Procurement method, system and device fusing fuzzy matching and multi-dimensional decision-making

Abstract

The application provides a procurement method, system and equipment fusing fuzzy matching and multi-dimensional decision, the method comprising: establishing a connection with an internal procurement system, obtaining heterogeneous data of the internal procurement system; identifying entity features in the heterogeneous data, and converting the entity features into numerical vectors; filtering out entity feature pairs pointing to the same real entity according to the numerical vectors, and attributing the entity feature pairs to the same cluster; wherein each cluster comprises an associated network graph about the real entity; constructing a decision model, the decision model being used to generate procurement strategies based on multiple dimensions; taking the associated network graph as a rule database, and taking historical strategies and / or random strategies as a training set to optimize the decision model, so that the procurement strategies generated by the decision model meet preset requirements; and generating multiple procurement strategies by using the optimized decision model. The application solves the problems of data fuzziness and decision dimension fragmentation in the procurement strategy generation process.

Description

Technical Field

[0001] This application relates to the field of computer technology, and in particular to procurement methods, systems and equipment that integrate fuzzy matching and multidimensional decision-making. Background Technology

[0002] As enterprise supply chains become increasingly complex and market competition intensifies, traditional methods for formulating annual procurement strategies face numerous challenges: These include: data heterogeneity issues—procurement data, supplier data, market data, patent technology data, etc., come from diverse sources, have varying formats, and differ in standards, resulting in a large amount of incomplete and inconsistent "dirty data"; information ambiguity issues—key information such as supplier capabilities, material technical parameters, and market risks inherently possesses ambiguity and uncertainty, making it difficult to quantify with precise values; fragmented decision-making dimensions—cost, quality, delivery, service, risk, and innovation are often considered in isolation, lacking systematic collaborative optimization; and static strategies—once formulated, traditional strategies are difficult to dynamically adjust in response to changes in the market environment, supply chain dynamics, and corporate strategy.

[0003] Therefore, there is an urgent need for a procurement method that can be applied to complex supply chains to meet the requirements for processing efficiency, accuracy, and dynamic adjustment in the procurement process. Summary of the Invention

[0004] In view of this, the purpose of this application is to propose a procurement method, system and equipment that integrates fuzzy matching and multi-dimensional decision-making, which solves the problems of data fuzziness and decision dimension fragmentation in the procurement strategy generation process.

[0005] To achieve one of the aforementioned objectives, this application provides a procurement method that integrates fuzzy matching and multi-dimensional decision-making, the method comprising:

[0006] Obtain heterogeneous data from the internal procurement system; Identify entity features in the heterogeneous data and convert the entity features into numerical vectors; Based on the numerical vector, entity feature pairs pointing to the same real-world entity are selected, and the entity feature pairs are grouped into the same cluster; wherein, each cluster includes an association network graph related to the real-world entity; A decision model for generating multi-dimensional procurement strategies is constructed. The correlation network graph is used as a rule database, and historical strategies and / or random strategies are used as a training set to optimize the decision model so that the procurement strategies generated by the decision model meet preset requirements. Real-time data from the internal procurement system is acquired and input into the optimized decision model, which is then used to generate multiple procurement strategies.

[0007] As a further improvement to one embodiment of this application, the step of identifying entity features in the heterogeneous data and converting the entity features into numerical vectors includes: Using natural language processing technology, the procurement text in the heterogeneous data is analyzed, entities in the procurement text are identified, entity features associated with the entities are extracted, and the entity features are converted into numerical vectors. The step of filtering entity feature pairs pointing to the same real-world entity based on the numerical vector includes: A similarity algorithm is used to perform a screening process to select candidate entity feature pairs. A secondary screening is performed using a composite similarity function that integrates edit distance, similarity algorithm, and semantic similarity based on domain knowledge base to select entity feature pairs from the candidate entity feature pairs; wherein the entity feature pairs point to the same real entity.

[0008] As a further improvement to one embodiment of this application, the entity feature pairs are grouped into the same cluster; wherein each cluster includes an association network graph of the real-world entities, including: The entity feature pairs are grouped into the same cluster, and unique master data is generated for each cluster; Based on the unique master data, the associated network graph is generated, which includes at least one of the following: the real entity's suppliers, supplied materials, related technologies, historical transactions, and performance evaluations.

[0009] As a further improvement to one embodiment of this application, the construction of a decision model, which is used to generate a procurement strategy based on multiple dimensions, includes: Construct the decision-making model from multiple dimensions; wherein, the multiple dimensions include at least the scope of procurement, procurement level, procurement method, and procurement organization form; Construct the strategy setpoints for the decision-making model; wherein the strategy setpoints include qualitative or quantitative methods. Construct a rule database for the decision-making model; wherein the rule database includes all the associated network graphs.

[0010] As a further improvement to one embodiment of this application, the step of using the associated network graph as a rule database and using historical strategies and / or random strategies as a training set to optimize the decision model so that the procurement strategy generated by the decision model meets preset requirements includes: The decision variables of the procurement strategy are encoded into the chromosome of the optimization algorithm; By incorporating the chromosome into the multidimensional perspective, the target value of the procurement strategy is calculated. The population of historical and / or random strategies is used as the initial scheme, and the target value is made to meet the preset requirements by evolving the population.

[0011] As a further improvement to one embodiment of this application, the step of making the procurement strategy generated by the decision model conform to preset requirements includes: The final decision objective is set as the preset requirement, and the final decision objective includes risk threshold, performance benchmark, minimization of total cost, and strategic weight. The decision-making model aims to ensure that the overall risk of the procurement strategy generated is lower than the risk threshold and the predicted performance of the procurement strategy is higher than the performance benchmark, with minimizing total cost as one of the optimization objectives, while also taking into account the calculation of the strategic weights.

[0012] As a further improvement to one embodiment of this application, the step of evolving the population to make the target value meet the preset requirements includes: The population is evolved by performing selection, crossover, and mutation operations using a non-dominated sorting genetic algorithm with an elitist strategy. The population diversity is maintained by using the crowding comparison operator in the non-dominated sorting genetic algorithm, and a Pareto optimal population is finally determined; wherein the target value of the procurement strategy in the Pareto optimal population meets the preset requirements.

[0013] As a further improvement to one embodiment of this application, after generating multiple procurement strategies using the optimized decision model, the process includes: Obtain the current preferences of decision-makers; Based on the current preferences, a recommended strategy is selected from multiple procurement strategies, and the recommended strategy is pushed to the interactive platform.

[0014] Based on the same inventive concept, this application also provides a procurement system that integrates fuzzy matching and multi-dimensional decision-making, including: The acquisition module is used to acquire heterogeneous data from the internal procurement system; A conversion module is used to identify entity features in the heterogeneous data and convert the entity features into numerical vectors. The filtering module is used to filter out entity feature pairs that point to the same real entity based on the numerical vector, and group the entity feature pairs into the same cluster; wherein, each cluster includes an association network graph about the real entity; An optimization module is used to construct a decision model for generating multi-dimensional procurement strategies. The model is optimized using the correlation network graph as a rule database and historical and / or random strategies as a training set, so that the procurement strategies generated by the decision model meet preset requirements. The generation module is used to acquire real-time data from the internal procurement system and generate multiple procurement strategies using an optimized decision-making model.

[0015] Based on the same inventive concept, this application also provides an electronic device, including: a processor and a memory; the memory stores a computer program, which, when executed by the processor, causes the processor to perform the steps of the above-described procurement method that integrates fuzzy matching and multidimensional decision-making.

[0016] Compared to existing technologies, the technical advantages of this invention are as follows: By constructing an entity association network graph through fuzzy matching, it not only serves procurement but also provides data and insight support for R&D sourcing, supply chain finance, and corporate strategic investment. Furthermore, by constructing and optimizing decision-making models, it transforms experience-based qualitative decisions into data- and model-based quantitative-qualitative decisions, effectively reducing subjective bias. This invention can provide large enterprises with complex supply chains, diverse procurement categories, and high cost and risk pressures with a complete digital and intelligent procurement strategy solution, enabling digital transformation and intelligent upgrading of the supply chain. Attached Figure Description

[0017] To more clearly illustrate the technical solutions in this application or related technologies, the drawings used in the description of the implementation methods or related technologies will be briefly introduced below. Obviously, the drawings described below are only the implementation methods of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 A flowchart illustrating a procurement method integrating fuzzy matching and multi-dimensional decision-making, provided for one embodiment of this application; Figure 2 A schematic diagram illustrating a clustering process of fuzzy features provided in one embodiment of this application; Figure 3 A scatter plot of clustered fuzzy features provided in one embodiment of this application; Figure 4 A schematic diagram of a decision model provided in one embodiment of this application; Figure 5 A schematic diagram of a procurement system that integrates fuzzy matching and multidimensional decision-making, provided as another embodiment of this application; Figure 6 This is a schematic diagram of the hardware structure of an electronic device provided for another embodiment of this application. Detailed Implementation

[0019] The present invention will now be described in detail with reference to the specific embodiments shown in the accompanying drawings. However, these embodiments do not limit the present invention, and any structural, methodological, or functional modifications made by those skilled in the art based on these embodiments are included within the scope of protection of the present invention.

[0020] It should be noted that, unless otherwise defined, the technical or scientific terms used in the embodiments of this application should have the ordinary meaning understood by those skilled in the art to which this application pertains. The terms "first," "second," and similar terms used in the embodiments of this application do not indicate any order, quantity, or importance, but are merely used to distinguish different components. Terms such as "comprising" or "including" mean that the element or object preceding the word encompasses the elements or objects listed following the word and their equivalents, without excluding other elements or objects.

[0021] Existing technologies include several procurement management systems and data analysis tools, but they typically suffer from the following shortcomings: they can only process structured data and are incapable of handling unstructured data such as text; they only provide post-event statistical analysis and lack pre-event prediction and strategy generation capabilities; or they employ simple weighted scoring methods, which cannot handle fuzzy decision-making and multi-objective conflict problems in complex situations. Therefore, there is an urgent need for a digital intelligence platform that can deeply integrate multi-source heterogeneous data, intelligently process fuzzy information, and generate dynamic optimization strategies under multi-dimensional constraints.

[0022] To address the aforementioned problems, this application provides a procurement method that integrates fuzzy matching and multi-dimensional decision-making, such as... Figure 1 As shown, the method includes the following steps: Step S100: Obtain heterogeneous data from the internal procurement system.

[0023] Specifically, before formulating a procurement strategy, it is essential to acquire as much diverse data as possible, including procurement data, supplier data, market data, and patent technology data. Due to the different formats, standards, or incompleteness of key information such as supplier capabilities, material technical parameters, and market risks, these data inherently possess ambiguity and uncertainty.

[0024] Step S200: Identify entity features in heterogeneous data and convert the entity features into numerical vectors.

[0025] Specifically, the process begins by connecting to the enterprise's internal ERP and other procurement systems. A fuzzy matching engine is then used to clean, align, and merge the extracted multi-source heterogeneous data, performing preliminary format standardization and missing value handling to create a usable database. Next, the data in the database is transformed into a unified numerical vector, representing the heterogeneous and fuzzy data in a consistent format. This achieves unified processing of multi-source heterogeneous data, breaking down data format barriers and addressing the pain points of fuzzy, scattered, and non-standardized procurement data. It provides a standardized, high-quality data foundation for subsequent entity matching and decision analysis, while simultaneously reducing the workload of manual data processing, minimizing errors during data cleaning, and improving data utilization efficiency.

[0026] In one possible implementation of this application, step S200 includes: Step S210: Using natural language processing technology, analyze the procurement text in heterogeneous data, identify entities in the procurement text, extract entity features associated with the entities, and convert the entity features into numerical vectors.

[0027] Specifically, the entities in the procurement document include suppliers, materials, and technical keywords. For supplier names, in addition to the name itself, other address keywords and product categories are extracted as features, and these features are converted into numerical vectors.

[0028] Step S300: Select entity feature pairs pointing to the same real entity based on the numerical vector, and group the entity feature pairs into the same cluster; wherein, each cluster includes a relational network graph about the real entity.

[0029] Specifically, based on the numerical vector obtained in step S200, a multi-level similarity filtering mechanism is used to remove redundant and duplicate entity features. Entity features pointing to the same real-world entity (such as the same supplier or the same material) are grouped into a cluster, forming entity feature pairs, such as... Figure 2 and Figure 3 As shown, these entity feature pairs are then grouped into the same cluster using a clustering algorithm to achieve entity disambiguation. At the same time, a network graph is constructed based on the association information of entities within the cluster, clearly presenting the inherent connections between entities and providing structured association data support for subsequent decision-making models.

[0030] In one possible implementation of this application, step S300 includes: Step S310: Use a similarity algorithm to perform a screening to select candidate entity feature pairs.

[0031] Step S320: A secondary screening is performed using a composite similarity function that incorporates edit distance, similarity algorithm, and semantic similarity based on domain knowledge base to select entity feature pairs from candidate entity feature pairs; wherein the entity feature pairs point to the same real entity.

[0032] Specifically, the fuzzy matching in this method employs a two-level matching strategy. The first level performs a rapid coarse screening based on similarity algorithms, such as the SimHash algorithm, to identify potentially identical candidate entity pairs. The second level, a secondary screening, refines the coarse screening results using a composite similarity function that integrates edit distance, Jaccard similarity, and semantic similarity based on a domain knowledge base. The semantic similarity calculation utilizes a Word2Vec model trained on procurement contracts and technical document corpora.

[0033] In one possible implementation of this application, step S300 further includes: Step S330: Group entity feature pairs into the same cluster and generate unique master data for each cluster.

[0034] Step S340: Based on the unique master data, generate a relational network graph, which includes at least one of the following: suppliers of real-world entities, supplied materials, related technologies, historical transactions, and performance evaluations.

[0035] Specifically, after identifying entity feature pairs, clustering is performed using the DBSCAN clustering algorithm. Records pointing to the same real-world entity are grouped into the same cluster based on fine similarity. A unique, accurate, and complete "golden record" is generated for each cluster as master data. Based on the master data, a network graph of associations is constructed, connecting "suppliers - supplied materials - related patented technologies - historical transactions - performance evaluations".

[0036] Step S400: Construct a decision model for generating multi-dimensional procurement strategies, using a relational network graph as a rule database and historical and / or random strategies as a training set to optimize the decision model so that the procurement strategies generated by the decision model meet preset requirements.

[0037] Specifically, based on the rules of the network graph, the training data of the decision-making model is ensured to be relevant and complete, avoiding the one-sidedness of model training; historical strategies are introduced as training sets, which can draw on past successful procurement experience to improve the rationality and feasibility of the generated strategies. At the same time, random strategies are added to enrich the training samples and prevent the model from getting stuck in local optima; the model is iteratively optimized through optimization algorithms so that the generated procurement strategies can meet the preset cost, risk, performance and other objectives, achieve optimal decision-making under multi-objective constraints, improve the robustness and competitiveness of the procurement strategies, and shorten the model optimization cycle and improve the efficiency of strategy generation.

[0038] Specifically, it breaks away from the single-dimensional limitations of traditional procurement strategy formulation, and achieves comprehensiveness, scientificity, and pertinence of procurement strategies by constructing decision models from multiple dimensions; it uses the relational network graph as the rule basis, enabling the decision model to make full use of entity relational information and avoid the one-sidedness of decision-making; at the same time, it adapts to the needs of different business scenarios by combining qualitative and quantitative strategy setting, promotes the transformation of procurement decision-making from experience-driven to data-driven, and improves the rationality and feasibility of procurement strategies.

[0039] In one possible implementation of this application, step S400 includes: Step S410: Construct a multi-dimensional perspective for the decision-making model; wherein, the multi-dimensional perspective includes at least the scope of procurement, procurement level, procurement method, and procurement organization form.

[0040] Step S420: Construct the strategy setpoints for the decision-making model; wherein the strategy setpoints include qualitative or quantitative ones.

[0041] Step S430: Construct a rule database for the decision model; wherein, the rule database includes all related network graphs.

[0042] Specifically, such as Figure 4 As shown, a decision model for annual procurement strategy based on four-dimensional dual-deterministic fuzzy data mining is constructed. The "four dimensions" are the perspectives of procurement strategy, namely "procurement scope", "procurement level", "procurement method" and "procurement organization form". "Dual determination" refers to the two sets of strategies due to different business natures, namely "qualitative" and "quantitative". "Fuzzy data mining" refers to the process and technology of searching for hidden information in a large amount of data through algorithms.

[0043] It should be noted that the decision-making model optimizes the mechanism for determining the scope, level, method, and organizational form of procurement, promoting a shift in procurement strategy from experience-based judgment to scientific decision-making, and ensuring the standardization and forward-looking nature of strategy formulation. Its multi-dimensional perspective enables intelligent recommendations for procurement scope, procurement level, procurement method, procurement organizational form, and intelligent prediction results for procurement strategies. The decision-making model, based on multi-dimensional perspectives, two strategy parameters, and fuzzy data mining, provides core technical support for the intelligent compilation and dynamic optimization of procurement catalogs, realizing a shift from "humans searching for data" to "data-assisted decision-making."

[0044] In one possible implementation of this application, step S400 includes: Step S440: Encode the decision variables of the procurement strategy into the chromosome of the optimization algorithm.

[0045] Step S450: Incorporate the chromosomes into a multidimensional perspective and calculate the target value of the procurement strategy.

[0046] Step S460: Use historical strategies and / or random strategies as the initial scheme population, and evolve the population to make the target value meet the preset requirements.

[0047] Specifically, the decision variables of the procurement strategy (such as procurement quotas from various suppliers, contract periods, and pricing models) are first encoded into chromosomes for the optimization algorithm. An initial solution population is formed by combining historically successful strategies and randomly generated strategies. Each strategy (chromosome) is then substituted into the aforementioned four-dimensional decision model to calculate its four objective values: total cost, overall risk, strategic fit, and predicted performance. Each procurement strategy (chromosome) is then substituted into the decision model to calculate its four objective values: total cost, overall risk, strategic fit, and predicted performance.

[0048] In one possible implementation of this application, step S460 includes: Step S461: Use a non-dominated sorting genetic algorithm with an elitist strategy to perform selection, crossover, and mutation operations on the population to evolve the population.

[0049] Step S462: The crowding comparison operator in the non-dominated sorting genetic algorithm is used to maintain the diversity of the population, and finally the Pareto optimal population is determined; wherein, the target value of the procurement strategy in the Pareto optimal population meets the preset requirements.

[0050] Specifically, in the multi-dimensional optimization process, a fast non-dominated sorting genetic algorithm with an elitist strategy is used for iterative optimization. The algorithm evolves the population through selection, crossover, and mutation operations, and maintains the diversity of the solution set using a crowding comparison operator, ultimately approximating or finding the Pareto optimal solution set. This solution set represents a set of balanced strategies that "cannot improve any objective without making other objectives worse."

[0051] In one possible implementation of this application, step S400, which involves making the procurement strategy generated by the decision model conform to preset requirements, further includes: Step S401: Set the final decision objective as a preset requirement. The final decision objective includes risk threshold, performance benchmark, minimizing total cost, and strategic weight.

[0052] Step S402 ensures that the overall risk of the procurement strategy generated by the decision model is lower than the risk threshold and the predicted performance of the procurement strategy is higher than the performance benchmark, with minimizing total cost as one of the optimization objectives, while also taking into account the calculation of strategic weights.

[0053] Specifically, the final decision objective function can be expressed as controlling the overall risk R. total Below the threshold θ, and the predicted performance P total Minimize the total cost C, given a value higher than the baseline δ. total At the same time, the strategic weight W sThat is: Minimize C total Subject to R total <θ,P total >δ, and Maximize Strategic Alignment (W s ).

[0054] Step S500: Obtain real-time data from the internal procurement system and input it into the optimized decision model. Use the optimized decision model to generate multiple procurement strategies.

[0055] In one possible implementation of this application, after step S500, the following is included: Step S601: Obtain the decision-maker's current preferences.

[0056] Step S602: Select a recommended strategy from multiple procurement strategies based on current preferences, and push the recommended strategy to the interactive platform.

[0057] Specifically, from the Pareto optimal population, 1-3 optimal strategies are recommended based on the decision-maker's current preferences (such as "cost-first" or "risk-averse"). Simultaneously, interpretability analysis of the strategies is provided, for example, using SHAP values ​​to demonstrate the degree of influence of each decision variable on the final cost or risk.

[0058] In one feasible implementation of this application, the procurement method integrating fuzzy matching and multidimensional decision-making offers the following advantages: The method effectively enhances the scientific rigor of decision-making: through fuzzy matching and intelligent optimization decision models, experience-based qualitative decisions are transformed into data- and model-based quantitative-qualitative decisions, effectively reducing subjective bias. The method significantly increases the efficiency of procurement strategy generation: automating data cleaning, model calculation, and solution optimization shortens the strategy formulation cycle from weeks to days or even hours, enabling procurement departments to respond quickly to market changes. The strategy possesses foresight and dynamism: through simulation and closed-loop learning, the strategy acquires the ability to "foresee the future" and "dynamically adjust," transforming from a static plan into a dynamic intelligent agent. The constructed supplier-material-technology network not only serves procurement but also provides data and insight support for R&D sourcing, supply chain finance, and corporate strategic investment.

[0059] In its implementation, a large equipment manufacturing company applied this method to formulate its annual core component procurement strategy. First, it integrated three years of internal transaction data with publicly available data and patent data from five external suppliers. Through fuzzy matching, it consolidated the original 1270 supplier records into 215 unique entities. Under two simulated scenarios—"a 15% expected increase in steel prices" and "potential supply chain disruptions in a certain region"—the platform-generated strategy, compared to traditional methods, predicted an 8.5% reduction in total costs while ensuring supply security. It also identified two potential suppliers with alternative innovative patents, enhancing supply chain resilience. Decision-makers adjusted risk preference parameters through a strategy simulation, instantly obtaining strategy options with different emphases, ultimately forming a procurement plan that combines robustness and cost competitiveness.

[0060] Another embodiment of this application discloses a digital intelligence platform that adopts a microservice architecture, including a data perception and fusion layer, an algorithm and model layer, a strategy generation and simulation layer, and an interactive application layer.

[0061] The data perception and fusion layer is responsible for connecting the enterprise's internal procurement systems such as ERP. Through the built-in fuzzy matching engine, it cleans, aligns, and merges the extracted multi-source heterogeneous data to form a unified "procurement data lake".

[0062] The algorithm and model layer includes: a fuzzy matching algorithm library, a multidimensional decision model library, and an intelligent optimization engine. The fuzzy matching algorithm library integrates an improved edit distance algorithm, a word vector-based semantic similarity algorithm, and a customized entity recognition and disambiguation model for the procurement domain. The multidimensional decision model library includes built-in strategic positioning models (such as the Karajek matrix), cost analysis models, risk assessment models (such as fuzzy DEMATEL), and performance prediction models. The intelligent optimization engine integrates genetic algorithms and particle swarm optimization algorithms to solve for optimal procurement solutions under multi-objective constraints.

[0063] The strategy generation and simulation layer automatically generates preliminary strategy plans based on the underlying models and algorithms, and conducts multi-scenario simulations (such as price surges and supply disruptions) in a "digital twin" environment to evaluate the resilience of the strategy.

[0064] The interactive application layer provides procurement decision-makers with intuitive strategic insights, recommended solutions, and risk warnings through visual dashboards, intelligent reports, and interactive "strategy sandboxes."

[0065] Another embodiment of this application discloses a procurement system that integrates fuzzy matching and multi-dimensional decision-making, such as... Figure 5 As shown, it includes: The acquisition module is used to acquire heterogeneous data from the internal procurement system. The conversion module is used to identify entity features in heterogeneous data and convert the entity features into numerical vectors. The filtering module is used to filter out entity feature pairs that point to the same real-world entity based on the numerical vectors, and group the entity feature pairs into the same cluster; wherein, each cluster includes a relational network graph about the real-world entity; The optimization module is used to build a decision model for generating multi-dimensional procurement strategies. It uses a network graph as a rule database and historical and / or random strategies as a training set to optimize the decision model so that the procurement strategies generated by the decision model meet the preset requirements. The generation module is used to acquire real-time data from the internal procurement system and generate multiple procurement strategies using an optimized decision-making model.

[0066] Figure 6 This diagram illustrates a more specific hardware structure of an electronic device provided in this embodiment. The device may include: a processor 1010, a memory 1020, an input / output interface 1030, a communication interface 1040, and a bus 1050. The processor 1010, memory 1020, input / output interface 1030, and communication interface 1040 are interconnected internally via the bus 1050.

[0067] The processor 1010 can be implemented using a general-purpose CPU (Central Processing Unit), microprocessor, application-specific integrated circuit (ASIC), or one or more integrated circuits, and is used to execute relevant programs to implement the technical solutions provided in the embodiments of this specification.

[0068] The memory 1020 can be implemented in the form of ROM (Read Only Memory), RAM (Random Access Memory), static storage device, dynamic storage device, etc. The memory 1020 can store the operating system and other application programs. When the technical solutions provided in the embodiments of this specification are implemented by software or firmware, the relevant program code is stored in the memory 1020 and is called and executed by the processor 1010.

[0069] The input / output interface 1030 is used to connect input / output modules to realize information input and output. The input / output modules can be configured as components in the device (not shown in the figure) or externally connected to the device to provide corresponding functions. Input devices may include keyboards, mice, touch screens, microphones, various sensors, etc., and output devices may include displays, speakers, vibrators, indicator lights, etc.

[0070] The communication interface 1040 is used to connect the communication module (not shown in the figure) to enable communication between this device and other devices. The communication module can communicate via wired means (such as USB, network cable, etc.) or wireless means (such as mobile network, WIFI, Bluetooth, radio (shortwave / ultra-shortwave) communication, satellite communication, data link communication, etc.).

[0071] Bus 1050 includes pathways for transmitting information between various components of the device, such as processor 1010, memory 1020, input / output interface 1030, and communication interface 1040.

[0072] It should be noted that although the above-described device only shows the processor 1010, memory 1020, input / output interface 1030, communication interface 1040, and bus 1050, in specific implementations, the device may also include other components necessary for normal operation. Furthermore, those skilled in the art will understand that the above-described device may only include the components necessary for implementing the embodiments described in this specification, and need not include all the components shown in the figures.

[0073] The electronic device described above is used to implement the procurement method that integrates fuzzy matching and multidimensional decision-making in any of the foregoing embodiments, and has the beneficial effects of the corresponding method implementation, which will not be elaborated here.

[0074] Based on the same inventive concept, corresponding to any of the above-described embodiments, this application also provides a non-transitory computer-readable storage medium storing computer instructions for causing the computer to execute the procurement method integrating fuzzy matching and multi-dimensional decision-making as described in any of the above embodiments.

[0075] The computer-readable medium in this embodiment includes permanent and non-permanent, removable and non-removable media, and information storage can be implemented by any method or technology. Information can be computer-readable instructions, data structures, program modules, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, magnetic magnetic disk storage or other magnetic storage devices, or any other non-transfer medium that can be used to store information accessible by a computing device.

[0076] The computer instructions stored in the storage medium of the above embodiments are used to cause the computer to execute the procurement method that integrates fuzzy matching and multi-dimensional decision-making as described in any of the above embodiments, and have the beneficial effects of the corresponding method implementations, which will not be repeated here.

[0077] Those skilled in the art should understand that the discussion of any of the above embodiments is merely exemplary and is not intended to imply that the scope of this application (including the claims) is limited to these examples; this manner of description is merely for clarity, and those skilled in the art should consider the specification as a whole. Within the framework of this application, the above embodiments or the technical features of different embodiments can also be appropriately combined, the steps can be implemented in any order, and there are many other variations of different aspects of the embodiments of this application as described above, which are not provided in the details for the sake of brevity.

[0078] Additionally, to simplify the description and discussion, and to avoid obscuring the embodiments of this application, the well-known power / ground connections to integrated circuit (IC) chips and other components may or may not be shown in the provided drawings. Furthermore, the apparatus may be shown in block diagram form to avoid obscuring the embodiments of this application, and this also takes into account the fact that the details of the implementation of these block diagram apparatuses are highly dependent on the platform on which the embodiments of this application will be implemented (i.e., these details should be entirely within the understanding of those skilled in the art). While specific details (e.g., circuits) are set forth to describe exemplary embodiments of this application, it will be apparent to those skilled in the art that the embodiments of this application can be implemented without these specific details or with variations thereof. Therefore, these descriptions should be considered illustrative rather than restrictive.

[0079] Although this application has been described in conjunction with specific embodiments thereof, many substitutions, modifications, and variations of these embodiments will be apparent to those skilled in the art from the foregoing description. For example, other memory architectures (e.g., dynamic RAM (DRAM)) may use the embodiments discussed.

[0080] The embodiments described herein are intended to cover all such substitutions, modifications, and variations that fall within the broad scope of the appended claims. Therefore, any omissions, modifications, equivalent substitutions, improvements, etc., made without departing from the spirit and principles of the embodiments described herein should be included within the protection scope of this application.

Claims

1. A procurement method integrating fuzzy matching and multidimensional decision-making, characterized in that, The method includes: Obtain heterogeneous data from the internal procurement system; Identify entity features in the heterogeneous data and convert the entity features into numerical vectors; Based on the numerical vector, entity feature pairs pointing to the same real-world entity are selected, and the entity feature pairs are grouped into the same cluster; wherein, each cluster includes an association network graph related to the real-world entity; A decision model for generating multi-dimensional procurement strategies is constructed. The correlation network graph is used as a rule database, and historical strategies and / or random strategies are used as a training set to optimize the decision model so that the procurement strategies generated by the decision model meet preset requirements. The system acquires real-time data from the internal procurement system and inputs it into the optimized decision model, which is then used to generate multiple procurement strategies.

2. The procurement method integrating fuzzy matching and multi-dimensional decision-making according to claim 1, characterized in that, The step of identifying entity features in the heterogeneous data and converting the entity features into numerical vectors includes: Using natural language processing technology, the procurement text in the heterogeneous data is analyzed, entities in the procurement text are identified, entity features associated with the entities are extracted, and the entity features are converted into numerical vectors. The step of filtering entity feature pairs pointing to the same real-world entity based on the numerical vector includes: A similarity algorithm is used to perform a screening process to select candidate entity feature pairs. A secondary screening is performed using a composite similarity function that integrates edit distance, similarity algorithm, and semantic similarity based on domain knowledge base to select entity feature pairs from the candidate entity feature pairs; wherein the entity feature pairs point to the same real entity.

3. The procurement method integrating fuzzy matching and multi-dimensional decision-making according to claim 2, characterized in that, The entity feature pairs are grouped into the same cluster; wherein each cluster includes a relational network graph of the real-world entities, including: The entity feature pairs are grouped into the same cluster, and unique master data is generated for each cluster; Based on the unique master data, the associated network graph is generated, which includes at least one of the following: the real entity's suppliers, supplied materials, related technologies, historical transactions, and performance evaluations.

4. The procurement method integrating fuzzy matching and multi-dimensional decision-making according to claim 1, characterized in that, The construction of the decision model for generating multi-dimensional procurement strategies includes: Construct the decision-making model from multiple dimensions; wherein, the multiple dimensions include at least the scope of procurement, procurement level, procurement method, and procurement organization form; Construct the strategy setpoints for the decision-making model; wherein the strategy setpoints include qualitative or quantitative methods. Construct a rule database for the decision-making model; wherein the rule database includes all the associated network graphs.

5. The procurement method integrating fuzzy matching and multi-dimensional decision-making according to claim 4, characterized in that, The step of using the associated network graph as a rule database and historical and / or random strategies as a training set to optimize the decision model, so that the procurement strategy generated by the decision model meets preset requirements, includes: The decision variables of the procurement strategy are encoded into the chromosome of the optimization algorithm; By incorporating the chromosome into the multidimensional perspective, the target value of the procurement strategy is calculated. The population of historical and / or random strategies is used as the initial scheme, and the target value is made to meet the preset requirements by evolving the population.

6. The procurement method integrating fuzzy matching and multi-dimensional decision-making according to claim 1, characterized in that, The step of making the procurement strategy generated by the decision model conform to preset requirements includes: The final decision objective is set as the preset requirement, and the final decision objective includes risk threshold, performance benchmark, minimization of total cost, and strategic weight. The decision-making model aims to ensure that the overall risk of the procurement strategy generated is lower than the risk threshold and the predicted performance of the procurement strategy is higher than the performance benchmark, with minimizing total cost as one of the optimization objectives, while also taking into account the calculation of the strategic weights.

7. The procurement method integrating fuzzy matching and multi-dimensional decision-making according to claim 5, characterized in that, The process of evolving the population to make the target value meet preset requirements includes: The population is evolved by performing selection, crossover, and mutation operations using a non-dominated sorting genetic algorithm with an elitist strategy. The population diversity is maintained by using the crowding comparison operator in the non-dominated sorting genetic algorithm, and a Pareto optimal population is finally determined; wherein the target value of the procurement strategy in the Pareto optimal population meets the preset requirements.

8. The procurement method integrating fuzzy matching and multi-dimensional decision-making according to claim 1, characterized in that, After generating multiple procurement strategies using the optimized decision model, the process includes: Obtain the current preferences of decision-makers; Based on the current preferences, a recommended strategy is selected from multiple procurement strategies, and the recommended strategy is pushed to the interactive platform.

9. A procurement system integrating fuzzy matching and multi-dimensional decision-making, characterized in that, The system includes: The acquisition module is used to acquire heterogeneous data from the internal procurement system; A conversion module is used to identify entity features in the heterogeneous data and convert the entity features into numerical vectors. The filtering module is used to filter out entity feature pairs that point to the same real entity based on the numerical vector, and group the entity feature pairs into the same cluster; wherein, each cluster includes an association network graph about the real entity; An optimization module is used to construct a decision model for generating multi-dimensional procurement strategies. The model is optimized using the correlation network graph as a rule database and historical and / or random strategies as a training set, so that the procurement strategies generated by the decision model meet preset requirements. The generation module is used to acquire real-time data from the internal procurement system and generate multiple procurement strategies using an optimized decision-making model.

10. An electronic device, characterized in that, include: Processor and memory; The memory stores a computer program that, when executed by the processor, causes the processor to perform the steps of the procurement method that integrates fuzzy matching and multidimensional decision-making as described in any one of claims 1 to 8.

Images