CN118365424A - Procurement sourcing method and system based on intelligent recommendation of supplier combination - Google Patents

Abstract

The embodiment of the present invention provides a procurement sourcing method and system based on intelligent recommendation of supplier combinations, which belongs to the technical field of procurement sourcing. The method includes: collecting procurement demand data, and determining candidate suppliers based on the procurement demand data to obtain candidate supplier combinations; respectively calculating the combined evaluation scores of each candidate supplier combination, and the basic evaluation scores of a single candidate supplier in each candidate supplier combination; respectively determining the procurement scores of each candidate supplier combination based on the combined evaluation scores of each candidate supplier combination and the basic evaluation scores of a single candidate supplier in each candidate supplier combination, and sorting the procurement scores of each candidate supplier combination; performing candidate supplier combination recommendation based on the sorting results, and performing procurement sourcing based on the recommended candidate supplier combination to obtain procurement sourcing results. The scheme of the present invention improves the quality and efficiency of sourcing.

Description

Procurement sourcing method and system based on intelligent recommendation of supplier combination

Technical Field

The present invention relates to the technical field of procurement sourcing, and in particular to a procurement sourcing method based on intelligent recommendation of supplier combination and a procurement sourcing system based on intelligent recommendation of supplier combination.

Background technique

Supplier relationship management (SRM) is a solution dedicated to establishing and maintaining a long-term, close partnership with suppliers. It is a new management mechanism that aims to improve the relationship between enterprises and suppliers. It is implemented in areas related to enterprise procurement business. The goal is to establish a long-term, close business relationship with suppliers and integrate the resources and competitive advantages of both parties to jointly develop the market, expand market demand and share, reduce the high initial cost of products, and achieve a win-win enterprise management model. At present, there is an intelligent procurement platform based on SRM, which can respond to the target procurement needs of the target procurement enterprise, parse the material name, specification and quantity contained in the target procurement needs, and select the target supplier corresponding to the minimum transaction cost required to purchase each material in the current transaction cycle. For example, the Chinese invention patent application with publication number CN116993444A can analyze the transaction prices of different suppliers for materials of the same specifications, select the target supplier with the lowest transaction cost for the procurement enterprise, and reduce the procurement cost.

However, the procurement needs of purchasing companies are often diverse. On the one hand, for needs with more demands and more complex project collaboration relationships, a single supplier is often unable to meet the needs. Therefore, for the same procurement demand, it may be necessary to divide it into multiple procurement lists for procurement sourcing. At this time, if the procurement list is not divided reasonably, it may cause difficulties in sourcing, or it may be difficult to find a truly suitable supplier. On the other hand, for procurement needs with scattered demand types and data, but few project collaboration relationships, if each scattered procurement demand is sourced separately, it will often increase procurement time and cost. In view of the problems of low sourcing accuracy and low sourcing efficiency of existing procurement sourcing solutions when facing the procurement needs of purchasing companies, a new procurement sourcing solution needs to be proposed.

Summary of the invention

The purpose of the embodiments of the present invention is to provide a procurement sourcing method and system based on intelligent recommendation of supplier combinations, so as to at least solve the problems of low sourcing accuracy and low sourcing efficiency in existing procurement sourcing solutions when facing the procurement needs of purchasing enterprises.

In order to achieve the above-mentioned objectives, the first aspect of the present invention provides a procurement sourcing method based on intelligent recommendation of supplier combinations, the method comprising: collecting procurement demand data, and determining candidate suppliers based on the procurement demand data to obtain candidate supplier combinations; respectively calculating the combined evaluation score of each candidate supplier combination, and the basic evaluation score of a single candidate supplier in each candidate supplier combination; respectively determining the procurement score of each candidate supplier combination based on the combined evaluation score of each candidate supplier combination and the basic evaluation score of a single candidate supplier in each candidate supplier combination, and sorting the procurement scores of each candidate supplier combination; performing candidate supplier combination recommendation based on the sorting result, and performing procurement sourcing based on the recommended candidate supplier combination to obtain procurement sourcing results.

Optionally, the procurement demand data is collected, and candidate suppliers are determined based on the procurement demand data to obtain a candidate supplier combination, including: generating a corresponding standardized total procurement demand list based on the procurement demand data; determining the candidate suppliers corresponding to each procurement target based on the standardized total procurement demand list to obtain a candidate supplier list; performing exhaustive search and matching based on the standardized total procurement demand list and the candidate supplier list to obtain all supplier combinations that meet the basic procurement demand corresponding to the procurement demand data as candidate supplier combinations.

Optionally, an exhaustive search and match is performed based on the standardized total procurement requirement list and the candidate supplier list to obtain all supplier combinations that meet the basic procurement requirements corresponding to the procurement requirement data as candidate supplier combinations, including: traversing all suppliers in the supplier list, considering whether to select or not select each supplier, thereby generating all possible supplier combinations; for each generated supplier combination, filtering out the supplier combinations that cannot meet the basic procurement requirements corresponding to the procurement requirement data, and taking the remaining supplier combinations as candidate supplier combinations.

Optionally, the respectively calculating of the combined evaluation score of each candidate supplier combination and the basic evaluation score of a single candidate supplier within each candidate supplier combination includes: calculating the matching degree between a single candidate supplier within each candidate supplier combination and the procurement demand data as the basic evaluation score of a single candidate supplier within each candidate supplier combination; and calculating the coordination degree between each single candidate supplier within each candidate supplier combination as the combined evaluation score of each candidate supplier combination.

Optionally, the calculation of the matching degree between a single candidate supplier in each candidate supplier combination and the procurement demand data as a basic evaluation score for the single candidate supplier in each candidate supplier combination includes: determining the priority of each procurement indicator based on the procurement demand data; wherein the procurement indicators include: procurement targets, procurement quantities, procurement quality requirements and required delivery times; performing a matching degree between each single candidate supplier and each procurement indicator based on a machine learning algorithm, and generating a matching degree score corresponding to each procurement indicator based on the determined matching degree; performing a weighted summation of the matching degree scores corresponding to each procurement indicator based on the priority of each procurement indicator to obtain a matching degree between the corresponding single candidate supplier and the procurement demand data; and traversing all single candidate suppliers to obtain the basic evaluation scores of the single candidate suppliers in each candidate supplier combination.

Optionally, the calculation of the degree of cooperation between each individual candidate supplier in each candidate supplier combination as the combined evaluation score corresponding to each candidate supplier combination includes: collecting the real-time supply capability information and retained procurement cooperation information of each individual candidate supplier in each candidate supplier combination; training the real-time supply capability information and the retained procurement cooperation information based on a pre-built combined evaluation score scoring model to obtain the combined evaluation score of each candidate supplier combination.

Optionally, the method also includes: executing pre-construction of a combination evaluation score scoring model, including: collecting historical procurement data, and reading historical cooperation information of each candidate supplier based on the historical procurement data; executing random combination of each candidate supplier to obtain multiple historical supplier combinations; pushing multiple historical supplier combinations to the user end, and based on the user end, recovering the user's combination evaluation score annotation results as the labeling information of each historical supplier combination to obtain labeled samples; performing sample expansion on the labeled samples to obtain expanded labeled samples; performing model training in a preselected neural network based on the expanded labeled samples to obtain a combination evaluation score scoring model.

Optionally, the procurement score of each candidate supplier combination is determined based on the combined evaluation score of each candidate supplier combination and the basic evaluation score of a single candidate supplier in each candidate supplier combination, including: based on the basic evaluation score of each single candidate supplier, performing a summation process in each candidate supplier combination to obtain the sum of the basic evaluation scores of each candidate supplier combination; performing a basic evaluation score and weighting of each candidate supplier combination and a combined evaluation score weighting of each candidate supplier combination based on a preset weight allocation relationship, performing a procurement score calculation for each candidate supplier combination based on the weighting result to obtain a procurement score for each candidate supplier combination.

Optionally, the recommendation of candidate supplier combinations based on the sorting results includes: based on the sorting results, determining the number of candidate supplier combinations whose procurement scores are greater than a preset recommended procurement score threshold; if the number of candidate supplier combinations whose procurement scores are greater than the preset recommended procurement score threshold is greater than a preset number threshold N, then taking the first N candidate supplier combinations equal to the preset number threshold N among the candidate supplier combinations whose procurement scores are greater than the preset recommended procurement score threshold as recommended candidate supplier combinations; if the number of candidate supplier combinations whose procurement scores are greater than the preset recommended procurement score threshold is not greater than the preset number threshold N, then taking all candidate supplier combinations whose procurement scores are greater than the preset recommended procurement score threshold as recommended candidate supplier combinations.

Optionally, the procurement sourcing is performed based on the recommended candidate supplier combinations to obtain procurement sourcing results, including: generating a recommended candidate supplier set corresponding to each candidate supplier combination based on each recommended candidate supplier combination; performing union processing on each recommended candidate supplier set to obtain all candidate recommended candidate supplier targets to be sourced; performing sourcing for each candidate recommended candidate supplier target to be sourced to obtain sourcing results for each recommended candidate supplier; and performing a sourcing result combination corresponding to each recommended candidate supplier under each recommended candidate supplier combination to obtain sourcing results for each recommended candidate supplier combination.

The second aspect of the present invention provides a procurement sourcing system based on intelligent recommendation of supplier combinations, the system comprising: a collection unit, used to collect procurement demand data, and determine candidate suppliers based on the procurement demand data to obtain candidate supplier combinations; a scoring unit, used to calculate the combined evaluation score of each candidate supplier combination, and the basic evaluation score of a single candidate supplier in each candidate supplier combination; a processing unit, used to determine the procurement score of each candidate supplier combination based on the combined evaluation score of each candidate supplier combination and the basic evaluation score of a single candidate supplier in each candidate supplier combination, and to sort the procurement scores of each candidate supplier combination; a sourcing unit, used to recommend candidate supplier combinations based on the sorting results, and perform procurement sourcing based on the recommended candidate supplier combinations to obtain procurement sourcing results.

On the other hand, the present invention provides a computer-readable storage medium having instructions stored thereon, which, when executed on a computer, enables the computer to execute the above-mentioned procurement sourcing method based on intelligent recommendation of supplier combinations.

Through the above technical solution, the solution of the present invention has the following beneficial effects:

1. Improve procurement efficiency: By intelligently recommending candidate supplier combinations, you can quickly determine the potential supplier selection range and save procurement personnel’s time and energy.

2. Optimize supplier selection: By calculating the combined evaluation score and the basic evaluation score, the overall performance of the supplier portfolio and the capabilities of individual suppliers can be comprehensively considered, which helps to select the optimal supplier portfolio.

3. Improve the accuracy of procurement decisions: Making recommendations based on procurement scoring ranking results can reduce the impact of subjective factors on supplier selection and improve the objectivity and accuracy of decision-making.

4. Enhance the reliability of procurement results: By executing procurement sourcing based on the recommended candidate supplier portfolio, it can ensure that the procurement process is consistent with the recommended results, thereby improving the reliability and consistency of procurement results.

Other features and advantages of the embodiments of the present invention will be described in detail in the following detailed description.

BRIEF DESCRIPTION OF THE DRAWINGS

The accompanying drawings are used to provide a further understanding of the embodiments of the present invention and constitute a part of the specification. Together with the following specific embodiments, they are used to explain the embodiments of the present invention, but do not constitute a limitation on the embodiments of the present invention. In the accompanying drawings:

FIG1 is a flowchart of the steps of a procurement sourcing method based on supplier combination intelligent recommendation provided by an embodiment of the present invention;

FIG2 is a system structure diagram of a procurement sourcing system based on supplier combination intelligent recommendation provided by an embodiment of the present invention.

Detailed ways

The specific implementation of the present invention is described in detail below in conjunction with the accompanying drawings. It should be understood that the specific implementation described herein is only used to illustrate and explain the present invention, and is not used to limit the present invention.

FIG1 is a method flow chart of a procurement sourcing method based on supplier combination intelligent recommendation provided by an embodiment of the present invention. As shown in FIG1 , an embodiment of the present invention provides a procurement sourcing method based on supplier combination intelligent recommendation, the method comprising:

Step S10: Collect procurement demand data, and determine candidate suppliers based on the procurement demand data to obtain a candidate supplier combination.

Specifically, a corresponding standardized total procurement requirement list is generated based on the procurement requirement data; based on the standardized total procurement requirement list, the candidate suppliers corresponding to each procurement target are determined to obtain a candidate supplier list; based on the standardized total procurement requirement list and the candidate supplier list, an exhaustive search and matching is performed to obtain all supplier combinations that meet the basic procurement requirements corresponding to the procurement requirement data as candidate supplier combinations.

In an embodiment of the present invention, candidate suppliers may be determined based on supplier portraits retained by historical cooperative suppliers, or may be determined based on potential suppliers mined from big data.

In an embodiment of the present invention, in a modern business environment, procurement management is a vital part of business operations. In order to effectively manage procurement needs and ensure the efficiency and transparency of the procurement process, many companies use automated and intelligent methods to process procurement demand data. Generating a corresponding standardized total procurement demand list based on the procurement demand data is a key task. This involves integrating and standardizing the procurement needs proposed by various departments or project teams in order to better manage and analyze these needs. Once the standardized total procurement demand list is obtained, the next step is to determine the candidate suppliers corresponding to each procurement target, thereby obtaining a list of candidate suppliers. This process usually involves screening and matching the supplier database to ensure that the selected supplier can meet the company's procurement needs and meet the relevant standards and requirements. The selection of candidate suppliers may be based on multiple factors, including the supplier's reputation, product quality, delivery capabilities, price competitiveness, etc.

Furthermore, exhaustive search and matching is performed based on the standardized total procurement demand list and the candidate supplier list to obtain all supplier combinations that meet the basic procurement demand corresponding to the procurement demand data as candidate supplier combinations. The corresponding supplier combination determination relationship proposed by the solution of the present invention ensures a comprehensive search and matching of all possible supplier combinations. Exhaustive search and matching can help enterprises find the best supplier combination, thereby minimizing procurement costs and maximizing procurement efficiency.

Furthermore, an exhaustive search and matching is performed based on the standardized total procurement requirement list and the candidate supplier list to obtain all supplier combinations that meet the basic procurement requirements corresponding to the procurement requirement data as candidate supplier combinations, including: traversing all suppliers in the supplier list, considering whether to select or not select each supplier, thereby generating all possible supplier combinations; for each generated supplier combination, filtering out the supplier combinations that cannot meet the basic procurement requirements corresponding to the procurement requirement data, and using the remaining supplier combinations as candidate supplier combinations.

In the embodiment of the present invention, a recursive algorithm is used to perform an exhaustive search and matching process for the generated standardized total procurement demand list and candidate supplier list. The recursive algorithm can traverse all suppliers in the supplier list and select or not select each supplier, thereby generating all possible supplier combinations. In each step, the algorithm considers whether the current supplier is selected or not, and then continues to recursively process the next supplier until all suppliers are traversed to form all possible supplier combinations.

Furthermore, for each generated supplier combination, it is necessary to match and filter the basic procurement requirements. This step can be achieved using conditional judgment and screening algorithms. For each supplier combination, the algorithm will check whether the suppliers therein can meet the basic procurement requirements corresponding to the procurement requirement data. If a supplier in a supplier combination cannot meet the basic procurement requirements, then the supplier combination will be filtered out and will not be included in the candidate supplier combination. The remaining supplier combinations are considered to be candidate supplier combinations that meet the basic procurement requirements.

The solution of the present invention can obtain all possible supplier combinations and screen out candidate supplier combinations that meet the basic procurement requirements by performing such an exhaustive search, matching and filtering process. The benefit of this method is that it can comprehensively consider the combination of all suppliers and ensure that each supplier combination is strictly matched with the basic procurement requirements, thereby improving the accuracy and reliability of the selected suppliers. At the same time, through the application of automated algorithms, human resources and time costs can be greatly saved, and the efficiency and accuracy of procurement decisions can be improved.

Step S20: Calculate the combined evaluation score of each candidate supplier combination and the basic evaluation score of each candidate supplier combination.

Specifically, the scheme of the present invention scores each candidate supplier in the candidate supplier combination to obtain a basic evaluation score, wherein the basic evaluation score represents the matching degree of the candidate supplier itself to the procurement demand; scores the combination capability of the candidate supplier combination to obtain a combined evaluation score, wherein the combined evaluation score represents the cooperation degree between the candidate suppliers; and combines the combined evaluation score of each candidate supplier combination and the basic score of each supplier in the candidate supplier combination to obtain the total score of each candidate supplier combination.

Preferably, the respectively calculating of the combined evaluation score of each candidate supplier combination and the basic evaluation score of a single candidate supplier in each candidate supplier combination includes: calculating the matching degree between a single candidate supplier in each candidate supplier combination and the procurement demand data as the basic evaluation score of a single candidate supplier in each candidate supplier combination; and calculating the coordination degree between each single candidate supplier in each candidate supplier combination as the combined evaluation score of each candidate supplier combination.

Specifically, the calculation of the matching degree between a single candidate supplier in each candidate supplier combination and the procurement demand data as the basic evaluation score of a single candidate supplier in each candidate supplier combination includes: determining the priority of each procurement indicator based on the procurement demand data; wherein the procurement indicators include: procurement target, procurement quantity, procurement quality requirements and required delivery time; performing the matching degree between each single candidate supplier and each procurement indicator based on a machine learning algorithm, and generating a matching degree score corresponding to each procurement indicator based on the determined matching degree; performing weighted summation on the matching degree score corresponding to each procurement indicator based on the priority of each procurement indicator to obtain the matching degree between the corresponding single candidate supplier and the procurement demand data; traversing all single candidate suppliers to obtain the basic evaluation score of a single candidate supplier in each candidate supplier combination. Specifically, the steps include:

1) Determine the priority of procurement indicators: Procurement indicators include procurement targets, procurement quantities, procurement quality requirements, and required delivery time. When processing procurement demand data, it is necessary to first determine the priority order of these procurement indicators so that the importance of each indicator can be better weighed in the subsequent matching and evaluation process.

2) Machine learning algorithm matches candidate suppliers with procurement indicators: Based on the machine learning algorithm, the matching degree between each candidate supplier and each procurement indicator is evaluated. This can be done by analyzing the supplier's historical data, qualifications, delivery capabilities and other information through the algorithm, and comparing them with the procurement indicators to determine the matching degree of each supplier under different indicators.

3) Generate matching scores for procurement indicators: Generate corresponding matching scores for each procurement indicator based on the determined matching degree. These scores reflect the performance of each supplier under different procurement indicators and provide a basis for subsequent evaluation and screening.

4) Calculate the matching degree between suppliers and procurement demand data by weighted summation: Based on the priority of each procurement indicator, the matching degree scores corresponding to each procurement indicator are weighted summation. Through this step, the importance of each indicator can be comprehensively considered to obtain the comprehensive matching degree score between each candidate supplier and procurement demand data.

5) Traverse all single candidate suppliers and obtain basic evaluation scores: Traverse all single candidate suppliers and calculate the basic evaluation scores for each single supplier in each supplier combination. This evaluation score comprehensively considers the matching degree of each supplier under each procurement indicator and the priority of each indicator, thus providing a comprehensive evaluation result for each supplier combination.

The scheme of the present invention can achieve comprehensive matching and evaluation of candidate suppliers, thereby generating a combination of candidate suppliers that meet the procurement demand data. This method can more accurately evaluate the adaptability of suppliers and improve the accuracy and reliability of selected suppliers. At the same time, through the application of machine learning algorithms, it is possible to quickly process and analyze a large amount of supplier data, save human resources and time costs, and improve the efficiency and quality of procurement management. This data-driven and algorithm-supported procurement management method helps enterprises optimize supply chain management, reduce procurement costs, and improve the accuracy and efficiency of procurement decisions, thereby enhancing the competitiveness and sustainable development capabilities of enterprises.

Furthermore, the calculation of the degree of cooperation between each single candidate supplier in each candidate supplier combination as the combined evaluation score corresponding to each candidate supplier combination includes: collecting the real-time supply capacity information and retained procurement cooperation information of each single candidate supplier in each candidate supplier combination; training the real-time supply capacity information and the retained procurement cooperation information based on a pre-built combined evaluation score scoring model to obtain the combined evaluation score of each candidate supplier combination. Specifically, the following steps are included:

1) Collect real-time supply capacity information and retain procurement cooperation information: First, it is necessary to collect real-time supply capacity information from each candidate supplier, including production capacity, inventory status, delivery speed and other data. At the same time, collect retained procurement cooperation information, such as historical transaction records, cooperation stability, etc., in order to more comprehensively evaluate the overall performance of suppliers.

2) Construct a combination evaluation score model: Based on the collected real-time supply capacity information and retained procurement cooperation information, establish a combination evaluation score model. This model can use machine learning algorithms, such as deep learning models or regression models, to train and learn data, so as to accurately predict and evaluate the performance and cooperation of the supplier portfolio.

3) Train the model and generate a combined evaluation score: After the model is built, use the existing real-time data and historical information to train the model. By training the model, it can learn the association rules and coordination between each candidate supplier in each candidate supplier combination. After the training is completed, the model will be able to generate a corresponding combined evaluation score for each candidate supplier combination.

4) Evaluate the performance of candidate supplier combinations: Use the generated combination evaluation scores to evaluate and rank the performance of each candidate supplier combination. These evaluation scores will reflect the degree of cooperation and comprehensive performance of different supplier combinations, helping the procurement team to better select the optimal supplier combination.

The solution of the present invention can achieve a comprehensive evaluation of the candidate supplier combination, which not only considers the matching degree between a single supplier and the procurement index, but also considers the cooperation and performance between suppliers. This comprehensive evaluation method can more accurately reflect the overall performance of the supplier combination and provide more powerful support and guidance for procurement decisions. At the same time, through the application of machine learning algorithms, it is possible to quickly process and analyze large amounts of data, improve the efficiency of procurement management and the accuracy of decision-making, thereby optimizing the company's supply chain management and procurement process.

Furthermore, the method also includes: executing pre-construction of a combination evaluation score scoring model, including: collecting historical procurement data, and reading historical cooperation information of each candidate supplier based on the historical procurement data; executing random combination of each candidate supplier to obtain multiple historical supplier combinations; pushing multiple historical supplier combinations to the user end, and based on the user end, recovering the user's combination evaluation score annotation results as the labeling information of each historical supplier combination to obtain labeled samples; performing sample expansion on the labeled samples to obtain expanded labeled samples; performing model training in a pre-selected neural network based on the expanded labeled samples to obtain a combination evaluation score scoring model.

Specifically, first, it is necessary to collect the real-time supply capacity information and retained procurement cooperation information of each individual candidate supplier in each candidate supplier combination. This includes real-time data such as production capacity, delivery on-time rate, product quality, and retained information such as historical transaction records and cooperation stability. Before building a combined evaluation score scoring model, a large amount of historical procurement data needs to be collected. These data include past purchase orders, supplier transaction records, delivery status, etc., which are used to analyze and evaluate the past performance of suppliers. Based on historical procurement data, the historical cooperation information of each candidate supplier is read. This includes analyzing past cooperation status, transaction frequency, problem feedback, etc., in order to understand the cooperation history between suppliers. By randomly combining candidate suppliers, multiple historical supplier combinations can be obtained, which helps to build a more comprehensive data set for training evaluation models. Multiple historical supplier combinations are pushed to the user end, and the user scores the combination points to obtain the target data of each sample data, and the trained model can obtain the ability to score independently. The sample expansion of the combined evaluation score annotation results collected by the user is carried out to increase the diversity and quantity of the data and improve the generalization ability and accuracy of the model. Based on the expanded labeled samples, model training is performed in a pre-selected neural network or other machine learning model. Through the training model, the degree of fit between each candidate supplier in each candidate supplier combination can be learned, and a combination evaluation score can be generated.

Step S30: Determine the procurement score of each candidate supplier combination based on the combined evaluation score of each candidate supplier combination and the basic evaluation score of each candidate supplier in each candidate supplier combination, and sort the procurement scores of each candidate supplier combination.

Specifically, based on the basic evaluation score of each individual candidate supplier, a summation process is performed within each candidate supplier combination to obtain the basic evaluation score and of each candidate supplier combination; based on a preset weight allocation relationship, the basic evaluation score and weighting of each candidate supplier combination and the combined evaluation score weighting of each candidate supplier combination are performed, and based on the weighted results, the procurement score calculation of each candidate supplier combination is performed to obtain the procurement score of each candidate supplier combination.

In the embodiment of the present invention, when the company faces urgent orders or short-term needs, the supplier's supply capacity becomes crucial. In this case, the ability to deliver on time and guarantee supply is the primary consideration. If the product quality and compliance requirements are high, the supplier's supply capacity directly affects the product quality and compliance. Therefore, the supplier's production capacity, quality control and compliance certification become crucial. When the company establishes long-term cooperation and strategic partnerships with suppliers, the combination relationship fit between suppliers becomes important, including cooperation stability, communication efficiency and the ability to develop together.

It can be seen that under different procurement needs, there are differences in the sensitivity to the supply capacity of a single supplier and the degree of cooperation between multiple suppliers. In order to adapt to this difference, when performing the procurement scoring calculation, the solution of the present invention performs adaptive weighting processing on the basic evaluation score and the combined evaluation score of each candidate supplier combination, so as to cope with the sensitivity requirements in different scenarios and ensure that the optimal supplier combination obtained is the supplier combination that meets the most sensitive needs of the current user.

Step S40: Recommend candidate supplier combinations based on the ranking results, and perform procurement sourcing based on the recommended candidate supplier combinations to obtain procurement sourcing results.

Specifically, the recommendation of candidate supplier combinations based on the sorting results includes: based on the sorting results, determining the number of candidate supplier combinations whose procurement scores are greater than a preset recommended procurement score threshold; if the number of candidate supplier combinations whose procurement scores are greater than the preset recommended procurement score threshold is greater than a preset number threshold N, then taking the first N candidate supplier combinations equal to the preset number threshold N among the candidate supplier combinations whose procurement scores are greater than the preset recommended procurement score threshold as recommended candidate supplier combinations; if the number of candidate supplier combinations whose procurement scores are greater than the preset recommended procurement score threshold is not greater than the preset number threshold N, then taking all candidate supplier combinations whose procurement scores are greater than the preset recommended procurement score threshold as recommended candidate supplier combinations.

Furthermore, the procurement sourcing is performed based on the recommended candidate supplier combinations to obtain procurement sourcing results, including: generating a recommended candidate supplier set corresponding to each candidate supplier combination based on each recommended candidate supplier combination; performing union processing on each recommended candidate supplier set to obtain all candidate recommended candidate supplier targets to be sourced; performing sourcing for each candidate recommended candidate supplier target to be sourced to obtain sourcing results for each recommended candidate supplier; and performing a sourcing result combination corresponding to each recommended candidate supplier under each recommended candidate supplier combination to obtain sourcing results for each recommended candidate supplier combination.

In the embodiment of the present invention, there is a high probability that the same supplier exists in each supplier combination. If sourcing is performed based on each recommended candidate supplier separately, the problem of repeated sourcing of the agreed supplier will inevitably occur. In order to avoid this problem, the solution of the present invention first performs a union process on each recommended candidate supplier combination to ensure that each candidate supplier is sourced only once. In order to facilitate customer review, after completing the sourcing, the solution of the present invention displays the sourcing results based on each candidate supplier combination for the user to choose.

Figure 2 is a system structure diagram of a procurement sourcing system based on intelligent recommendation of supplier combinations provided by an embodiment of the present invention. As shown in Figure 2, an embodiment of the present invention provides a procurement sourcing system based on intelligent recommendation of supplier combinations, the system comprising: a collection unit, used to collect procurement demand data, and determine candidate suppliers based on the procurement demand data to obtain candidate supplier combinations; a scoring unit, used to calculate the combined evaluation scores of each candidate supplier combination, and the basic evaluation scores of a single candidate supplier in each candidate supplier combination; a processing unit, used to determine the procurement scores of each candidate supplier combination based on the combined evaluation scores of each candidate supplier combination and the basic evaluation scores of a single candidate supplier in each candidate supplier combination, and to sort the procurement scores of each candidate supplier combination; a sourcing unit, used to recommend candidate supplier combinations based on the sorting results, and perform procurement sourcing based on the recommended candidate supplier combinations to obtain procurement sourcing results.

The embodiment of the present invention further provides a computer-readable storage medium, on which instructions are stored, which, when executed on a computer, enable the computer to execute the above-mentioned procurement sourcing based on intelligent recommendation of supplier combinations.

Those skilled in the art will appreciate that all or part of the steps in the method for implementing the above-mentioned embodiments can be completed by instructing the relevant hardware through a program, and the program is stored in a storage medium, including several instructions for making a single-chip microcomputer, a chip or a processor (processor) execute all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk and other media that can store program codes.

The optional embodiments of the present invention are described in detail above in conjunction with the accompanying drawings. However, the embodiments of the present invention are not limited to the specific details in the above embodiments. Within the technical concept of the embodiments of the present invention, the technical scheme of the embodiments of the present invention can be subjected to a variety of simple modifications, and these simple modifications all belong to the protection scope of the embodiments of the present invention. It should also be noted that the various specific technical features described in the above specific embodiments can be combined in any suitable manner without contradiction. In order to avoid unnecessary repetition, the embodiments of the present invention will not further describe various possible combinations.

In addition, various embodiments of the present invention may be arbitrarily combined, and as long as they do not violate the concept of the embodiments of the present invention, they should also be regarded as the contents disclosed in the embodiments of the present invention.

Claims (12)

1. A purchasing and sourcing method based on vendor combination intelligent recommendation, the method comprising:

Collecting purchasing demand data, and determining candidate suppliers based on the purchasing demand data to obtain candidate supplier combinations;

Calculating the combined evaluation score of each candidate provider combination respectively, and corresponding to the basic evaluation scores of the single candidate providers in each candidate provider combination;

Determining purchase scores of each candidate provider combination based on the combined evaluation score of each candidate provider combination and the base evaluation score of the individual candidate provider within each candidate provider combination, respectively, and performing ranking on the purchase scores of each candidate provider combination;

And performing candidate supplier combination recommendation based on the sorting result, and performing purchase source searching based on the recommended candidate supplier combination to obtain a purchase source searching result.

2. The method of claim 1, wherein the collecting purchase demand data and determining candidate suppliers based on the purchase demand data to obtain candidate supplier combinations comprises:

Generating a corresponding standardized total purchasing demand list based on the purchasing demand data;

determining candidate suppliers corresponding to each purchasing target based on the standardized total purchasing demand list to obtain a candidate supplier list;

And performing exhaustive search matching based on the standardized total purchasing demand list and the candidate supplier list to obtain all supplier combinations meeting the basic purchasing demands corresponding to the purchasing demand data as candidate supplier combinations.

3. The method of claim 2, wherein the performing an exhaustive search match based on the normalized total purchase demand list and the candidate supplier list to obtain, as candidate supplier combinations, all supplier combinations that satisfy the base purchase demand corresponding to the purchase demand data comprises:

Traversing all suppliers in the supplier list, considering for each supplier whether selected or unselected, thereby generating all possible combinations of suppliers;

And filtering out the supplier combinations in which the basic purchasing requirements corresponding to the purchasing requirement data cannot be met for each generated supplier combination, and taking the rest supplier combinations as candidate supplier combinations.

4. The method of claim 1, wherein the separately calculating the combined score for each candidate supplier combination and the base score for the individual candidate suppliers within each candidate supplier combination comprises:

calculating the matching degree between the single candidate provider in each candidate provider combination and the purchasing demand data as the basic evaluation score of the single candidate provider in each candidate provider combination;

And calculating the coordination degree between each single candidate provider in each candidate provider combination as the combination evaluation score of each candidate provider combination.

5. The method of claim 4, wherein said calculating a degree of match between individual candidate suppliers within each candidate supplier combination and said purchase demand data as a base rating score for individual candidate suppliers within each candidate supplier combination comprises:

Determining the priority of each purchasing index based on the purchasing demand data; wherein,

The purchasing index comprises: purchasing objective, purchasing quantity, purchasing quality requirement and demand delivery time;

matching degree between each single candidate provider and each purchasing index is carried out based on a machine learning algorithm, and matching degree scores corresponding to the purchasing indexes are generated based on the determined matching degree;

Performing weighted summation on matching degree scores corresponding to all the purchasing indexes based on the priority of all the purchasing indexes to obtain matching degrees between corresponding single candidate suppliers and the purchasing demand data;

traversing all individual candidate suppliers to obtain individual candidate suppliers in each candidate supplier combination

A base rating score.

6. The method of claim 4, wherein calculating the degree of fit between individual candidate suppliers within each candidate supplier combination as a combination score for each candidate supplier combination comprises:

collecting real-time supply capacity information and reserved purchasing cooperation information of each single candidate provider in each candidate provider combination;

And training the real-time supply capacity information and the reserved purchasing cooperation information based on the pre-constructed combined evaluation score scoring model to obtain combined evaluation scores of all candidate supplier combinations.

7. The method of claim 6, wherein the method further comprises:

performing a combined evaluation score scoring model pre-build comprising:

Collecting historical purchase data, and executing historical cooperation information reading of each candidate provider based on the historical purchase data;

Executing random combinations of candidate suppliers to obtain a plurality of historical supplier combinations;

pushing a plurality of history provider combinations to a user side, and recovering a combination evaluation score annotation result of a user based on the user side to obtain annotation samples as annotation information of each history provider combination;

Sample expansion is carried out on the labeling sample, and an expanded labeling sample is obtained;

And performing model training in a preselected neural network based on the expanded labeling sample to obtain a combined evaluation score scoring model.

8. The method of claim 1, wherein the determining the purchase score for each candidate supplier combination based on the combined score for each candidate supplier combination and the base score for the individual candidate suppliers within each candidate supplier combination, respectively, comprises:

performing a summation process within each candidate provider combination based on the base evaluation scores of each individual candidate provider to obtain a base evaluation score sum for each candidate provider combination;

And performing basic evaluation scores and weights of the candidate supplier combinations and combined evaluation score weights of the candidate supplier combinations based on a preset weight distribution relation, and performing purchase score calculation of the candidate supplier combinations based on the weight results to obtain purchase scores of the candidate supplier combinations.

9. The method of claim 1, wherein the performing candidate supplier combination recommendation based on the ranking result comprises:

based on the sorting result, judging the number of candidate supplier combinations with purchase scores greater than a preset recommended purchase score threshold;

If the number of the candidate supplier combinations with the purchase scores larger than the preset recommended purchase score threshold is larger than the preset number threshold N, taking the first N candidate supplier combinations with the purchase scores larger than the preset recommended purchase score threshold as recommended candidate supplier combinations;

And if the number of the candidate supplier combinations with the purchase scores greater than the preset recommended purchase score threshold is not greater than the preset number threshold N, taking all candidate supplier combinations with the purchase scores greater than the preset recommended purchase score threshold as recommended candidate supplier combinations.

10. The method of claim 1, wherein the performing purchase sourcing based on the recommended candidate provider combination to obtain a purchase sourcing result comprises:

Generating a recommended candidate provider set corresponding to each candidate provider combination based on each recommended candidate provider combination;

performing union processing on each recommendation candidate provider set to obtain all recommendation candidate provider targets to be searched;

performing target source searching of each candidate recommendation candidate provider to be searched to obtain a source searching result of each candidate recommendation provider;

And executing the source searching result combination corresponding to each recommended candidate provider under each recommended candidate provider combination to obtain the source searching result of each recommended candidate provider combination.

11. A procurement sourcing system based on vendor-combined intelligent recommendations, the system comprising:

The acquisition unit is used for acquiring purchasing demand data, determining candidate suppliers based on the purchasing demand data and obtaining candidate supplier combinations;

A scoring unit for calculating a combined score of each candidate provider combination and a base score of a single candidate provider in each candidate provider combination;

A processing unit for determining purchase scores of each candidate provider combination based on the combined evaluation score of each candidate provider combination and the basic evaluation score of the individual candidate provider within each candidate provider combination, respectively, and performing ranking on the purchase scores of each candidate provider combination;

And the source searching unit is used for performing candidate supplier combination recommendation based on the sorting result, and performing purchase source searching based on the recommended candidate supplier combination to obtain a purchase source searching result.

12. A computer readable storage medium having instructions stored thereon which when run on a computer cause the computer to perform the supplier-based combined intelligent recommended procurement sourcing method of any of claims 1-10.