A method for dynamic configuration and flexible calculation of recruitment indicators based on business scenarios

Abstract

This invention relates to the field of intelligent data processing technology for procurement, specifically a method for dynamic configuration and flexible calculation of procurement indicators based on business scenarios. The method includes: receiving procurement business data streams and extracting features to identify business scenario types; matching corresponding calculation rule templates from an indicator library according to the scenario type; obtaining dynamic weight configuration parameters including time decay factors and business urgency coefficients, and constructing an indicator calculation model for the current scenario based on these parameters; using the model to calculate results from real-time business data and performing business rationality verification; triggering rule adjustments if the verification fails. This technology achieves intelligent matching of indicator calculation rules with changing business scenarios, and through dynamic parameters, the calculation results adapt to changes in time and business urgency, improving the real-time performance, adaptability, and management accuracy of indicator calculation.

Description

Technical Field

[0001] This invention relates to the field of intelligent processing technology for procurement data, and in particular to a method for dynamic configuration and flexible calculation of procurement indicators based on business scenarios. Background Technology

[0002] Existing procurement indicator calculation systems generally employ predefined and fixed calculation rules and weighting models. These models are typically set based on general management logic or historical benchmarks, and once their calculation dimensions, formula structures, and parameter weights are determined, they are applied indiscriminately to various procurement activities. This approach directly binds complex business data flows to a single, static calculation logic.

[0003] The drawback of this static rule system lies in its lack of scenario adaptability and dynamic responsiveness. Because it cannot identify the business intent and context behind specific procurement activities, the same rigid model struggles to accurately reflect the core management needs under different business objectives, resulting in a weak correlation between the output results and real-time business status. Furthermore, its weighting system is fixed and cannot adjust itself based on the time attributes of the data or the urgency of business needs, making it impossible for the calculated indicators to effectively capture and quantify dynamic changes in the business process.

[0004] A technical solution is needed that enables the indicator calculation rules to intelligently adapt to changing business scenarios and allows its key parameters to dynamically evolve based on time effects and business priorities, thereby improving the real-time relevance and management guidance value of the indicator results. Summary of the Invention

[0005] The purpose of this invention is to address the shortcomings of existing technologies by proposing a dynamic configuration and flexible calculation method for procurement indicators based on business scenarios.

[0006] To achieve the above objectives, the present invention adopts the following technical solution: a method for dynamic configuration and flexible calculation of procurement indicators based on business scenarios, comprising:

[0007] Receive procurement business data streams from different business systems, wherein the procurement business data streams include supplier information, material information, price information and transaction history information;

[0008] The business scenario features of the procurement business data stream are extracted to identify the business scenario type to which the current procurement business belongs;

[0009] Based on the identified business scenario type, the corresponding procurement indicator calculation rule template is matched from the indicator library;

[0010] Obtain dynamic weight configuration parameters for the current business scenario, including time decay factor and business urgency coefficient;

[0011] Based on the matched procurement indicator calculation rule template and dynamic weight configuration parameters, a procurement indicator calculation model for the current business scenario is constructed.

[0012] The procurement indicator calculation model is used to calculate real-time procurement business data and generate procurement indicator calculation results.

[0013] The business rationality of the calculation results of the procurement indicators is verified. If the verification fails, the indicator calculation rule adjustment process is triggered.

[0014] As a further aspect of the present invention, business scenario features are extracted from the procurement business data stream to identify the business scenario type to which the current procurement business belongs, specifically including:

[0015] Key business characteristics are extracted from the procurement business data stream, including procurement amount range, material classification characteristics, supplier level characteristics, and procurement cycle characteristics.

[0016] Establish a business scenario feature vector, and quantify the key business features into the various dimensional components of the feature vector;

[0017] Calculate the similarity between the business scenario feature vector and each template vector in the preset scenario feature template library;

[0018] The business scenario type corresponding to the scenario feature template with the highest similarity is selected as the recognition result.

[0019] As a further aspect of the present invention, based on the identified business scenario type, a corresponding procurement indicator calculation rule template is matched from the indicator library, specifically including:

[0020] Establish a mapping table between business scenario types and procurement indicator calculation rule templates in the indicator library;

[0021] Query the mapping relationship table based on the identified business scenario type to obtain the corresponding basic indicator calculation rule set;

[0022] Extract indicators from historical procurement data for similar business scenarios to calculate optimization parameters;

[0023] The basic indicator calculation rule set and indicator calculation optimization parameters are integrated to generate a procurement indicator calculation rule template.

[0024] As a further aspect of the present invention, obtaining dynamic weight configuration parameters under the current business scenario specifically includes:

[0025] Analyze the time characteristics of the current business scenario and calculate the time decay factor from the reference time;

[0026] Assess the urgency of the current procurement process and calculate the urgency coefficient based on the business priority rules;

[0027] Obtain the volatility index of the current market environment as an external environment adjustment factor;

[0028] The time decay factor, business urgency coefficient, and external environment adjustment factor are combined to form a dynamic weight configuration parameter set.

[0029] As a further aspect of the present invention, based on the matched procurement indicator calculation rule template and dynamic weight configuration parameters, a procurement indicator calculation model for the current business scenario is constructed, specifically including:

[0030] The logical relationships of indicator calculation in the procurement indicator calculation rule template are analyzed.

[0031] Each parameter in the dynamic weight configuration parameter set is embedded into the indicator calculation logic relationship;

[0032] Establish a dynamic adjustment mechanism for indicator calculation parameters so that the model can automatically adjust the calculation parameters according to changes in business data;

[0033] The logical integrity of the constructed procurement indicator calculation model is verified to ensure the closed-loop nature of the model calculation process.

[0034] As a further aspect of the present invention, the procurement index calculation model is used to calculate real-time procurement business data to generate procurement index calculation results, specifically including:

[0035] Perform data cleaning and standardization on real-time procurement business data to ensure that the data format meets the model input requirements;

[0036] The processed real-time procurement business data is input into the procurement indicator calculation model;

[0037] Calculate the value of each sub-indicator layer by layer according to the logical relationship of the indicators in the model;

[0038] The comprehensive procurement index value is obtained by weighting and synthesizing each sub-index based on dynamic weight configuration parameters.

[0039] The intermediate and final results of the calculation process are recorded to form a complete calculation trajectory.

[0040] As a further aspect of the present invention, the calculation results of the procurement indicators are subjected to business rationality verification, specifically including:

[0041] Establish a reasonable value range library for procurement indicator results, which includes reasonable threshold values ​​for indicators under different business scenarios;

[0042] Compare the current procurement indicators with the reasonable thresholds for the corresponding business scenarios;

[0043] When the calculated result of the indicator exceeds the reasonable threshold range, the abnormal result analysis process is initiated.

[0044] Analyze the specific reasons for abnormal indicator calculation results, including data abnormalities, rule abnormalities, or parameter abnormalities.

[0045] As a further aspect of the present invention, when the verification fails, a process for adjusting the indicator calculation rules is triggered, specifically including:

[0046] Based on the analysis of abnormal results and the resulting causes of the abnormalities, determine the indicator calculation rule components that need to be adjusted.

[0047] Search the historical adjustment record database for rule adjustment schemes in similar scenarios;

[0048] The found rule adjustment schemes are adaptively optimized based on the current business characteristics;

[0049] Generate suggestions for adjusting the indicator calculation rules, including adjustment values ​​for rule parameters and optimization points for calculation logic.

[0050] As a further aspect of the present invention, after triggering the indicator calculation rule adjustment process, the following is further included:

[0051] The adjusted indicator calculation rules were simulated and tested, and the adjustment effect was verified using historical procurement business data.

[0052] When the simulation test results meet expectations, the adjusted indicator calculation rules will be updated in the indicator library.

[0053] Establish a rule version management mechanism to record detailed information and effective time of each rule adjustment;

[0054] The system notifies relevant business systems that the indicator calculation rules have been updated and provides a document explaining the rule changes.

[0055] As a further aspect of the present invention, after updating the index calculation rules, it further includes:

[0056] Monitor the effectiveness of the new rules in actual business operations and collect feedback information on business usage;

[0057] Regularly analyze the adaptability of the indicator calculation rules and assess the degree of matching between the rules and business scenarios;

[0058] Establish a rule optimization and iteration mechanism to continuously improve the calculation rules for procurement indicators based on business changes;

[0059] Maintain lifecycle management records for indicator calculation rules to ensure the traceability of rule changes.

[0060] Compared with the prior art, the advantages and positive effects of the present invention are as follows:

[0061] By automating feature analysis and scenario type identification of real-time procurement data, the system achieves intelligent association between calculation rules and specific business contexts. Based on the identified scenarios, the system matches and calls corresponding dedicated calculation rule templates from the indicator library, replacing the mechanical application of general rules. This process enables indicator calculations to automatically adapt to the inherent logic and management objectives of different procurement activities, generating evaluation results that closely align with the actual management intent of specific scenarios, thus enhancing the relevance and interpretability of the analysis.

[0062] By introducing a dynamically configurable time decay factor and a business urgency coefficient into the computational model, the core computational parameters gain the ability to evolve dynamically. The time decay factor automatically adjusts its influence weight based on the timeliness of the data, ensuring that the model's calculation results continuously reflect the latest business performance trends and reduce the interference of outdated data. The business urgency coefficient, on the other hand, adjusts the weight of relevant evaluation dimensions in real time based on the urgency of specific tasks. The synergistic effect of these two mechanisms drives the indicator calculation results to respond sensitively to dynamic changes in the business environment and time dimension, outputting quantitative evaluations with time sensitivity and priority awareness, providing more timely and context-relevant data for decision-making. Attached Figure Description

[0063] Figure 1 This is a flowchart of the dynamic configuration and flexible calculation method for procurement indicators based on business scenarios as described in this invention;

[0064] Figure 2 A flowchart for feature extraction from business scenarios;

[0065] Figure 3 A flowchart for matching the template of the rules for calculating procurement indicators. Detailed Implementation

[0066] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0067] In the description of this invention, it should be understood that the terms "length," "width," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," and "outer," etc., indicating orientation or positional relationships, are based on the orientation or positional relationships shown in the accompanying drawings and are only for the convenience of describing the invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of the invention. Furthermore, in the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified.

[0068] See Figure 1 The system receives procurement data streams from different business systems, including supplier information, material information, price information, and transaction history. It extracts business scenario features from the received data streams to identify the current procurement scenario type. Based on the identified scenario type, it matches the corresponding procurement indicator calculation rule template from a pre-built indicator library. Simultaneously, it acquires dynamic weight configuration parameters for the current business scenario, including time decay factors and business urgency coefficients. Based on the matched procurement indicator calculation rule template and the acquired dynamic weight configuration parameters, it constructs a procurement indicator calculation model for the current business scenario. The constructed model is used to calculate the real-time incoming procurement data, generating procurement indicator calculation results. The generated results undergo business rationality verification; if the verification fails, the system triggers an indicator calculation rule adjustment process.

[0069] See Figure 2In one embodiment of the present invention, the process of extracting key business features from the procurement business data stream involves parsing and transforming the original data. For example, a procurement business data stream for purchasing office supplies includes supplier information "Supplier A", material information "printing paper", price information "unit price 10 yuan" and transaction history information "number of purchases in the past year 5 times". The extraction of key business features is based on the specific values ​​and classification of these data. The procurement amount range is determined by mapping the total procurement amount to preset amount segments such as "0-1000 yuan" and "1001-5000 yuan". The material classification feature maps "printing paper" to the "office consumables" category according to the material coding system. The supplier level feature classifies "Supplier A" as a "first-level supplier" according to the supplier performance score. The procurement cycle feature is calculated as "7 days" based on the number of days from order generation to expected delivery. When establishing a business scenario feature vector, the key business features mentioned above are quantified into various dimensional components of the feature vector. For example, the purchase amount range is represented as a binary vector component using one-hot encoding, the material classification feature is represented as an integer component using category encoding, the supplier level feature is represented as a numerical component using ordinal encoding, and the purchase cycle feature is directly represented as a continuous component using the number of days. These components together constitute a multi-dimensional business scenario feature vector, which is used to characterize the overall scenario attributes of the current procurement business.

[0070] In some embodiments, the preset scenario feature template library contains template vectors corresponding to multiple predefined business scenario types. For example, the template vector for the "daily office procurement" scenario has a range of component values, while the template vector for the "emergency production material procurement" scenario has a different distribution of component values. When calculating the similarity between the business scenario feature vector and each template vector in the preset scenario feature template library, cosine similarity is used to measure the directional consistency between the two vectors. The similarity calculation formula is as follows:

[0071]

[0072] in: Represents the feature vector of the business scenario. This represents a preset scene feature template vector. This represents the total number of dimensions of the feature vector. and Representing vectors respectively and The Each component. Data comparison is reflected in the similarity calculation between the feature vectors processed from data of different procurement businesses and the template vectors. For example, the feature vector of a business scenario with a purchase amount of "500 yuan", material category of "office consumables", supplier level of "level one", and procurement cycle of "7 days" has a similarity calculation result of 0.92 with the template vector of "daily office procurement" and 0.31 with the template vector of "emergency production material procurement". The numerical comparison can intuitively show the degree of similarity between the vectors.

[0073] Optionally, after the similarity calculation is completed, the system iterates through the similarity values ​​of all template vectors and sorts them, selecting the business scenario type corresponding to the scenario feature template with the highest similarity as the recognition result. For example, among the similarity values ​​mentioned above, 0.92 is the highest value and corresponds to the "daily office procurement" template. Therefore, the business scenario type of the current procurement business is identified as "daily office procurement". In some embodiments, the quantization process of the feature vector also includes normalization processing, converting components with different dimensions into a unified scale. For example, the number of days for the procurement cycle feature is divided by a baseline cycle value to convert it into a value between 0 and 1, and the ordinal code for the supplier level feature is divided by the maximum level number to convert it into a proportional value. This helps to improve the accuracy and stability of the similarity calculation.

[0074] It is understandable that the dimensional component design of the business scenario feature vector needs to cover the core attributes of the procurement business. Purchase amount range, material classification characteristics, supplier level characteristics, and procurement cycle characteristics serve as basic dimensions, and other dimensions such as geographical location characteristics or contract type characteristics can be added to enhance the granularity of identification. In specific implementation, the construction of the preset scenario feature template library is based on cluster analysis of historical procurement data. The template vectors represent the central or typical characteristics of various business scenarios. Similarity calculation uses distance or angle measures in the vector space model. Cosine similarity is one of the commonly used methods, but it can also be replaced by the reciprocal of Euclidean distance or other similarity functions. Optionally, weighting coefficients are introduced during the similarity calculation process to adjust the importance of different feature dimensions. For example, the weight of the purchase amount range dimension is set to 0.3, the weight of the material classification characteristic dimension is set to 0.2, the weight of the supplier level characteristic dimension is set to 0.25, and the weight of the procurement cycle characteristic dimension is set to 0.25. The weighted similarity calculation can better reflect actual business priorities.

[0075] See Figure 3In one embodiment of the present invention, a mapping table is established in the indicator library to map business scenario types to procurement indicator calculation rule templates. The mapping table is stored in a structured manner. For example, one table maps the "daily office procurement" business scenario type to a basic indicator calculation rule set named "routine procurement assessment," and the "emergency equipment repair procurement" business scenario type to a basic indicator calculation rule set named "emergency procurement assessment." The mapping table is queried based on the identified business scenario type to obtain the corresponding basic indicator calculation rule set. For example, if the system identifies the current business scenario type as "emergency equipment repair procurement," the basic indicator calculation rule set named "emergency procurement assessment" is obtained after querying the mapping table. This basic indicator calculation rule set includes the calculation logic and original formulas for basic indicators such as price fluctuation tolerance, supplier response speed, and historical cooperation stability. Optimization parameters for indicator calculation under similar business scenarios are extracted from historical procurement data. For example, from historical data of all "emergency equipment repair procurement" scenarios in the past year, statistical analysis reveals that the correction coefficient for price fluctuation tolerance is 1.2, the base weight value for supplier response speed is 0.4, and the threshold parameter for historical cooperation stability is 0.7. These statistical values ​​constitute the optimization parameters for indicator calculation. The basic indicator calculation rule set and indicator calculation optimization parameters are integrated to generate a procurement indicator calculation rule template. The integration operation involves substituting the specific values ​​of the indicator calculation optimization parameters into the original formulas and logic of the basic indicator calculation rule set. For example, the coefficient in the original calculation formula of price fluctuation tolerance is replaced with 1.2, and the weight value of supplier response speed is set to 0.4, thereby generating a procurement indicator calculation rule template with preliminary parameter optimization for the "emergency equipment repair and procurement" scenario.

[0076] In some embodiments, the time characteristics of the current business scenario are analyzed and a time decay factor from a reference time is calculated. The reference time can be set as a fixed date, such as the start date of a fiscal year, or a dynamic date, such as the date of the most recent purchase of similar materials. The time decay factor is obtained by calculating the time difference between the current time and the reference time and applying a decay function. For example, if the reference time is set to "January 1, 2023" and the current time is "October 1, 2023", the time difference is 273 days, and an exponential decay function is used. Perform calculations, where Indicates the time decay factor. This represents the difference in days between the current time and the reference time. The decay rate coefficient is a preset constant. Data comparison reflects the difference in the time decay factor value caused by different time differences. For example, the time decay factor calculated when the time difference is 273 days is 0.76, and the time decay factor calculated when the time difference is 30 days is 0.95. The system assesses the urgency of the current procurement business and calculates the business urgency coefficient based on business priority rules. The business priority rule can be a mapping table that maps business attributes such as "production line shutdown" and "safety risk level" to numerical values. For example, the rule specifies that the urgency of "procurement that causes production line shutdown" is 0.9, and the urgency of "routine spare parts procurement" is 0.3. The system matches the rules with the description information of the current procurement business and obtains the business urgency coefficient.

[0077] Optionally, the volatility index of the current market environment can be obtained as an external environment adjustment factor. This index can be obtained from an external economic data interface or calculated using the variance of price fluctuations of major materials monitored internally. For example, if the current steel price volatility index is 1.15 and the copper price volatility index is 1.08 obtained from an external interface, and the current procurement business involves steel, then the external environment adjustment factor is set to 1.15. The time decay factor, business urgency coefficient, and external environment adjustment factor are combined to form a dynamic weight configuration parameter set. This combination can be achieved by storing the three factors in vector form. For example, the dynamic weight configuration parameter set can be represented as follows:

[0078]

[0079] in: This represents the set of dynamic weight configuration parameters. Indicates the time decay factor. This indicates the urgency level of the business. This represents the external environment adjustment factor. This parameter set will be used for dynamic weight adjustment in the subsequent indicator calculation model.

[0080] It is understandable that the establishment of the mapping relationship table relies on the prior business scenario analysis and rule definition. The basic indicator calculation rule set contains the static framework for indicator calculation, while the indicator calculation optimization parameters come from the mining and learning of historical data. The fusion process is essentially injecting empirical data into the static framework to achieve scenario-based customization of rules. In specific implementation, the calculation function of the time decay factor is not limited to exponential decay; linear decay or other monotonically decreasing functions can also be used. The selection strategy of the base time directly affects the meaning of the time decay factor. For example, if the contract validity period start date is selected as the base time, the time decay factor reflects the progress of the contract life cycle. The calculation of the business urgency coefficient relies on clear and executable business priority rules, which need to transform vague business descriptions into quantifiable values.

[0081] In some embodiments, the extraction process of optimization parameters for indicator calculation may involve machine learning models, such as using regression models to analyze the relationship between various indicator parameters in historical data and final procurement performance, thereby deriving optimization parameter values. It is understood that the combination factors of the dynamic weight configuration parameter set can be expanded according to business complexity, such as adding seasonal adjustment factors or geographical region factors, but the core factors—time decay factors and business urgency coefficients—form the basis for dynamic adjustment. The introduction of external environment adjustment factors enables the calculation of procurement indicators to respond to changes in the external market. For example, during periods of sharp fluctuations in raw material prices, the value of external environment adjustment factors increases, thereby amplifying the weight of price stability-related indicators in the indicator calculation.

[0082] Optionally, each factor in the dynamic weight configuration parameter set can be normalized before participating in subsequent calculations, so that the numerical range of all factors is on the same dimension. For example, the values ​​of the time decay factor, business urgency coefficient, and external environment adjustment factor can all be scaled to the range of 0 to 1. Data comparison is also reflected in the differences in the values ​​of the dynamic weight configuration parameter set obtained under different business scenarios. For example, a "routine office procurement" scenario may calculate a time decay factor of 0.95, a business urgency coefficient of 0.3, and an external environment adjustment factor of 1.02, while a "urgent equipment repair procurement" scenario calculates a business urgency coefficient of 0.9, and other factors may also be different. This difference directly reflects the dynamism of the configuration parameters and their relevance to the scenario.

[0083] In one embodiment of the present invention, parsing the indicator calculation logic relationship in the procurement indicator calculation rule template involves interpreting the structured content of the template. For example, a procurement indicator calculation rule template for "comprehensive supplier evaluation" includes three sub-indicators: price score, on-time delivery rate score, and quality pass rate score. The indicator calculation logic relationship is defined as adding the price score and the on-time delivery rate score, and then multiplying it by the quality pass rate score to obtain a preliminary comprehensive score. The parsing process involves identifying the operation logic and order of this addition and multiplication. Various parameters from the dynamic weight configuration parameter set are embedded into the indicator calculation logic relationship. For example, the dynamic weight configuration parameter set includes a time decay factor. Business urgency coefficient and external environmental regulators Embedding operations are to incorporate the business urgency coefficient. Multiplying the on-time delivery rate score amplifies the importance of delivery speed during urgent procurement, while adjusting for external environmental factors. Multiply by the price score to reflect the impact of market price fluctuations, and incorporate a time decay factor. This serves as a decay coefficient for historical transaction data when calculating the quality pass rate score. A dynamic adjustment mechanism for the indicator calculation parameters is established, enabling the model to automatically adjust these parameters based on changes in business data. For example, when real-time inflow price data exceeds 20% of the historical average price for several consecutive periods, the stability weight parameter in the price score calculation formula is automatically increased. From 0.5 to 0.7, if all recent delivery data from a supplier is on time, the penalty coefficient for historical lateness records in the delivery on-time rate calculation will be automatically reduced. The logical integrity of the constructed procurement indicator calculation model is verified. The verification process checks whether all input data items have corresponding calculation paths to reach the final output, and whether all embedded dynamic parameters are correctly referenced in the calculation logic. This ensures that there are no undefined variables or unreachable calculation branches in the model, thereby guaranteeing the closed-loop nature of the model's calculation process.

[0084] In some embodiments, real-time procurement data undergoes data cleaning and standardization. Data cleaning includes removing obviously erroneous data records, such as negative prices or incorrect date formats, and filling in missing field values, such as filling missing price information with the average price of similar materials. Standardization includes converting the currencies of quotations from different suppliers to a base currency and converting the quantities of materials in different units of measurement to a standard unit. The processed real-time procurement data is then input into the procurement indicator calculation model. The input data format is consistent with the input interface defined by the model; for example, if the model requires input of price, delivery date, and historical quality inspection report number, the processed data is provided in this order and type. The values ​​of each sub-indicator are calculated layer by layer according to the indicator calculation logic in the model. For example, the input price data, base price, and external environment adjustment factors in the dynamic weight configuration parameter set are first used as the basis for calculation. Calculate price rating Then, based on the input delivery date data, promised delivery date, and business urgency coefficient... Calculate on-time delivery rate score Next, based on the entered historical quality inspection report number, the historical pass rate is retrieved and the time decay factor is applied. Calculate the quality pass rate score The comprehensive procurement index value is obtained by weighting and combining the various sub-indicators based on dynamic weight configuration parameters. The weighted combination formula is as follows:

[0085]

[0086] in: This represents the comprehensive procurement index value. Indicates price rating, The on-time delivery rate score is indicated. This indicates the quality pass rate score. This indicates the weight of the price rating. The weighting of the on-time delivery rate score indicates the weighting of the on-time delivery rate score. The weights represent the scores for the quality pass rate. It may be related to the business urgency coefficient. Positive correlation. Intermediate and final results are recorded during the calculation process to form a complete calculation trajectory. The recorded content includes, but is not limited to, the original and cleaned input data, the intermediate values ​​obtained from each sub-indicator calculation, the specific values ​​of the dynamic parameters used, the calculation results of each step of the weighted synthesis formula, and the final comprehensive procurement index value. These records are stored in a structured log format.

[0087] Optionally, the dynamic adjustment mechanism for the indicator calculation parameters can be based on threshold trigger rules, such as when the quality pass rate score is... If the result of three consecutive calculations is below the threshold of 60, an adjustment mechanism will be automatically triggered to increase the weight of the quality pass rate score in the weighted synthesis. Alternatively, the query time range for high-quality test reports can be adjusted. Logical integrity verification also includes checking for circular references, such as verifying whether the price score calculation incorrectly references variables that depend on the final comprehensive procurement index value. Standardization processing during data cleaning also involves data normalization, scaling raw data of different dimensions to the [0,1] interval to facilitate model calculations.

[0088] It is understandable that calculating the values ​​of each sub-indicator layer by layer is a concrete manifestation of executing the model's calculation logic. Each layer's calculation strictly adheres to the parsed logical relationships and may utilize embedded dynamic parameters. The calculation order of the sub-indicators is determined by the dependencies in the logical relationships. Weighted synthesis is a crucial step in integrating multiple sub-indicator evaluation dimensions to obtain a single comprehensive evaluation. Dynamic weight configuration parameters influence the weights... , , The value is used to dynamically adjust the computation strategy. Recording the complete computation trajectory is not only used for auditing and tracing, but also provides detailed data support for subsequent model optimization and anomaly analysis.

[0089] In practical implementation, data comparison in example scenarios can be reflected in the differences in calculation results for the same supplier under different dynamic parameters. For example, for supplier X, in a routine procurement, the business urgency coefficient... The calculated on-time delivery rate score The comprehensive procurement index value is 85. The urgency coefficient was 80, while in an emergency procurement, the business urgency coefficient was... With other input data remaining unchanged, the calculated on-time delivery rate score The coefficient amplification causes significant changes, which in turn affects the overall procurement index value. The value rose to 88, demonstrating the direct impact of dynamic weight configuration parameters on the calculation results. It's understandable that data cleaning and standardization are prerequisites for ensuring the accuracy of model calculations; invalid or poorly formatted data can lead to errors in sub-indicator calculations, resulting in meaningless comprehensive procurement indicator values.

[0090] In one embodiment of the present invention, a reasonable value range library for procurement indicator results is established. This library is stored in the form of a structured data table, containing reasonable threshold values ​​for indicators under different business scenarios. For example, for the "daily office procurement" business scenario, the reasonable threshold range for the price competitiveness indicator is set to [60, 90], and the reasonable threshold range for the on-time delivery rate indicator is set to [70, 95]. For the "emergency equipment repair procurement" business scenario, the reasonable threshold range for the price competitiveness indicator may be widened to [50, 85], while the reasonable threshold range for the on-time delivery rate indicator is tightened to [85, 99]. The current procurement indicator calculation result is compared with the reasonable threshold value for the corresponding business scenario. For example, in an "emergency equipment repair procurement" business scenario, the calculated comprehensive procurement indicator value for Supplier A is 45. However, the system queries the reasonable value range library and finds that the lower limit of the reasonable threshold value for this comprehensive indicator under the "emergency equipment repair procurement" scenario is 50. The comparison reveals that the value 45 is lower than the lower limit of 50. When the calculated result of an indicator exceeds a reasonable threshold range, an anomaly analysis process is initiated. The process first identifies the abnormal indicator, such as an abnormal value in the comprehensive procurement indicator. Then, it traces the constituent sub-indicators of this comprehensive indicator, analyzing the values ​​of sub-indicators such as price competitiveness and on-time delivery rate, and comparing them with their respective reasonable thresholds to pinpoint the root cause of the anomaly. The specific reasons for the abnormal indicator calculation results are analyzed. These reasons may include data anomalies, rule anomalies, or parameter anomalies. Data anomalies refer to errors or extreme values ​​in the basic procurement business data input into the model. Rule anomalies refer to the calculation logic in the procurement indicator calculation rule template not conforming to the current business reality. Parameter anomalies refer to inappropriate parameter values ​​in the dynamic weight configuration parameter set. See Table 1.

[0091] Table 1: Reasonable Thresholds for Procurement Indicators under Different Business Scenarios

[0092]

[0093] In some embodiments, the anomaly analysis determines the component of the indicator calculation rule that needs adjustment. For example, if the anomaly is determined to be a parameter anomaly, specifically, the business urgency coefficient in the dynamic weight configuration parameter set is set to 0.3 in the current "emergency equipment repair and procurement" scenario, while in successful cases of similar scenarios in the past, this coefficient is generally higher than 0.7. The component that needs adjustment is the calculation rule of the business urgency coefficient. The system searches for rule adjustment solutions for similar scenarios in the historical adjustment record database. This database stores rule adjustment measures taken in the past for various anomalies and their effects. For example, if a record of "parameter anomaly: business urgency coefficient is too low" is found, the corresponding adjustment solution is "increase the base value in the business urgency coefficient calculation formula by 0.4". The found rule adjustment solution is then adaptively optimized based on the current business characteristics. For example, if the current business characteristics show that this "emergency equipment repair and procurement" involves a critical production line and the business urgency is higher than in historical cases, the found adjustment solution "increase by 0.4" is optimized to "increase by 0.5". Generate suggestions for adjusting the indicator calculation rules. These suggestions include adjustments to rule parameters and optimization points for the calculation logic. For example, a suggested adjustment might be: "Adjust the base value parameters in the business urgency coefficient calculation rule." "Adjusted from 0.3 to 0.8", and "Added recognition of the 'critical production line' label in the calculation logic and added additional weights".

[0094] Optionally, the establishment of a reasonable threshold range library can be based on statistical analysis of historical procurement indicator results. For example, the 10th and 90th percentiles of all comprehensive indicator values ​​for the "routine office procurement" scenario over the past three years can be calculated as the lower and upper limits of the reasonable range, respectively. When comparing the current procurement indicator calculation results with the reasonable thresholds, various judgment logics can be used to calculate the degree of deviation. :

[0095]

[0096] in: This indicates the current procurement indicator calculation results. This represents the lower limit of a reasonable threshold range. This represents the upper limit of a reasonable threshold range, when or or If the result exceeds a certain set threshold, it is considered to be outside the reasonable range. The abnormal result analysis process can be automated. The system sequentially checks the integrity of the input data, the version of the calculation rules, and the parameter configuration, and compares it with historical normal calculation records to locate the most likely cause.

[0097] It's understandable that data anomalies might stem from data collection errors or transmission distortions, such as price data being incorrectly labeled as "ten thousand yuan" when it should actually be "yuan." Rule anomalies might arise from changes in business scenarios rendering old rules inapplicable; for example, new environmental regulations might have rendered existing supplier evaluation rules inapplicable due to a lack of consideration for "green supplier" certification indicators. Parameter anomalies might be caused by sudden changes in the external environment, rendering preset parameters ineffective; for example, a sudden international logistics crisis might have rendered existing on-time delivery rate weighting coefficients unable to reflect actual risks. Searching for solutions in the historical adjustment record database is a case-based reasoning process; the system needs to match the characteristics of the current anomaly with the characteristics of anomalies in historical records.

[0098] In practice, data comparison is reflected in the direct numerical comparison between outliers and reasonable thresholds, as well as in the comparison of parameter values ​​before and after adjustment. For example, the business urgency coefficient might be adjusted from 0.3 at the time of anomaly detection to a suggested 0.8. When generating rule adjustment suggestions, in addition to directly modifying parameter values, it may also be suggested to add new calculation dimensions. For example, when parameter anomalies occur frequently and are strongly correlated with "supply chain risk level," the adjustment suggestion might include "adding a 'supply chain risk coefficient' sub-indicator to the comprehensive procurement indicator calculation model." It is understandable that adaptive optimization based on current business characteristics is necessary, because directly applying historical adjustment schemes may not be a perfect match. The optimization process needs to combine the specific business context information of this anomaly, such as the purchase amount, material criticality, and supplier location, to calibrate the magnitude or scope of historical adjustment schemes.

[0099] In one embodiment of the present invention, the adjusted indicator calculation rules are simulated and tested, and the adjustment effect is verified using historical procurement business data. For example, based on the adjustment suggestion of "adjusting the base value parameter in the business urgency coefficient calculation rule from 0.3 to 0.8", the adjusted procurement indicator calculation rules are generated. The simulation test selects historical procurement business data of all "emergency equipment repair procurement" scenarios in the past six months as input, and recalculates the comprehensive procurement indicator value of each historical procurement event using the rules before and after adjustment. The adjustment effect is evaluated by comparing the distribution of the calculation results before and after adjustment with the consistency of historical actual procurement decisions. For example, in the historical data, 100 records had indicator values ​​below the reasonable threshold calculated under the rules before adjustment, but the actual procurement result was a successful cooperation. Under the rules after adjustment, the indicator values ​​of these records rose to the reasonable threshold range, and the consistency improved. Calculate an adjustment effect evaluation value:

[0100]

[0101] in: This indicates an adjustment to the effect evaluation value. This represents the total number of historical data records used for simulation testing. Indicates the first The record uses the indicator value calculated using the rules before adjustment. Indicates the first Each record uses the adjusted rules to calculate the indicator value. When the simulation test results meet expectations, the adjusted indicator calculation rules are updated in the indicator library. The standard for meeting expectations can be the adjustment effect evaluation value. If the target threshold is exceeded and the accuracy improvement rate reaches the target, the update operation involves finding the corresponding procurement indicator calculation rule template for the "emergency equipment repair and procurement" scenario in the indicator library, and replacing the original template with the new rule that includes parameter adjustments. A rule version management mechanism is established to record detailed information and the effective time of each rule adjustment. This mechanism maintains a version history list for each rule template, generating a new version number each time it is updated (e.g., upgrading from V1.2 to V1.3). Recorded information includes the version number, description of the adjustment, reason for the adjustment, summary of simulation test results, effective time, and the operator. The relevant business systems are notified that the indicator calculation rules have been updated and a rule change documentation is provided. Notification is delivered via inter-system message queues or API calls. The rule change documentation details the adjustment points, reasons for the adjustment, expected impact, and how to use the new version of the rule.

[0102] In some embodiments, the effectiveness of new rules in actual business operations is monitored and feedback information on business usage is collected. Monitoring effectiveness involves continuously calculating the distribution of indicator values ​​and the frequency of abnormal triggers generated after the new rules are applied to real-time business data, and comparing this with the performance of the old rules during the same period. For example, within one week of the new rules' implementation, the proportion of comprehensive procurement indicator values ​​for the "emergency equipment repair and procurement" scenario falling within a reasonable threshold range is statistically analyzed. Feedback information on business usage is collected through system interfaces, receiving confirmations or questions from procurement specialists regarding the indicator calculation results, or periodically collecting feedback from questionnaire surveys. The adaptability of the indicator calculation rules is analyzed periodically, and the degree of matching between the rules and business scenarios is evaluated. This periodic analysis can be conducted monthly or quarterly, and the rule matching degree is calculated during the analysis process.

[0103]

[0104] in: Indicates the rule matching degree. This indicates the number of times the indicator value calculated by the rules was adopted by the business system or did not cause any abnormal disputes within the analysis period. This represents the total number of times the rule was invoked within the analysis period. A rule optimization and iteration mechanism is established to continuously improve the procurement indicator calculation rules based on business changes. The rule optimization and iteration mechanism defines when the rule matching degree... When the data falls below a certain threshold, or a certain number of specific types of feedback are received, or a significant shift in business scenario characteristics is detected, a new round of rule analysis and adjustment process will be automatically triggered.

[0105] Optionally, the historical procurement data used in the simulation test needs to be time-sliced ​​to ensure that the data used for verification is time-independent from the data analyzed when generating the adjustment suggestions. Updating the adjusted indicator calculation rules to the indicator library can be set to take effect after manual review or automatically. The manual review process requires relevant domain experts to confirm the simulation test results and adjustment suggestions. The rule version management mechanism supports version rollback; if unforeseen major problems occur after a new rule is launched, it can be quickly switched back to the previous stable version.

[0106] Understandably, simulation testing is a crucial step in verifying the effectiveness and security of rule adjustments. Using historical data allows for the assessment of the impact of adjustments without affecting the actual production environment. The rule version management mechanism ensures the traceability and manageability of rule changes, facilitating troubleshooting and auditing. Notifying relevant business systems ensures that upstream and downstream systems are aware of changes and may adjust the behavior of their dependent rules. Monitoring the effectiveness of new rules is an ongoing process aimed at promptly identifying new deviations between rules and business practices. Regular analysis provides a quantitative assessment of rule health. The rule optimization and iteration mechanism enables the entire procurement indicator calculation method to adapt to the continuously changing business environment.

[0107] In practical implementation, data comparison is reflected in the calculation of differences in indicator values ​​before and after adjustments during the simulation testing phase, and also in the comparison between monitoring data and historical baseline data after the new rules are implemented. For example, in the first month after the new rules were implemented, the number of abnormal indicator value alarms decreased by 40% compared to the same period under the old rules. Maintaining a lifecycle management record for indicator calculation rules and ensuring the traceability of rule changes is crucial. The lifecycle management record is a complete log that records all events from rule creation, each adjustment, effectiveness, monitoring, to potential retirement. Each event includes a timestamp, event type, event details, and associated version number, ensuring that the rule status and its change history at any point in time are queryable. The specific triggering conditions for the rule optimization iteration mechanism can be configured, such as when the rule matching degree is monitored for three consecutive analysis cycles. If all values ​​are below 0.85, or if the number of similar feedback reports regarding "unreasonable indicator calculation results" in the business feedback system exceeds 10 times in a single month, the system will automatically create a rule optimization task. It can be understood that lifecycle management records are an extension and supplement to the version management mechanism; they cover a broader range of rule state changes, not just version update events. Collecting business usage feedback information can establish channels for collecting both positive and negative feedback. Positive feedback includes cases where rule calculation results are highly consistent with manual evaluations, while negative feedback includes objections to indicator calculation results submitted by system users.

[0108] The above are merely preferred embodiments of the present invention and are not intended to limit the present invention in any other way. Any person skilled in the art may make changes or modifications to the above-disclosed technical content to create equivalent embodiments that can be applied to other fields. However, any simple modifications, equivalent changes, and modifications made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the protection scope of the present invention.

Claims

1. A method for dynamically configuring and flexibly calculating procurement indicators based on business scenarios, characterized in that, Includes the following steps: Receive procurement business data streams from different business systems, wherein the procurement business data streams include supplier information, material information, price information and transaction history information; Business scenario features are extracted from the procurement business data stream to identify the business scenario type to which the current procurement business belongs; Based on the identified business scenario type, the corresponding procurement indicator calculation rule template is matched from the indicator library; Obtain dynamic weight configuration parameters for the current business scenario, including time decay factor and business urgency coefficient; Based on the matched procurement indicator calculation rule template and dynamic weight configuration parameters, a procurement indicator calculation model for the current business scenario is constructed. The procurement indicator calculation model is used to calculate real-time procurement business data and generate procurement indicator calculation results. The business rationality of the calculation results of the procurement indicators is verified. If the verification fails, the indicator calculation rule adjustment process is triggered.

2. The method for dynamic configuration and flexible calculation of procurement indicators based on business scenarios according to claim 1, characterized in that, The business scenario features of the procurement business data stream are extracted to identify the business scenario type to which the current procurement business belongs, specifically including: Key business characteristics are extracted from the procurement business data stream, including procurement amount range, material classification characteristics, supplier level characteristics, and procurement cycle characteristics. Establish a business scenario feature vector, and quantify the key business features into the various dimensional components of the feature vector; Calculate the similarity between the business scenario feature vector and each template vector in the preset scenario feature template library; The business scenario type corresponding to the scenario feature template with the highest similarity is selected as the recognition result.

3. The method for dynamic configuration and flexible calculation of procurement indicators based on business scenarios according to claim 2, characterized in that, Based on the identified business scenario type, the corresponding procurement indicator calculation rule template is matched from the indicator library, specifically including: Establish a mapping table between business scenario types and procurement indicator calculation rule templates in the indicator library; Query the mapping relationship table based on the identified business scenario type to obtain the corresponding basic indicator calculation rule set; Extract indicators from historical procurement data for similar business scenarios to calculate optimization parameters; The basic indicator calculation rule set and indicator calculation optimization parameters are integrated to generate a procurement indicator calculation rule template.

4. The method for dynamic configuration and flexible calculation of procurement indicators based on business scenarios according to claim 3, characterized in that, Retrieve dynamic weight configuration parameters for the current business scenario, specifically including: Analyze the time characteristics of the current business scenario and calculate the time decay factor from the reference time; Assess the urgency of the current procurement process and calculate the urgency coefficient based on the business priority rules; Obtain the volatility index of the current market environment as an external environment adjustment factor; The time decay factor, business urgency coefficient, and external environment adjustment factor are combined to form a dynamic weight configuration parameter set.

5. The method for dynamic configuration and flexible calculation of procurement indicators based on business scenarios according to claim 4, characterized in that, Based on the matched procurement indicator calculation rule template and dynamic weight configuration parameters, a procurement indicator calculation model for the current business scenario is constructed, specifically including: The logical relationships of indicator calculation in the procurement indicator calculation rule template are analyzed. Each parameter in the dynamic weight configuration parameter set is embedded into the indicator calculation logic relationship; Establish a dynamic adjustment mechanism for indicator calculation parameters so that the model can automatically adjust the indicator calculation parameters according to changes in business data; The logical integrity of the constructed procurement indicator calculation model is verified to ensure the closed-loop nature of the model calculation process.

6. The method for dynamic configuration and flexible calculation of procurement indicators based on business scenarios according to claim 5, characterized in that, The procurement indicator calculation model is used to calculate real-time procurement business data and generate procurement indicator calculation results, specifically including: Perform data cleaning and standardization on real-time procurement business data to ensure that the data format meets the model input requirements; The processed real-time procurement business data is input into the procurement indicator calculation model; Calculate the value of each sub-indicator layer by layer according to the logical relationship of the indicators in the model; The comprehensive procurement index value is obtained by weighting and synthesizing each sub-index based on dynamic weight configuration parameters. The intermediate and final results of the calculation process are recorded to form a complete calculation trajectory.

7. The method for dynamic configuration and flexible calculation of procurement indicators based on business scenarios according to claim 6, characterized in that, The business rationality verification of the calculation results of the procurement indicators includes: Establish a reasonable value range library for procurement indicator results, which includes reasonable threshold values ​​for indicators under different business scenarios; Compare the current procurement indicators with the reasonable thresholds for the corresponding business scenarios; When the calculated result of the indicator exceeds the reasonable threshold range, the abnormal result analysis process is initiated. Analyze the specific reasons for abnormal indicator calculation results, including data abnormalities, rule abnormalities, or parameter abnormalities.

8. The method for dynamic configuration and flexible calculation of procurement indicators based on business scenarios according to claim 7, characterized in that, When the verification fails, the indicator calculation rule adjustment process is triggered, which includes: Based on the analysis of abnormal results and the resulting causes of the abnormalities, determine the indicator calculation rule components that need to be adjusted. Search the historical adjustment record database for rule adjustment schemes in similar scenarios; The found rule adjustment schemes are adaptively optimized based on the current business characteristics; Generate suggestions for adjusting the indicator calculation rules, including adjustment values ​​for rule parameters and optimization points for calculation logic.

9. The method for dynamic configuration and flexible calculation of procurement indicators based on business scenarios according to claim 8, characterized in that, After triggering the indicator calculation rule adjustment process, the following further applies: The adjusted indicator calculation rules were simulated and tested, and the adjustment effect was verified using historical procurement business data. When the simulation test results meet expectations, the adjusted indicator calculation rules will be updated in the indicator library. Establish a rule version management mechanism to record detailed information and effective time of each rule adjustment; The system notifies relevant business systems that the indicator calculation rules have been updated and provides a document explaining the rule changes.

10. A method for dynamic configuration and flexible calculation of procurement indicators based on business scenarios according to claim 9, characterized in that, After updating the indicator calculation rules, the following further applies: Monitor the effectiveness of the new rules in actual business operations and collect feedback information on business usage; Regularly analyze the adaptability of the indicator calculation rules and assess the degree of matching between the rules and business scenarios; Establish a rule optimization and iteration mechanism to continuously improve the calculation rules for procurement indicators based on business changes; Maintain lifecycle management records for indicator calculation rules to ensure the traceability of rule changes.

Images