A table tennis domestic competition public opinion analysis method

Abstract

The application discloses a table tennis domestic competition public opinion analysis method, and belongs to the field of sports information engineering and intelligent analysis technology. The application realizes the method as follows: 1, structurally pre-processing the historical competition data of international competition players, and screening out competition samples in China; 2, enhancing the competition contribution index; 3, enhancing the contribution index by adopting time attenuation; 4, setting a competition weight factor according to the competition category; 5, forming a single competition contribution index; 6, obtaining a competition index; 7, obtaining a public opinion analysis index; 8, constructing an ability vector representation model; 9, dynamically distributing the weight of the ability dimension; 10, forming a basic confrontation score; 11, constructing a comprehensive confrontation score; 12, converting a match win rate prediction value; 13, outputting the match win rate prediction value, the match win rate and the energy vector dimension. Compared with the prior art, the application solves the problem of the competition public opinion analysis of the international table tennis competition combined with the confrontation situation to provide the technical barrier of the competition for the domestic players.

Description

Technical Field

[0001] This invention relates to a method for analyzing domestic competitive public opinion in table tennis, belonging to the field of sports information engineering and intelligent analysis technology, and applied to intelligent processing of sports data and prediction of competitive risks. Background Technology

[0002] With the continuous improvement of the international table tennis tournament system, competitive data from various open tournaments, tour events, and comprehensive competitions are becoming increasingly frequent and diverse. Domestic sports event organizations, in their preparation and public opinion analysis, comprehensively evaluate overseas players using world rankings, historical win-loss records, and media reviews to assist in training arrangements and team formation decisions for domestic players. In recent years, data-driven analysis methods have been gradually applied to competitive preparation, but overall, it still relies primarily on experience-based judgment and comparisons of single indicators.

[0003] However, existing methods for analyzing public opinion in table tennis events generally suffer from the following shortcomings: First, the analysis process often relies on static rankings or simple win rate statistics, lacking comprehensive consideration of the competitive context, differences in event level, and time factors, making it difficult for the evaluation results to reflect the true competitive risks; Second, some models only evaluate the ability of individual athletes, failing to construct a dynamic interactive structure under the condition of two-way competition, making it difficult to characterize the changes in the probability of winning or losing under specific match relationships; Third, when the data sample distribution is uneven or the sample size is small, existing analysis systems are prone to problems such as large fluctuations in results or insufficient generalization ability, affecting the reliability of decision-making.

[0004] Therefore, how to combine international table tennis competition with a competitive context to provide technical barriers for domestic players through competitive public opinion analysis has become an urgent problem to be solved. Summary of the Invention

[0005] The purpose of this invention is to address the technical problems in analyzing competitive public opinion regarding international table tennis competitions and their competitive contexts, thereby creating a technical barrier for domestic players. This invention proposes a method for analyzing domestic competitive public opinion in table tennis. Addressing the limitations of existing competitive public opinion analyses, which can only assess static indicators, struggle to reflect the risks of specific competitive scenarios, lack the ability to predict win rates, and have an incomplete overall decision support structure, this invention establishes a complete analytical path from historical match data processing and risk index calculation to specific match win rate predictions. This improves the systematic nature, stability, and interpretability of competitive public opinion analysis.

[0006] The objective of this invention is achieved through the following technical solution:

[0007] This invention discloses a method for analyzing domestic competitive public opinion regarding table tennis, comprising the following steps:

[0008] Step 1: Perform structured preprocessing on the historical competition data of international athletes using data cleaning and standardized mapping, and then select samples of competitions against Chinese athletes.

[0009] Step 1.1: Extract the historical match data of players in international tournaments, which consists of match time, opponent, match result, tournament category, and opponent's world ranking;

[0010] Step 1.2: Remove missing and abnormally formatted data from historical competition data and construct a standardized mapping with player names to form preprocessed historical competition data;

[0011] Step 1.3: Extract the preprocessed data of historical matches featuring Chinese players to form a sample of matches against Chinese players;

[0012] Step 2: The competitive contribution index of the competition samples ranked in descending order against China is enhanced by a nonlinear mapping as shown in Equation (1);

[0013] (1)

[0014] Where r is the world ranking of China's rival, and k, a, and b are preset parameters; To enhance the competitive contribution of high-ranking Chinese players when they are defeated;

[0015] Step 3: The contribution index of the competition samples against China selected through recent time is enhanced by time decay as shown in formula (2);

[0016] (2)

[0017] in, λ represents the time difference between the match time and the current time; λ is the time decay coefficient. This is used to reflect the higher contribution of recent matches to the competitive index;

[0018] Step 4: Set the event weighting factor based on the event category to determine the event level. ;

[0019] Step 5: The contribution index of the recent descending ranking of the matches against China is fused by nonlinear mapping, time decay and event weight factor as shown in Equation (3) to form a single match contribution index;

[0020] (3)

[0021] in, For ranking mapping values, This is the time decay value. As a weighting factor for the event;

[0022] Step 6: Obtain the competitive index by fusing the sample size stability correction of the China-based match samples with the cumulative single-match contribution index;

[0023] Step 6.1: The stability of the sample size of the competition samples against China is corrected using the method shown in Equation (4);

[0024] (4)

[0025] Where n is the number of samples in matches against China; For stability adjustment parameters;

[0026] Step 6.2: Accumulate the single-match contribution index of the sample of matches against China and integrate it with the stability correction to form the competitive index as shown in Equation (5). ;

[0027] (5)

[0028] in, To accumulate the contribution values ​​of all individual games;

[0029] Step 7: Perform oversaturation processing on the competition index using the exponential compression method as shown in equation (6) to obtain the public opinion analysis index;

[0030] (6)

[0031] in, It is the upper limit constant of the exponent; The saturation coefficient; The compression index is [value]; the public opinion analysis index is [value]. ;

[0032] Step 8: Construct a capability vector representation model with dimensional indicators energy vector dimension in the form of a dimensional index containing public opinion analysis index, and use the enhanced mapping function to nonlinearly enhance the distinguishability of the energy vector dimension in the model;

[0033] Step 8.1: Construct a public opinion analysis index The energy vector dimension, composed of win rate stability index, key game handling ability index, psychological stress resistance index, competition experience index, and specific confrontation situation control index, is normalized and mapped to form an ability vector; the ability vector is then used to construct an ability vector representation model in a dimensional manner.

[0034] Step 8.2: Use the capability vector representation model to obtain the corresponding energy vector dimension difference between Chinese and foreign athletes as shown in equation (7);

[0035] (7)

[0036] in, This indicates the dimension of the energy vector corresponding to the relevant indicator;

[0037] Step 8.3: Use the enhancement mapping function g(·) as shown in equation (8) to perform nonlinear amplification on the energy vector dimension difference according to the monotonic function, thereby increasing the distinguishability of energy vector dimension differences;

[0038] (8)

[0039] Where g(·) is a monotonic continuous function;

[0040] Step 9: Dynamically assign capability dimension weights based on the event category as the energy vector dimension. ;

[0041] Step 9.1: When the competition category is upgraded by one level, the weights of the corresponding ability dimensions of the psychological stress resistance index and the competition experience index will be increased proportionally.

[0042] Step 9.2: When the event category is downgraded by one level, the public opinion analysis index will be adjusted. The weights of the capability dimensions corresponding to the win rate stability index are increased proportionally.

[0043] Step 10: Adaptively weight and fuse the dynamically assigned capability dimension weights and the energy vector dimension differentiation to form the basic adversarial score as shown in Equation (9);

[0044] (9)

[0045] Step 11: Construct a comprehensive confrontation score using a basic confrontation score and a comprehensive correction item consisting of risk trigger correction items and tactical style matching correction items;

[0046] Step 11.1: Construct a correction term based on the key local capability differences Psychological resistance modification items Differences in control of specific adversarial situations - correction terms Composition of risk trigger correction items Matching modification items with technical and tactical styles;

[0047] Step 11.2: Use the technical and tactical style compatibility matrix as shown in equation (10) to modify the technical and tactical style matching correction term;

[0048] (10)

[0049] in, This represents the correction coefficient for the corresponding style combination in the tactical style compatibility matrix; This refers to the technical and tactical style of Chinese athletes; The types of technical and tactical styles of foreign athletes;

[0050] Step 11.3: Construct a comprehensive correction term as shown in Equation (11) using the risk-triggered correction term and the corrected tactical style matching correction term. ;

[0051] (11)

[0052] in, ; The key session capability difference correction item is calculated based on the difference in key session processing capability indicators between the two parties; The psychological resistance correction item is calculated based on the difference in psychological resilience indicators between the two parties; The correction term for control differences in specific adversarial situations is calculated based on the differences in control indicators between the two sides in specific adversarial situations;

[0053] Step 11.4: Construct the comprehensive adversarial score as shown in Equation (12) using the basic adversarial score and the comprehensive correction term;

[0054] (12)

[0055] Step 12: Use the probability mapping function to convert the comprehensive adversarial score into the predicted win rate value as shown in Equation (13);

[0056] (13)

[0057] in, For monotony A type mapping function is used to map the score value to the interval [0,1].

[0058] Step 13: Obtain the public opinion analysis index of a single domestic player from the historical match data of international players using the methods described in Steps 1 to 7, and obtain the comprehensive confrontation score through the ability vector difference calculation method and comprehensive correction described in Steps 8 to 11; after the probability mapping described in Step 12, convert the comprehensive confrontation score into the predicted value of the match win rate, the match win rate and the energy vector dimension for visualization output.

[0059] Compared with existing technologies, it has the following beneficial effects:

[0060] 1. This invention constructs a complete modeling path from historical competition data processing to public opinion analysis index output, realizing a systematic and quantitative assessment of the competitive risks of foreign athletes to China, and improving the overall structural integrity of public opinion analysis; compared with traditional analysis methods based on simple win rate statistics, this invention can improve the accuracy of competitive risk identification by about 3% to 5%.

[0061] 2. Based on the calculation of public opinion analysis index, this invention further establishes a dynamic modeling mechanism for adversarial scenarios. This mechanism can contextually weight and fuse the differences in abilities between the two sides under specific competition conditions, improving the ability to characterize specific matchups. This invention improves the stability of matchup prediction by approximately 2%.

[0062] 3. This invention introduces an adversarial compatibility correction structure. By comprehensively processing differences in ability, psychological resilience, and tactical style matching in key games, the model can reflect the game characteristics of actual matches, improving the rationality and interpretability of the prediction results. The accuracy rate in predicting changes in the outcome of key games is improved by approximately 4% to 6%.

[0063] 4. This invention constructs a win rate mapping model, converting the comprehensive adversarial score into a predicted win rate value, and further derives the score probability distribution of multi-set matches, achieving multi-level output from risk assessment to win rate prediction, thus enhancing the application value of the model. This invention improves the stability of win rate prediction in multi-set matches by approximately 5%.

[0064] 5. This invention improves the model's stability and generalization ability under varying sample sizes by employing sample stability correction and difference enhancement mapping mechanisms, thus avoiding result fluctuations under small sample conditions. Under smaller sample sizes, the fluctuation range of prediction results is reduced by approximately 6% to 8%.

[0065] 6. This invention outputs a public opinion analysis index, match win rate, score probability distribution, and lineup decision ranking results, which can provide data support for competitive preparation analysis and lineup selection, and improve the decision-making assistance capability of competitive public opinion analysis. Through simulated lineup selection experiments, the prediction accuracy of lineup decisions based on the model of this invention is improved by approximately 5% to 8%. Attached Figure Description

[0066] Figure 1 is a schematic diagram of the process of the present invention;

[0067] Figure 2 This is a schematic diagram of the calculation structure for sports public opinion analysis and match win rate prediction. Detailed Implementation

[0068] To better illustrate the purpose and advantages of this invention, the invention will be further described below with reference to the accompanying drawings and examples. It should be noted that the implementation of this invention is not limited to the following embodiments, and any modifications or alterations made to this invention will fall within the scope of protection of this invention.

[0069] The public opinion analysis index calculation method of this invention includes steps such as historical data acquisition, screening of samples related to China, nonlinear mapping calculation, time decay processing, event weight calculation, single-event contribution fusion, stability correction, and index saturation compression output. Its public opinion analysis index calculation structure consists of a data input layer, a nonlinear mapping layer, a single-event contribution calculation layer, a stability correction layer, and an index compression output layer. These layers are interconnected through functional relationships to achieve multi-factor fusion calculation.

[0070] This embodiment uses a foreign table tennis player, A, as an example to specifically illustrate the method of the present invention. First, all match data for player A over the past two years, totaling 120 matches, is obtained, including 8 matches against Chinese players. The match data includes match time, opponent, match result, event category, and opponent's world ranking information.

[0071] Example

[0072] like Figure 1 As shown in the figure, the specific implementation steps of the table tennis domestic competitive public opinion analysis method in this embodiment are as follows:

[0073] Step 1: Perform structured preprocessing on the historical competition data of international athletes using data cleaning and standardized mapping, and then select samples of competitions against Chinese athletes.

[0074] Step 1.1: Extract the historical match data of players in international tournaments, which consists of match time, opponent, match result, tournament category, and opponent's world ranking;

[0075] Step 1.2: Remove missing and abnormally formatted data from historical competition data and construct a standardized mapping with player names to form preprocessed historical competition data;

[0076] Step 1.3: Extract the preprocessed data of historical matches featuring Chinese players to form a sample of matches against Chinese players;

[0077] Step 2: The competitive contribution index of the competition samples ranked in descending order against China is enhanced by a nonlinear mapping as shown in Equation (1);

[0078] (1)

[0079] Where r is the world ranking of China's rival, and k, a, and b are preset parameters; To enhance the competitive contribution of high-ranking Chinese players when they are defeated;

[0080] In this embodiment, historical match data is filtered, retaining only the sample set of matches against Chinese players. In 8 matches against Chinese players, player A defeated Chinese players 3 times. This invention calculates the threat contribution value only for matches where the player defeated a Chinese player. According to... Figure 2 The nonlinear mapping layer shown is configured with the following world rankings for the three wins against Chinese opponents: r1 = 2, r2 = 5, r3 = 18. The parameters of the nonlinear mapping function are set as: k = 12, a = 1, b = 1.

[0081] The non-linear ranking mapping function is: Substituting into the calculation, we get:

[0082] f(2) = 12 / ln(3) ≈ 10.93

[0083] f(5) = 12 / ln(6) ≈ 6.70

[0084] f(18) = 12 / ln(19) ≈ 4.07

[0085] Step 3: The contribution index of the competition samples against China selected through recent time is enhanced by time decay as shown in formula (2);

[0086] (2)

[0087] in, λ represents the time difference between the match time and the current time; λ is the time decay coefficient. This is used to reflect the higher contribution of recent matches to the competitive index;

[0088] In the embodiments, according to Figure 2 The time decay layer shown has a time decay coefficient λ = 0.04.

[0089] The time differences between the three matches are as follows:

[0090] Δt1 = 3 months

[0091] Δt2 = 9 months

[0092] Δt3 = 14 months

[0093] The time decay function is: Substituting into the calculation, we get:

[0094] w1 ≈ 0.886

[0095] w2 ≈ 0.697

[0096] w3 ≈ 0.571

[0097] Step 4: Set the event weighting factor based on the event category to determine the event level. ;

[0098] Step 5: The contribution index of the recent descending ranking of the matches against China is fused by nonlinear mapping, time decay and event weight factor as shown in Equation (3) to form a single match contribution index;

[0099] (3)

[0100] in, For ranking mapping values, This is the time decay value. As a weighting factor for the event;

[0101] In the embodiments, according to Figure 2 The tournament weights shown are as follows: World Championship weight = 1.5, WTT Championship weight = 1.3, and Regular Open weight = 1.0. The corresponding weights for the three wins are 1.5, 1.3, and 1.0, respectively.

[0102] according to Figure 2 The single-game contribution calculation layer shown has the following formula: Substituting into the calculation, we get:

[0103] C1 ≈ 14.52

[0104] C2 ≈ 6.06

[0105] C3 ≈ 2.32

[0106] Original threat value: X = 14.52 + 6.06 + 2.32 = 22.90

[0107] Step 6: Obtain the competitive index by fusing the sample size stability correction of the China-based match samples with the cumulative single-match contribution index;

[0108] Step 6.1: The stability of the sample size of the competition samples against China is corrected using the method shown in Equation (4);

[0109] (4)

[0110] Where n is the number of samples in matches against China; For stability adjustment parameters;

[0111] Step 6.2: Accumulate the single-match contribution index of the sample of matches against China and integrate it with the stability correction to form the competitive index as shown in Equation (5). ;

[0112] (5)

[0113] in, To accumulate the contribution values ​​of all individual games;

[0114] In the embodiments, according to Figure 2 The stability correction layer shown has the following parameters: number of samples from matches against China: n = 8; stability parameter: τ = 6. The stability function is: Substituting the values, we get: S ≈ 0.736, and the corrected threat value is: X' = X × S ≈ 16.85;

[0115] Step 7: Perform oversaturation processing on the competition index using the exponential compression method as shown in equation (6) to obtain the public opinion analysis index;

[0116] (6)

[0117] in, It is the upper limit constant of the exponent; The saturation coefficient; The compression index is [value]; the public opinion analysis index is [value]. ;

[0118] In the embodiments, according to Figure 2 The exponentially compressed output layer shown is given by M = 100. = 0.25, γ = 1.15, the exponential compression function is: Substituting the values, we get: T ≈ 98.3. The final output T is the public opinion analysis index for contestant A.

[0119] Step 8: Construct a capability vector representation model with dimensional indicators energy vector dimension in the form of a dimensional index containing public opinion analysis index, and use the enhanced mapping function to nonlinearly enhance the distinguishability of the energy vector dimension in the model;

[0120] Step 8.1: Construct a public opinion analysis index The energy vector dimension, composed of win rate stability index, key game handling ability index, psychological stress resistance index, competition experience index, and specific confrontation situation control index, is normalized and mapped to form an ability vector; the ability vector is then used to construct an ability vector representation model in a dimensional manner.

[0121] In the example, after obtaining contestant A's public opinion analysis index T=98.3, according to Figure 2The ability vector modeling and adversarial scoring calculation structure shown further analyzes the adversarial relationship between the Chinese player and the foreign player. First, an ability vector model is constructed. According to step 8.1, the public opinion analysis index, win rate stability index, key game handling ability index, psychological resilience index, tournament experience index, and specific adversarial situation control index are used as six energy vector dimensions, and a normalized mapping is used to form an ability vector representation. Let the ability vectors of Chinese player B and foreign player A be respectively:

[0122] Chinese athlete's B ability vector: ;

[0123] Foreign player A's ability vector: ;

[0124] Step 8.2: Use the capability vector representation model to obtain the corresponding energy vector dimension difference between Chinese and foreign athletes as shown in equation (7);

[0125] X i =X i (China)-X i (Foreign) (7)

[0126] Among them, X i This indicates the dimension of the energy vector corresponding to the relevant indicator;

[0127] In this embodiment, according to step 8.2, the dimensional difference between the capability vectors of both parties is calculated: The differences were obtained in six dimensions: public opinion analysis index was -0.06; win rate stability was 0.08; ability to handle key games was 0.10; psychological resilience was 0.12; tournament experience was 0.03; and control of competitive situations was 0.02.

[0128] Step 8.3: Use the enhancement mapping function g(·) as shown in equation (8) to perform nonlinear amplification on the energy vector dimension difference according to the monotonic function, thereby increasing the distinguishability of energy vector dimension differences;

[0129] (8)

[0130] Where g(·) is a monotonic continuous function;

[0131] In this embodiment, according to step 8.3, the above difference is nonlinearly amplified using the enhancement mapping function g(·) to enhance the discriminative power of differences across different dimensions. Let the enhancement function be: After mapping, the enhanced difference vector is obtained as G = (−0.065, 0.092, 0.115, 0.138, 0.034, 0.022).

[0132] Step 9: Dynamically assign capability dimension weights based on the event category as the energy vector dimension. ;

[0133] Step 9.1: When the competition category is upgraded by one level, the weights of the corresponding ability dimensions of the psychological stress resistance index and the competition experience index will be increased proportionally.

[0134] Step 9.2: When the event category is downgraded by one level, the public opinion analysis index will be adjusted. The weights of the capability dimensions corresponding to the win rate stability index are increased proportionally.

[0135] In this embodiment, the weights of the capability dimensions are then dynamically adjusted according to the event category, based on step 9. If the current event is a world championship-level event, the weights of the psychological resilience and major competition experience dimensions are increased accordingly. The weights of the six dimensions are set as follows: public opinion analysis index = 0.18; win rate stability = 0.16; critical game handling ability = 0.17; psychological resilience = 0.21; major competition experience = 0.16; and competitive situation control = 0.12.

[0136] Step 10: Adaptively weight and fuse the dynamically assigned capability dimension weights and the energy vector dimension differentiation to form the basic adversarial score as shown in Equation (9);

[0137] (9)

[0138] In this embodiment, according to step 10, a weighted fusion calculation of the enhanced capability difference vector is performed to calculate the basic adversarial score: Substituting into the calculation, we get: ;

[0139] Step 11: Construct a comprehensive confrontation score using a basic confrontation score and a comprehensive correction item consisting of risk trigger correction items and tactical style matching correction items;

[0140] Step 11.1: Construct a correction term based on the key local capability differences Psychological resistance modification items Differences in control of specific adversarial situations - correction terms Composition of risk trigger correction items Matching modification items with technical and tactical styles;

[0141] In this embodiment, according to step 11, a comprehensive correction term is introduced to adjust the basic confrontation score. The risk-triggered correction term includes a key situation capability difference correction term, a psychological confrontation correction term, and a specific confrontation situation control difference correction term. Based on the differences in indicators between the two sides, the following is set: Key Situation Capability Correction Term ; Psychological resistance modification item ; Context control correction item The risk trigger correction item is: ;

[0142] Step 11.2: Use the technical and tactical style compatibility matrix as shown in equation (10) to modify the technical and tactical style matching correction term;

[0143] (10)

[0144] in, This represents the correction coefficient for the corresponding style combination in the tactical style compatibility matrix; This refers to the technical and tactical style of Chinese athletes; The types of technical and tactical styles of foreign athletes;

[0145] In this embodiment, according to step 11.2, the style compatibility matrix is ​​used to calculate the style matching correction term. Let Chinese player B be a "penhold two-sided loop drive player" and foreign player A be a "backhand speed attack player". The corresponding correction coefficients are obtained from the style compatibility matrix: ;

[0146] Step 11.3: Construct a comprehensive correction term as shown in Equation (11) using the risk-triggered correction term and the corrected tactical style matching correction term. ;

[0147] (11)

[0148] in, ; The key session capability difference correction item is calculated based on the difference in key session processing capability indicators between the two parties; The psychological resistance correction item is calculated based on the difference in psychological resilience indicators between the two parties; The correction term for control differences in specific adversarial situations is calculated based on the differences in control indicators between the two sides in specific adversarial situations;

[0149] In the embodiment, the comprehensive correction term is: Substituting into the calculation, we get: ;

[0150] Step 11.4: Construct the comprehensive adversarial score as shown in Equation (12) using the basic adversarial score and the comprehensive correction term;

[0151] (12)

[0152] In this embodiment, according to step 11.4, the comprehensive adversarial score is as follows: The calculation yielded: ;

[0153] Step 12: Use the probability mapping function to convert the comprehensive adversarial score into the predicted win rate value as shown in Equation (13);

[0154] (13)

[0155] in, For monotony A type mapping function is used to map the score value to the interval [0,1].

[0156] In this embodiment, finally, according to step 12, the comprehensive adversarial score is converted into a predicted win rate value through a probability mapping function: Let the probability mapping function be the Logistic function: Substituting into the calculation, we get: This means that Chinese player B's predicted winning percentage under these matchup conditions is approximately 60%.

[0157] Step 13: Obtain the public opinion analysis index of a single domestic player from the historical match data of international players using the methods described in Steps 1 to 7, and obtain the comprehensive confrontation score through the ability vector difference calculation method and comprehensive correction described in Steps 8 to 11; after the probability mapping described in Step 12, convert the comprehensive confrontation score into the predicted value of the match win rate, the match win rate and the energy vector dimension for visualization output.

[0158] In this embodiment, the final system output includes: the public opinion analysis index of overseas players, the difference in ability vectors between the two sides, the comprehensive confrontation score, and the prediction result of the winning rate of the match, thereby realizing a comprehensive analysis of the relationship between competitive risk and winning rate in the context of international table tennis confrontation.

Claims

1. A method for analyzing domestic competitive public opinion regarding table tennis, characterized in that: Includes the following steps, Step 1: Perform structured preprocessing on the historical competition data of international athletes using data cleaning and standardized mapping, and then select samples of competitions against Chinese athletes. Step 2: The competitive contribution index of the competition samples ranked in descending order against China is enhanced by a nonlinear mapping as shown in Equation (1); （1） Where r is the world ranking of China's rival, and k, a, and b are preset parameters; To enhance the competitive contribution of high-ranking Chinese players when they are defeated; Step 3: The contribution index of the competition samples against China selected through recent time is enhanced by time decay as shown in formula (2); （2） in, λ represents the time difference between the match time and the current time; λ is the time decay coefficient. This is used to reflect the higher contribution of recent matches to the competitive index; Step 4: Set the event weighting factor based on the event category to determine the event level. ; Step 5: The contribution index of the recent descending ranking of the matches against China is fused by nonlinear mapping, time decay and event weight factor as shown in Equation (3) to form a single match contribution index; （3） in, For ranking mapping values, This is the time decay value. As a weighting factor for the event; Step 6: Obtain the competitive index by fusing the sample size stability correction of the China-based match samples with the cumulative single-match contribution index; Step 7: Perform oversaturation processing on the competition index using the exponential compression method as shown in equation (6) to obtain the public opinion analysis index; （6） in, It is the upper limit constant of the exponent; The saturation coefficient; The compression index is [value]; the public opinion analysis index is [value]. ; Step 8: Construct a capability vector representation model with dimensional indicators energy vector dimension in the form of a dimensional index containing public opinion analysis index, and use the enhanced mapping function to nonlinearly enhance the distinguishability of the energy vector dimension in the model; Step 8.1: Construct a public opinion analysis index The energy vector dimension, composed of win rate stability index, key game handling ability index, psychological stress resistance index, competition experience index, and specific confrontation situation control index, is normalized and mapped to form an ability vector; the ability vector is then used to construct an ability vector representation model in a dimensional manner. Step 8.2: Use the capability vector representation model to obtain the corresponding energy vector dimension difference between Chinese and foreign athletes as shown in equation (7); X i =X i (China)-X i (Foreign) (7) Among them, X i This indicates the dimension of the energy vector corresponding to the relevant indicator; Step 8.3: Use the enhancement mapping function g(·) as shown in equation (8) to perform nonlinear amplification on the energy vector dimension difference according to the monotonic function, thereby increasing the distinguishability of energy vector dimension differences; （8） Where g(·) is a monotonic continuous function; Step 9: Dynamically assign capability dimension weights based on the event category as the energy vector dimension. ; Step 10: Adaptively weight and fuse the dynamically assigned capability dimension weights and the energy vector dimension differentiation to form the basic adversarial score as shown in Equation (10); （9） Step 11: Construct a comprehensive confrontation score using a basic confrontation score and a comprehensive correction item consisting of risk trigger correction items and tactical style matching correction items; Step 11.1: Construct a correction term based on the key local capability differences Psychological resistance modification items Differences in control of specific adversarial situations - correction terms Composition of risk trigger correction items Matching modification items with technical and tactical styles; Step 11.2: Use the technical and tactical style compatibility matrix as shown in equation (11) to modify the technical and tactical style matching correction term; （10） in, This represents the correction coefficient for the corresponding style combination in the tactical style compatibility matrix; This refers to the technical and tactical style of Chinese athletes; The types of technical and tactical styles of foreign athletes; Step 11.3: Construct a comprehensive correction term as shown in Equation (12) using the risk-triggered correction term and the corrected tactical style matching correction term. ; （11） in, ; The key session capability difference correction item is calculated based on the difference in key session processing capability indicators between the two parties; The psychological resistance correction item is calculated based on the difference in psychological resilience indicators between the two parties; The correction term for control differences in specific adversarial situations is calculated based on the differences in control indicators between the two sides in specific adversarial situations; Step 11.4: Construct the comprehensive adversarial score as shown in equation (13) using the basic adversarial score and the comprehensive correction term; （12） Step 12: Use the probability mapping function to convert the comprehensive adversarial score into the predicted win rate value as shown in Equation (14); （13） in, For monotony A type mapping function is used to map the score value to the interval [0,1]. Step 13: Obtain the public opinion analysis index of a single domestic player from the historical match data of international players using the methods described in Steps 1 to 7, and obtain the comprehensive confrontation score through the ability vector difference calculation method and comprehensive correction described in Steps 8 to 11; after the probability mapping described in Step 12, convert the comprehensive confrontation score into the predicted value of the match win rate, the match win rate and the energy vector dimension for visualization output.

2. The method for analyzing domestic competitive public opinion regarding table tennis as described in claim 1, characterized in that: Step 1 is implemented as follows: Step 1.1: Extract the historical match data of players in international tournaments, which consists of match time, opponent, match result, tournament category, and opponent's world ranking; Step 1.2: Remove missing and abnormally formatted data from historical competition data and construct a standardized mapping with player names to form preprocessed historical competition data; Step 1.3: Extract the preprocessed data of historical matches featuring Chinese players to form a sample of matches against Chinese players.

3. The method for analyzing domestic competitive public opinion regarding table tennis as described in claim 1, characterized in that: Step 6 is implemented as follows: Step 6.1: The stability of the sample size of the competition samples against China is corrected using the method shown in Equation (4); （4） Where n is the number of samples in matches against China; For stability adjustment parameters; Step 6.2: Accumulate the single-match contribution index of the sample of matches against China and integrate it with the stability correction to form the competitive index as shown in Equation (5). ; （5） in, This is to accumulate the contribution values ​​of all individual games.

4. The method for analyzing domestic competitive public opinion regarding table tennis as described in claim 1, characterized in that: Step 9 is implemented as follows: Step 9.1: When the competition category is upgraded by one level, the weights of the corresponding ability dimensions of the psychological stress resistance index and the competition experience index will be increased proportionally. Step 9.2: When the event category is downgraded by one level, the public opinion analysis index will be adjusted. The weights of the capability dimensions corresponding to the win rate stability index are increased proportionally.

Images