Probiotic efficacy evaluation method based on knowledge graph

Abstract

The invention relates to the field of information retrieval and data processing, in particular to a probiotic efficacy evaluation method based on a knowledge graph. The method comprises the following steps: collecting probiotic data and effect evaluation data, and carrying out unified standardization processing to obtain standardized probiotic data and effect evaluation data; constructing a knowledge graph containing probiotic nodes and effect nodes, calculating a relation weight between the nodes, and searching an optimal path and a suboptimal path; calculating the evaluation scores of the probiotic effects of the optimal path and the suboptimal path based on the relation weight between the nodes; and calculating a final recommendation score of the probiotics based on the evaluation score of the effect of the probiotics. The problems that a traditional probiotic efficacy evaluation method neglects a complex nonlinear relation between probiotics and effects, lacks multi-dimensional comprehensive consideration, is not comprehensive in evaluation result, and is difficult to realize personalized recommendation are solved; and the problems of resource waste and influence on efficiency caused by redundant calculation in the calculation complexity are solved.

Description

Technical Field

[0001] The present invention relates to the fields of information retrieval and data processing, and in particular to a method for evaluating the efficacy of probiotics based on a knowledge graph. Background Art

[0002] As a beneficial microbial population, probiotics are widely used in the fields of health promotion and disease prevention. In recent years, with the in-depth study of probiotics, more and more studies have shown that probiotics have positive effects in improving intestinal health, enhancing the immune system, relieving allergies, improving indigestion, etc. However, despite the broad application prospects of probiotics, how to scientifically and accurately evaluate the effects of probiotics has always been a difficult problem in research.

[0003] Currently, the evaluation of probiotic efficacy mainly relies on traditional experimental designs and data analysis methods. These methods usually set up experimental groups and control groups to monitor the impact of probiotics on specific health indicators over a certain period of time. However, traditional probiotic efficacy evaluation methods often face problems such as a large amount of complex data, a cumbersome data processing process, and a lack of personalized recommendations. Moreover, they mostly use static models and fail to fully consider the complex dynamic relationship between probiotics and effects, which results in limited timeliness and accuracy of evaluation results.

[0004] With the continuous development of big data and artificial intelligence technologies, the knowledge graph, as a powerful data structure and analysis tool, has gradually been applied in multiple fields. By constructing a knowledge graph, scattered data can be effectively integrated, the relationships between different elements can be revealed, and thus a scientific basis can be provided for decision-making. In the field of probiotic efficacy evaluation, constructing a knowledge graph can not only comprehensively consider multi-dimensional experimental data but also achieve dynamic adjustment of relationships and personalized recommendations, thereby greatly improving the accuracy and practicality of evaluation.

[0005] In summary, the traditional probiotic efficacy evaluation methods have the following technical problems: they ignore the complex non-linear relationship between probiotics and effects, mostly rely on single experimental results or single-dimensional analysis, lack multi-dimensional comprehensive consideration, resulting in incomplete or biased evaluation results and difficulty in achieving personalized recommendations; usually there is redundant calculation in computational complexity, leading to waste of resources and affecting efficiency. Summary of the Invention

[0006] The present invention provides a method for evaluating the efficacy of probiotics based on a knowledge graph to solve the problems that traditional probiotic efficacy evaluation methods ignore the complex non-linear relationship between probiotics and effects, mostly rely on single experimental results or single-dimensional analysis, lack multi-dimensional comprehensive consideration, resulting in incomplete or biased evaluation results and difficulty in achieving personalized recommendations; usually there is redundant calculation in computational complexity, leading to waste of resources and affecting efficiency.

[0007] A method for evaluating the efficacy of probiotics based on a knowledge graph specifically includes the following technical solutions: A method for evaluating the efficacy of probiotics based on a knowledge graph includes the following steps: S1. Collect probiotic data and efficacy evaluation data, and perform unified standardization processing to obtain standardized probiotic data and standardized efficacy evaluation data; based on the standardized probiotic data and standardized efficacy evaluation data, construct a knowledge graph including probiotic nodes and efficacy nodes, calculate the relationship weights between the nodes, and search for the optimal path and sub-optimal path; S2. Based on the relationship weights between the nodes, calculate the evaluation scores of the probiotic efficacy of the optimal path and sub-optimal path; based on the evaluation scores of the probiotic efficacy, calculate the final recommended score of the probiotics.

[0008] Preferably, the S1 specifically includes: Based on the standardized probiotic data and standardized efficacy evaluation data, introduce a time-weighted dynamic adjustment mechanism to adjust the relationship weights between the nodes.

[0009] Preferably, the S1 specifically includes: The time-weighted dynamic adjustment mechanism calculates the relationship weights between the nodes by non-linearly adjusting the differences between the standardized probiotic data and standardized efficacy evaluation data and combining time factors.

[0010] Preferably, the S1 specifically includes: Based on the relationship weights between the nodes, calculate the shortest path length between the probiotic nodes and efficacy nodes, and sort according to the principle of minimizing the cumulative weight, and select the optimal path and sub-optimal path.

[0011] Preferably, the S2 specifically includes: Based on the relationship weights between the nodes on the path, combine the similarity of the nodes, and introduce a time-weighted coefficient to calculate the evaluation scores of the probiotic efficacy on the optimal path and all sub-optimal paths.

[0012] Preferably, the S2 specifically includes: The specific calculation formula for the evaluation score of the probiotic efficacy is as follows: , where, is the evaluation score of the th probiotic node and the th efficacy node, indicating the evaluation score of the probiotic efficacy; represents from the th probiotic node to the The relationship weight of the intermediate node at time ; is the number of intermediate nodes in the path; represents the -th intermediate node in the path from the -th probiotic node to the -th effect node; The similarity between the eigenvector of the intermediate node and the eigenvector of the target effect; is the eigenvector of the -th intermediate node; is the eigenvector of the target effect, obtained by weighted aggregation of all intermediate nodes in the path; is the time weighting coefficient.

[0013] Preferably, the S2 specifically includes: Select the path with the highest evaluation score of the probiotic effect, combine the user characteristics, calculate the final recommendation score of the probiotic by weighting, and perform personalized probiotic recommendation.

[0014] Preferably, the S2 specifically includes: Based on the shortest path length and relationship weight between the probiotic node and the effect node, calculate the path cost; based on the path cost, combine the standardized probiotic data and the standardized effect evaluation data to calculate the total computational complexity; based on the total computational complexity, control the allocation of computing resources to avoid redundant calculations; the specific calculation formula of the total computational complexity is: , where is the total computational complexity; is the number of probiotic nodes; is the number of effect nodes; is the relationship weight between the -th probiotic node and the -th effect node at time ; represents the shortest path length from the -th probiotic node to the -th effect node; represents the path cost; is the similarity between the -th standardized probiotic data and the -th standardized effect evaluation data .

[0015] The beneficial effects of the technical solution of the present invention are: 1. The present invention comprehensively evaluates the efficacy by introducing various attributes of probiotic data (such as the type of strain, dosage, usage method, storage conditions, etc.) and various evaluation indicators of efficacy evaluation data (such as changes in gut microbiota, immune response, etc.), and comprehensively considers the multi-dimensional relationship between various attributes and effects of probiotics, providing more accurate and systematic support for the scientific evaluation and selection of optimal probiotics.

[0016] 2. The present invention incorporates a time adjustment factor, which can dynamically adjust the relationship weight between probiotic nodes and effect nodes according to timeliness, taking into account the possible effect fluctuations over time, making the efficacy evaluation more in line with the actual situation and improving the application accuracy in long-term use and clinical scenarios; through shortest path calculation and path weight optimization, the present invention can select the optimal path in the knowledge graph, thereby reducing the spread of redundant information and improving the efficiency of information flow, making large-scale data processing and real-time recommendation possible.

[0017] 3. The present invention recommends the probiotics that best meet the user's needs and preferences by combining the user characteristics and the evaluation scores of probiotic effects, ensuring the pertinence and effectiveness of the recommendation, and effectively controlling the computing resource allocation of the probiotic recommendation system by designing a reasonable complexity calculation formula, avoiding redundant calculations, thus ensuring the efficient operation of the probiotic recommendation system, while ensuring the accuracy of the final recommendation result, being able to efficiently process complex data and computing requirements in a big data environment and having good scalability. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 It is a flowchart of a method for evaluating the efficacy of probiotics based on a knowledge graph according to the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0019] In order to further elaborate on the technical means and effects adopted by the present invention to achieve the predetermined invention purpose, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0020] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the technical field to which the present invention belongs.

[0021] The following specifically describes the specific solution of a method for evaluating the efficacy of probiotics based on a knowledge graph provided by the present invention in conjunction with the accompanying drawings.

[0022] Refer to the appendixFigure 1 , which shows a flowchart of a probiotic efficacy evaluation method provided by an embodiment of the present invention. The method includes the following steps: S1. Collect probiotic data and effect evaluation data, and perform unified standardization processing to obtain standardized probiotic data and standardized effect evaluation data; based on the standardized probiotic data and standardized effect evaluation data, construct a knowledge graph containing probiotic nodes and effect nodes, calculate the relationship weights between nodes, and search for the optimal path and sub-optimal path; Collect probiotic data and effect evaluation data, and perform unified standardization processing to obtain standardized probiotic data and standardized effect evaluation data, providing sufficient support for subsequent knowledge graph construction, information retrieval, and efficacy evaluation. Specifically, the core content of data collection includes the following two aspects: Probiotic data collection: mainly involves the basic information of probiotics, including the type of strain, strain number, dose, usage method (such as oral, topical, etc.), storage conditions, etc. In addition, the usage environment data of probiotics, such as the culture medium, temperature, humidity, etc. used in the experiment, should also be recorded to consider the environmental impact subsequently. All probiotic data needs to be collected and stored in the form of structured data, and its integrity and accuracy should be ensured; Effect evaluation data collection: mainly includes the efficacy data after the use of probiotics, usually in the form of experimental results, observation indicators, and effect indicators. For example, the indicators that may be involved in the experiment include changes in the intestinal flora, immune response, improvement of the digestive system, etc. The key point of effect evaluation data collection is to regularly record the changes in experimental indicators through experimental design, setting of treatment groups and control groups, and generate effect evaluation data.

[0023] Perform unified standardization processing on all collected probiotic data and effect evaluation data, remove the influence of different units and dimensions, and convert all input data into standardized data with a mean of zero and a variance of one, so that different types of data can be processed and compared with the same standard, ensuring that subsequent analysis is not affected by the difference in data dimensions, thereby improving the fairness and reliability of efficacy evaluation; the mathematical expression of the standardization processing is: , where, is the th standardized probiotic data; is the th standardized effect evaluation data; represents the th data point in the probiotic data set is the effect evaluation data set the th data point; is the mean of the probiotic dataset ; is the mean of the effect evaluation dataset ; is the standard deviation of the probiotic dataset ; is the standard deviation of the effect evaluation dataset ;

[0024] Use a knowledge graph to model the potential relationship between probiotics and effects. The nodes of the knowledge graph include probiotic nodes and effect nodes; each probiotic is represented as an independent node, and probiotic data is recorded as node attributes, such as the type, characteristics, usage environment data, etc. of the probiotic; each effect (i.e., the efficacy of the probiotic) is also represented as an independent node, and the effect node records attributes such as the type of effect and relevant health improvement data; the edge represents the potential relationship between the probiotic and the effect, for example, the enhancement effect of the probiotic on the immune system; each edge has a weight value, and the weight of the edge represents the degree of correlation between the probiotic and the effect. The initial relationship weight is determined based on inferences from historical data and literature; over time, it is necessary to adjust the relationship weight according to the newly input data (i.e., the standardized probiotic data and the standardized effect evaluation data) through a time-weighted dynamic adjustment mechanism. The formula is: , where, is the relationship weight between the th probiotic node and the th effect node at time ; is a coefficient used to control the relationship strength, obtained through experiments; is the time adjustment factor, reflecting the impact of timeliness on the relationship weight, obtained through experiments. The time-weighted dynamic adjustment mechanism dynamically adjusts the relationship weight by non-linearly adjusting the difference between the standardized probiotic data and the standardized effect evaluation data and combining the time factor. It not only depends on static data but also can take into account the possible relationship changes over time, thereby improving the accuracy of efficacy evaluation.

[0025] Search for the optimal path in the knowledge graph based on the dynamically adjusted relationship weights; the optimal path is the path with the shortest length and the optimal relationship weight between a certain probiotic node and a certain effect node. The selection of the optimal path depends on the shortest path calculation of the knowledge graph, which is obtained by accumulating the relationship weights of the nodes on the path in the knowledge graph and minimizing them; the path will go through multiple intermediate nodes during information propagation, and the relationship weight of each intermediate node will affect the overall path evaluation. Therefore, all nodes of each possible path must be comprehensively considered to ensure that the path with the minimum cost is selected. The calculation formula for the optimal path is as follows: , where, represents the shortest path length from the -th probiotic node to the -th effect node, indicating the optimized relationship distance between the two; represents the relationship weight from the -th probiotic node to the -th intermediate node at time ; represents the relationship weight from the -th intermediate node to the -th effect node at time ; is the number of intermediate nodes on the path. The shortest path is selected through the principle of minimizing the accumulated weight to ensure that the optimized result of the path conforms to the strategy of the lowest cost. Redundant and noisy information in the information propagation process is reduced through path optimization, thereby ensuring the efficiency and accuracy of the efficacy evaluation. Sort according to the principle of minimizing the accumulated weight, and select the first paths except the optimal path as the sub-optimal paths.

[0026] S2. Calculate the evaluation scores of the probiotic effects of the optimal path and the sub-optimal paths based on the relationship weights between the nodes; calculate the final recommendation score of the probiotic based on the evaluation scores of the probiotic effects.

[0027] After obtaining the optimal path, perform a weighted sum on the data in the knowledge graph path to calculate the contributions of the optimal path and all sub-optimal paths to the evaluation scores of the probiotic effects; the importance of the path depends on the product of the similarity of each node on the path and the relationship weight. Integrate the similarity and weight information of the nodes on the path through the weighted average formula, and introduce a time weighting coefficient to dynamically respond to external changes to obtain the evaluation score of the probiotic effect, thereby providing a more accurate efficacy evaluation; the specific formula is as follows: , where, is the -th probiotic node and the The evaluation score of an effect node, representing the evaluation score of the probiotic effect; Indicates the th probiotic node to the th effect node in the th intermediate node's feature vector The similarity between the feature vector of the target effect ; Is the th intermediate node's feature vector; Is the feature vector of the target effect, obtained by weighted aggregation of all intermediate nodes in the path; Is the time-weighted coefficient, used to adjust the importance of the th intermediate node in the path, obtained through experiments.

[0028] Select the path with the highest score from the evaluation scores of the probiotic effects of all obtained paths, combine user characteristics and the evaluation scores of the probiotic effects, and perform personalized probiotic recommendations. Calculate the final recommendation score of the probiotics through a weighted method to provide personalized probiotic recommendations for users, thereby meeting the individualized needs and taste preferences of users. The calculation formula for the final recommendation score of the probiotics is: , where is the final recommendation score of the th probiotic; Is the number of probiotic nodes; is the th weight coefficient of the probiotic node, obtained through experiments; is the th similarity between the feature vector of the probiotic node and the user characteristics ; user characteristics are from an existing database.

[0029] To optimize computing resources, it is necessary to calculate the computational complexity of the probiotic recommendation system, that is, the total computational complexity, to precisely control the allocation of computing resources, avoid redundant calculations, and ensure the efficient operation of the recommendation process. The specific calculation formula for the total computational complexity is: , where is the total computational complexity; is the number of effect nodes; Represents the path cost, reflecting the ratio of the path length (i.e., the relationship complexity between the probiotic and the effect) to the relationship weight (i.e., the path strength). The higher the path cost, the lower the utility of the path; Is the The similarity between the probiotic data and the standardized efficacy evaluation data for the

[0030] th effect evaluation. By calculating the cost of each path and combining the similarity between the standardized probiotic data and the standardized efficacy evaluation data, a balance between the computational effort and the recommendation quality is ensured, enabling the probiotic recommendation system to be both efficient and not affected by excessive computational effort in terms of the accuracy of the final recommendation result.

[0031] In summary, a method for evaluating the efficacy of probiotics based on a knowledge graph has been completed.

[0032] The order of the invention embodiments is only for description and does not represent the superiority or inferiority of the embodiments. The processes depicted in the drawings do not necessarily require the specific order or consecutive order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0033] The above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements for some of the technical features; and these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be included in the protection scope of the present invention.

Claims

1. A probiotic efficacy evaluation method based on knowledge graph, characterized in that: The following steps are involved: S1. Collect probiotic data and effect evaluation data, and perform unified standardization processing to obtain standardized probiotic data and standardized effect evaluation data; based on the standardized probiotic data and standardized effect evaluation data, construct a knowledge graph containing probiotic nodes and effect nodes, calculate the relationship weights between nodes, and search for optimal paths and suboptimal paths; S2. Calculate the evaluation scores of the probiotic effects of the optimal path and the suboptimal path based on the relationship weights between the nodes; Based on the evaluation scores of probiotic effects, the final recommendation scores for probiotics were calculated.

2. A probiotics efficacy evaluation method based on knowledge graph according to claim 1, characterized in that: The S1 specifically includes: Based on the standardized probiotic data and standardized effect evaluation data, a time-weighted dynamic adjustment mechanism is introduced to adjust the relationship weights between nodes.

3. A probiotic efficacy evaluation method based on knowledge graph according to claim 2, characterized in that: The S1 specifically includes: The time-weighted dynamic adjustment mechanism calculates the relationship weights between nodes by nonlinearly adjusting the difference between the standardized probiotic data and the standardized effect evaluation data and combining the time factor.

4. A probiotic efficacy evaluation method based on knowledge graph according to claim 3, characterized in that: The S1 specifically includes: Based on the relationship weights between nodes, the shortest path length between probiotic nodes and effect nodes is calculated, and they are sorted according to the principle of minimizing the cumulative weight to select the optimal path and the suboptimal path.

5. A probiotics efficacy evaluation method based on knowledge graph according to claim 1, characterized in that: The S2 specifically includes: Based on the relationship weights between nodes on the path, combined with the similarity of the nodes, and the introduction of the time weighting coefficient, the evaluation scores of the probiotic effects on the optimal path and all suboptimal paths are calculated.

6. A probiotics efficacy evaluation method based on knowledge graph according to claim 5, characterized in that: The S2 specifically includes: The specific calculation formula for the evaluation score of the probiotic effect is as follows: ， in, For the probiotics node and The evaluation score of each effect node represents the evaluation score of the probiotic effect; Indicates that from Probiotics nodes to The intermediate nodes at time The relationship weight of is the number of intermediate nodes in the path; Indicates Probiotics nodes to The first effect node in the path The feature vector of the intermediate node Characteristic vector with target effect The similarity between It is The feature vector of the intermediate nodes; is the characteristic vector of the target effect, obtained by weighted aggregation of all intermediate nodes in the path; is the time weighting factor.

7. A probiotics efficacy evaluation method based on knowledge graph according to claim 6, characterized in that: The S2 specifically includes: Select the path with the highest evaluation score of the probiotic effect, combine the user characteristics, calculate the final recommendation score of the probiotics in a weighted manner, and make personalized probiotic recommendations.

8. A probiotics efficacy evaluation method based on knowledge graph according to claim 7, characterized in that: The S2 specifically includes: Based on the shortest path length and relationship weight of the probiotic node and the effect node, the path cost is calculated; based on the path cost, the total computational complexity is calculated by combining the standardized probiotic data and the standardized effect evaluation data; based on the total computational complexity, the allocation of computing resources is controlled to avoid redundant calculations; the specific calculation formula for the total computational complexity is: ， in, is the total computational complexity; is the number of probiotic nodes; is the number of effect nodes; It is probiotics node and Between effect nodes at time The relationship weight of Indicates that from Probiotics nodes to The shortest path length of the effect nodes; represents the path cost; After standardization Probiotics data and the standardized Effect evaluation data The similarity between .

Images