the benchmark coefficient range is limited to 1.2-1.5, with the benchmark coefficient for member customers being the highest. The specific settings are initially set by the management personnel; for member customers, the goal is to ensure maximum profit while guaranteeing channel quality.

[0029] The duration of customer cooperation is converted into cumulative months of cooperation. The mapping rule between the cumulative number of months of cooperation and the duration value coefficient is used to convert it into a duration value coefficient; This involves setting up a range of cooperation months corresponding to the cumulative number of cooperation months, and each range of months corresponds to a set of duration value coefficients. The duration value coefficient is limited to a range of 0.97-1. The higher the cumulative number of cooperation months, the higher the probability of matching 0.97. The customer's cumulative number of cooperation months is matched with the corresponding range of cooperation months to determine the duration value coefficient.

[0030] Extract the total number of SMS messages sent and the amount of payment made within a set time window (nearly 6 months) before the current point in time from the customer's historical cooperation records; Calculate the percentage of a customer's total SMS messages sent within the set time window as the percentage of total SMS messages sent by all customers, and calculate the percentage of a customer's total payment amount within the set time window as the percentage of total revenue, and calculate the percentage of revenue contribution. Extract the agreed payment deadlines from each cooperation record with the current client, identify the cooperations where the actual payment time is earlier than the corresponding payment deadline, and record them as timely payment cooperations. Calculate the percentage of timely payment cooperations in the total number of cooperations to obtain the timely payment percentage. The cooperation performance index of current customers is obtained by comprehensively processing the proportion of cooperation contribution, revenue contribution, and timely payment of current customers. Cooperation Performance Index ,in , This indicates the percentage of contribution from collaboration, the percentage of revenue contribution, and the percentage of timely payments. , as well as These are preset weighting coefficients, and their sum is one.

[0031] The mapping rule between the cooperation performance index and the customer incentive coefficient is used to convert it into a customer incentive coefficient; This involves setting up index ranges for each group corresponding to the cooperation performance index, with each index range corresponding to a set of customer incentive coefficients. The range of customer incentive coefficients is limited to 0.95-1. The higher the cooperation performance index, the higher the probability of matching 0.95. The customer's cooperation performance index is matched with the corresponding index range to determine the customer incentive coefficient.

[0032] Premium coefficient = Benchmark coefficient × Duration value coefficient × Customer incentive coefficient; Gross profit margin is obtained by subtracting the average unit cost from the quoted unit price and then dividing by the quoted unit price. Preset the minimum gross profit margin requirements corresponding to different customer contract levels; Minimum gross profit margin requirement for member customers > Standard customers > Customers with low cost requirements.

[0033] The determined gross profit margin is compared with the preset minimum gross profit margin corresponding to the customer's contract level. If it is higher than the minimum gross profit margin, the verification result is determined to meet the gross profit requirement; otherwise, the verification result is determined to not meet the gross profit requirement.

[0034] The scheduling and execution self-healing module is used to execute SMS sending tasks according to the selected primary and backup channels after the customer determines the unit price in the quotation plan. It monitors the operating status of the primary channel in real time, identifies the degree of abnormality and makes a judgment, and selectively performs automatic channel switching to ensure the stability and continuity of the sending task. Specifically: Monitor the degree of anomalies in the main channel by sending real-time status and channel concurrency ratio; Based on the customer's current SMS sending request, preset the window size and window step size of the sliding window; The real-time status of the sent data includes the window send success rate, the window send arrival rate, and the window response timeout rate. The number of SMS messages that successfully received receipts in the main channel within the current sliding window (i.e., received "sent successfully" receipts from the operator) is counted, and the percentage of these messages in the total number of SMS messages submitted for sending in the main channel within the current sliding window is calculated to obtain the window sending success rate. The number of SMS messages successfully delivered to end users in the current sliding window is counted, and its proportion in the total number of SMS messages submitted for sending in the current sliding window is calculated to obtain the window delivery rate. The window delivery rate reflects the final reach quality of the channel and is judged in conjunction with the delivery success rate. If both decrease at the same time, it indicates that the main channel is highly abnormal. Within the sliding window, count the number of times the response time of a single request in the main channel exceeds the preset timeout threshold (3 seconds), calculate its proportion in the total number of submissions within the window, and obtain the window response timeout rate; Within the sliding window, the average actual concurrency per unit time of the main channel is statistically analyzed, and its proportion in the preset maximum concurrency capacity of the main channel is calculated to obtain the concurrency proportion, which reflects the load pressure of the channel. The abnormal state coefficient of the main channel is output by comprehensively processing the window sending success rate, window sending arrival rate, window response timeout rate, and concurrency ratio. Abnormal state coefficient ,in , , as well as This represents the window send success rate, window send arrival rate, window response timeout rate, and concurrency percentage. , , as well as The weighting coefficients are set, and their sum is one.

[0035] If the abnormal state coefficient of the main channel in the current sliding window is higher than the preset abnormal threshold coefficient, the pre-edited diversion strategy will be executed in the next sliding window. A preset proportion (30%-50%) of the task volume required to be sent in the next sliding window main channel is evenly distributed to the first and second backup channels for execution. After the distribution, the abnormal state coefficient of the next sliding window main channel is monitored. If the abnormal state coefficient of the next sliding window main channel is lower than the abnormal threshold coefficient, the distribution strategy is lifted in the next sliding window. If it is still higher than the abnormal threshold coefficient, the preset proportion is adjusted based on the difference in the degree of abnormality. Additional explanation: The preset ratio is 30%-50%, which is determined by the magnitude of the abnormal state coefficient. That is, the difference between the abnormal state coefficient and the abnormal threshold coefficient is calculated and recorded as the abnormality degree difference. The system presets the difference ranges for each group of abnormality levels, and each difference range corresponds to a preset ratio. The larger the difference in abnormality level, the greater the probability of a match (50%).

[0036] After completing the SMS sending task, update the actual average unit cost and actual gross profit margin of this sending.

[0037] The above formulas are all dimensionless calculations. Dimensionless calculations can be performed using various methods such as standardization, which will not be elaborated here. The formulas are derived from software simulations based on a large amount of collected data, and the preset parameters in the formulas can be set by those skilled in the art according to the actual situation.

[0038] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, as a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, all or part of the processes or functions described in the embodiments of this application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that includes one or more sets of available media. The available medium can be a magnetic medium (e.g., floppy disk, ATA hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium. The semiconductor medium can be a solid-state ATA hard disk.

[0039] It should be understood that in the various embodiments of this application, the order of the above-mentioned processes does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0040] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0041] In the several embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between apparatuses or units may be electrical, mechanical, or other forms.

[0042] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment, depending on actual needs.

[0043] In addition, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0044] If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable ATA hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0045] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A short message channel intelligent matching system based on multi-dimension weight distribution, characterized in that, include: The channel hierarchical filtering module is used to obtain SMS sending requests initiated by customers, combine customer attributes with preset filtering rules, conduct a preliminary screening of all available channel processes, and lock the available channel pool. The dimension weight matching module is used to output the channel performance coefficient based on multi-dimensional indicators and weight allocation for all channels in the available channel pool, and select the primary and backup transmission channels according to the channel performance coefficient. The multi-dimensional indicators include cost indicators, quality indicators, timeliness indicators and load balancing indicators. The cost accounting and control module is used to output a quotation scheme based on the SMS unit price of the selected primary and backup channels, combined with customer cooperation data, after comprehensive processing. The quotation plan includes the unit price charged to the customer, the gross profit margin, and the gross profit verification results; The scheduling and execution self-healing module is used to execute SMS sending tasks according to the selected primary and backup channels after the customer determines the unit price in the quotation plan. It monitors the operating status of the primary channel in real time, identifies the degree of abnormality, makes a judgment, and selectively performs automatic channel switching.

2. The SMS channel intelligent matching system based on multi-dimensional weight allocation according to claim 1, characterized in that: Customer attributes include the customer's industry affiliation, the type of SMS sent, and the complaint rate of the requested channel. Each of the fully available channels is pre-marked with the supported customer industry affiliation and SMS type; The system can search all available channels and simultaneously support channels based on the current customer's industry affiliation and SMS sending type. In the search results, channels with complaint rates lower than the customer's required complaint rate are retained as an available channel pool.

3. The SMS channel intelligent matching system based on multi-dimensional weight allocation according to claim 1, characterized in that: Filtering logic for primary and backup transmission channels; For all channels in the available channel pool, cost indicators, quality indicators, timeliness indicators, and load balancing indicators are extracted and combined with the weight set corresponding to the customer's contract level. After comprehensive processing, the channel performance coefficient is obtained. The customer contract level includes member customers, standard customers, and low-cost demand customers. A weight set is set for each different customer contract level. For all channels in the available channel pool, sort the channels from high to low according to their performance coefficients. Starting from the left, select the channel ranked 1st as the primary sending channel for this sending request, and select the channels ranked 2nd-3rd as backup sending channels.

4. The SMS channel intelligent matching system based on multi-dimensional weight allocation according to claim 3, characterized in that: Multi-dimensional indicator calculation logic; Obtain the pre-marked unit SMS price of the channel as a cost indicator; Extract the total number of SMS messages sent within a set time zone before the current time point; count the number of SMS messages successfully delivered to end users within the total number of SMS messages sent, and calculate its proportion in the total number of messages sent, as the channel arrival rate; count the number of user complaints within the total number of SMS messages sent, and calculate its proportion in the total number of messages sent, as the channel complaint rate; combine the channel arrival rate and channel complaint rate with preset arrival weights and complaint weights for comprehensive processing to obtain quality indicators; The average time taken to send SMS messages within a set time zone is used as a timeliness indicator. The maximum concurrent capacity of the preset channel is calculated, the actual transmission volume of the channel within a unit of time before the current time point is counted, and the ratio between the actual transmission volume and the maximum concurrent capacity is calculated as a load balancing indicator.

5. The SMS channel intelligent matching system based on multi-dimensional weight allocation according to claim 1, characterized in that: The logic for obtaining the unit price of the quotation received by the customer; The application message sending quantity in the customer short message sending request is parsed, a preset sending quantity distribution ratio and a channel unit short message unit price are combined, and a formula outputs the average unit cost of the current customer short message sending request ; wherein a unit short message price of the main channel, a preset distribution sending amount of the main channel, obtained by multiplying the main channel execution proportion in the distribution proportion of the application short message sending amount and the preset sending amount. and unit price of short message for two standby channels, and bottom line sending amount allocated for two standby channels; obtained by multiplying the application short message sending amount and the standby channel execution proportion in the preset sending amount allocation proportion; the quoted unit price applied to the current customer short message sending request is obtained by multiplying the average unit cost and the premium coefficient.

6. The SMS channel intelligent matching system based on multi-dimensional weight allocation according to claim 5, characterized in that: Logic for obtaining the premium coefficient; The premium coefficient is determined by combining customer contract level, historical cooperation records, and cooperation duration from customer cooperation data; Analyze the customer's contract level and set a set of benchmark coefficients for member customers, standard customers, and low-cost demand customers respectively; The duration of customer cooperation is converted into cumulative months of cooperation. The mapping rule between the cumulative number of months of cooperation and the duration value coefficient is used to convert it into a duration value coefficient; Extract the current customer's cooperation contribution percentage, revenue contribution percentage, and timely payment percentage from the customer's historical cooperation records and process them comprehensively to obtain the current customer's cooperation performance index. The mapping rule between the cooperation performance index and the customer incentive coefficient is used to convert it into a customer incentive coefficient; Premium coefficient = Benchmark coefficient × Duration value coefficient × Customer incentive coefficient.

7. The SMS channel intelligent matching system based on multi-dimensional weight allocation according to claim 5, characterized in that: Gross profit margin is obtained by subtracting the average unit cost from the quoted unit price and then dividing by the quoted unit price. Preset the minimum gross profit margin requirements corresponding to different customer contract levels; The determined gross profit margin is compared with the preset minimum gross profit margin corresponding to the customer's contract level. If it is higher than the minimum gross profit margin, the verification result is determined to meet the gross profit requirement; otherwise, the verification result is determined to not meet the gross profit requirement.

8. The SMS channel intelligent matching system based on multi-dimensional weight allocation according to claim 1, characterized in that: Logic for identifying the degree of abnormality in the main channel's operating status; Monitor the degree of anomalies in the main channel by sending real-time status and channel concurrency ratio; The real-time status of the sent data includes the window send success rate, the window send arrival rate, and the window response timeout rate. The abnormal state coefficient of the main channel is output by comprehensively processing the window sending success rate, window sending arrival rate, window response timeout rate, and concurrency ratio.

9. The SMS channel intelligent matching system based on multi-dimensional weight allocation according to claim 8, characterized in that: The calculation logic for sending real-time status and the proportion of channel concurrency; The number of SMS messages that successfully received receipts in the main channel within the current sliding window is counted, and their proportion in the total number of SMS messages submitted for sending in the main channel within the current sliding window is calculated to obtain the window sending success rate. The number of SMS messages successfully delivered to end users within the current sliding window is counted, and the percentage of these SMS messages in the total number of SMS messages submitted for sending within the current sliding window is calculated to obtain the window delivery rate. Within the sliding window, count the number of times the response time of a single request in the main channel exceeds the preset timeout threshold, calculate its proportion in the total number of submissions within the window, and obtain the window response timeout rate; Within the sliding window, the average actual concurrency per unit time of the main channel is statistically analyzed, and its proportion in the preset maximum concurrency capacity of the main channel is calculated to obtain the concurrency proportion.

10. The SMS channel intelligent matching system based on multi-dimensional weight allocation according to claim 7, characterized in that: The specific process of automated channel switching; If the abnormal state coefficient of the main channel in the current sliding window is higher than the preset abnormal threshold coefficient, the pre-edited diversion strategy will be executed in the next sliding window. The task volume of the next sliding window main channel is evenly distributed to the first and second backup channels for execution, based on a preset proportion of the task volume to be sent.

Images