Product supplier and consumer matching method and system based on information axiom

Abstract

The invention relates to the technical field of agricultural information, in particular to a product supplier and consumer matching method and system based on information axiom, and the method comprises the steps: recognizing the attributes of agricultural products; defining a system range and a design range for each attribute; for each matching scheme between the consumer and the supplier, checking whether all hard attributes are met or not, and if not, filtering out the matching scheme; for the soft attributes, calculating the information amount from the perspective of consumers and the perspective of suppliers: for the quantitative soft attributes, calculating the information amount according to the types of the quantitative soft attributes; for qualitative soft attributes, applying a fuzzy mathematical method, and then calculating information amount; respectively calculating the total information amount of the consumer perspective and the total information amount of the supplier perspective; establishing a multi-objective optimization model; for all possible matching schemes, calculating the total information amount of consumers and suppliers, and determining respective maximum values and minimum values; for each matching scheme, calculating a normalized membership function value by comparing the total information amount with the maximum value and the minimum value of all the matching schemes; the matching scheme enabling the weighted comprehensive score of the membership function to be the highest is selected. The invention provides an intelligent matching method which can comprehensively consider unique attributes, dynamic transaction demands and fuzzy language features of agricultural products, so as to optimize supply and demand matching and promote agricultural product sales and sustainable economic development.

Description

Technical Field

[0001] The present invention relates to the field of agricultural information technology, and in particular to a method and system for matching product suppliers and consumers based on information axioms. Background Art

[0002] Agriculture is a vital pillar of many countries' economies, particularly in rural areas, where small-scale farmers are the primary producers of agricultural products. In China, for example, as of 2022, there were still approximately 240 million small-scale farmers nationwide, who primarily rely on growing and selling agricultural products for income. However, due to limitations such as small-scale operations, insufficient capital, and outdated technology, smallholder farmers often adopt a family-based production model, resulting in fragmented production and sales, making it difficult to generate stable returns from agricultural product cultivation and trading. This fragmented agricultural production and management model presents significant market challenges for smallholder farmers, hindering the sustainable development of the rural economy and narrowing the urban-rural gap.

[0003] To facilitate the sale of agricultural products by smallholder farmers, various solutions have been developed. The establishment of agricultural product information platforms is a common approach. These platforms provide supply and demand information, connecting smallholder farmers with consumers or intermediaries. Leveraging internet technology, these platforms have transformed traditional sales models, reshaped the relationship between suppliers and consumers, and provided new sales channels for smallholder farmers. Additionally, some research has attempted to address the supply-demand matching problem by integrating supply chains, developing trading platforms, or forecasting logistics needs.

[0004] Although existing technologies have made some progress in facilitating agricultural product transactions, they still have the following limitations:

[0005] 1. Information asymmetry between supply and demand: Existing agricultural product information platforms mainly provide information sharing and transaction functions. However, it is difficult for consumers to find agricultural product solutions that fully meet their needs, and it is also difficult for suppliers to accurately match suitable consumers, resulting in low transaction efficiency.

[0006] 2. Ignoring the unique attributes of agricultural products: Existing research often addresses supply-demand matching from the perspective of supply chain integration or logistics forecasting, with less attention paid to the unique attributes of agricultural products, such as freshness, maturity, seasonality, and origin. These attributes significantly impact trade matching and transaction volume due to the perishability of agricultural products and the high logistics costs.

[0007] 3. Inadequate Applicability of Matching Methods: Existing matching algorithms each have specific application scenarios and struggle to simultaneously address the following key characteristics of agricultural product transactions: bilateral matching (consumer-supplier), consideration of both quantitative and qualitative attributes, and one-time transaction requirements. Existing methods are unable to fully adapt to the dynamic and ambiguous nature of agricultural product transactions.

[0008] 4. Market disadvantages of small farmers: Due to the lack of effective marketing channels and bargaining power, small farmers usually rely on intermediaries or residents of nearby towns to sell their products, resulting in opaque prices, compressed profits, and difficulty in achieving sustainable income growth.

[0009] In summary, existing technologies still have significant shortcomings in solving the difficulties faced by small farmers in selling agricultural products, especially in how to quickly and intelligently match suppliers and consumers to improve transaction efficiency and matching. There is a lack of systematic and targeted solutions. Summary of the Invention

[0010] In view of the above technical problems, the present invention provides a method and system for matching product suppliers and consumers based on information axioms, so as to provide an intelligent matching method that can comprehensively consider the unique attributes of agricultural products, dynamic transaction needs and fuzzy language characteristics, so as to optimize supply and demand matching and promote agricultural product sales and sustainable economic development.

[0011] Other features and advantages of the present invention will become apparent from the following detailed description, or may be learned in part by practice of the present invention.

[0012] According to one aspect of the present invention, a method for matching product suppliers and consumers based on information axioms is proposed, the method comprising:

[0013] Identify the attributes of agricultural products, including general attributes and specific attributes. General attributes include brand, logistics distance, product grade, and price. Specific attributes include freshness, maturity, product certification, seasonality, origin, and product safety. Categorize the general attributes and specific attributes into mandatory hard attributes and flexible soft attributes, as well as quantitative attributes and qualitative attributes. Soft attributes are categorized into interval-based, benefit-based, and cost-based types based on consumer satisfaction. The interval-based type requires attribute values to be close to a specific range, the benefit-based type requires attribute values to be as high as possible, and the cost-based type requires attribute values to be as low as possible.

[0014] For each attribute, define the system range and design range, where the system range is based on the attribute value range provided by the supplier and the design range is based on the attribute value range required by the consumer;

[0015] For each matching scheme between a consumer and a supplier, check whether all hard attributes are met. If not, filter out the matching scheme. For soft attributes, calculate the information content from both the consumer and supplier perspectives. For quantitative soft attributes, calculate the information content based on their type, and ensure that the information content is non-negative by comparing the degree of overlap between the system scope and the design scope. For qualitative soft attributes, apply fuzzy mathematics methods to quantify the linguistic description into numerical values using triangular or trapezoidal membership functions, and then calculate the information content. Calculate the total information content from the consumer perspective and the total information content from the supplier perspective, respectively. The total information content is the sum of the information content of all soft attributes in the corresponding matching scheme.

[0016] Establish a multi-objective optimization model, the goal of which is to minimize the sum of the total amount of information from the consumer's perspective and the sum of the total amount of information from the supplier's perspective in all matches, with the constraints that each supplier in the matching scheme is matched with at most one consumer, each consumer is matched with at most one supplier, and for each matched consumer-supplier pair, the amount of information from the consumer's perspective does not exceed a preset upper limit for the consumer, and the amount of information from the supplier's perspective does not exceed a preset upper limit for the supplier;

[0017] For all possible matching schemes, the total amount of information of consumers and suppliers is calculated, and the respective maximum and minimum values are determined; for each matching scheme, its normalized membership function value is calculated by comparing its total amount of information with the maximum and minimum values of all matching schemes; the matching scheme with the highest weighted comprehensive score of the membership function is selected, where the weight is a predefined value.

[0018] Furthermore, for quantitative soft attributes, the calculation of information quantity is based on a range comparison of attribute types, which include interval type, benefit type, or cost type.

[0019] Furthermore, for interval-type quantitative attributes, the calculation of information quantity takes into account the boundary relationship between the system scope and the design scope:

[0020] If the system range is completely outside the design range, the amount of information is infinite;

[0021] If the system range partially overlaps the design range, the information content is calculated based on the difference in range widths using a logarithmic function;

[0022] If the system scope is completely contained within the design scope, the amount of information is zero.

[0023] Furthermore, for qualitative soft attributes, the information amount is calculated using fuzzy membership functions and quantitative language descriptions.

[0024] Furthermore, the total amount of information matched between consumers and suppliers is the sum of the amount of information of all soft attributes in the matching scheme.

[0025] Furthermore, the membership function of the supplier is calculated by comparing the total information amount of the supplier in the corresponding matching solution with the maximum and minimum total information amounts in all the matching solutions, and calculating a normalized value.

[0026] Furthermore, the comprehensive score of each matching solution is calculated by performing weighted summation on the membership function values of the supplier and the consumer, wherein the weights are predefined values and the sum is 1.

[0027] Furthermore, the system scope and the design scope of each attribute are defined according to requirements provided by suppliers and consumers, respectively.

[0028] According to another aspect of the present invention, a product supplier and consumer matching system based on information axioms is provided, comprising:

[0029] An attribute identification and classification module is used to identify the attributes of agricultural products, including general attributes and specific attributes. General attributes include brand, logistics distance, product grade, and price, and specific attributes include freshness, maturity, product certification, seasonality, origin, and product safety. The module classifies the general attributes and specific attributes into mandatory hard attributes and flexible soft attributes, as well as quantitative attributes and qualitative attributes. Soft attributes are classified into interval-based, benefit-based, and cost-based types based on consumer satisfaction. The interval-based type requires attribute values to be close to a specific range, the benefit-based type requires attribute values to be as high as possible, and the cost-based type requires attribute values to be as low as possible.

[0030] A range definition module, configured to define, for each attribute, a system range and a design range, wherein the system range is based on an attribute value range provided by a supplier, and the design range is based on an attribute value range required by a consumer;

[0031] The information volume calculation module checks whether all hard attributes are met for each matching solution between consumers and suppliers. If not, the matching solution is filtered out. For soft attributes, the information volume is calculated from the consumer perspective and the supplier perspective respectively. For quantitative soft attributes, the information volume is calculated according to their type, and the degree of overlap between the system scope and the design scope is compared to ensure that the information volume is non-negative. For qualitative soft attributes, fuzzy mathematical methods are applied to quantify the language description into numerical values through triangular or trapezoidal membership functions, and then the information volume is calculated. The total information volume from the consumer perspective and the total information volume from the supplier perspective are calculated respectively. The total information volume is the sum of the information volume of all soft attributes in the corresponding matching solution.

[0032] a model optimization module for establishing a multi-objective optimization model, wherein the objective is to minimize the sum of the total amount of information from the consumer's perspective and the sum of the total amount of information from the supplier's perspective in all matches, subject to the constraints that each supplier in the matching scheme is matched with at most one consumer, each consumer is matched with at most one supplier, and for each matched consumer-supplier pair, the amount of information from the consumer's perspective does not exceed a preset upper limit for the consumer, and the amount of information from the supplier's perspective does not exceed a preset upper limit for the supplier;

[0033] The scheme selection module is used to calculate the total amount of information of consumers and suppliers for all possible matching schemes and determine their respective maximum and minimum values; for each matching scheme, calculate its normalized membership function value by comparing its total amount of information with the maximum and minimum values of all matching schemes; and select the matching scheme with the highest weighted comprehensive score of the membership function, where the weight is a predefined value.

[0034] The technical solution of the present invention has the following beneficial effects:

[0035] 1. Improve supply-demand matching efficiency: This invention comprehensively considers the general attributes (such as price, brand, and logistics distance) and specific attributes (such as freshness, maturity, and origin) of agricultural products, and categorizes them into hard and soft attributes, thereby accurately identifying and matching the needs of consumers and suppliers. Compared to existing agricultural product information platforms that only provide information sharing, this invention quantifies the amount of attribute information and applies a multi-objective optimization model to quickly screen matching solutions that meet hard attribute requirements and are optimal in terms of soft attributes, thereby significantly improving the efficiency of supply-demand matching and reducing information asymmetry during transactions.

[0036] 2. Enhance the adaptability of matching schemes: This invention employs information axioms and fuzzy mathematics to process quantitative and qualitative attributes separately, overcoming the limitations of existing matching algorithms (such as Gale-Shapley and SMAA) in handling agricultural product transactions, which are characterized by insufficient dynamism and fuzziness. For quantitative attributes, the amount of information is calculated by comparing the degree of overlap between the system scope and the design scope; for qualitative attributes, triangular or trapezoidal membership functions are used to quantify the language description. This comprehensive approach enables matching schemes to adapt to the unique attributes of agricultural products (such as seasonality and certification) and the diverse trading requirements (such as one-time transactions and bilateral matching), thereby improving the flexibility and applicability of matching schemes.

[0037] 3. Optimizing Mutual Satisfaction of Consumers and Suppliers: This method calculates information from both the consumer and supplier perspectives and minimizes the total information shared by both parties through a multi-objective optimization model. This ensures that the matching solution simultaneously meets both consumer demands for product quality and price and supplier expectations for market opportunities and revenue. A case study demonstrated that this matching method achieved a high overall matching degree (0.863) in a banana transaction. Compared to traditional methods that only consider unilateral needs, this method balances the interests of both parties, improving both transaction satisfaction and transaction success rates.

[0038] 4. Promote the economic development of smallholder farmers: This invention uses intelligent matching methods to provide smallholder farmers with direct-to-consumer sales channels, reducing reliance on intermediaries, lowering transaction costs, and increasing sales profits. In existing technologies, smallholder farmers often face profit squeezes due to a lack of effective marketing channels and bargaining power. Through precise matching and a transparent transaction process, this invention enhances the market competitiveness of smallholder farmers, provides them with a sustainable source of income, helps narrow the urban-rural economic gap, and promotes the sustainable development of the rural economy.

[0039] 5. Support for Dynamic Trading Scenarios: By defining the system scope (provided by suppliers) and the design scope (consumer requirements), and optimizing the matching scheme within constraints (each entity is matched at most once, and the amount of information does not exceed an upper limit), this system can adapt to the dynamic changes in agricultural product trading, such as seasonal fluctuations, price changes, and logistics requirements. Compared to the static or single matching methods used in existing technologies, this system can respond to market changes in real time, providing more stable and reliable trading support.

[0040] 6. Improved transaction transparency and fairness: This invention selects the optimal matching solution by normalizing membership function values and applying a weighted comprehensive score. This ensures that the matching process is based on objective information evaluation and avoids human intervention or subjective bias. Compared to the price opacity and information asymmetry issues in traditional agricultural product transactions, this invention provides a standardized matching process, enhances transaction transparency and fairness, and creates a fair trading environment for consumers and suppliers. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] Figure 1 This is a flow chart of a method for matching product suppliers and consumers based on information axioms in an embodiment of this specification;

[0042] Figure 2 This is a structural block diagram of a product supplier and consumer matching system based on information axioms in an embodiment of this specification. DETAILED DESCRIPTION

[0043] Example embodiments will now be described more fully with reference to the accompanying drawings. However, example embodiments can be implemented in many forms and should not be construed as limited to the examples set forth herein; rather, these embodiments are provided so that the present invention will be more comprehensive and complete and the concepts of the example embodiments will be fully conveyed to those skilled in the art. The described features, structures, or characteristics may be combined in any suitable manner in one or more embodiments. In the following description, many specific details are provided to provide a full understanding of the embodiments of the present invention. However, those skilled in the art will appreciate that the technical solutions of the present invention may be practiced while omitting one or more of the specific details, or that other methods, components, devices, steps, etc. may be employed. In other cases, well-known technical solutions are not shown or described in detail to avoid obscuring various aspects of the present invention.

[0044] The accompanying drawings are merely schematic illustrations of the present invention. Identical reference numerals in the drawings denote identical or similar components, and thus repetitive descriptions thereof will be omitted. Some of the blocks shown in the accompanying drawings represent functional entities that do not necessarily correspond to physically or logically independent entities. These functional entities may be implemented in software, in one or more hardware modules or integrated circuits, or in different networks and / or processor devices and / or microcontroller devices.

[0045] The present invention provides a method for matching product suppliers and consumers based on information axioms. Figure 1 The figure shows a flow chart of a method for matching product suppliers and consumers based on information axioms according to an embodiment of the present invention. The method can be applied to electronic devices such as personal computers and servers. The method can be performed by a device, which can be implemented by software and / or hardware. The method can specifically include the following steps S101 to S105:

[0046] In step S101, the attributes of agricultural products are identified, including general attributes and specific attributes, wherein the general attributes include brand, logistics distance, product grade, and price, and the specific attributes include freshness, maturity, product certification, seasonality, origin, and product safety; the contents of the general attributes and the specific attributes are classified into hard attributes that must be met and soft attributes with flexibility, as well as quantitative attributes and qualitative attributes; for the soft attributes, the types of the soft attributes are divided into interval type, benefit type, and cost type according to consumer satisfaction, wherein the interval type requires the attribute value to be close to a specific range, the benefit type requires the attribute value to be as high as possible, and the cost type requires the attribute value to be as low as possible.

[0047] In the process of matching supply and demand for agricultural products, it is first necessary to identify the key attributes that influence transaction decisions. This embodiment analyzes the characteristics of agricultural product transactions and identifies two types of attributes: general attributes and specific attributes. General attributes are characteristics that apply to most agricultural products and can directly influence the choices of both parties involved in the transaction. General attributes include brand, logistics distance, product grade, and price. For example, brand reflects the supplier's market reputation, logistics distance determines transportation costs and time, product grade indicates quality level, and price is a core consideration in transactions. Specific attributes target the unique characteristics of agricultural products, especially those with high perishability and regional requirements. Specific attributes include freshness, maturity, product certification, seasonality, origin, and product safety. For example, freshness affects the shelf life of fruits and vegetables, maturity determines the eating experience, product certification (such as organic certification) meets consumer safety needs, seasonality limits supply time, origin influences consumer preferences, and product safety ensures the absence of chemical residues. The identification of these attributes is based on actual market transaction needs to ensure comprehensive coverage of factors influencing matching.

[0048] After identifying the attributes, this embodiment classifies general attributes and specific attributes into hard attributes that must be met and soft attributes that are flexible. Hard attributes refer to non-negotiable requirements in a transaction. If they are not met, the possibility of matching is directly excluded. For example, the packaging type may be specified as a hard attribute. If the consumer requires plastic packaging and the supplier provides paper packaging, the match is invalid. Soft attributes allow a certain degree of flexibility, and consumers and suppliers can negotiate within a certain range. For example, price is usually classified as a soft attribute because consumers may accept a price that is slightly higher than expected but still reasonable. The selection of hard attributes is based on the mandatory requirements of the transaction, while the selection of soft attributes takes into account the optimization space of the transaction. In actual operation, the division of hard attributes and soft attributes is determined by the specific needs of the two parties to the transaction. For example, in a banana transaction, consumers can set product certification (such as organic certification) as a hard attribute and logistics distance as a soft attribute, allowing a certain range of fluctuations.

[0049] This embodiment further classifies common attributes and specific attributes into quantitative attributes and qualitative attributes to support subsequent information volume calculations. Quantitative attributes refer to characteristics that can be directly measured by numerical values and have clear ranges or units. For example, logistics distance (in kilometers), price (in yuan) and product grade (expressed in standard grades 1 to 5) are quantitative attributes. Qualitative attributes involve subjective evaluation or non-numerical descriptions, usually expressed in language. For example, brand (such as "a well-known brand" or "no brand"), freshness (such as "very fresh" or "generally fresh") and origin (such as "local" or "imported") are qualitative attributes. The processing of quantitative attributes is based on numerical range comparison, while qualitative attributes need to be quantified by fuzzy mathematical methods (such as membership functions). In implementation, the data of quantitative attributes are provided directly by suppliers and consumers, and the language description of qualitative attributes is collected through questionnaires or platform input to ensure the accuracy and operability of the classification.

[0050] For attributes classified as soft attributes, this embodiment further divides them into interval-type, benefit-type, and cost-type based on different requirements for consumer satisfaction to optimize matching solutions. Interval-type soft attributes require that the attribute value is close to a certain range. Deviating from this range will reduce satisfaction. For example, ripeness may be defined as an interval-type attribute. Consumers hope that the ripeness of bananas is in the "moderate" range (for example, a ripeness score between 4 and 6). Too low or too high is not ideal. Benefit-type soft attributes require that the attribute value is as high as possible, reflecting consumers' preference for high quality. For example, product grade is a typical benefit-type attribute. The higher the grade (such as grade 5 is better than grade 3), the higher the consumer satisfaction. Cost-type soft attributes require that the attribute value is as low as possible, which is usually related to cost or burden. For example, price and logistics distance are cost-type attributes. Lower prices or shorter logistics distances can improve consumer satisfaction. This classification is based on consumers' perception of attribute value, and the specific type is determined by the transaction scenario and product characteristics.

[0051] In step S102 , for each attribute, a system range and a design range are defined, wherein the system range is based on the attribute value range provided by the supplier, and the design range is based on the attribute value range required by the consumer.

[0052] To support subsequent information volume calculations, this embodiment defines a system range and a design range for each attribute. The system range refers to the range of attribute values provided by suppliers or the range accepted by consumers. For example, the price range of bananas offered by suppliers is between 3.5 and 6 yuan per kilogram. The design range refers to the range of attribute values requested by consumers or the range expected by suppliers. For example, consumers prefer a price between 3 and 5 yuan per kilogram. The public range is the overlap between the system range and the design range. For example, the public price range is between 3.5 and 5 yuan.

[0053] In implementation, data on system scope and design scope is collected through the trading platform. For example, consumers enter their desired price range through the interface, and suppliers upload product parameters. This information is calculated to provide a quantitative basis for subsequent matching optimization.

[0054] To verify the identification and classification of the above attributes, this embodiment takes banana trading as an example. The specific implementation steps are as follows:

[0055] Attribute recognition: Identify common attributes (price, logistics distance, product grade, brand) and specific attributes (freshness, maturity, origin, product certification). For example, consumers require a price between 3 and 5 yuan, a logistics distance less than 50 kilometers, a freshness rating of "very fresh," and an origin of "local."

[0056] Classification of hard and soft attributes: Product certification (must be organic) and packaging type (must be plastic packaging) are set as hard attributes; price, logistics distance, freshness and maturity are set as soft attributes.

[0057] Quantitative and qualitative attributes: Price and logistics distance are quantitative attributes, while freshness and brand are qualitative attributes. The range of quantitative attributes is input by numerical value, while qualitative attributes are quantified by options (for example, "very fresh" corresponds to a membership of 0.8).

[0058] Soft attribute type classification: price and logistics distance are cost-type (the lower the better), product grade is benefit-type (the higher the better), and maturity is interval-type (scores 4 to 6 are best).

[0059] Scope definition: Define the system scope and design scope for each attribute. For example, the price range provided by the supplier is 3.5 yuan to 6 yuan, the price range required by the consumer is 3 yuan to 5 yuan, and the public range is 3.5 yuan to 5 yuan.

[0060] Information content calculation: For quantitative attributes (such as price), if the system range is completely contained in the design range, the information content is zero; for qualitative attributes (such as freshness), the information content is calculated by quantifying "very fresh" to 0.8 through the triangular membership function.

[0061] Through the above steps, this embodiment completes the identification and classification of attributes in banana transactions, providing a basis for subsequent matching optimization.

[0062] In step S103, for each matching scheme between the consumer and the supplier, check whether all hard attributes are met. If not, filter out the matching scheme; for soft attributes, calculate the amount of information from the consumer perspective and the supplier perspective respectively: for quantitative soft attributes, calculate the amount of information according to their type, and ensure that the amount of information is non-negative by comparing the degree of overlap between the system scope and the design scope; for qualitative soft attributes, apply fuzzy mathematics methods, quantify the language description into a numerical value through a triangular or trapezoidal membership function, and then calculate the amount of information; calculate the total amount of information from the consumer perspective and the total amount of information from the supplier perspective respectively, and the total amount of information is the sum of the information amounts of all soft attributes in the corresponding matching scheme.

[0063] During the supply and demand matching process for agricultural products, step S103 first performs a hard attribute check on each potential match between the consumer and supplier to ensure that they meet the mandatory transaction requirements. Hard attributes are attributes that must be fully met. If any hard attribute is not met, the match is filtered out and cannot proceed to the subsequent information calculation stage. The hard attribute check is based on whether the attribute values provided by the supplier meet the consumer's design range. For example, in a banana transaction, the consumer may specify the packaging type as a hard attribute, requiring plastic packaging. If the supplier provides paper packaging, the match is eliminated as it does not meet the hard attribute requirement. In practice, the hard attribute check is automatically performed by the transaction platform: the consumer enters the hard attribute requirements (e.g., "Packaging type = plastic") on the platform, and the supplier uploads the product parameters (e.g., "Packaging type = paper"). The system compares the two and marks the match invalid if any inconsistency is found. This embodiment significantly reduces unnecessary calculations and improves matching efficiency through hard attribute filtering.

[0064] For matching solutions that pass the hard attribute check, this embodiment calculates the information content of quantitative soft attributes to quantify the degree of match between consumers and suppliers. The information content of quantitative soft attributes (such as price and logistics distance) is calculated based on their type (interval, benefit, or cost) by comparing the degree of overlap between the system scope and the design scope, ensuring that the information content is non-negative. The information content is calculated based on the degree of overlap between the scopes, and the specific formula is as follows:

[0065] For interval-type quantitative attributes, the calculation formula for information quantity I is:

[0066]

[0067] Among them, f 1 、f 2 is the system-wide boundary, d 1 d 2 is the boundary of the design range, the formula expresses:

[0068] If the system range is completely outside the design range (no overlap), the amount of information is infinite, indicating that matching is not feasible;

[0069] If the system range partially overlaps the design range, the information content is calculated based on the difference in range widths using a logarithmic function;

[0070] If the system scope is completely contained within the design scope, the amount of information is zero, indicating that the requirements are fully met.

[0071] For example, in a banana transaction, a consumer demands a price (cost-based attribute) between 3 and 5 yuan (design range), while the supplier offers a price between 3.5 and 6 yuan (system range). Given a common range of 3.5 to 5 yuan, the information content calculated by the formula is zero, indicating that the price fully meets the consumer's needs. In practice, data on the system range and design range are collected through the trading platform, and the information content calculation is automatically performed by the algorithm.

[0072] For qualitative soft attributes (such as freshness and brand), this embodiment applies fuzzy mathematics methods to quantify verbal descriptions into numerical values using triangular or trapezoidal membership functions, and then calculates the amount of information. Verbal descriptions of qualitative attributes (such as "very fresh" and "generally fresh") are fuzzy and cannot be directly compared, so they need to be converted into numerical values through membership functions. The membership function defines the mapping relationship between verbal descriptions and numerical values. For example, "very fresh" corresponds to a membership degree of 0.8, and "generally fresh" corresponds to 0.5. In implementation, consumers and suppliers select descriptions of qualitative attributes through the trading platform, and the system quantizes them into values between 0 and 1 according to a predefined membership function (such as a triangular membership function). The amount of information is then calculated based on the deviation between the quantized values and the designed range. For example, if a consumer requires "very fresh" (membership degree 0.8) for freshness and the supplier provides "generally fresh" (membership degree 0.5), the amount of information is calculated using the deviation (0.8-0.5=0.3), which reflects the degree of imperfection of the match. This embodiment uses fuzzy mathematics methods to address the uncertainty of qualitative attributes and improve the accuracy of matching.

[0073] This embodiment calculates the total information content of each matching solution from both the consumer and supplier perspectives to support bilateral matching optimization. The total information content from the consumer's perspective is the sum of all soft attribute information about the supplier in the matching solution, while the total information content from the supplier's perspective is the sum of all soft attribute information about the consumer. The information content from the consumer's perspective represents the consumer-supplier matching bias, while the information content from the supplier's perspective represents the supplier-consumer matching bias. For example, in a banana transaction, the soft attributes from the consumer's perspective include price (information content 0), logistics distance (information content 0.2), and freshness (information content 0.3), with a total information content of 0 + 0.2 + 0.3 = 0.5. The soft attributes from the supplier's perspective may include the consumer's acceptable price range and logistics requirements, and the total information content is calculated similarly. In implementation, the trading platform uses an algorithm to calculate the information content of each matching solution's soft attributes one by one and sum them to obtain the total information content from both perspectives. This total information content provides a quantitative basis for subsequent multi-objective optimization, ensuring that the matching solution simultaneously considers the needs of both parties.

[0074] To verify the above-mentioned hard attribute filtering and soft attribute information calculation technology, this embodiment takes banana trading as an example. The specific implementation steps are as follows:

[0075] Hard attribute check: The consumer requires plastic packaging (hard attribute). Supplier A provides plastic packaging, while Supplier B provides paper packaging. Supplier B's matching solution is eliminated because it does not meet the hard attribute, and Supplier A enters the subsequent calculation.

[0076] Quantitative soft attribute calculation: For supplier A, the consumer requires a price between 3 and 5 yuan (design range), the supplier provides 3.5 to 5 yuan (system range), and the information content is zero (complete overlap). The logistics distance requirement is less than 50 kilometers (design range), the supplier provides 40 to 60 kilometers (system range), and the public range is 40 to 50 kilometers. The information content is calculated as 0.2 using the formula;

[0077] Qualitative soft attribute calculation: The consumer requires freshness to be "very fresh" (membership degree 0.8), and supplier A provides "normally fresh" (membership degree 0.5). The information content is calculated as 0.3 based on the deviation (0.8-0.5);

[0078] Total information from both perspectives: From the consumer perspective, the total information is price (0) + logistics distance (0.2) + freshness (0.3) = 0.5. From the supplier perspective, assuming that consumers accept a price range of 3 to 5.5 yuan (information 0.1), the total information is calculated similarly.

[0079] Data collection and calculation: The trading platform collects attribute data of consumers and suppliers, automatically performs hard attribute checks and information volume calculations, and outputs the total information volume of each matching solution from two perspectives.

[0080] Through the above steps, this embodiment completes the hard attribute filtering and soft attribute information calculation in the banana transaction, providing data support for subsequent optimization.

[0081] In step S104, a multi-objective optimization model is established, the goal of which is to minimize the sum of the total amount of information from the consumer's perspective and the sum of the total amount of information from the supplier's perspective in all matches. The constraints are that each supplier in the matching scheme is matched with at most one consumer, and each consumer is matched with at most one supplier. For each matched consumer-supplier pair, the amount of information from the consumer's perspective does not exceed the preset upper limit of the consumer, and the amount of information from the supplier's perspective does not exceed the preset upper limit of the supplier.

[0082] In the process of matching the supply and demand of agricultural products, this step optimizes the matching scheme between consumers and suppliers by establishing a multi-objective optimization model. This embodiment first defines the objective function of the model, which aims to minimize the sum of the total information from the consumer's perspective and the sum of the total information from the supplier's perspective in all matching. The sum of the total information from the consumer's perspective reflects the overall matching deviation between consumers and suppliers, while the sum of the total information from the supplier's perspective reflects the overall matching deviation between suppliers and consumers. The mathematical expression of the objective function is as follows:

[0083] Z1=∑ i ∑ j I ij ·x ij ;

[0084] Z2=∑ i ∑ j I ji ·x ij ;

[0085] Among them, I ij is the sum of the soft attribute information of consumer i about supplier j, I ji is the sum of the soft attribute information of supplier j to consumer i, x ij is a 0-1 decision variable indicating whether consumer i is a match with supplier j (1 for match, 0 for no match). The goal is to minimize Z1 (the sum of the total information from the consumer’s perspective) and Z2 (the sum of the total information from the supplier’s perspective). In practice, the information I ij and I ji As calculated in step S103, the trading platform automatically aggregates the information of all matching solutions through an algorithm, constructs an objective function, and provides a quantitative target for subsequent optimization.

[0086] To ensure the feasibility of the matching solution, this embodiment defines constraints of the multi-objective optimization model, including matching restrictions and an upper limit on the amount of information.

[0087] The single matching constraint is: each supplier is matched with at most one consumer, and each consumer is matched with at most one supplier. The mathematical expression is:

[0088] ∑ i x ij ≤1 for each supplier j;

[0089] ∑ j x ij ≤1 for each consumer i;

[0090] These constraints ensure that the matching scheme is one-to-one and avoids duplicate assignments.

[0091] Information limit constraint: For each matching consumer-supplier pair, the amount of information from the consumer's perspective does not exceed the consumer's preset upper limit, and the amount of information from the supplier's perspective does not exceed the supplier's preset upper limit. Mathematically expressed as:

[0092] I ij ·x ij ≤t i For each matching (i,j);

[0093] I ji ·x ij ≤t j For each matching (i,j);

[0094] In practice, both parties enter their own upper limits through the platform. For example, a consumer sets a price limit of no more than 0.5, while a supplier sets a distance limit of no more than 0.3. The platform uses an algorithm to check whether each matching solution meets these constraints, eliminating any infeasible solutions.

[0095] Since multi-objective optimization involves minimizing the total amount of information of consumers and suppliers at the same time, this embodiment uses a linear weighted sum method to convert the multi-objective problem into a single-objective problem to simplify the solution. The model selects the optimal matching solution through normalized membership function and weighted comprehensive score. Although step S104 focuses on model establishment, the solution logic is closely related to the model, so it is explained here. The normalized membership function is used to evaluate the relative pros and cons of the total amount of information of each matching solution relative to the maximum and minimum values, and the comprehensive score is calculated by weighted summation. Although the specific formula of the membership function is defined in the subsequent step (S105), the model solution depends on these scores. In implementation, the trading platform uses an optimization algorithm (such as linear programming or heuristic algorithm) to solve the model, and the input includes:

[0096] Objective function: the sum of the total information of consumers and suppliers;

[0097] Constraints: single matching and information limit;

[0098] Parameters: information volume data (obtained from step S103), preset upper limit and weight.

[0099] For example, the platform may set the weight of the consumer perspective to 0.6 and the weight of the supplier perspective to 0.4, and solve the matching solution that maximizes the weighted comprehensive score. In this embodiment, the model solution can be implemented using standard optimization tools to ensure efficiency and accuracy.

[0100] To verify the technology for establishing a multi-objective optimization model, this embodiment takes banana trading as an example. The specific implementation steps are as follows:

[0101] Data preparation: Assume that there are 6 consumers and 11 suppliers. Step S103 has calculated the consumer perspective information volume I of each matching solution. ij and supplier perspective information I ii For example, Consumer 1's I to Supplier 2 12 =0.5, I of supplier 2 to consumer 1 21 =0.3.

[0102] Objective function construction: Calculate Z1=∑ i ∑ j I ij ·x ij and Z2=∑ i ∑ j I ji ·x ii For example, if consumer 1 matches supplier 2, the amount of information contributed is 0.5 (from the consumer's perspective) and 0.3 (from the supplier's perspective).

[0103] Constraint setting: Single matching, ensuring that consumer 1 is matched with only one supplier, and supplier 2 is matched with only one consumer; Information volume upper limit, the preset upper limit for consumer 1 is 0.6, and the preset upper limit for supplier 2 is 0.4.

[0104] Model solution: Use linear programming algorithm to solve, input information data, constraints and weights (for example, consumer perspective weight 0.6, supplier perspective weight 0.4), and output the optimal matching solution.

[0105] Platform implementation: The trading platform stores information and constraint data in a database, calls optimization modules (such as Python's SciPy library) to perform calculations, and generates matching solutions.

[0106] The multi-objective optimization model established in this embodiment has the following technical advantages:

[0107] Bilateral Optimization: By simultaneously minimizing the total information volume of both consumers and suppliers, this approach balances the interests of both parties, overcoming the limitations of traditional methods that only consider the needs of one party. Constrained Completeness: Single matching and information volume upper bound constraints ensure the feasibility of matching solutions, adapting to the dynamic nature of agricultural product transactions. Efficient Solution: The linear weighted sum method simplifies multi-objective problems and, combined with standard optimization algorithms, enables fast computation.

[0108] In step S105, for all possible matching schemes, the total amount of information of consumers and suppliers is calculated, and the respective maximum and minimum values are determined; for each matching scheme, its normalized membership function value is calculated by comparing its total amount of information with the maximum and minimum values of all matching schemes; the matching scheme with the highest weighted comprehensive score of the membership function is selected, where the weight is a predefined value.

[0109] In the process of matching supply and demand for agricultural products, step S105 first calculates the total information from the consumer's perspective and the supplier's perspective for all possible matching scenarios to assess the matching quality of each scenario. The total information from the consumer's perspective is the sum of the soft attribute information of the consumer on the supplier across all matches, while the total information from the supplier's perspective is the sum of the soft attribute information of the supplier on the consumer across all matches. In practice, the information amount is a 0-1 decision variable (1 indicates a match between consumer i and supplier j, 0 indicates no match). In practice, the information amount is calculated in step S103, and the trading platform uses an algorithm to traverse all possible matching scenarios (satisfying the constraints of step S104). For example, in a banana transaction, a matching scenario may include matching consumer 1 with supplier 2, and consumer 2 with supplier 3. The platform aggregates the information amount of these matches to obtain the corresponding scenario's value. For example, in a banana transaction, a matching scenario may include matching consumer 1 with supplier 1, and consumer 2 with supplier 3. The platform aggregates the information amount of these matches to obtain the corresponding scenario's value. This embodiment, through systematic calculation, provides a quantitative basis for subsequent normalization.

[0110] To evaluate the relative merits of each matching solution, this embodiment determines the maximum and minimum values for the total information volume from the consumer's perspective and the total information volume from the supplier's perspective for all matching solutions. The maximum and minimum values are obtained by comparing the total information volume of all feasible matching solutions. A feasible matching solution is one that meets the constraints of step S104 (single matching and information volume upper limit). During implementation, the trading platform algorithm traverses all feasible solutions and records the maximum and minimum values for the total information volume. For example, in a banana transaction, assuming there are 100 feasible matching solutions, the platform calculates the total information volume of each solution and obtains a specific value. These specific values are used to calculate the normalized membership function to ensure the fairness and comparability of the evaluation.

[0111] This embodiment calculates a normalized membership function value for each matching solution to quantify its relative merits from the perspectives of consumers and suppliers. The membership function formula is as follows:

[0112]

[0113]

[0114] Among them, μ(Z1) is the membership function value from the consumer's perspective, μ(Z2) is the membership function value from the supplier's perspective, Z1 and Z2 are the total information content of the solution, and maxZ1, minZ1, minZ2, and maxZ2 are the maximum and minimum values determined in step S105. This formula normalizes the total information content to a value between 0 and 1. The closer the value is to 1, the better the match (the lower the information content). In implementation, the trading platform applies the above formula to each feasible matching solution. For example, if Z1=4, Z1=4, maxZ1=10, and minZ1=2 for a solution, then:

[0115]

[0116] μ(Z2) is calculated similarly. This embodiment eliminates the influence of the dimension of information quantity through normalization and provides a standardized evaluation index.

[0117] To determine the optimal matching solution, this embodiment calculates the weighted comprehensive score of the membership function and selects the solution with the highest score:

[0118] Z=α1·μ(Z1)+α2·μ(Z2);

[0119] Where Z is the comprehensive score, μ(Z1) and μ(Z2) are the membership function values from the perspectives of consumers and suppliers, α1 and α2 are predefined weights, and α1+α2=1. The weights reflect the relative importance of consumers and suppliers in matching. For example, α1=0.6 and α2=0.4 indicate a greater emphasis on consumer satisfaction. In implementation, the trading platform calculates μ(Z1) and μ(Z2) for each matching solution, applies the weights to calculate Z, and selects the solution with the largest Z as the optimal match. For example, if μ(Z1)=0.75 and μ(Z2)=0.8 for a solution, and the weights are α1=0.6 and α2=0.4, then:

[0120] Z=0.6·0.75+0.4·0.8=0.45+0.32=0.77;

[0121] The platform compares the Z of all solutions and outputs the solution with the highest score. This embodiment achieves a balanced optimization of consumer and supplier needs through weighted comprehensive scoring. The technology of this embodiment has the following advantages:

[0122] Standardized evaluation: Normalized membership function values eliminate the impact of information differences and provide fair matching evaluation;

[0123] Bilateral balance: A weighted comprehensive score takes into account the needs of both consumers and suppliers to ensure mutual satisfaction;

[0124] Efficient optimization: Algorithmic calculations support rapid screening of optimal solutions and adapt to large-scale transaction scenarios.

[0125] In implementation, the trading platform collects weight settings (such as consumer priority or supplier priority) through the user interface and automatically performs total information calculation, normalization and score selection.

[0126] In one embodiment, for quantitative soft attributes, the information amount is calculated based on a range comparison of attribute types, including interval type, benefit type, or cost type.

[0127] As a supplement, for interval-type quantitative attributes, the calculation of information quantity takes into account the boundary relationship between the system scope and the design scope:

[0128] If the system range is completely outside the design range, the amount of information is infinite; if the system range partially overlaps the design range, the amount of information is calculated based on the difference in range widths using a logarithmic function; if the system range is completely contained within the design range, the amount of information is zero.

[0129] Among them, in the supply and demand matching of agricultural products, the information volume calculation of quantitative soft attributes is used to quantify the matching deviation between consumers and suppliers. This embodiment calculates the information volume by comparing the system range provided by the supplier and the design range required by the consumer to ensure that the result is a non-negative value. Quantitative soft attributes are directly measurable attributes (such as price, logistics distance), and their information volume reflects the degree of deviation between the attribute value and the expectation. The information volume calculation is based on the attribute type, including interval type (the attribute value is best when it is close to a specific range), benefit type (the higher the attribute value, the better) and cost type (the lower the attribute value, the better). Through range comparison, the system evaluates whether the attribute value meets consumer demand and provides a basis for matching optimization.

[0130] In one embodiment, for qualitative soft attributes, the information amount is calculated using a fuzzy membership function and is calculated through a quantitative language description.

[0131] Among them, in the supply and demand matching process of agricultural products, the information volume calculation of qualitative soft attributes is used to quantify the matching deviation between consumers and suppliers, especially for subjective attributes that cannot be directly measured. Qualitative soft attributes (such as freshness, brand) are usually expressed in language descriptions, such as "very fresh" or "well-known brand", which are fuzzy and subjective. This embodiment converts these language descriptions into numerical values through fuzzy membership functions to calculate the amount of information. The fuzzy membership function can effectively handle the uncertainty of qualitative attributes. By quantifying the language description, it generates a membership degree between 0 and 1, reflecting the degree of closeness between the attribute value and the consumer's expectations. The amount of information is calculated based on the quantified numerical deviation, providing a basis for matching optimization. This embodiment solves the non-numerical problem of qualitative attributes through fuzzy mathematics methods and improves the accuracy of matching.

[0132] This embodiment uses a triangular or trapezoidal membership function to quantify the linguistic description of qualitative soft attributes into numerical values. Common linguistic descriptions (such as "very fresh", "generally fresh", and "not too fresh") are mapped to an interval of 0 to 1 through a predefined membership function. For example, the membership function of the freshness attribute may be defined as: "very fresh" corresponds to a membership of 0.8, "generally fresh" corresponds to 0.5, and "not too fresh" corresponds to 0.2. During the quantification process, consumers select the desired linguistic description (such as "very fresh") through the trading platform, and suppliers provide the actual linguistic description (such as "generally fresh"). The platform calculates the difference in membership between the two based on the membership function, for example, 0.8-0.5=0.3, and calculates the amount of information based on this difference to reflect the degree of imperfection of the match. This embodiment ensures the repeatability and consistency of the quantification results through the standardized membership function design.

[0133] This embodiment uses banana trading as an example to illustrate the implementation of qualitative soft attribute information volume calculation: the consumer requires freshness to be "very fresh" (membership degree 0.8), and the supplier provides "generally fresh" (membership degree 0.5). The trading platform uses a triangular membership function to quantify the two into numerical values, and calculates the difference of 0.3 as the basis of information volume, reflecting the matching deviation of freshness. Similarly, the brand attribute may be defined as "well-known brand" (membership degree 0.9) versus "ordinary brand" (membership degree 0.4), and the information volume is calculated based on the difference of 0.5. In implementation, the trading platform collects the consumer's expected description and the supplier's actual description through a drop-down menu or questionnaire, calls a predefined membership function library (such as Python's skfuzzy library) to perform quantification, and automatically outputs the information volume.

[0134] In one embodiment, the total information content of the matching between consumers and suppliers is the sum of the information content of all soft attributes in the matching solution.

[0135] In one embodiment, the membership function of a supplier is calculated by comparing the total information content of the supplier in the corresponding matching solution with the maximum and minimum total information content in all the matching solutions, and calculating a normalized value.

[0136] In one embodiment, the comprehensive score of each matching solution is calculated by performing a weighted summation on the membership function values of the supplier and the consumer, wherein the weights are predefined values and the sum is 1.

[0137] In one embodiment, the system scope and the design scope of each attribute are defined according to requirements provided by suppliers and consumers, respectively.

[0138] Based on the same idea, as shown in Figure A, a product supplier and consumer matching system based on information axiom is provided, including:

[0139] Attribute identification and classification module 201 is used to identify the attributes of agricultural products, including general attributes and specific attributes. General attributes include brand, logistics distance, product grade, and price. Specific attributes include freshness, maturity, product certification, seasonality, origin, and product safety. The general attributes and specific attributes are classified into mandatory hard attributes and flexible soft attributes, as well as quantitative attributes and qualitative attributes. Soft attributes are classified into interval type, benefit type, and cost type based on consumer satisfaction. The interval type requires attribute values to be close to a specific range, the benefit type requires attribute values to be as high as possible, and the cost type requires attribute values to be as low as possible.

[0140] A range definition module 202 is configured to define, for each attribute, a system range and a design range, wherein the system range is based on the attribute value range provided by the supplier, and the design range is based on the attribute value range required by the customer;

[0141] The information quantity calculation module 203 checks, for each matching solution between a consumer and a supplier, whether all hard attributes are satisfied. If not, the matching solution is filtered out. For soft attributes, the information quantity is calculated from both the consumer and supplier perspectives. For quantitative soft attributes, the information quantity is calculated based on their type, and the degree of overlap between the system scope and the design scope is compared to ensure that the information quantity is non-negative. For qualitative soft attributes, fuzzy mathematical methods are applied to quantify the linguistic description into numerical values using triangular or trapezoidal membership functions, and then the information quantity is calculated. The total information quantity from the consumer perspective and the total information quantity from the supplier perspective are calculated separately. The total information quantity is the sum of the information quantities of all soft attributes in the corresponding matching solution.

[0142] A model optimization module 204 is configured to establish a multi-objective optimization model, wherein the objective is to minimize the sum of the total amount of information from the consumer's perspective and the sum of the total amount of information from the supplier's perspective in all matches, subject to the constraints that each supplier in the matching scheme is matched with at most one consumer, each consumer is matched with at most one supplier, and for each matched consumer-supplier pair, the amount of information from the consumer's perspective does not exceed a preset upper limit for the consumer, and the amount of information from the supplier's perspective does not exceed a preset upper limit for the supplier;

[0143] The scheme selection module 205 is used to calculate the total amount of information of consumers and suppliers for all possible matching schemes and determine their respective maximum and minimum values; for each matching scheme, calculate its normalized membership function value by comparing its total amount of information with the maximum and minimum values of all matching schemes; and select the matching scheme with the highest weighted comprehensive score of the membership function, where the weight is a predefined value.

[0144] From the above system, it can be seen that this embodiment has the following beneficial effects:

[0145] 1. Improve supply-demand matching efficiency: This invention comprehensively considers the general attributes (such as price, brand, and logistics distance) and specific attributes (such as freshness, maturity, and origin) of agricultural products, and categorizes them into hard and soft attributes, thereby accurately identifying and matching the needs of consumers and suppliers. Compared to existing agricultural product information platforms that only provide information sharing, this invention quantifies the amount of attribute information and applies a multi-objective optimization model to quickly screen matching solutions that meet hard attribute requirements and are optimal in terms of soft attributes, thereby significantly improving the efficiency of supply-demand matching and reducing information asymmetry during transactions.

[0146] 2. Enhance the adaptability of matching schemes: This invention employs information axioms and fuzzy mathematics to process quantitative and qualitative attributes separately, overcoming the limitations of existing matching algorithms (such as Gale-Shapley and SMAA) in handling agricultural product transactions, which are characterized by insufficient dynamism and fuzziness. For quantitative attributes, the amount of information is calculated by comparing the degree of overlap between the system scope and the design scope; for qualitative attributes, triangular or trapezoidal membership functions are used to quantify the language description. This comprehensive approach enables matching schemes to adapt to the unique attributes of agricultural products (such as seasonality and certification) and the diverse trading requirements (such as one-time transactions and bilateral matching), thereby improving the flexibility and applicability of matching schemes.

[0147] 3. Optimizing Mutual Satisfaction of Consumers and Suppliers: This method calculates information from both the consumer and supplier perspectives and minimizes the total information shared by both parties through a multi-objective optimization model. This ensures that the matching solution simultaneously meets both consumer demands for product quality and price and supplier expectations for market opportunities and revenue. A case study demonstrated that this matching method achieved a high overall matching degree (0.863) in a banana transaction. Compared to traditional methods that only consider unilateral needs, this method balances the interests of both parties, improving both transaction satisfaction and transaction success rates.

[0148] 4. Promote the economic development of smallholder farmers: This invention uses intelligent matching methods to provide smallholder farmers with direct-to-consumer sales channels, reducing reliance on intermediaries, lowering transaction costs, and increasing sales profits. In existing technologies, smallholder farmers often face profit squeezes due to a lack of effective marketing channels and bargaining power. Through precise matching and a transparent transaction process, this invention enhances the market competitiveness of smallholder farmers, provides them with a sustainable source of income, helps narrow the urban-rural economic gap, and promotes the sustainable development of the rural economy.

[0149] 5. Support for Dynamic Trading Scenarios: By defining the system scope (provided by suppliers) and the design scope (consumer requirements), and optimizing the matching scheme within constraints (each entity is matched at most once, and the amount of information does not exceed an upper limit), this system can adapt to the dynamic changes in agricultural product trading, such as seasonal fluctuations, price changes, and logistics requirements. Compared to the static or single matching methods used in existing technologies, this system can respond to market changes in real time, providing more stable and reliable trading support.

[0150] 6. Improved transaction transparency and fairness: This invention selects the optimal matching solution by normalizing membership function values and applying a weighted comprehensive score. This ensures that the matching process is based on objective information evaluation and avoids human intervention or subjective bias. Compared to the price opacity and information asymmetry issues in traditional agricultural product transactions, this invention provides a standardized matching process, enhances transaction transparency and fairness, and creates a fair trading environment for consumers and suppliers.

[0151] The specific details of each module in the above system have been described in detail in the implementation method part. For details not disclosed, please refer to the implementation method part, and they will not be repeated here.

[0152] Through the description of the above embodiments, it is easy for those skilled in the art to understand that the example embodiments described herein can be implemented by software or by combining software with necessary hardware. Therefore, the technical solution according to the embodiment of the present invention can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (which can be a CD-ROM, a USB flash drive, a mobile hard disk, etc.) or on a network, and includes a number of instructions to enable a computing device (which can be a personal computer, a server, a terminal device, or a network device, etc.) to execute the method according to the exemplary embodiment of the present invention.

[0153] Furthermore, the figures above are merely illustrative of the processes included in the method according to exemplary embodiments of the present invention and are not intended to be limiting. It is readily understood that the processes illustrated in the figures above do not indicate or limit the temporal order of these processes. Furthermore, it is readily understood that these processes may be executed synchronously or asynchronously, for example, in multiple modules.

[0154] It should be noted that, although several modules or units of the device for action execution are mentioned in the above detailed description, this division is not mandatory. In fact, according to an exemplary embodiment of the present invention, the features and functions of two or more modules or units described above can be concretized in one module or unit. Conversely, the features and functions of one module or unit described above can be further divided into multiple modules or units to be concretized.

[0155] Those skilled in the art will readily identify other embodiments of the present invention after considering the specification and practicing the invention disclosed herein. This application is intended to cover any variations, uses, or adaptations of the present invention that follow the general principles of the invention and include common knowledge or customary techniques in the art not disclosed herein. The description and embodiments are to be considered as exemplary only, with the true scope and spirit of the invention being indicated by the claims.

[0156] It should be understood that the present invention is not limited to the exact construction described above and shown in the drawings, and that various modifications and changes may be made without departing from the scope thereof. The scope of the present invention is limited only by the appended claims.

Claims

1. A method for matching product suppliers and consumers based on information axiom, characterized in that: The method comprises: Identify the attributes of agricultural products, including general attributes and specific attributes. General attributes include brand, logistics distance, product grade, and price. Specific attributes include freshness, maturity, product certification, seasonality, origin, and product safety. Categorize the general attributes and specific attributes into mandatory hard attributes and flexible soft attributes, as well as quantitative attributes and qualitative attributes. Soft attributes are categorized into interval-based, benefit-based, and cost-based types based on consumer satisfaction. The interval-based type requires attribute values to be close to a specific range, the benefit-based type requires attribute values to be as high as possible, and the cost-based type requires attribute values to be as low as possible. For each attribute, define the system range and design range, where the system range is based on the attribute value range provided by the supplier and the design range is based on the attribute value range required by the consumer; For each matching scheme between a consumer and a supplier, check whether all hard attributes are met. If not, filter out the matching scheme. For soft attributes, calculate the information content from both the consumer and supplier perspectives. For quantitative soft attributes, calculate the information content based on their type, and ensure that the information content is non-negative by comparing the degree of overlap between the system scope and the design scope. For qualitative soft attributes, apply fuzzy mathematics methods to quantify the linguistic description into numerical values using triangular or trapezoidal membership functions, and then calculate the information content. Calculate the total information content from the consumer perspective and the total information content from the supplier perspective, respectively. The total information content is the sum of the information content of all soft attributes in the corresponding matching scheme. Establish a multi-objective optimization model, the goal of which is to minimize the sum of the total amount of information from the consumer's perspective and the sum of the total amount of information from the supplier's perspective in all matches, with the constraints that each supplier in the matching scheme is matched with at most one consumer, each consumer is matched with at most one supplier, and for each matched consumer-supplier pair, the amount of information from the consumer's perspective does not exceed a preset upper limit for the consumer, and the amount of information from the supplier's perspective does not exceed a preset upper limit for the supplier; For all possible matching schemes, the total amount of information of consumers and suppliers is calculated, and the respective maximum and minimum values are determined; for each matching scheme, its normalized membership function value is calculated by comparing its total amount of information with the maximum and minimum values of all matching schemes; the matching scheme with the highest weighted comprehensive score of the membership function is selected, where the weight is a predefined value.

2. The method for matching product suppliers and consumers based on information axioms according to claim 1, characterized in that: For quantitative soft attributes, the information amount is calculated based on a range comparison of attribute types, including interval type, benefit type, or cost type.

3. The method for matching product suppliers and consumers based on information axioms according to claim 2, characterized in that: For interval-type quantitative attributes, the calculation of information quantity takes into account the boundary relationship between the system scope and the design scope: If the system range is completely outside the design range, the amount of information is infinite; If the system range partially overlaps the design range, the information content is calculated based on the difference in range widths using a logarithmic function; If the system scope is completely contained within the design scope, the amount of information is zero.

4. The method for matching product suppliers and consumers based on information axioms according to claim 1, characterized in that: For qualitative soft attributes, the amount of information is calculated using fuzzy membership functions and quantitative language descriptions.

5. The method for matching product suppliers and consumers based on information axioms according to claim 1, characterized in that: The total information content of the matching between consumers and suppliers is the sum of the information content of all soft attributes in the matching scheme.

6. The method for matching product suppliers and consumers based on information axioms according to claim 1, characterized in that: The membership function of a supplier is calculated by comparing the total information amount of the supplier in the corresponding matching solution with the maximum and minimum total information amounts in all the matching solutions, and calculating a normalized value.

7. The method for matching product suppliers and consumers based on information axioms according to claim 1, characterized in that: The comprehensive score of each matching solution is calculated by weighted summing the membership function values of suppliers and consumers, wherein the weights are predefined values and the sum is 1.

8. The method for matching product suppliers and consumers based on information axioms according to claim 1, characterized in that: The system scope and the design scope of each attribute are defined according to requirements provided by suppliers and consumers, respectively.

9. A product supplier and consumer matching system based on information axiom, characterized by: include: An attribute identification and classification module is used to identify the attributes of agricultural products, including general attributes and specific attributes. General attributes include brand, logistics distance, product grade, and price, and specific attributes include freshness, maturity, product certification, seasonality, origin, and product safety. The module classifies the general attributes and specific attributes into mandatory hard attributes and flexible soft attributes, as well as quantitative attributes and qualitative attributes. Soft attributes are classified into interval-based, benefit-based, and cost-based types based on consumer satisfaction. The interval-based type requires attribute values to be close to a specific range, the benefit-based type requires attribute values to be as high as possible, and the cost-based type requires attribute values to be as low as possible. A range definition module, configured to define, for each attribute, a system range and a design range, wherein the system range is based on an attribute value range provided by a supplier, and the design range is based on an attribute value range required by a consumer; An information quantity calculation module checks, for each matching solution between a consumer and a supplier, whether all hard attributes are satisfied. If not, the matching solution is filtered out. For soft attributes, the information volume is calculated from both the consumer and supplier perspectives. For quantitative soft attributes, the information volume is calculated based on their type, and the degree of overlap between the system scope and the design scope is compared to ensure that the information volume is non-negative. For qualitative soft attributes, fuzzy mathematical methods are applied to quantify the linguistic description into numerical values using triangular or trapezoidal membership functions, and then the information volume is calculated. The total information volume from the consumer perspective and the total information volume from the supplier perspective are calculated separately. The total information volume is the sum of the information volumes of all soft attributes in the corresponding matching solution. a model optimization module for establishing a multi-objective optimization model, wherein the objective is to minimize the sum of the total amount of information from the consumer's perspective and the sum of the total amount of information from the supplier's perspective in all matches, subject to the constraints that each supplier in the matching scheme is matched with at most one consumer, each consumer is matched with at most one supplier, and for each matched consumer-supplier pair, the amount of information from the consumer's perspective does not exceed a preset upper limit for the consumer, and the amount of information from the supplier's perspective does not exceed a preset upper limit for the supplier; A solution selection module, configured to calculate the total information amount of consumers and suppliers for all possible matching solutions and determine the maximum and minimum values of each; For each matching solution, its normalized membership function value is calculated by comparing its total information content with the maximum and minimum values of all matching solutions; the matching solution with the highest weighted comprehensive score of the membership function is selected, where the weight is a predefined value.

Images