Method and system for determining product design decision based on information efficiency

Abstract

The invention discloses a method and system for determining a product design decision based on information efficiency, and particularly relates to the technical field of product design decisions, and the method comprises the following steps: S1, information source collection and establishment, S2, knowledge graph construction, S3, dynamic information entropy weight distribution, S4, reverse information flow verification, S5, reverse engineering data integration, S6, information fusion processing, S7, decision making, S8, fault-tolerant mechanism, and S9, fault-tolerant mechanism determination. And performing reverse verification. According to the method, the dynamic knowledge graph capable of supporting semantic association retrieval and context reasoning is constructed by determining the multi-dimensional information source, the information weight is dynamically adjusted according to the information generation time and the association degree, inspiration is provided for innovative design, and the decision accuracy is improved; and the decision model can be corrected by using a failure case library, and the knowledge graph logic relationship is corrected through product version test feedback, so that the decision process is optimized, and continuous innovation of the product is promoted.

Description

Technical Field

[0001] The present invention relates to the technical field of product design decision-making, and in particular to a method and system for determining product design decisions based on information effectiveness. Background Art

[0002] In today's highly competitive market, product design decisions play a vital role in a company's development. Accurate and scientific design decisions can help products better meet market demands and enhance a company's competitiveness. However, irrational decisions can lead to product failure and waste resources.

[0003] Existing product design decision-making methods, which mostly rely on the designer's experience or simple statistical analysis, fail to fully tap the potential value behind massive amounts of information, nor can they cope with the complex and volatile market environment and diverse user needs. Furthermore, the analysis and utilization of failure cases during the product design process are insufficient. When a product encounters a problem or fails to meet expected performance, it is often impossible to effectively trace back to the information gaps in the design phase, resulting in similar design flaws reappearing in subsequent products.

[0004] Therefore, it is urgent to design a method that can integrate multi-dimensional information, effectively utilize failure cases and competitive product information, and make scientific and reasonable product design decisions. Summary of the Invention

[0005] The main purpose of the present invention is to provide a method and system for determining product design decisions based on information effectiveness, which can effectively solve the problems mentioned in the background technology.

[0006] To achieve the above object, the technical solution adopted by the present invention is:

[0007] A method for determining product design decisions based on information effectiveness includes the following steps:

[0008] S1, determine the multi-dimensional information sources related to product design, including but not limited to user demand data, technical parameters, market data, and environmental constraints;

[0009] S2, constructs a dynamic knowledge graph based on information sources. At the same time, based on historical cases and collected multi-dimensional information sources, it builds a knowledge graph that includes multi-dimensional relationships such as technology, market, and users;

[0010] S3, dynamically adjusts the importance of information based on the time when the information was generated in the dynamic knowledge graph, and the newer the information, the higher the weight;

[0011] S4, establish a product failure case library and trace back the original information gaps corresponding to design defects;

[0012] S5: Establish a reverse engineering library for competitor information to expand the dynamic knowledge graph. Use a crawler to capture public patents, teardown reports, and user reviews. Combined with reverse analysis tools, extract competitor design parameters and performance indicators to construct a competitor feature matrix.

[0013] S6, uses the powerful semantic reasoning ability of knowledge graph to analyze and judge the data collected from different information sources;

[0014] S7 converts the rich entity relationships in the knowledge graph into specific decision rules, combines them with the NSGA-III algorithm to achieve multi-objective optimization, and prioritizes the Pareto solutions that meet the knowledge constraints;

[0015] S8, uses the failed product case library to modify the decision model. When it is detected that a certain type of information frequently causes design deviations, its weight is reduced and a manual review process is triggered;

[0016] S9, test two or more versions of the product and compare and analyze the test results with the performance data of competing products.

[0017] Preferably, in S2, the dynamic knowledge graph established supports semantic association retrieval and contextual reasoning.

[0018] Preferably, in S6, the knowledge graph is combined with a multi-dimensional performance balance algorithm and a weighted compromise planning method is adopted to find a compromise solution between conflicting requirements according to the importance of different dimensions.

[0019] Preferably, in said S8, during the manual review process, professionals are organized to conduct a detailed review of the information source and the relevant decision-making process, analyze the causes of the problems, and make timely corrections.

[0020] Preferably, in S9, each version of the product is presented to different user groups, user feedback data is collected, and the advantages and disadvantages of different versions are evaluated.

[0021] Preferably, in S9, it is identified whether there is an over-reliance on past experience and data and a situation where new changes are ignored.

[0022] Preferably, in said S4, the information collection rules and the performance evaluation threshold are dynamically updated according to the result obtained by reverse tracing.

[0023] A product design decision determination system based on information effectiveness, comprising:

[0024] S1, information source collection and establishment module, determines the multi-dimensional information sources related to product design, including but not limited to user demand data, technical parameters, market data, and environmental constraints;

[0025] S2, the knowledge graph construction module, constructs a dynamic knowledge graph based on information sources. It also builds a knowledge graph containing multi-dimensional relationships such as technology, market, and users based on historical cases and collected multi-dimensional information sources.

[0026] S3, the dynamic information entropy weight distribution module, dynamically adjusts the importance of information based on the time when the information was generated in the dynamic knowledge graph, and the newer the information, the higher the weight;

[0027] S4, the reverse information flow verification module, establishes a product failure case library, reversely traces the original information gaps corresponding to design defects, and dynamically updates information collection rules and performance evaluation thresholds;

[0028] S5, the reverse engineering data integration module, establishes a reverse engineering database for competitor information to expand the dynamic knowledge graph. It crawls public patents, disassembly reports, and user reviews, and uses reverse analysis tools to extract competitor design parameters and performance indicators to construct a competitor feature matrix.

[0029] S6, the information fusion processing module, uses the powerful semantic reasoning ability of the knowledge graph to analyze and judge the data collected from different information sources;

[0030] S7, the decision module, transforms the rich entity relationships in the knowledge graph into specific decision rules, combines the NSGA-III algorithm to achieve multi-objective optimization, and prioritizes the Pareto solutions that meet the knowledge constraints;

[0031] S8, the fault-tolerance mechanism module, uses a library of failed product cases to revise the decision model. When it detects that a certain type of information frequently causes design deviations, it reduces its weight and triggers a manual review process.

[0032] S9, the reverse verification module, tests two or more versions of a product and compares and analyzes the test results with the performance data of competing products.

[0033] Compared with the prior art, the present invention has the following beneficial effects:

[0034] 1. The present invention determines multi-dimensional information sources to construct a dynamic knowledge graph that supports semantic association retrieval and contextual reasoning. It dynamically adjusts the information weight according to the time when the information is generated and the degree of relevance, so as to adapt to market changes, provide inspiration for innovative design and improve the accuracy of decision-making. At the same time, the establishment of a product failure case library can trace back to the root cause of the problem to avoid repeating the same mistakes in subsequent designs; the establishment of a reverse engineering library of competitor information can realize the expansion of the knowledge graph. In the face of conflicting data, the knowledge graph can use semantic reasoning and multi-dimensional performance balance algorithms to find a compromise solution. The entity relationship is converted into a decision rule and combined with the NSGA-III algorithm to achieve multi-objective optimization and provide scientific decision-making. In addition, the decision model is corrected using the failure case library, and the logical relationship of the knowledge graph is corrected through product version test feedback, so as to optimize the decision-making process, promote continuous product innovation, and ultimately achieve the beneficial effect of comprehensively improving the scientificity and effectiveness of product design decisions and better adapting to market changes. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] Figure 1 It is a schematic diagram of the overall process of the present invention. DETAILED DESCRIPTION

[0036] In order to make the technical means, creative features, objectives and effects achieved by the present invention easier to understand, the present invention is further described below in conjunction with specific implementation methods.

[0037] Embodiment 1, a method for determining product design decisions based on information effectiveness, includes:

[0038] S1. Identify multi-dimensional information sources related to product design, including user demand data (questionnaires, behavior logs), technical parameters (material properties, manufacturing processes), market data (competitive product analysis, trend forecasts), and environmental constraints (regulations, sustainability requirements).

[0039] S2 builds a dynamic knowledge graph based on information sources. Based on historical cases and collected multi-dimensional information sources, it establishes a knowledge graph containing multi-dimensional relationships such as technology, market, and users. This supports semantic association retrieval and contextual reasoning. When designing new products, it can quickly find relevant technical cases and market demand information through semantic retrieval, providing rich inspiration and basis for innovative design.

[0040] S3, dynamically adjusts the importance of information based on the time when the information was generated in the dynamic knowledge graph, with newer information having higher weight to adapt to the rapidly changing market environment;

[0041] At the same time, the information relevance coefficient is defined to measure the closeness of the association between information by calculating the semantic similarity and business logic association between information. Information with higher relevance will receive additional bonus in weight distribution.

[0042] Real-time calculation of information source weights. For example, for user demand data, as market demand rapidly changes, this ensures that more valuable and relevant information can dominate the decision-making process at different times, improving the speed and accuracy of decision-making responses to market changes and avoiding decision-making errors caused by the use of outdated or irrelevant information.

[0043] S4. Establish a product failure case library and reversely trace the original information gaps corresponding to design defects. When a product encounters problems in the market or fails to achieve expected performance, reverse tracing can be used to find the original information gaps corresponding to the design defects. For example, if a product fails in a specific environment, analysis may reveal that the failure was due to the failure to identify the requirements of a certain environmental regulation during the design phase. Based on the discovered issues, the information collection rules can be dynamically updated to include information related to the environmental regulation in the information collection scope. The performance evaluation threshold can also be adjusted to increase the attention and weight of similar important information to avoid the same problems in subsequent product designs.

[0044] S5, establish a reverse engineering library of competitor information to achieve the expansion of dynamic knowledge graphs, use crawler technology to go deep into major patent databases, industry information websites and e-commerce platforms, capture public patent information, and analyze in detail the technical innovations, structural design and claims in the patent specifications; the disassembly report is obtained through cooperation with professional disassembly agencies, or an internal team is formed to physically disassemble the competitor, recording the material, process and connection method of each component; user comments are collected from major e-commerce platforms, social media and professional product forums, and natural language processing technology is used to perform sentiment analysis and keyword extraction on the comments to explore user satisfaction, pain points and expectations for the product. Combined with reverse analysis tools, the competitor's physical product is scanned and modeled to extract key design parameters, such as size specifications, appearance features, etc., as well as performance indicators, such as the battery life of electronic products, processor speed, load capacity of mechanical products, and operating accuracy. Finally, these rich data are sorted and constructed into a comprehensive and detailed competitor feature matrix. Each row of the matrix represents a competitor, and each column corresponds to a different feature dimension;

[0045] S6. Leverage the powerful semantic reasoning capabilities of the knowledge graph to analyze and judge data collected from different information sources. When conflicting data arises, for example, a user expects a product to be lightweight for portability, but from a material strength perspective, lightweight materials may not meet the product's strength requirements. The knowledge graph can automatically identify this contradiction based on its defined semantic relationships and logical rules. Then, combined with a multi-dimensional performance balance algorithm and a weighted compromise planning method, it can find a compromise solution between conflicting demands based on the importance of different dimensions (such as the weights of product performance, cost, user experience, etc.). By comprehensively considering multiple conflicting dimensions, it calculates an optimal balance solution so that the performance of each dimension is within an acceptable range.

[0046] S7 converts the rich entity relationships in the knowledge graph into specific decision rules. When the product design involves the selection of high-cost materials, the decision engine can automatically trigger the search process for low-cost alternative materials based on this rule. Combined with the NSGA-III algorithm to achieve multi-objective optimization, the NSGA-III algorithm is an efficient multi-objective evolutionary algorithm that can find the optimal solution among multiple conflicting objectives (such as maximizing product performance, minimizing costs, and shortening the R&D cycle). During the optimization process, the Pareto solution that meets the knowledge constraints defined in the knowledge graph is retained first. The Pareto solution means that under multiple objectives, there is no other solution that can improve the performance of at least one objective without reducing the performance of other objectives;

[0047] When using the NSGA-III algorithm, the target optimization problem is defined as follows:

[0048] min[f1(x),f2(x),...,f k (x)]x∈Ω

[0049] Pareto optimal solution: the solution x* satisfies that for all i, there is no x' such that

[0050] f i (x′)≤f i (x * )

[0051] and at least one

[0052] f j (x′) <f j (x * )

[0053] NSGA-III divides the target space by reference points to maintain the diversity of solutions.

[0054] Reference points refer to a set of uniformly distributed virtual points pre-defined in the target space, which are used to divide the target space into sub-regions;

[0055] For example, suppose the optimization problem has two objectives (M=2), the reference points are generated as p=4, and the target space is divided into 5 reference points. The algorithm ensures that each subregion has solution coverage by assigning the solution to the nearest reference point.

[0056] Provide scientific and intelligent decision support for product decisions. Under complex multi-objective constraints, quickly generate a series of reasonable decision-making plans and ensure the feasibility and rationality of the plans through knowledge constraints, helping enterprises find the product design and development direction that best suits their own interests among many options.

[0057] S8: Decision models are revised using a database of failed product cases. When a certain type of information is detected to frequently lead to design deviations, such as user demand information obtained from a specific channel that consistently differs significantly from actual market feedback, the weight of that information source is reduced, minimizing its influence in the decision-making process. Simultaneously, a manual review process is triggered, with professionals conducting a detailed review of the information source and related decision-making processes, analyzing the causes of the problem (e.g., improper information collection methods, data processing errors, etc.), and implementing timely corrections.

[0058] When correcting the weight of an information source, if a deviation is detected:

[0059] w new =w old (1-η·coefficient of deviation)

[0060] Where η refers to the learning rate;

[0061] The deviation coefficient is obtained by normalizing the two values: the proportion of exceeding the threshold number and the weighted average of the exceeding threshold error.

[0062] S9, by presenting two or more versions of a product to different user groups, collects user feedback data and evaluates the pros and cons of each version. The test results are then compared and analyzed with the performance data of competing products. During this comparison process, local overfitting of information effectiveness is identified. For example, in product design, over-reliance on historical data will lead to insufficient product innovation. While performing well in certain historical scenarios, it may lack competitiveness in the face of new market demands and competitive environments. These analysis results are fed back into the knowledge graph to modify the logical relationships within the graph, such as adjusting the correlation strength between certain technologies and market demands, or supplementing the relationship between new market trends and product features.

[0063] For example, when testing and comparing the differences between two groups of user feedback, use:

[0064]

[0065] Calculate the result.

[0066] Among them, μ A ,μ B : The sample mean of user feedback from two groups (Group A and Group B);

[0067] S 2 A , S 2 B : The variance of the two groups of samples (measure the degree of dispersion of the data);

[0068] n A , n B : The sample size of the two groups (i.e., the number of users in each group);

[0069] t: used to determine whether the difference between two groups is significant.

[0070] Through the reverse verification closed loop, we continuously optimize the knowledge graph and decision-making process, improve the accuracy of information efficiency evaluation and the scientific nature of decision-making, avoid product design falling into local optimal solutions, promote continuous product innovation and optimization, and better adapt to market changes.

[0071] like Figure 1 As shown, a system for determining product design decisions based on information effectiveness includes:

[0072] S1, information source collection and establishment module, determines the multi-dimensional information sources related to product design, including but not limited to user demand data, technical parameters, market data, and environmental constraints;

[0073] S2, the knowledge graph construction module, constructs a dynamic knowledge graph based on information sources. It also builds a knowledge graph containing multi-dimensional relationships such as technology, market, and users based on historical cases and collected multi-dimensional information sources.

[0074] S3, the dynamic information entropy weight distribution module, dynamically adjusts the importance of information based on the time when the information was generated in the dynamic knowledge graph, and the newer the information, the higher the weight;

[0075] S4, the reverse information flow verification module, establishes a product failure case library, reversely traces the original information gaps corresponding to design defects, and dynamically updates information collection rules and performance evaluation thresholds;

[0076] S5, the reverse engineering data integration module, establishes a reverse engineering database for competitor information to expand the dynamic knowledge graph. It crawls public patents, disassembly reports, and user reviews, and uses reverse analysis tools to extract competitor design parameters and performance indicators to construct a competitor feature matrix.

[0077] S6, the information fusion processing module, uses the powerful semantic reasoning ability of the knowledge graph to analyze and judge the data collected from different information sources;

[0078] S7, the decision module, transforms the rich entity relationships in the knowledge graph into specific decision rules, combines the NSGA-III algorithm to achieve multi-objective optimization, and prioritizes the Pareto solutions that meet the knowledge constraints;

[0079] S8, the fault-tolerance mechanism module, uses a library of failed product cases to revise the decision model. When it detects that a certain type of information frequently causes design deviations, it reduces its weight and triggers a manual review process.

[0080] S9, the reverse verification module, tests two or more versions of a product and compares and analyzes the test results with the performance data of competing products.

[0081] Therefore, this solution identifies multi-dimensional information sources and constructs a dynamic knowledge graph that supports semantic association retrieval and contextual reasoning. This approach dynamically adjusts information weighting based on information generation time and relevance, adapting to market changes, providing inspiration for innovative designs, and improving decision accuracy. Furthermore, a library of product failure cases is established to trace back to the root causes of problems, preventing subsequent designs from repeating the same mistakes. A reverse engineering library of competitor information is also established to expand the knowledge graph. Faced with conflicting data, the knowledge graph can leverage semantic reasoning and a multi-dimensional performance balance algorithm to find compromises. Entity relationships are converted into decision rules and combined with the NSGA-III algorithm to achieve multi-objective optimization and provide scientific decision-making. Furthermore, the failure case library is used to refine the decision model, and feedback from product version testing is used to correct the logical relationships in the knowledge graph. This optimizes the decision-making process, promotes continuous product innovation, and ultimately achieves the beneficial effect of comprehensively improving the scientific nature and effectiveness of product design decisions and better adapting to market changes.

[0082] The basic principles, main features, and advantages of the present invention are shown and described above. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The above embodiments and descriptions are merely illustrative of the principles of the present invention. Various changes and modifications may be made to the present invention without departing from the spirit and scope of the present invention. Such changes and modifications are intended to fall within the scope of the present invention. The scope of protection claimed in the present invention is defined by the appended claims and their equivalents.

Claims

1. A method for determining product design decisions based on information effectiveness, comprising the following steps: S1, determine the multi-dimensional information sources related to product design, including but not limited to user demand data, technical parameters, market data, and environmental constraints; S2, constructs a dynamic knowledge graph based on information sources. At the same time, based on historical cases and collected multi-dimensional information sources, it builds a knowledge graph that includes multi-dimensional relationships such as technology, market, and users; S3, dynamically adjusts the importance of information based on the time when the information was generated in the dynamic knowledge graph, and the newer the information, the higher the weight; S4, establish a product failure case library and trace back the original information gaps corresponding to design defects; S5: Establish a reverse engineering library for competitor information to expand the dynamic knowledge graph. Use a crawler to capture public patents, teardown reports, and user reviews. Combined with reverse analysis tools, extract competitor design parameters and performance indicators to construct a competitor feature matrix. S6, uses the powerful semantic reasoning ability of knowledge graph to analyze and judge the data collected from different information sources; S7 converts the rich entity relationships in the knowledge graph into specific decision rules, combines them with the NSGA-III algorithm to achieve multi-objective optimization, and prioritizes the Pareto solutions that meet the knowledge constraints; S8, uses the failed product case library to modify the decision model. When it is detected that a certain type of information frequently causes design deviations, its weight is reduced and a manual review process is triggered; S9, test two or more versions of the product and compare and analyze the test results with the performance data of competing products.

2. The method for determining product design decisions based on information effectiveness according to claim 1, characterized in that: In S2, the dynamic knowledge graph established supports semantic association retrieval and contextual reasoning.

3. The method for determining product design decisions based on information effectiveness according to claim 1, characterized in that: In S6, the knowledge graph is combined with a multi-dimensional performance balance algorithm and a weighted compromise planning method is adopted to find a compromise solution between conflicting requirements based on the importance of different dimensions.

4. The method for determining product design decisions based on information effectiveness according to claim 1, characterized in that: In the manual review process in S8, professionals are organized to conduct a detailed review of the information source and related decision-making processes, analyze the causes of the problems, and make timely corrections.

5. The method for determining product design decisions based on information effectiveness according to claim 1, characterized in that: In S9, each version of the product is presented to different user groups, user feedback data is collected, and the advantages and disadvantages of different versions are evaluated.

6. The method for determining product design decisions based on information effectiveness according to claim 5, characterized in that: In S9, it is identified whether there is an over-reliance on past experience and data and a situation where new changes are ignored.

7. The method and system for determining product design decisions based on information effectiveness according to claim 1, characterized in that: In S4, the information collection rules and the performance evaluation threshold are dynamically updated according to the results obtained by reverse tracing.

8. The system used in the method for determining product design decisions based on information effectiveness according to claims 1-7, characterized in that: include: S1, information source collection and establishment module, determines the multi-dimensional information sources related to product design, including but not limited to user demand data, technical parameters, market data, and environmental constraints; S2, the knowledge graph construction module, constructs a dynamic knowledge graph based on information sources. It also builds a knowledge graph containing multi-dimensional relationships such as technology, market, and users based on historical cases and collected multi-dimensional information sources. S3, the dynamic information entropy weight distribution module, dynamically adjusts the importance of information based on the time when the information was generated in the dynamic knowledge graph, and the newer the information, the higher the weight; S4, the reverse information flow verification module, establishes a product failure case library, reversely traces the original information gaps corresponding to design defects, and dynamically updates information collection rules and performance evaluation thresholds; S5, the reverse engineering data integration module, establishes a reverse engineering database for competitor information to expand the dynamic knowledge graph. It crawls public patents, disassembly reports, and user reviews, and uses reverse analysis tools to extract competitor design parameters and performance indicators to construct a competitor feature matrix. S6, the information fusion processing module, uses the powerful semantic reasoning ability of the knowledge graph to analyze and judge the data collected from different information sources; S7, the decision module, transforms the rich entity relationships in the knowledge graph into specific decision rules, combines the NSGA-III algorithm to achieve multi-objective optimization, and prioritizes the Pareto solutions that meet the knowledge constraints; S8, the fault-tolerance mechanism module, uses a library of failed product cases to revise the decision model. When it detects that a certain type of information frequently causes design deviations, it reduces its weight and triggers a manual review process. S9, the reverse verification module, tests two or more versions of a product and compares and analyzes the test results with the performance data of competing products.

Images