Tobacco material purchasing supervision compliance management method based on big data

Abstract

The invention discloses a tobacco material purchasing supervision compliance management method based on big data, and relates to the technical field of material purchasing management, and the method comprises the steps: building an evaluation knowledge graph according to competition evaluation data and cooperation evaluation data of a target supplier for material purchasing, the method comprises the following steps: extracting an evaluation trend feature map to obtain a reference feature map, evaluating competition association degrees between a target supplier and a competition supplier and between the target supplier and a cooperation supplier to obtain a competition association value and a cooperation association value, and performing evaluation trend statistics according to the competition association value and the cooperation association value to obtain a to-be-detected trend feature 1 and a to-be-detected trend feature 2; and carrying out statistics on the to-be-tested trend features 1 and 2 to obtain pre-processing feature data, and according to the historical transaction data and the pre-processing feature data, evaluating the compliance degree of the target supplier for purchasing in the tobacco industry to obtain the material purchasing compliance degree. According to the big data-based tobacco material purchasing supervision compliance management method provided by the invention, the purchasing compliance management efficiency can be improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of material procurement management, and particularly relates to a tobacco material procurement supervision compliance management method based on big data. BACKGROUND

[0002] With the rapid development of information technology, the amount of data is growing explosively, and big data technology emerges as the times require. Big data has the characteristics of large data volume, various types and fast processing speed, and can collect, store, analyze and mine massive tobacco material procurement data, so as to discover potential laws and problems and provide strong support for supervision compliance management. For example, by analyzing historical procurement data, a supplier evaluation model can be established to screen out high-quality suppliers and reduce procurement risks.

[0003] In the traditional technology, the supplier selected for tobacco material procurement is only intuitively determined according to the industry evaluation, customer evaluation and historical performance provided in the background of the profile, and the product quality and reputation of the supplier are determined, and no further interference data filtering processing is performed on whether there is false evaluation data, so that the final selection of the cooperative supplier for material procurement is not in compliance. SUMMARY

[0004] In order to overcome the shortcomings of the prior art, the present application provides a tobacco material procurement supervision compliance management method based on big data.

[0005] The tobacco material procurement supervision compliance management method based on big data provided by the present application comprises:

[0006] Step S1, according to the competitive evaluation data and cooperative evaluation data of the target supplier for material procurement, an evaluation knowledge graph is established, and a reference feature map is obtained by extracting the evaluation trend feature map;

[0007] Step S2, the competitive correlation value and the cooperative correlation value are obtained by evaluating the competitive correlation degree between the target supplier and the competitive supplier and the cooperative supplier, respectively, the evaluation trend statistics are performed according to the competitive correlation value and the cooperative correlation value to obtain the measured trend feature one and the measured trend feature two, the pre-processing evaluation data one to which the measured trend feature two conforms is reserved, the evaluation data to which the measured trend feature one belongs and the customer evaluation data of the target supplier are comprehensively evaluated to obtain the pre-processing evaluation data two, and the pre-processing feature data is counted according to the pre-processing evaluation data one, the pre-processing evaluation data two and the measured trend feature one;

[0008] Step S3, the historical transaction data of the target supplier is obtained, and the historical transaction data and the pre-processing feature data are used to evaluate the compliance degree of the target supplier for tobacco industry procurement to obtain the material procurement compliance degree.

[0009] Preferably, the industry evaluation data of the target supplier is obtained to obtain the to-be-evaluated industry evaluation data, and the to-be-evaluated industry evaluation data is classified into the competitive evaluation data and the cooperative evaluation data according to the existence of the competitive relationship and the existence of the cooperative relationship.

[0010] According to the competitive supplier to which the competitive evaluation data belongs, other competitive suppliers having a competitive relationship with the competitive supplier are counted to obtain the other competitive suppliers, and the competitive correlation degree between the competitive supplier and the other competitive suppliers in different historical periods is evaluated to obtain a competitive correlation data set.

[0011] The industry evaluation data of the other competitive suppliers by the competitive supplier in different historical periods is obtained to obtain auxiliary evaluation data, and the evaluation knowledge graph between the competitive supplier and the other competitive suppliers is established according to the competitive correlation data set and the auxiliary evaluation data to obtain an auxiliary evaluation knowledge graph.

[0012] Preferably, the auxiliary evaluation knowledge graph is subjected to comprehensive curve trend statistics to obtain an auxiliary evaluation trend feature map.

[0013] A stable median line and a stable amplitude threshold are preset, and an evaluation trend feature map having an alternating variation amplitude on the stable median line less than or equal to the stable amplitude threshold is extracted from the auxiliary evaluation trend feature map, and is integrated to obtain a first type of reserved feature map.

[0014] According to the cooperative evaluation data, the cooperative supplier to which the cooperative evaluation data belongs, the other suppliers having a cooperative relationship with the cooperative supplier, and the first type of reserved feature map, the second type of reserved feature map to which the cooperative evaluation data belongs is counted, and the first type of reserved feature map and the second type of reserved feature map are combined into a reference feature map.

[0015] Preferably, the competitive correlation degree between the target supplier and the competitive supplier is evaluated to obtain a competitive correlation value, and the cooperative correlation degree between the target supplier and the cooperative supplier is evaluated to obtain a cooperative correlation value.

[0016] The competitive correlation value and the competitive evaluation data are subjected to evaluation trend statistics to obtain a to-be-tested trend feature one, the cooperative correlation value and the cooperative evaluation data are subjected to evaluation trend statistics to obtain a to-be-tested trend feature two, and the cooperative evaluation data to which the evaluation trend of the to-be-tested trend feature two conforms is reserved to obtain preprocessed evaluation data one.

[0017] Preferably, the development trends of the target supplier and the competitive supplier in different historical periods are counted to obtain a target development trend and a competitive development trend.

[0018] The target development trend and the competitive development trend are compared to obtain a trend comparison result, if the trend comparison result is that the trends are opposite, and the measured trend characteristic one does not conform to the reference characteristic map, then the competitive evaluation data to which the measured trend characteristic one belongs is removed;

[0019] If the trend comparison result is that the trends are the same, and the measured trend characteristic one does not conform to the reference characteristic map, or if the trend comparison result is that the trends are opposite, and the measured trend characteristic one conforms to the reference characteristic map, then the customer evaluation data to which the target supplier belongs is obtained, the evaluation data to which the measured trend characteristic one belongs and the customer evaluation data are comprehensively evaluated to obtain preprocessed evaluation data two.

[0020] Preferably, if the trend comparison result is that the trends are the same, and the measured trend characteristic one conforms to the reference characteristic map, then the competitive evaluation data to which the measured trend characteristic one belongs is retained, and preprocessed evaluation data three is output.

[0021] The historical transaction data of the target supplier is obtained, and the historical transaction data, the customer evaluation data, the preprocessed evaluation data one, the preprocessed evaluation data two and the preprocessed evaluation data three are combined to obtain preprocessed characteristic data.

[0022] Preferably, the historical transaction data of the target supplier is obtained, and the delivery capability value of the target supplier is evaluated according to the historical transaction data to obtain a delivery capability value.

[0023] The market material procurement hierarchical classification price is obtained, the price of the target supplier is compared with the market price to obtain a price rationality degree, and the compliance degree of the target supplier for tobacco industry procurement is evaluated according to the preprocessed characteristic data, the delivery capability value and the price rationality degree to obtain a material procurement compliance degree.

[0024] Compared with the prior art, the present application has the following characteristics and beneficial effects:

[0025] The industry evaluation data of the target supplier is classified according to the competition relationship and the cooperation relationship, so as to facilitate subsequent differentiated detection of whether there is false evaluation data filtering, avoid unified judgment of all industry evaluation data, and cause data filtering to be excessive, increase the error of data processing results, and improve the orderliness of the whole data processing. The competition evaluation data and the cooperation evaluation data are classified, and the evaluation data knowledge graph is established between the competition suppliers and other competition relationship suppliers and between the cooperation suppliers and other cooperation relationship suppliers. In order to reduce the interference of false and irregular evaluation data of the competition suppliers and the cooperation suppliers, the evaluation trend stable feature graph is extracted from the established evaluation knowledge graph, that is, the reference feature graph is counted, which is used for subsequent feature comparison and elimination of false evaluation data of the competition evaluation data and the cooperation evaluation data, and the real cooperation evaluation data, that is, the preprocessed evaluation data, is retained. In order to improve the filtering effectiveness of the competition relationship evaluation data, the development trend of the target supplier and the development trend of the competition supplier are counted, so as to know whether the competition relationship between the two is enhanced. If it is enhanced, there is a great possibility of malicious false evaluation. Therefore, according to the comparison of the development trend relationship between the two, the competition evaluation data is further eliminated, and finally, the historical transaction data and the market material procurement price of the target supplier are obtained to comprehensively evaluate the compliance degree of the target supplier. Through the above processing mode, the diversity data is fully classified and processed, and a large amount of evaluation data is filtered. The accuracy of the final material procurement compliance evaluation result is improved. BRIEF DESCRIPTION OF DRAWINGS

[0026] Figure 1 is a step block diagram of a tobacco material procurement supervision compliance management method based on big data, which is mainly embodied in the present embodiment. DETAILED DESCRIPTION

[0027] The present application will be further described in detail below in combination with the following embodiments.

[0028] REFERENCE Figure 1 A tobacco material procurement supervision compliance management method based on big data, the method comprising the following steps:

[0029] Step S1, according to the competition evaluation data and the cooperation evaluation data of the target supplier, an evaluation knowledge graph is established, and an evaluation trend feature graph is extracted to obtain a reference feature graph.

[0030] Step S2, the target supplier is respectively evaluated with the competition association degree between the competitive suppliers and the cooperative suppliers to obtain the competition association value and the cooperation association value, the evaluation trend statistics are obtained according to the competition association value and the cooperation association value respectively, the evaluation trend of the pre-processing evaluation data one is retained, which is consistent with the reference feature map, the evaluation data of the evaluation trend of the target supplier and the customer evaluation data of the target supplier are comprehensively evaluated to obtain the pre-processing evaluation data two, and the pre-processing feature data is obtained according to the pre-processing evaluation data one, the pre-processing evaluation data two and the evaluation trend feature one.

[0031] Step S3, the historical transaction data of the target supplier is obtained, and the compliance degree of the target supplier for tobacco industry procurement is evaluated according to the historical transaction data and the pre-processing feature data to obtain the material procurement compliance degree.

[0032] Specifically, the industry evaluation data of the target supplier for material procurement is classified according to the competition relationship and the cooperation relationship, so as to facilitate subsequent differentiated detection of whether there is false evaluation data filtering, avoid the filtering processing of all industry evaluation data in a unified judgment mode, reduce the data filtering error of the data processing result, improve the orderliness of the whole data processing, and establish the evaluation data knowledge graph between the competitive suppliers and other competitive relationship suppliers and between the cooperative suppliers and other cooperative relationship suppliers through the classified competitive evaluation data and cooperative evaluation data, in order to reduce the interference influence of the false and irregular evaluation data of the competitive suppliers and the cooperative suppliers, the feature map with stable evaluation trend is extracted from the established evaluation knowledge graph, that is, the reference feature map is obtained, which is used for subsequent feature comparison and elimination of false evaluation data of the competitive evaluation data and the cooperative evaluation data, and the real cooperative evaluation data, that is, the pre-processing evaluation data one, is retained. In order to improve the filtering effectiveness of the competitive relationship evaluation data, the development trend of the target supplier and the development trend of the competitive supplier are also counted, so as to know whether the competition between the two is enhanced. If it is enhanced, there is a great possibility of malicious false evaluation. Therefore, according to the comparison of the development trend relationship between the two, the competitive evaluation data is further removed from the false evaluation data. Finally, the historical transaction data and the market procurement price of the target supplier are obtained to comprehensively evaluate the compliance degree of the target supplier for tobacco industry procurement. Through the above processing mode, the diversity data is fully classified and processed, the interference data of a large amount of evaluation data is filtered, and the accuracy of the final material procurement compliance degree evaluation result is improved.

[0033] The specific step S1 includes the following substeps:

[0034] The evaluation data of the target suppliers to which the materials are to be procured is obtained to obtain the evaluation data of the industry to be evaluated. The evaluation data of the industry to be evaluated is then classified into competitive evaluation data and cooperative evaluation data, based on whether there is a competitive relationship or a cooperative relationship.

[0035] Based on the competitive suppliers to which the competitive evaluation data belongs, other suppliers that compete with the competitive suppliers are identified to obtain other competitive suppliers. The degree of competitive relationship between competitive suppliers and other competitive suppliers in different historical periods is evaluated to obtain a competitive relationship dataset.

[0036] By acquiring industry evaluation data of competing suppliers from different historical periods, auxiliary evaluation data is obtained. Based on the aforementioned competitive association dataset and auxiliary evaluation data, an evaluation knowledge graph between competing suppliers and other competing suppliers is established, resulting in an auxiliary evaluation knowledge graph.

[0037] A trend feature map of auxiliary evaluation is obtained by statistically analyzing the trend of the auxiliary evaluation knowledge graph.

[0038] The system presets a stable median line and a stable amplitude threshold. It then extracts evaluation trend feature maps from the auxiliary evaluation trend feature maps where the alternating fluctuations above and below the stable median line are less than or equal to the stable amplitude threshold, and integrates them to obtain the first type of reserved feature map.

[0039] Based on the cooperation evaluation data, the cooperating suppliers to which the cooperation evaluation data belongs, other suppliers with cooperative relationships with the cooperating suppliers, and the first type of reserved feature map, the second type of reserved feature map to which the cooperation evaluation data belongs is calculated. The first type of reserved feature map and the second type of reserved feature map are combined to form a reference feature map.

[0040] Specifically, as the industry evaluation data to be evaluated (here mainly refers to the key quality characteristics of the supplier delivery link (evaluation level is 1-6 grades, and the evaluation degree decreases in turn: 1 grade is quality evaluation index 100%, 2 grade is quality evaluation index 80%, 3 grade is quality evaluation index 60%, 4 grade is quality evaluation index 40%, 5 grade is quality evaluation index 20%, 6 grade is quality evaluation index 0, and below 3 grades is a poor evaluation), the comprehensive content of evaluation includes: key quality characteristic evaluation data of delivery link (for example, product quality at delivery, delivery compliance, delivery timeliness, problem response and processing are evaluated, and the final evaluation result is the comprehensive evaluation index of key quality characteristics of delivery link), competitive evaluation data and cooperative evaluation data (competitive evaluation data refers to the competitive relationship between the user to which the competitive evaluation data belongs and the target supplier, which is the product evaluation between suppliers in the same industry. Cooperative evaluation data refers to the cooperative relationship between the user to which the cooperative evaluation data belongs and the target supplier, which is the product evaluation between cooperative suppliers. Because there are malicious bad reviews in competitive evaluation data and false good reviews in cooperative evaluation data, in order to reduce the interference of false data, subsequent differentiated filtering analysis of these two types of data is required), competitive suppliers (if G1 (it is necessary to explain that there are multiple competitive suppliers, and an example is given here to facilitate understanding), other competitive suppliers (if g1, g2, g3, g4, g5), competitive correlation data set (such as the overlap area ratio between the covered market area of G1 transaction and the covered market area of g1 transaction, if the overlap area ratio is 1 / 5, the competitive correlation degree between G1 and g1 is 0.2, g2, g3, g4, g5 and so on), auxiliary measurement evaluation data (referring to the product evaluation data of G1 on g1, g2, g3, g4 and g5 respectively), auxiliary measurement evaluation knowledge graph (such as G1 and g1 as connecting nodes, the length of the connecting line between G1 and g1 is determined by the competitive correlation data (such as 1 / 5, if the maximum connecting length is set as R, then the length of the connecting line between G1 and g1 is R*1 / 5), the evaluation frequency of G1 on g1 is p1, that is, according to the conversion ratio J / p of the evaluation frequency and the correlation inclination angle, the inclination horizontal angle of the connecting line between G1 and g1 is p1*J / p, which is j1, the comprehensive evaluation or poor evaluation of the auxiliary measurement evaluation data of G1 on g1 is carried out, if it is good, the direction of the evaluation knowledge graph connecting line between G1 and g1 is upward, if it is poor, the direction of the evaluation knowledge graph connecting line between G1 and g1 is downward, if it is good, then according to the connecting line length between G1 and g1, j1 and the auxiliary measurement evaluation data, the evaluation knowledge graph between G1 and g1 is constructed, and subsequently, g1 and g2 are taken as two nodes, and the knowledge graph is continuously constructed from g1: the competitive correlation data between G1 and g2 is 0.4 (then the connecting length between g1 and g2 is R*0.4), if the evaluation frequency of G1 to g2 is p2 (the inclination level angle of the connecting line between g1 and g2 is p2 * j / p if it is j2), the auxiliary evaluation data is a poor evaluation (the continuous line between G1 and g1 is downward), the evaluation knowledge graph between G1 and g1, g2 is constructed, and so on (taking g2 and g3 as nodes, g3 and g4 as nodes, and g4 and g5 as nodes), the evaluation knowledge graph between G1 and g1, g2, g3, g4 and g5 is constructed, that is, the auxiliary evaluation knowledge graph), the auxiliary evaluation trend feature graph (such as the evaluation knowledge graph between G1 and g1, g2, g3, g4 and g5 in different historical periods is statistically counted, and the overall curve trend of several evaluation knowledge graphs in the auxiliary evaluation trend feature graph is counted (such as a fluctuation curve graph)), a preset stable median line and a stable amplitude threshold (there are several auxiliary evaluation trend feature graphs, such as establishing x and y axes, G1 as the origin, the preset stable median line being x axis, the stable amplitude threshold (if it is "+y1"-"-y1", "+y1" is above the x axis, and "-y1" is below the x axis), if the auxiliary evaluation trend feature graph is a fluctuation curve away from the x and y axes, the single direction is downward, and the fluctuation amplitude is greater than y1, it is judged that the competitive evaluation data corresponding to the auxiliary evaluation trend feature graph is false evaluation data, and is removed, if the auxiliary evaluation trend feature graph is a fluctuation curve crossing the x axis and away from the y axis, the alternating direction is upward and downward, and the fluctuation amplitude is between "+y1" and "-y1", it is judged that the auxiliary evaluation trend feature graph is a first type of reserved feature graph, the competitive evaluation data corresponding to the first type of reserved feature graph is real evaluation data, so the first type of reserved feature graph is used as a reference comparison feature graph for subsequent evaluation data judgment, and several auxiliary evaluation trend feature graphs in the first type of reserved feature graph are integrated in different historical periods (for example, the knowledge graph between G1 and g1 in different historical periods is processed to obtain an evaluation trend feature graph, that is, the first type of reserved feature graph).

[0041] The specific step S2 includes the following sub-steps:

[0042] The competition correlation value is obtained by evaluating the competition correlation degree between the target supplier and the competitive supplier, and the cooperation correlation value is obtained by evaluating the cooperation correlation degree between the target supplier and the cooperative supplier.

[0043] The evaluation trend statistics of the competition correlation value and the competition evaluation data obtain a to-be-tested trend feature one, the evaluation trend statistics of the cooperation correlation value and the cooperation evaluation data obtain a to-be-tested trend feature two, and the cooperation evaluation data to which the to-be-tested trend feature two conforms is reserved to obtain preprocessed evaluation data one.

[0044] The development trends of the target supplier and the competitive supplier in different historical periods are statistically obtained to obtain a target development trend and a competitive development trend.

[0045] The target development trend and the competitive development trend are compared to obtain a trend comparison result, if the trend comparison result is that the trends are opposite, and the to-be-tested trend feature one does not conform to the reference feature map, the competition evaluation data to which the to-be-tested trend feature one belongs is eliminated.

[0046] If the trend comparison result is that the trends are the same, and the to-be-tested trend feature one does not conform to the reference feature map, or if the trend comparison result is that the trends are opposite, and the to-be-tested trend feature one conforms to the reference feature map, the customer evaluation data to which the target supplier belongs is obtained, the evaluation data to which the to-be-tested trend feature one belongs and the customer evaluation data are comprehensively evaluated to obtain preprocessed evaluation data two.

[0047] If the trend comparison result is that the trends are the same, and the to-be-tested trend feature one conforms to the reference feature map, the competition evaluation data to which the to-be-tested trend feature one belongs is reserved, and preprocessed evaluation data three is output.

[0048] The historical transaction data of the target supplier is obtained, and the historical transaction data, the customer evaluation data, the preprocessed evaluation data one, the preprocessed evaluation data two and the preprocessed evaluation data three are combined to obtain preprocessed feature data.

[0049] Specifically, like the competition correlation value, the cooperation correlation value (the same as the explanation of the competition correlation dataset, if the target supplier is G0, the competition correlation degree between G0 and G1 is evaluated as L1, and if the cooperation supplier is G2, the cooperation correlation degree between G0 and G2 is evaluated as L2), the first to-be-measured trend feature (the same as the explanation of the auxiliary evaluation trend feature map, first construct the evaluation knowledge graph between G0 and G1, and then perform comprehensive curve trend statistics to obtain the first to-be-measured trend feature, except that here only G0 and G1 are two nodes, and G1 is the starting connection point for evaluation knowledge graph construction), the second to-be-measured trend feature (comprehensive curve trend of the evaluation knowledge graph between G0 and G2), the first preprocessed evaluation data (curve comparison between the second to-be-measured trend feature and the reference feature map: that is, if the cooperation correlation degree between G2 and another cooperation supplier in the reference feature map is L2, and the feature map T is T1, whether the second to-be-measured trend feature coincides with T1 is compared, if the second to-be-measured trend feature coincides with T1, the cooperation evaluation data corresponding to the second to-be-measured trend feature is retained as real evaluation data, that is, the first preprocessed evaluation data), the target development trend and the competition development trend (such as the product sales development trend of G0 and G1 is counted (such as the development trend of different time periods is upward or downward or fluctuates), the development curve is counted, if the target development trend is Q1 and the competition development trend is Q2), the competition evaluation data corresponding to the first to-be-measured trend feature is excluded (if the trend change direction of the curves of Q1 and Q2 is opposite, it indicates that the product sales development trends of G0 and G1 are not synchronized, at this time, the competition interference correlation degree between them is increased, and the feature map T2 corresponding to the competition correlation degree L1 between G1 and another competition supplier (g1, g2, g3, g4, g5) in the reference feature map is compared with the first to-be-measured trend feature to determine whether they coincide, if T2 and the first to-be-measured trend feature do not coincide, under the condition that both of the two judgment conditions are not met, it can be judged that the competition evaluation data corresponding to the first to-be-measured trend feature is malicious fake evaluation data, and the data is filtered), the second preprocessed evaluation data (whether the trend comparison result is the same direction, whether the first to-be-measured trend feature meets the reference feature map, if only one of the two conditions is met, it cannot be directly judged whether the corresponding competition evaluation data is false or real evaluation data, in order to minimize the error of the final data processing, the data is neutralized, that is, the customer evaluation data and the competition evaluation data corresponding to the first to-be-measured trend feature are comprehensively evaluated (such as the comprehensive evaluation result of the customer evaluation data is a quality evaluation index of 80%, and the competition evaluation data corresponding to the first to-be-measured trend feature is a quality evaluation index of 40%), the final neutralization evaluation result is: a quality evaluation index of 60%, that is, the second preprocessed evaluation data), if the trend comparison result is the same direction (if the trend change direction of the curves of Q1 and Q2 is the same,If the product sales development trend of G0 and G1 is synchronized, the competition interference correlation between the two is reduced, and it is most likely that the market demand changes, at this time, the authenticity of the competition evaluation data can be further judged according to whether the to-be-tested trend feature one conforms to the reference feature map (the same as comparing whether the to-be-tested trend feature one coincides with the feature map in which the competition correlation degree between G1 and one of the other competitive suppliers (g1, g2, g3, g4, g5) in the reference feature map is L1, and the feature map is T2), if the to-be-tested trend feature one conforms to the reference feature map, it can be determined that the to-be-tested trend feature one corresponds to the true evaluation data of G1 to G0, that is, the pre-processing evaluation data three), historical transaction data (including the historical product sales performance data of G0).

[0050] The specific step S3 includes the following sub-steps:

[0051] Obtain the historical transaction data of the target supplier, and evaluate the delivery capability value of the target supplier according to the historical transaction data to obtain the delivery capability value.

[0052] Obtain the market material procurement hierarchical classification price, compare the price of the target supplier with the market price to obtain the price rationality degree, and evaluate the compliance degree of the target supplier in the tobacco industry procurement according to the pre-processing feature data, the delivery capability value and the price rationality degree to obtain the material procurement compliance degree.

[0053] Specifically, like historical transaction data (also including historical time period delivery record, order satisfaction record, etc. data), delivery capability value (for example, according to the on-time delivery rate (such as collecting the delivery record of the supplier in the past period (such as the past year). For each delivery, compare the delivery date agreed upon in the contract and the actual delivery date. On-time delivery rate = (on-time delivery times ÷ total delivery times) × 100%. For example, a certain supplier has 20 deliveries in the past year, of which 18 are on-time deliveries, so its on-time delivery rate is (18 ÷ 20) × 100% = 90%) and order fulfillment rate (such as the order fulfillment rate refers to the proportion of the supplier's ability to fully meet the order quantity and specification requirements. The calculation method is order fulfillment rate = (delivery times that fully meet the order quantity ÷ total delivery times) × 100%. For example, the enterprise placed 10 orders with the supplier, of which 8 were delivered according to the order quantity and specification, so the order fulfillment rate is (8 ÷ 10) × 100% = 80%) to conduct comprehensive evaluation, if 85%)). Price reasonableness (if the planned procurement price (i.e. average price) is N, and the market material procurement price of G0 is n, then n / N if Z, Z is closer to 1, the more reasonable), material procurement compliance degree (for example, using weighted scoring method: principle: assign a weight to each evaluation dimension, then give a score according to the supplier's performance in that dimension, and finally calculate the weighted total score. Steps: determine the weight of each dimension (can be determined by expert scoring, historical data analysis, etc.). Score each supplier's performance in each dimension (can use five-point system, ten-point system, etc.). Calculate the weighted total score: weighted total score = Σ (dimension weight × dimension score). Suppose the evaluation dimensions (pre-processed feature data (historical transaction data, customer evaluation data and pre-processed evaluation data), delivery capability value and price reasonableness) are five, and the weights are 0.2, 0.2, 0.2, 0.2, 0.2 (here simplified as equal weight, actual application can be adjusted according to importance). The scores of a certain supplier in each dimension are 4, 5, 3, 4, and 5, respectively. Then the weighted total score = (0.2 × 4) + (0.2 × 5) + (0.2 × 3) + (0.2 × 4) + (0.2 × 5) = 4.2).

[0054] The above are preferred embodiments of the present application, and are not intended to limit the protection scope of the present application, therefore: any equivalent changes made according to the structure, shape, principle of the present application shall be covered within the protection scope of the present application.

Claims

1. A method for supervising and managing compliance of tobacco material procurement based on big data, characterized in that, Comprise the following steps: Step S1, according to the target supplier of the material procurement belongs to the competition evaluation data and cooperation evaluation data, to establish the evaluation knowledge graph, the extraction of the evaluation trend characteristic graph obtains the reference characteristic graph; Step S2, the competition association value and the cooperation association value are obtained by respectively evaluating the competition association degree between the target supplier and the competitive supplier and the cooperative supplier, the evaluation trend statistics are carried out according to the competition association value and the cooperation association value respectively to obtain the first trend characteristic and the second trend characteristic, the preprocessing evaluation data one of the evaluation trend characteristic two is consistent with the reference characteristic graph is reserved, the preprocessing evaluation data two is obtained by comprehensively evaluating the evaluation data of the first trend characteristic and the customer evaluation data of the target supplier, the preprocessing characteristic data is counted according to the preprocessing evaluation data one, preprocessing evaluation data two and the first trend characteristic; Step S3, the historical transaction data of the target supplier is obtained, and the compliance degree of the target supplier in the tobacco industry procurement is evaluated according to the historical transaction data and the preprocessing characteristic data to obtain the material procurement compliance degree.

2. The method according to claim 1, wherein, Step S1 comprises: The industry evaluation data of the target supplier of the material procurement is obtained to obtain the evaluation industry data, the evaluation industry data is classified into competition evaluation data and cooperation evaluation data according to the industry evaluation data of the competition relationship and the cooperation relationship; According to the competition supplier of the competition evaluation data, other competition suppliers existing competition relationship with the competition supplier are obtained to obtain the competition supplier, and the competition association degree between the competition supplier and other competition suppliers in different historical periods is evaluated to obtain the competition association data set; The industry evaluation data of the competition supplier to other competition suppliers in different historical periods is obtained to obtain the auxiliary evaluation data, and the evaluation knowledge graph between the competition supplier and other competition suppliers is established according to the competition association data set and the auxiliary evaluation data to obtain the auxiliary evaluation knowledge graph.

3. The method according to claim 2, wherein, Step S1 further comprises: The auxiliary evaluation trend characteristic graph is obtained by comprehensive curve trend statistics of the auxiliary evaluation knowledge graph; The stable median line and the stable amplitude threshold are preset, the evaluation trend characteristic graph with the alternating variation amplitude less than or equal to the stable amplitude threshold on the stable median line is extracted from the auxiliary evaluation trend characteristic graph, and the first type of reserved characteristic graph is obtained by integrating; According to the cooperation evaluation data, the cooperation supplier of the cooperation evaluation data, the other suppliers existing cooperation relationship with the cooperation supplier and the first type of reserved characteristic graph, the second type of reserved characteristic graph of the cooperation evaluation data is counted, and the first type of reserved characteristic graph and the second type of reserved characteristic graph are combined into the reference characteristic graph.

4. The method according to claim 3, wherein, Step S2 comprises: The competition association value is obtained by evaluating the competition association degree between the target supplier and the competitive supplier, and the cooperation association value is obtained by evaluating the cooperation association degree between the target supplier and the cooperative supplier; The competition correlation value and competition evaluation data are evaluated to obtain a to-be-tested trend feature one, the cooperation correlation value and cooperation evaluation data are evaluated to obtain a to-be-tested trend feature two, and the cooperation evaluation data corresponding to the to-be-tested trend feature two that is consistent with the reference feature map is reserved to obtain preprocessed evaluation data one.

5. The method for supervision and compliance management of tobacco material procurement based on big data according to claim 4, characterized in that, Step S2 further includes: The development trends of the target supplier and the competitive suppliers in different historical periods are counted to obtain a target development trend and a competitive development trend; The target development trend and the competitive development trend are compared to obtain a comparison result, if the comparison result is that the trends are opposite and the to-be-tested trend feature one is inconsistent with the reference feature map, the competition evaluation data corresponding to the to-be-tested trend feature one is excluded; If the comparison result is that the trends are the same and the to-be-tested trend feature one is inconsistent with the reference feature map, or if the comparison result is that the trends are opposite and the to-be-tested trend feature one is consistent with the reference feature map, the customer evaluation data of the target supplier is obtained, the evaluation data corresponding to the to-be-tested trend feature one and the customer evaluation data are comprehensively evaluated to obtain preprocessed evaluation data two.

6. The method for supervision and compliance management of tobacco material procurement based on big data according to claim 5, characterized in that, Step S2 further includes: If the comparison result is that the trends are the same and the to-be-tested trend feature one is consistent with the reference feature map, the competition evaluation data corresponding to the to-be-tested trend feature one is reserved, and preprocessed evaluation data three is output, The historical transaction data of the target supplier is obtained, and the historical transaction data, the customer evaluation data, the preprocessed evaluation data one, the preprocessed evaluation data two, and the preprocessed evaluation data three are combined to obtain preprocessed feature data.

7. The method for supervision and compliance management of tobacco material procurement based on big data according to claim 6, characterized in that, Step S3 includes: The historical transaction data of the target supplier is obtained, and the delivery capability value of the target supplier is evaluated according to the historical transaction data to obtain a delivery capability value; The market material procurement hierarchical classification price is obtained, the price of the target supplier and the market price are compared to obtain a price rationality degree, and the preprocessed feature data, the delivery capability value, and the price rationality degree are used to evaluate the compliance degree of the target supplier in tobacco industry procurement to obtain a material procurement compliance degree.

Images