Cosmetic formula development method, system and equipment and storage medium

Abstract

The invention provides a cosmetic formula development method, system and device and a storage medium, and the method comprises the steps: obtaining a target demand, and screening out key components meeting the target demand based on a pre-constructed cosmetic knowledge graph; constructing a mathematical model for describing the relationship between the component proportion of the key component and the target performance index by using a response surface method, and carrying out synergistic / antagonistic effect analysis based on the mathematical model to determine a combination formula with a synergistic effect; the combination formula comprises the combination of key components and the component concentration; and performing a uniform test on the combined formula to obtain uniform test data, taking the uniform test data as input of an efficacy prediction model based on a support vector regression machine, and outputting an optimal target formula based on the efficacy prediction model. According to the method, the design and development efficiency of new cosmetic products can be greatly improved, and particularly, the effect and quality of cosmetics are strictly and accurately controlled.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent cosmetics development, and particularly to a method, system, device and storage medium for developing a cosmetics formula. Background Art

[0002] The main problems faced in current cosmetics development include that the innovation cycle is prolonged due to high dependence on historical formula data, and the R & D speed and innovation ability of new products are limited by excessive dependence on experience; the trial - and - error cost is high and the development direction is not clear, resulting in low product development efficiency and increased economic burden; the efficacy evaluation is too single, and traditional methods only evaluate the effect of cosmetics through a single index, lacking a multi - dimensional comprehensive evaluation system that can comprehensively reflect the actual performance of products. At the same time, existing computer - aided formula technologies also have limitations. For example, linear models are difficult to handle complex non - linear interactions between multiple components and have a low accuracy rate in stability prediction. These problems jointly restrict the innovative development of the cosmetics industry. Summary of the Invention

[0003] Embodiments of the present invention provide a method, system, device and storage medium for developing a cosmetics formula to solve the problems existing in the related technologies. The technical solutions are as follows:

[0004] In a first aspect, embodiments of the present invention provide a method for developing a cosmetics formula, including:

[0005] Obtain target requirements, and screen out key ingredients that meet the target requirements based on a pre - constructed cosmetics knowledge graph;

[0006] Use the response surface method to construct a mathematical model describing the relationship between the ingredient ratios of key ingredients and target performance indicators, and perform synergistic / antagonistic effect analysis based on the mathematical model to determine a combined formula with a synergistic effect; the combined formula includes a combination of key ingredients and ingredient concentrations;

[0007] Conduct a uniform experiment on the combined formula to obtain uniform experimental data, use the uniform experimental data as the input of an efficacy prediction model based on a support vector regression machine, and output an optimal target formula through the efficacy prediction model.

[0008] In an implementation manner, it further includes:

[0009] Obtain a data source, import the data source into a graph database, where the graph database includes nodes and edges connecting the corresponding nodes;

[0010] Based on predefined entities and the relationships between them, map the primary key fields describing the entities in the data source to the nodes in the graph database, and map the fields describing the relationships between entities to the edges in the graph database to connect the relevant nodes, so as to convert the data source into a graph structure and form a cosmetics knowledge graph.

[0011] In one embodiment, constructing a mathematical model that describes the relationship between the component ratios of key components and the target performance indicators by using the response surface method includes:

[0012] Generating an experimental plan based on the key components and experimental resources, where the experimental plan includes the specific contents of each key component in each group of experiments;

[0013] Obtaining the experimental data obtained by executing the experimental plan, and constructing a second-order polynomial model that describes the relationship between the component ratios of key components and the target performance indicators by using regression analysis based on the experimental data.

[0014] In one embodiment, it further includes:

[0015] Using the second-order polynomial model to predict the target performance indicators corresponding to different component ratios, and visually plotting different component ratios and their corresponding target performance indicators to obtain a response surface plot.

[0016] In one embodiment, performing synergy / antagonism effect analysis based on the mathematical model includes:

[0017] Analyzing the change characteristics of the target performance indicators within the component concentration range based on the response surface plot, and determining that there is a synergy effect among multiple key components when the target performance indicators show a gradually increasing trend along the component concentration range.

[0018] In one embodiment, performing synergy / antagonism effect analysis based on the mathematical model includes:

[0019] Determining the interaction term coefficient according to the second-order polynomial model, determining that there is a synergy effect among multiple key components when the interaction term coefficient is positive, and determining that there is an antagonism effect among multiple key components when the interaction term coefficient is negative.

[0020] In one embodiment, outputting the optimal target formulation through the efficacy prediction model includes:

[0021] Obtaining the selected multi-objective optimization strategy to construct a bi-objective model, where the objective function of the bi-objective model includes an efficacy comprehensive index and the total cost;

[0022] Verifying the prediction results output by the efficacy prediction model through the bi-objective model, and finally determining the optimal target formulation.

[0023] In a second aspect, an embodiment of the present invention provides a cosmetic formulation development system that executes the cosmetic formulation development method as described above.

[0024] In a third aspect, an embodiment of the present invention provides an electronic device, which includes a memory and a processor. The memory and the processor communicate with each other through an internal connection path. The memory is used to store instructions, and the processor is used to execute the instructions stored in the memory. When the processor executes the instructions stored in the memory, the processor is caused to execute the method in any one of the above aspects and any one of the implementation manners.

[0025] In a fourth aspect, an embodiment of the present invention provides a computer-readable storage medium. The computer-readable storage medium stores a computer program. When the computer program runs on a computer, the method in any one of the above aspects and any one of the implementation manners is executed.

[0026] The advantages or beneficial effects in the above technical solutions at least include:

[0027] Based on the preliminary screening of the efficacy components of cosmetics, the present invention constructs a mathematical model describing the relationship between the component ratios of key components and the target performance indicators through the response surface method, analyzes the synergistic / antagonistic effects between components, thereby selects a combination formula with a synergistic effect, and then designs new cosmetics and predicts the effects by using the experimental design method and the efficacy prediction model, which can greatly improve the design and development efficiency of new cosmetics, especially strictly and precisely control the effects and quality of cosmetics.

[0028] The above summary is only for the purpose of the specification and is not intended to be limiting in any way. In addition to the above-described illustrative aspects, implementation manners, and features, further aspects, implementation manners, and features of the present invention will be readily apparent by reference to the drawings and the following detailed description. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] In the drawings, unless otherwise specified, the same reference numerals throughout the several views denote the same or similar components or elements. These drawings are not necessarily drawn to scale. It should be understood that these drawings only depict some implementation manners disclosed according to the present invention and should not be regarded as limiting the scope of the present invention.

[0030] Figure 1 It is a schematic flowchart of the method for developing a cosmetic formula of the present invention;

[0031] Figure 2 It is a structural block diagram of an electronic device according to an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0032] In the following, only some exemplary embodiments are simply described. As those skilled in the art can recognize, the described embodiments can be modified in various different ways without departing from the spirit or scope of the present invention. Therefore, the drawings and the description are considered to be exemplary in nature rather than restrictive.

[0033] This embodiment provides a method for developing a cosmetic formula. By using this solution to pre-design the product before actual development, the effects and quality of cosmetics can be strictly and precisely controlled, thereby greatly improving the design and development efficiency of new cosmetic products.

[0034] This embodiment provides a method for developing a cosmetic formula, which realizes pre-designing the product before actual development and improves the product development efficiency. As Figure 1 shown, the method for developing a cosmetic formula includes the following steps:

[0035] Step S1: Obtain the target requirements, and screen out the key ingredients that meet the target requirements based on the pre-constructed cosmetic knowledge graph.

[0036] It should be noted that the cosmetic knowledge graph is pre-constructed before actual development. Before constructing the cosmetic knowledge graph, it is necessary to clarify the goals and scope of the knowledge graph. Suppose a cosmetic knowledge graph with whitening efficacy needs to be developed. The goal of constructing the knowledge graph is to construct a knowledge graph containing information such as whitening ingredients, mechanism of action, efficacy, safety, and suitable skin types. The corresponding scope is to determine the whitening ingredients to be covered (such as vitamin C, niacinamide, arbutin, etc.) and related attributes (such as chemical structure, mechanism of action, research literature, product applications, etc.).

[0037] After clarifying the goals and scope of the cosmetic knowledge graph, it also includes:

[0038] Step S11: Obtain the data source, and import the data source into the graph database, which contains nodes and edges connecting the corresponding nodes.

[0039] Among them, the data source can be obtained from literature databases, cosmetic ingredient databases, or even third-party platforms. Literature databases such as PubMed and ScienceDirect are used to obtain scientific research data on whitening ingredients. Cosmetic ingredient databases such as INCI and Coslng are used to obtain ingredient function and safety information. The third-party platforms can be Meili Xiuxing, CosDNA, etc., which are used to obtain ingredient actual applications and user feedback.

[0040] The types of data sources obtained include structured data and unstructured data. Structured data includes ingredient names, chemical formulas, efficacy classifications, etc.; unstructured data includes research literature, user comments, etc.

[0041] Subsequently, the data source is processed, such as data cleaning, that is, removing duplicate, incorrect or incomplete data to unify the data format (such as standardizing the names of ingredients); data extraction is performed on unstructured data, that is, extracting key information from the literature (such as the mechanism of action of ingredients, experimental data); then the data is integrated, and the data from different sources is integrated into a unified database for subsequent construction of the knowledge graph.

[0042] Step S12: Based on the predefined entities and their relationships, map the primary key fields describing the entities in the data source to the nodes in the graph database, and map the fields describing the relationships between the entities to the edges in the graph database to connect the relevant nodes, so as to convert the data source into a graph structure and form a cosmetics knowledge graph.

[0043] To construct a cosmetics knowledge graph, it is necessary to pre-define what the core entities are and their relationships. Suppose a knowledge graph of whitening ingredients is to be constructed, then its entities are whitening ingredients (such as vitamin C, niacinamide, etc.), mechanism of action (such as inhibiting melanin production, promoting stratum corneum renewal, etc.), efficacy (such as reducing age spots, brightening skin tone, etc.), safety (such as the irritation level to the skin), skin types suitable for use (such as dry, oily, sensitive, etc.), research literature (relevant scientific research papers or reports), products (specific cosmetics containing certain whitening ingredients), etc. And the relationships include ingredient - mechanism of action, which describes the action mode of a certain whitening ingredient; ingredient - efficacy, which shows the specific effects that a certain ingredient can achieve; ingredient - safety, which provides information about the safety of the ingredient; ingredient - skin type suitable for use, which indicates which skin types the ingredient is suitable for; ingredient / product - research literature, which associates the scientific basis of the ingredient or product; product - ingredient, which lists the specific ingredients contained in the product.

[0044] Based on the defined entities and relationships, create corresponding nodes (representing entities) and edges (representing relationships) in the graph database. For example, create a node for each whitening ingredient and connect it to the corresponding nodes of its mechanism of action, efficacy, safety, etc. through edges.

[0045] Use the batch import tool provided by the graph database or write a script to import the prepared data into the database. Usually, each row or each record in the data source corresponds to an entity (node); for example, the data source of whitening ingredients may contain fields such as ingredient name, mechanism of action, efficacy, etc., and these fields will be parsed and mapped to the node attributes in the graph database. And the fields describing the relationships between entities (such as "mechanism of action" or "skin type suitable for use") will be mapped to edges to connect the relevant nodes; for example, "vitamin C - inhibits melanin production" will create two nodes (vitamin C and inhibits melanin production) and represent their relationship through an edge.

[0046] During the import process, scripts or tools (such as Cypher statements in Neo4j or ETL tools) are used to match the data source fields with the nodes and edges in the graph database, ensuring that each entity and relationship can be correctly transformed into a graph structure, and finally forming a cosmetics knowledge graph.

[0047] For different product effects, different knowledge graphs can be pre-constructed for different product effects respectively through the methods of step S11 and step S12. For example, a whitening knowledge graph is constructed for the whitening effect, and an anti-aging knowledge graph is constructed for the anti-aging effect.

[0048] In this embodiment, after determining the target requirements for the cosmetics to be developed this time, according to the entities (such as whitening ingredients, mechanisms of action, effects, etc.) and their relationships (such as "ingredient - effect", "ingredient - suitable skin type") in the cosmetics knowledge graph, key ingredients that meet the target requirements are screened out. Suppose the target requirement is to develop cosmetics with whitening effect. Combining the whitening and freckle - removing principles of various efficacy substances, multiple relatively safe and effective key ingredients are screened out from the whitening knowledge graph for subsequent optimization of the ingredient ratio.

[0049] Step S2: Use the response surface method to construct a mathematical model describing the relationship between the ingredient ratios of key ingredients and the target performance indicators, and conduct a synergistic / antagonistic effect analysis based on the mathematical model to determine the combined formula with synergistic effects; the combined formula includes the combination of key ingredients and the ingredient concentrations.

[0050] It should be noted that the response surface method (Response Surface Methodology, RSM) is a statistical method used to optimize the ratios of multiple ingredients to obtain the best response.

[0051] In this embodiment, according to the target requirements of the product, the target performance indicators (response variables) to be optimized are clarified, and these indicators can be extracted or supplemented from the knowledge graph; for example, whitening effect (such as chromaticity change), stability (such as the degradation rate of active ingredients during storage), safety (such as the skin irritation score). At the same time, clarify the goals to be achieved, such as maximizing the whitening effect, minimizing irritation, and maintaining a certain stability, etc.

[0052] Based on the information in the cosmetics knowledge graph, determine the ratio range of each key ingredient (such as niacinamide 0.5% - 3%, vitamin C 0.1% - 2%, etc.), and use the response surface method to explore the relationship between the input variables (which key ingredients and their ratios) and the response variables (one or more target performance indicators) through a series of designed experiments. Specifically:

[0053] Step S21: Generate an experimental plan according to the key ingredients and experimental resources.

[0054] Step S22: Obtain the experimental data obtained from implementing the experimental plan, and based on the experimental data, use regression analysis to construct a second-order polynomial model that describes the relationship between the proportion of key components and the target performance indicators.

[0055] In this embodiment, the experimental design method in the response surface method (such as the Box-Behnken experimental design plan) is used to design and arrange the experiments to ensure coverage of all variable spaces of interest. According to the selected experimental design method, a specific experimental plan is generated to clarify the specific content of each key component in each group of experiments. Samples with different component ratios are prepared according to the ratios in the experimental design, and each sample is tested to record the actual values of its response variables (such as whitening effect, stability, etc.). For example, a colorimeter is used to measure the whitening effect, and the stability is evaluated through an accelerated aging test, etc., so as to obtain the experimental data.

[0056] Based on the experimental data, use regression analysis to construct a mathematical model that describes the relationship between the input variables (component ratios) and the response variables (target performance indicators). It is equivalent to fitting a response surface through a series of deterministic "trials" to simulate the true limit state surface, and in this embodiment, a second-order model is used to approximate the response surface:

[0057]

[0058] Among them, y is the response variable (dependent variable), representing the experimental result or output; β0 is the constant term, representing the value of the response variable when all independent variables are zero; β i is the linear term coefficient, representing the linear influence of the independent variable X i on the response variable; β ii is the quadratic form coefficient, representing the non-linear influence of the independent variable X i on the response variable; β ij is the interaction coefficient, representing the influence of the interaction between the independent variables X i and X j on the response variable; X i and X j are independent variables, representing the inputs or control factors in the experiment; k is the number of independent variables, and ε is the random error term. The above X i and X j are the contents of k whitening components. The range of the contents can be determined from literature materials, and the design of the contents can be determined through central composite design. y is the value of the whitening efficacy index, which can be determined according to literature materials such as the Chinese Cosmetic Raw Material Regulations Database.

[0059] Through mathematical model analysis, identify which components and their ratios have the greatest impact on the response variable. For example, through analysis, it can be known that the ratio of niacinamide and vitamin C has a significant synergistic effect on the whitening effect.

[0060] Meanwhile, a mathematical model is used to predict the response variable values under different ratios, and the different ratios and their corresponding response variable values are visually plotted to obtain a response surface plot. Among them, the X-axis and Y-axis of the response surface plot represent two independent input variables respectively, and the Z-axis represents the value of the response variable. In the example of cosmetics development, the X-axis and Y-axis are the concentrations of two different active ingredients, and the height of the Z-axis reflects the trend of the response variable changing with the input variables. The response surface plot shows how the response variable changes as the input variables on the X-axis and Y-axis change, and the shape of the surface can reveal the interaction between the input variables and how they jointly affect the response variable. For example, the rising of the surface may mean that increasing the value of the input variable will improve the effect of the response variable; on the contrary, the falling of the surface means that decreasing the value of the input variable may be more beneficial to the response variable.

[0061] Step S23: Conduct synergistic / antagonistic effect analysis based on the mathematical model.

[0062] In this embodiment, the method for synergistic / antagonistic effect analysis can analyze the regression coefficients (especially the interaction term coefficients) in the model to quantify the synergistic or antagonistic effects between components. That is, the interaction term coefficients are determined according to the second-order polynomial model. When the interaction term coefficient is positive, it is determined that there is a synergistic effect between multiple key components. When the interaction term coefficient is negative, it is determined that there is an antagonistic effect between multiple key components.

[0063] Or, based on the visual response surface plot, analyze the change characteristics of the target performance index within the component concentration range to determine the synergistic / antagonistic effect. If the response variable value of the response surface plot has a gradually rising trend along the component concentration range, it indicates that there is an interactive promotion effect between the corresponding multiple components. If the surface of the response surface plot is flat, it indicates that the effect between the corresponding multiple components is not significant.

[0064] In this embodiment, a mathematical model is constructed by the response surface method to quantitatively evaluate the interaction between components, and then combined with the synergistic / antagonistic effect analysis to find the component ratio combination that makes the target performance index reach the optimal and has a synergistic effect, so as to obtain the corresponding one. For example, through model prediction, the best ratio is 2% niacinamide, 1% vitamin C, and 0.5% arbutin.

[0065] Step S3: Conduct a uniform experiment on the combined formula to obtain uniform experiment data, and use the uniform experiment data as the input of the efficacy prediction model based on the support vector regression machine, and output the optimal target formula through the efficacy prediction model.

[0066] In this embodiment, the combined formula is experimentally verified through uniform experimental design to confirm whether the performance of the combined formula in actual application meets expectations. Compared with other experimental design methods with the same number of factor levels, uniform experimental design requires fewer experimental runs. The core of uniform experiment is to make the experimental points distribute as evenly as possible in the experimental space, so as to capture the overall characteristics of the system with fewer experimental runs, and more rely on interpolation methods or global approximation models (such as Kriging model) to predict the response variable. Select a suitable uniform experimental design table according to the number of components in the combined formula to arrange actual experiments. Each experimental scheme corresponds to a set of specific input variables (such as the concentration combination of efficacy components), test each experimental scheme, and record the corresponding response variable (such as whitening effect, stability, etc.) to obtain uniform experimental data.

[0067] Use regression analysis to fit the uniform experimental data, establish a efficacy prediction model between the input variable and the response variable, and verify the prediction results output by the efficacy prediction model through a two-objective model to finally determine the optimal target formula.

[0068] Since the Support Vector Regression (SVR) has good prediction performance, this embodiment uses the support vector regression machine to predict the product effect. In the support vector regression machine, the selection of the kernel function is crucial. In this embodiment, the kernel function K(x i , x j ) with the best prediction effect is selected from the linear kernel, polynomial kernel, RBF kernel and Simoid kernel according to the prediction error and the coefficient of determination.

[0069] Taking the research and development of whitening cosmetics as an example, this embodiment establishes a whitening efficacy prediction model based on the support vector classification and regression machine, and its expression is:

[0070]

[0071] Among them, α i and α i * are Lagrange multipliers; K(x i , x j ) is the kernel function; b is the bias term. The content of each whitening component in the whitening product is used as the input variable x i , i = 1, 2... k, and each whitening index value representing the whitening efficacy is used as the output variable f(x). Use the constructed efficacy prediction module to predict the response variable (such as whitening effect) under different ratios, and further analyze the interaction between whitening efficacy substances.

[0072] During the construction of the efficacy prediction model, the leave-one-out method can be used to test and evaluate the model. In addition, to make up for the shortage of actual data volume, model pre-training can also be carried out through transfer learning. Transfer learning can learn general features from data in other related fields, thereby improving the generalization ability of the model.

[0073] After constructing the efficacy prediction model in this embodiment, the efficacy prediction model is integrated into the dynamic weight bi-objective model to find the optimal formulation scheme. Specifically:

[0074] Obtain the selected multi-objective optimization strategy to construct a bi-objective model. The objective function of the bi-objective model includes the comprehensive efficacy index and the total cost. The objective function expression is:

[0075] minF(x)=α·(1 / Efficacy)+β·Cost

[0076] Where, Efficacy is the comprehensive efficacy index, and the comprehensive efficacy index is calculated by the weighted average of each objective performance index; Cost is the total cost;

[0077] The constraint conditions are:

[0078] 0.5%≤x i ≤5%(Safety concentration constraint)

[0079] ∑x i =100%(Formulation integrity)

[0080] α+β=1, α=f(t)(Time-varying weight coefficient)

[0081] Use the efficacy prediction model to predict the whitening effect, cost and process difficulty under different ingredient formulations; based on these prediction results, through the dynamic weight bi-objective model, the best balance can be flexibly found between the whitening efficacy and the cost + process difficulty, and finally the optimal target formulation can be determined.

[0082] In this embodiment, the dynamic weight bi-objective model is constructed considering that the emphasis on cost and efficacy is different in different stages of the development cycle. For example, in some cell experiments or the primary stage, it may be desired to screen out those with better efficacy, and when it comes to human experiments or later experiments, the test cost needs to be considered due to the increased cost.

[0083] Based on the preliminary screening of cosmetic efficacy ingredients, this embodiment constructs a mathematical model describing the relationship between the ingredient ratios of key ingredients and the target performance indicators through the response surface method, analyzes the synergistic / antagonistic effects between ingredients, thereby selects a combined formulation with synergistic effects, and then designs new cosmetics using the experimental design method and the efficacy prediction model and predicts the effects, which can greatly improve the design and development efficiency of new cosmetic products, especially strictly and precisely control the effects and quality of cosmetics.

[0084] In another embodiment, a cosmetic formulation development system is provided, which executes the cosmetic formulation development method as described in the first embodiment.

[0085] It should be noted that the system functions and principles implemented by the cosmetic formulation development system can be referred to the corresponding descriptions in the above method, and will not be elaborated here.

[0086] In another embodiment, an electronic device is provided. Figure 2 A structural block diagram of an electronic device according to an embodiment of the present invention is shown. As Figure 2 shown, the electronic device includes: a memory 100 and a processor 200, and a computer program that can run on the processor 200 is stored in the memory 100. When the processor 200 executes the computer program, the cosmetic formulation development method in the above embodiment is implemented. The number of the memory 100 and the processor 200 can be one or more.

[0087] The electronic device further includes:

[0088] A communication interface 300, which is used to communicate with external devices and perform data interaction and transmission.

[0089] If the memory 100, the processor 200, and the communication interface 300 are implemented independently, the memory 100, the processor 200, and the communication interface 300 can be connected to each other through a bus and complete communication with each other. The bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, or an Extended Industry Standard Architecture (EISA) bus, etc. The bus can be divided into an address bus, a data bus, a control bus, etc.

[0090] Optionally, in a specific implementation, if the memory 100, the processor 200, and the communication interface 300 are integrated on a chip, the memory 100, the processor 200, and the communication interface 300 can complete communication with each other through an internal interface.

[0091] An embodiment of the present invention provides a computer-readable storage medium, which stores a computer program, and when the program is executed by a processor, the method provided in the embodiment of the present invention is implemented.

[0092] An embodiment of the present invention also provides a chip, which includes a processor for calling and running instructions stored in a memory, so that a communication device equipped with the chip executes the method provided by the embodiment of the present invention.

[0093] An embodiment of the present invention also provides a chip, including: an input interface, an output interface, a processor, and a memory. The input interface, the output interface, the processor, and the memory are connected through an internal connection path. The processor is configured to execute code in the memory. When the code is executed, the processor is configured to execute the method provided by the embodiment of the invention.

[0094] It should be understood that the above-mentioned processor may be a central processing unit (CPU), or may also be other general-purpose processors, digital signal processors (DSP), application specific integrated circuits (ASIC), field programmable gate arrays (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or any conventional processor, etc. It is worth noting that the processor may be a processor supporting the advanced RISC machines (ARM) architecture.

[0095] Further, optionally, the above-mentioned memory may include a read-only memory and a random access memory, and may further include a non-volatile random access memory. The memory may be a volatile memory or a non-volatile memory, or may include both volatile and non-volatile memories. Among them, the non-volatile memory may include a read-only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable programmable read-only memory (EEPROM), or a flash memory. The volatile memory may include a random access memory (RAM), which is used as an external cache. By way of example but not limitation, many forms of RAM are available. For example, static random access memory (SRAM), dynamic random access memory (DRAM), synchronous dynamic random access memory (SDRAM), double data rate synchronous dynamic random access memory (DDR SDRAM), enhanced synchronous dynamic random access memory (ESDRAM), synchronous link dynamic random access memory (SLDRAM), and direct rambus random access memory (DR RAM).

[0096] In the above embodiments, it can be implemented in whole or in part by software, hardware, firmware, or any combination thereof. When implemented using software, it can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, the processes or functions according to the present invention are generated in whole or in part. The computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices. The computer instructions may be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium.

[0097] In the description of this specification, the description with reference to terms such as "one embodiment", "some embodiments", "examples", "specific examples", or "some examples" means that the specific features, structures, materials, or characteristics described in connection with the embodiment or example are included in at least one embodiment or example of the present invention. Moreover, the specific features, structures, materials, or characteristics described may be combined in a suitable manner in any one or more embodiments or examples. In addition, without contradiction, those skilled in the art can combine and combine the different embodiments or examples described in this specification and the features of different embodiments or examples.

[0098] In addition, the terms "first" and "second" are only used for descriptive purposes and cannot be construed as indicating or implying relative importance or implicitly specifying the quantity of the indicated technical features. Thus, the features defined with "first" and "second" may explicitly or implicitly include at least one of such features. In the description of the present invention, "a plurality" means two or more unless otherwise specifically defined.

[0099] As described above, the above is only the specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present invention can easily think of various changes or substitutions, and these should all be covered by the protection scope of the present invention. Therefore, the protection scope of the present invention shall be subject to the protection scope of the claims.

Claims

1. A method for developing a cosmetic formulation, characterized in that, Including: Obtain a target requirement, and screen out key ingredients that meet the target requirement based on a pre-constructed cosmetics knowledge graph; Use the response surface method to construct a mathematical model describing the relationship between the ingredient proportions of the key ingredients and the target performance indicators, and perform synergistic / antagonistic effect analysis based on the mathematical model to determine a combined formula with a synergistic effect; the combined formula includes a combination of key ingredients and ingredient concentrations; Conduct a uniform experiment on the combined formula to obtain uniform experiment data, use the uniform experiment data as the input of an efficacy prediction model based on a support vector regression machine, and output an optimal target formula based on the efficacy prediction model.

2. The method for developing a cosmetic formulation according to claim 1, wherein It also includes: Obtain a data source, and import the data source into a graph database, where the graph database contains nodes and edges connecting the corresponding nodes; Based on predefined entities and the relationships between them, map the primary key fields describing the entities in the data source to the nodes in the graph database, and map the fields describing the relationships between entities to the edges in the graph database to connect the relevant nodes, so as to convert the data source into a graph structure and form the cosmetics knowledge graph.

3. The method for developing a cosmetic formulation according to claim 1, characterized in that, The construction of the mathematical model using the response surface method to describe the relationship between the ingredient proportions of the key ingredients and the target performance indicators includes: Generate an experimental plan according to the key ingredients and experimental resources, and the experimental plan includes the specific content of each key ingredient in each group of experiments; Obtain the experimental data obtained by executing the experimental plan, and use regression analysis based on the experimental data to construct a second-order polynomial model describing the relationship between the ingredient proportions of the key ingredients and the target performance indicators.

4. The method for developing a cosmetic formulation according to claim 3, wherein, It also includes: Use the second-order polynomial model to predict the target performance indicators corresponding to different ingredient proportions, and perform visual plotting on different ingredient proportions and their corresponding target performance indicators to obtain a response surface plot.

5. The method for developing a cosmetic formulation according to claim 4, wherein The synergistic / antagonistic effect analysis based on the mathematical model includes: Analyze the change characteristics of the target performance indicators within the ingredient concentration range based on the response surface plot. When the target performance indicators show a gradually rising trend along the ingredient concentration range, it is determined that there is a synergistic effect among the multiple key ingredients.

6. The method for developing a cosmetic formulation according to claim 3, characterized in that, The synergistic / antagonistic effect analysis based on the mathematical model includes: Determine the interaction term coefficients according to the second-order polynomial model. When the interaction term coefficients are positive, it is determined that there is a synergistic effect among the multiple key ingredients. When the interaction term coefficients are negative, it is determined that there is an antagonistic effect among the multiple key ingredients.

7. The method for developing a cosmetic formulation according to claim 1, wherein The output of the optimal target formula based on the efficacy prediction model includes: Obtain a selected multi-objective optimization strategy to construct a bi-objective model, and the objective function of the bi-objective model includes an efficacy comprehensive index and a total cost; Verify the prediction results output by the efficacy prediction model through the bi-objective model, and finally determine the optimal target formula.

8. A cosmetic formulation development system, characterized in that, Execute the cosmetics formula development method according to any one of claims 1 to 7.

9. An electronic device, characterized in that, Including: A processor and a memory, where instructions are stored in the memory, and the instructions are loaded and executed by the processor to implement the cosmetics formula development method according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, it implements the cosmetic formulation development method according to any one of claims 1 to 7.

Images