Game content customization method and system based on user behaviors

Abstract

The invention discloses a game content customization method and system based on user behaviors, and relates to the technical field of game development. The method comprises the following steps: acquiring audio data and player behavior data; environment surge features, intention ripple features and sound wave melody features are constructed according to the preprocessed audio data and player behavior data; a silence score is obtained through calculation according to the environment surge characteristics, a true speech score is obtained through calculation according to the intention ripple characteristics, and a melody dynamic score is obtained through calculation according to the sound wave melody characteristics; fusing the silence score, the true speech score and the melody dynamic score to construct a comprehensive score; a game context atmosphere, NPC dialogue behavior, and task progress award mechanism are dynamically adjusted based on the comprehensive score and silence score, truth score, and melody dynamic score. By analyzing the voice features and behavior modes of players in real time, the game situation atmosphere, NPC interaction behaviors and personalized contents are dynamically adjusted, so that the immersion is enhanced, the interactivity is improved, and the personalized experience is optimized.

Description

Technical Field

[0001] The present invention relates to the technical field of game development, and particularly to a method and system for customizing game content based on user behavior. Background Art

[0002] With the development of the game industry, personalized game experiences have gradually become an important direction for enhancing player immersion and game satisfaction. Traditional game content mainly relies on preset rules and fixed storylines to drive, making it difficult to meet the personalized needs of players. In recent years, the progress of technologies such as artificial intelligence, audio analysis, and behavior recognition has made it possible to adaptively adjust game content based on user behavior.

[0003] During the game interaction process, the player's voice, volume changes, behavior patterns, and emotional states can all reflect their current game experience and psychological state. However, most existing methods for adaptively adjusting game content focus on basic game data analysis, such as operation behaviors, task completion status, etc., and rarely consider the subtle manifestations of players in aspects such as voice, emotion, and rhythm changes. How to achieve more accurate game content customization through more comprehensive user behavior analysis, especially audio feature extraction and calculation, remains a challenge in current game design. Summary of the Invention

[0004] Based on the above-mentioned disadvantages of the prior art, the purpose of the present invention is to provide a method and system for customizing game content based on user behavior to solve the above technical problems.

[0005] To achieve the above purpose, the present invention provides the following technical solution: A method for customizing game content based on user behavior, including:

[0006] Obtain audio data and player behavior data, and perform preprocessing;

[0007] Construct environmental surging features, intention ripple features, and sonic melody features based on the preprocessed audio data and player behavior data;

[0008] Calculate a silence score based on the environmental surging features, calculate a truth score based on the intention ripple features, and calculate a melody dynamic score based on the sonic melody features;

[0009] Fuse the silence score, truth score, and melody dynamic score to construct a comprehensive score;

[0010] Dynamically adjust the game scenario atmosphere, NPC dialogue behavior, and task progress reward mechanism based on the comprehensive score and the silence score, truth score, and melody dynamic score.

[0011] The present invention is further configured such that the environmental surging features include instantaneous volume, short-term energy change rate, average background noise, spectral analysis, instantaneous volume entropy, life rhythm perception, instantaneous sound sparsity, and silence stability index;

[0012] The intention ripple features include speech sharpness index, speech consistency index, emotional reverberation effect, sound space oscillation index, intention prediction oscillation, and interaction rhythm imbalance index;

[0013] The acoustic wave melody features include dynamic emotion regulation, pitch dimension expansion, amplitude volatility, melody time sequence correlation, melody energy growth rate, and note interval isolation;

[0014] The silence score is calculated based on the instantaneous volume, short-term energy change rate, average background noise, spectral analysis, instantaneous volume entropy, life rhythm perception, instantaneous sound sparsity, and silence stability index;

[0015] The truth score is calculated based on the speech sharpness index, speech consistency index, emotional reverberation effect, sound space oscillation index, intention prediction oscillation, and interaction rhythm imbalance index;

[0016] The melody dynamic score is calculated based on the dynamic emotion regulation, pitch dimension expansion, amplitude volatility, melody time sequence correlation, melody energy growth rate, and note interval isolation.

[0017] The present invention is further configured such that the calculation logic of the instantaneous volume is: where L(t) is the instantaneous volume at time point t, N is the sampling window size, s i is the sampling value of the i-th audio sample. Based on the above calculation logic, a smoothing operation is performed on L(t) to obtain The calculation logic of is: where is the smoothed volume, and α is the smoothing coefficient;

[0018] The calculation logic of the short-term energy change rate is: where EV(t) is the short-term energy change rate at the current time t;

[0019] The calculation logic of the average background noise is: BNL(t) = β·BNL(t - 1)+(1 - β)·L env (t), where BNL(t) is the background noise at the current time t, BNL(t - 1) is the background noise at the previous moment, L env (t) is the environmental noise measured at the current time t, and β is the weighting factor;

[0020] The calculation logic of the spectral analysis is: Among them, HF(t) is the spectrum analysis at time point t, P(f) is the power spectral density at frequency f, and f max is the maximum frequency;

[0021] The calculation logic of the instantaneous entropy of volume is as follows: Among them, LE(t) is the instantaneous entropy of volume at the current time t, and G i is the probability distribution of the i-th sampled volume point, and the calculation logic of G i is as follows: Among them, L i is the volume sampling value,

[0022] The calculation logic of the perception of the rhythm of life is as follows: Among them, BPI(t) is the perception of the rhythm of life at the current time t;

[0023] The calculation logic of the silence stability index is as follows: Among them, SSI(t) is the silence stability index at the current time t, T is the length of the historical time under investigation, and κ is the sensitivity adjustment parameter;

[0024] The calculation logic of the silence score is as follows: Among them, S(t) is the silence score at the current time t, α1 is the volume attenuation coefficient, β1 is the energy change weight, γ is the balance factor, η is the noise suppression threshold, and δ is the noise influence coefficient.

[0025] The present invention is further set such that the calculation logic of the speech sharpness index is as follows: Among them, SUI(t) is the speech sharpness index at the current time t, and P i is the pitch of the i-th sampling point, PD i is the pause time of the i-th sampling point, SR j is the value of the speech rate at the j-th moment, U is the number of sampling points of the speech data within a given time window, and P max is the maximum volume, PD max is the maximum pause time, and SR max is the maximum speech rate value;

[0026] The calculation logic of the speech consistency index is as follows: Among them, SCI(t) is the speech consistency index at the current time t;

[0027] The calculation logic of the emotional reverberation effect is as follows: Among them, ERE(t) is the emotional reverberation effect at the current time t, R is the number of sampling points of the emotional intensity or pitch and volume change within a given time window, and S iis the i-th emotional intensity sample, PD j is the pause time at the j-th sampling point;

[0028] The calculation logic of the sound space oscillation index is as follows: where VSI(t) is the sound space oscillation index at the current time t, V i is the volume of the i-th audio sample, P min is the minimum value of the pitch, V min is the minimum value of the volume;

[0029] The calculation logic of the intention prediction oscillation is as follows: where IPO(t) is the intention prediction oscillation at the current time t, O is the number of time segment sampling points in the voice interaction between the player and the NPC, EI i is the emotional intensity at the i-th moment, P target is the target voice frequency;

[0030] The calculation logic of the interaction rhythm imbalance index is as follows: where IRI(t) is the interaction rhythm imbalance index at the current time t, SR i is the player's speaking speed, SR NPC is the NPC's speaking speed, PD NPC is the NPC's pause time;

[0031] The calculation logic of the true word score is: LS(t) = (w1·SUI(t) + w2·SCI(t) + w3·EREt + w4·VSIt + w5·IPOt + w6·IRIt·1 + VtVmax, where LSt is the true word score at the current time t, V(t) is the current emotional fluctuation amplitude of the player, V max is the maximum emotional fluctuation value, w1, w2, w3, w4, w5 and w6 are weight coefficients.

[0032] The present invention is further set that the calculation logic of the dynamic emotion regulation is as follows: where DEt is the dynamic emotion regulation at the current time t, Y is the number of notes within a given time window, Δf i is the pitch change amplitude, Δv i is the volume change amplitude, τ i is the duration of the i-th note, α2, β2, γ1 and δ1 are adjustment parameters;

[0033] The calculation logic of the pitch dimension expansion is as follows: where PD(t) is the pitch dimension expansion at the current time t, is the pitch value of the i-th note in the k-th dimension, m is the number of pitch space dimensions, λ is the non-linear control factor;

[0034] The calculation logic of amplitude volatility is as follows: Among them, AF(t) is the amplitude volatility at the current time t, σ(f i ) is the non-linear transformation function of pitch, and ∈ and μ are adjustment parameters;

[0035] The calculation logic of the melodic temporal correlation degree is as follows: Among them, MTCt is the melodic temporal correlation degree at the current time t, υi is the starting time of the i-th note, and κ1 and μ1 are parameters for controlling the temporal difference sensing;

[0036] The calculation logic of the melodic energy growth rate is as follows: Among them, MEGR(t) is the melodic energy growth rate at the current time t, and ζ and η1 are energy growth control parameters;

[0037] The calculation logic of the note interval degree is as follows: Among them, NID(t) is the note interval degree at the current time t, is the non-linear transformation parameter of the interval;

[0038] The calculation logic of the melodic dynamic score is: OEI(t) = A1·DE(t) + A2·PD(t) + A3·AF(t) + A4·MTC(t) + A5·MEGR(t) + A6·NID(t), where OEI(t) is the melodic dynamic score at the current time t, and A1, A2, A3, A4, A5 and A6 are weight coefficients.

[0039] The present invention is further configured that the calculation logic of the comprehensive score is:

[0040]

[0041] Among them, CS(t) is the comprehensive score at the current time t; S(t) is the silence score at the current time t, LS(t) is the truth score at the current time t, OEI(t) is the melodic dynamic score at the current time t, S(t)0 is the silence score reference value, ε0, ε1 and ε2 are weight coefficients, λ s , θ t , α m and δ m are non-linear adjustment parameters.

[0042] The present invention is further configured that the calculation logic of adjusting the scenario atmosphere is: E = α e·(1 - e^(-βe·St - St0 + θs·1 + γs·sin(ω1·LSt) + δe·(1 - tanh(φm·OEIt)), where E is the scenario atmosphere, α e and β e are the coefficients for adjusting the reflectivity of the silence score to the environmental tension, θ s and γ s are the coefficients for adjusting the influence of the truth score on the scenario atmosphere, δ e and φ m are the coefficients for controlling the influence of the melody dynamics score on the scenario atmosphere, and ω1 is the coefficient for adjusting the periodic fluctuation of the melody dynamics score on the environment.

[0043] The present invention is further configured such that the calculation logic for adjusting the NPC dialogue behavior is:

[0044] where R NPC is the NPC dialogue behavior, ω2 is the adjustment factor for the NPC response period by the silence score, α t and β t are the coefficients for controlling the NPC response behavior by the truth score, γ m and δ t are the coefficients for adjusting the emotional responsiveness of the NPC by the melody dynamics score.

[0045] The present invention is further configured such that the calculation logic for adjusting the task progress reward mechanism is: where T reward is the adjustment of the task progress reward mechanism, μ t and ζ t are the coefficients for adjusting the bonus effect of the truth score on the reward, θ m is the coefficient for adjusting the influence of the melody score, α a and δ a are the coefficients for adjusting the non - linear influence of the melody dynamics score, λ a is the coefficient for adjusting the influence of the silence score on the task reward.

[0046] The present invention also provides a game content customization system based on user behavior, and the system includes:

[0047] Data acquisition module: used to acquire audio data and player behavior data and perform pre - processing;

[0048] Feature construction module: used to construct environmental surge features, intention ripple features, and sound wave melody features based on the pre - processed audio data and player behavior data;

[0049] The first calculation module: used to calculate the silence score based on the environmental surging characteristics, calculate the truth score based on the intention ripple characteristics, and calculate the melody dynamic score based on the sound wave melody characteristics;

[0050] The second calculation module: used to fuse the silence score, the truth score, and the melody dynamic score to construct a comprehensive score;

[0051] The adjustment module: used to dynamically adjust the game scenario atmosphere, NPC dialogue behavior, and task progress reward mechanism based on the comprehensive score, as well as the silence score, the truth score, and the melody dynamic score.

[0052] The present invention provides a method and system for customizing game content based on user behavior. The method includes obtaining audio data and player behavior data and performing preprocessing; constructing environmental surging characteristics, intention ripple characteristics, and sound wave melody characteristics based on the preprocessed audio data and player behavior data; calculating the silence score based on the environmental surging characteristics, calculating the truth score based on the intention ripple characteristics, and calculating the melody dynamic score based on the sound wave melody characteristics; fusing the silence score, the truth score, and the melody dynamic score to construct a comprehensive score; dynamically adjusting the game scenario atmosphere, NPC dialogue behavior, and task progress reward mechanism based on the comprehensive score, as well as the silence score, the truth score, and the melody dynamic score. The beneficial effects generated include:

[0053] 1. Enhance game immersion: By real-time analyzing the voice characteristics and behavior patterns of players, dynamically adjusting the game scenario atmosphere, making the game content more in line with the emotions and states of players, and enhancing the immersive experience;

[0054] 2. Improve game interactivity: By calculating the truth score and the melody dynamic score, adjusting the dialogue behavior of NPCs, enabling them to make more natural and context-compliant responses based on the voice intonation, emotional fluctuations, and rhythm changes of players, and improving the interaction realism between NPCs and players;

[0055] 3. Precise personalized experience: Adopting multi-dimensional audio data analysis methods, including instantaneous volume, volume entropy, emotional reverberation effect, etc., to achieve more delicate player emotion perception and behavior prediction, so as to provide personalized game content adjustment for different players.

[0056] The above description is only an overview of the technical solution of this application. In order to be able to understand the technical means of this application more clearly, it can be implemented according to the content of the description. And in order to make the above and other purposes, features, and advantages of this application more obvious and understandable, the following specifically gives the specific implementation manners of this application. Brief Description of the Drawings

[0057] To more clearly illustrate the technical solutions in the embodiments of the present invention, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings. In the drawings:

[0058] Figure 1 A flowchart of a game content customization method based on user behavior shown in an exemplary embodiment of the present invention;

[0059] Figure 2 A schematic structural diagram of a game content customization system based on user behavior shown in an exemplary embodiment of the present invention. Detailed implementation manners

[0060] The following will describe the embodiments of the present invention with reference to the drawings and preferred embodiments. Those skilled in the art can easily understand other advantages and effects of the present invention from the content disclosed in this specification. The present invention can also be implemented or applied through other different specific implementation manners. Various details in this specification can also be modified or changed based on different viewpoints and applications without departing from the spirit of the present invention. It should be understood that the preferred embodiments are only for explaining the present invention, rather than for limiting the protection scope of the present invention.

[0061] It should be noted that the diagrams provided in the following embodiments only illustrate the basic concept of the present invention in a schematic manner. Therefore, only the components related to the present invention are shown in the diagrams, rather than being drawn according to the number, shape, and size of the components in actual implementation. The type, quantity, and ratio of each component in actual implementation can be arbitrarily changed, and the component layout type may also be more complex.

[0062] In the following description, a large number of details are explored to provide a more thorough explanation of the embodiments of the present invention. However, it is obvious to those skilled in the art that the embodiments of the present invention can be implemented without these specific details. In other embodiments, well-known structures and devices are shown in the form of block diagrams rather than in detail to avoid making the embodiments of the present invention difficult to understand.

[0063] Embodiment 1

[0064] A game content customization method based on user behavior, as Figure 1 shown, includes:

[0065] Obtain audio data and player behavior data, and perform preprocessing;

[0066] Construct environmental surging features, intention ripple features, and sound wave melody features according to the preprocessed audio data and player behavior data;

[0067] The silence score is calculated based on the characteristics of environmental surges, the truth score is calculated based on the characteristics of intention ripples, and the melody dynamic score is calculated based on the characteristics of acoustic melodies;

[0068] The silence score, the truth score, and the melody dynamic score are fused to construct a comprehensive score;

[0069] Based on the comprehensive score and the silence score, the truth score, and the melody dynamic score, the game scenario atmosphere, the NPC dialogue behavior, and the task progress reward mechanism are dynamically adjusted.

[0070] The present invention is further configured such that the characteristics of environmental surges include instantaneous volume, short-term energy change rate, average background noise, spectral analysis, instantaneous volume entropy, perception of life rhythm, instantaneous sound sparsity, and silence stability index; The present invention is further configured such that the calculation logic of the instantaneous volume is: where L(t) is the instantaneous volume at time point t, N is the size of the sampling window, and s i is the sampling value of the i-th audio sample. Based on the above calculation logic, a smoothing operation is performed on L(t) to obtain The calculation logic of is: where is the smoothed volume and α is the smoothing coefficient; The calculation logic of the short-term energy change rate is: where EV(t) is the short-term energy change rate at the current time t; The calculation logic of the average background noise is: BNL(t) = β·BNL(t - 1)+(1 - β)·L env (t), where BNL(t) is the background noise at the current time t, BNL(t - 1) is the background noise at the previous moment, and L env (t) is the environmental noise measured at the current time t, and β is the weighting factor; The calculation logic of the spectral analysis is: where HF(t) is the spectral analysis at time point t, P(f) is the power spectral density at frequency f, and f max is the maximum frequency; The calculation logic of the instantaneous volume entropy is: where LE(t) is the instantaneous volume entropy at the current time t, and G i is the probability distribution of the i-th sampling volume point, and the calculation logic of G i is: where L i is the volume sampling value; The calculation logic of the perception of life rhythm is: where BPI(t) is the perception of life rhythm at the current time t; The calculation logic of the silence stability index is: Among them, SSI(t) is the silence stability index at the current time t, T is the length of the historical time under investigation, and κ is the sensitivity adjustment parameter; specifically, the instantaneous volume is used to measure the intensity of the audio signal at the current moment, and a smoothing operation is used to reduce the impact of short-term fluctuations; the short-term energy change is used to detect energy changes, capture sharp changes in the audio signal, identify "sudden sounds" or "sudden silences", and avoid mis-triggering game effects due to instantaneous silences; the average background noise is used to reduce the interference of environmental noise on player voice analysis; spectral analysis is used to distinguish voice features from background noise. When HF(t) is below the set threshold, it indicates that the current sound is mainly environmental noise rather than the player's voice; the instantaneous entropy of volume is used to measure the uncertainty of volume distribution, helping to identify the clarity and stability of speech. When LE(t) is below the set threshold, it indicates a long-term stable silence and can trigger the silence trial; the perception of the life rhythm is used to evaluate the naturalness and fluency of speech. The player's breathing may still produce slight noises in a quiet state. The low-frequency band is used to detect whether the player holds their breath. When BPI(t) is approximately equal to 0, it indicates that the player holds their breath and is very likely in a true silent state; the silence stability index is used to evaluate the smoothness of volume changes, avoid misjudging short pauses as silent states, improve the recognition accuracy of true silent states, and combine historical data analysis to measure the similarity between the current audio state and the silent state in the past period of time to prevent mis-triggering due to short-term silences. When SSI(t) is greater than a certain threshold and lasts for a period of time, it can be determined that the player is truly in a silent state; α is used to affect the smoothness of the instantaneous volume, and its value range is [0.8, 0.9]; β is used to control the smoothness of background noise, and its value range is [0.3, 0.9]; κ is used to control the judgment threshold of silence stability, and its value range is [0.1, 1]; the game scenario is dynamically adjusted through voice feature analysis to make the game content more in line with the player's current state, and by combining indicators such as the perception of the life rhythm, more delicate prediction of the player's state is achieved to optimize the game experience;

[0071] The intention ripple features include the voice sharpness index, the voice consistency index, the emotional reverberation effect, the sound space oscillation index, the intention prediction oscillation, and the interaction rhythm imbalance index; the present invention is further set such that the calculation logic of the voice sharpness index is: Among them, SUI(t) is the voice sharpness index at the current time t, P i is the pitch at the i-th sampling point, PD i is the pause time at the i-th sampling point, SR j is the value of the speech rate at the j-th moment, U is the number of sampling points of the voice data within a given time window, P max is the maximum volume, PD max is the maximum pause time, SR max is the maximum speech rate value;

[0072] The calculation logic of the speech consistency index is as follows: Among them, SCI(t) is the speech consistency index at the current time t; the calculation logic of the emotional swing effect is as follows: Among them, ERE(t) is the emotional swing effect at the current time t, R is the number of sampling points of the emotional intensity or pitch and volume change within a given time window, S i is the i-th emotional intensity sample, PD j is the pause time at the j-th sampling point; the calculation logic of the sound space oscillation index is as follows: Among them, VSI(t) is the sound space oscillation index at the current time t, V i is the volume of the i-th audio sample, P min is the minimum value of the pitch, V min is the minimum value of the volume; the calculation logic of the intention prediction oscillation is as follows: Among them, IPO(t) is the intention prediction oscillation at the current time t, O is the number of sampling points of the time segment in the voice interaction between the player and the NPC, EI i is the emotional intensity at the i-th moment, P target is the target voice frequency; the calculation logic of the interaction rhythm imbalance index is as follows: Among them, IRI(t) is the interaction rhythm imbalance index at the current time t, SR i is the player's speech rate, SR NPC is the NPC's speech rate, PD NPCis the NPC pause time; specifically, the voice sharpness index is used to measure the suddenness characteristics of the voice within a time window, for analyzing the player's urgent emotions, intention fluctuations, and drastic changes in the interaction rhythm. The higher the SUI(t), the more rapid the speech and the shorter the pauses, which may reflect tension or excitement in emotions; the voice consistency index is used to measure the smoothness of the voice within consecutive time windows. When the SCI(t) value is larger, it indicates higher voice consistency, meaning a stable speaking style and smaller fluctuations in pitch and speaking speed; the emotional reverberation effect is used to reflect the periodic change characteristics of voice emotions over time, and can measure the stability of voice emotional fluctuations. When the emotional changes are relatively consistent within a certain time window, the ERE(t) value is higher, showing a certain reverberation pattern; the voice space oscillation index is used to measure the fluctuation amplitude and its stability in pitch and volume of the voice. When the VSI(t) value is larger, it means the greater the voice variation of the player and the stronger the expressed voice dynamics; the intention prediction oscillation is used to measure the emotional fluctuations and voice intention changes in the voice interaction between the player and the NPC; the interaction rhythm imbalance index is used to measure the matching degree of the voice rhythms of the player and the NPC. When the IRI(t) value is larger, it indicates a significant difference in the voice rhythms between the player and the NPC, which may lead to unnatural interactions; is the smoothing factor. This form is used to reduce the influence of outliers, making small changes have less impact on the result, while significant changes have a more prominent impact on the result. Its specific manifestation in the above calculation logic is: and is the exponential normalization factor. This function is used to smooth outliers and prevent a certain feature, including extreme pitch, from having too much impact on the calculation result. The specific manifestation of this type of function in the above calculation logic is: By analyzing the above various indices, the accuracy of voice emotion analysis can be improved, the voice interaction experience can be optimized, the naturalness of NPC interaction can be enhanced, and the coherence and accuracy of voice recognition and synthesis can be increased;

[0073] The sonic melody features include dynamic emotion regulation, pitch dimension expansion, amplitude volatility, melody time series correlation, melody energy growth rate, and note separation degree; The present invention is further configured such that the calculation logic of dynamic emotion regulation is: wherein, D E (t) is the dynamic emotion regulation at the current time t, Y is the number of notes within a given time window, Δf i is the pitch change amplitude, Δv i is the volume change amplitude, τ i is the duration of the i-th note, and α2, β2, γ1, and δ1 are adjustment parameters; The calculation logic of pitch dimension expansion is: wherein, PD(t) is the pitch dimension expansion at the current time t, is the pitch value of the i-th note in the k-th dimension, m is the number of pitch space dimensions, and λ is the non-linear control factor; the calculation logic of amplitude volatility is as follows: where AF(t) is the amplitude volatility at the current time t, and σ(f i ) is the non-linear transformation function of pitch, and ∈ and μ are adjustment parameters; the calculation logic of melody temporal correlation is as follows: where MTC(t) is the melody temporal correlation at the current time t, and υ i is the start time of the i-th note, and κ1 and μ1 are parameters for controlling the temporal difference sensing; the calculation logic of melody energy growth rate is as follows: where MEGR(t) is the melody energy growth rate at the current time t, and ζ and η1 are energy growth control parameters; the calculation logic of note isolation is as follows: where NID(t) is the note isolation at the current time t, is the non-linear transformation parameter for the interval; specifically, dynamic emotion regulation is used to measure the degree of emotional change of speech or music within a certain time window; pitch dimension expansion is used to measure the degree of change of speech or music in a multi-dimensional pitch space; amplitude volatility is used to measure the degree of change of amplitude; melody temporal correlation is used to measure the correlation of note time intervals, that is, the relationship between the start time of the note and the volume change, reflecting the coherence of the melody; melody energy growth rate is used to measure the energy change trend of the melody over time; note isolation is used to reflect the degree of discreteness of pitch changes between notes, measuring whether the pitch changes between notes are uniform; α2 is used to weigh the influence of pitch on emotional change, and its value range is [0,1]; β2 is used to weigh the influence of volume on emotional change, and its value range is [0,1]; γ1 is used to determine the starting weight of time decay, and its value range is (0,1]; δ1 affects the decay rate of the duration on emotional change, and its value range is (0,1]; λ is used to adjust the synthesis method of pitch changes in different dimensions, determining the sensitivity of the calculation result, and its value range is [0.5,3]; ∈ is used to control the decay amplitude of volume change, and its value range is [0,1]; μ is used to control the decay speed of volume change, and its value range is [0,1]; κ1 is used to control the amplitude of time correlation, determining the influence degree of time interval on melody coherence, and its value range is [0.5,1]; μ1 is used to control the exponential decay rate of time interval, determining the influence degree of long time interval on melody temporal correlation, and its value range is [0.1,5]; ζ is used to overall adjust the amplitude of energy calculation, preventing the energy growth value from being too large or too small, and its value range is [0.1,2]; η1 is used to control the influence degree of pitch on melody energy growth, determining the amplification effect of energy growth in the high pitch part, and its value range is [0,0.1]; The calculation method for adjusting the isolation between notes is such that the calculation result is more in line with the human ear's perception of the smoothness of the melody, and the value range is [0.5, 2]; by analyzing features such as the emotional fluctuations, temporal correlations, pitch changes, and note intervals of the melody, it can optimize speech recognition, speech synthesis, music generation and analysis, improve the emotional perception ability, pitch naturalness, and music style recognition accuracy, and then trigger the game plot, enhancing the player experience and immersion;

[0074] The silence score is calculated based on the instantaneous volume, short-term energy change rate, average background noise, spectral analysis, instantaneous volume entropy, life rhythm perception, instantaneous sound sparsity, and silence stability index; the calculation logic of the silence score is: Among them, S(t) is the silence score at the current time t, α1 is the volume attenuation coefficient, β1 is the energy change weight, γ is the balance factor, η is the noise suppression threshold, and δ is the noise influence coefficient; specifically, the silence score can reflect the "silent" degree of the audio at a specific time point and evaluate the balance between the silent period and the dynamic period in the audio; α1 is used to determine the speed of volume attenuation, and the value range is [0, 1]; β1 is used to reflect the influence of energy change on the score, and the value range is [0, 1]; γ is used to balance different terms in the model, and the value range is [0, 10]; η is used to control the influence of background noise on the silence score, and the value range is [0, 10]; δ is used to control the influence of noise on the silence score, and the value range is [0, 1]; through the silence score, it is possible to identify when in a silent state, which helps to improve the accuracy of speech recognition, reduce the interference of background noise, and enhance the immersion in the virtual world;

[0075] The truth score is calculated based on the speech sharpness index, speech consistency index, emotional reverberation effect, sound space oscillation index, intention prediction oscillation, and interaction rhythm imbalance index; the calculation logic of the truth score is: LS(t) = (w1·SUI(t) + w2·SCI(t) + w3·ERE(t) + w4·VSI(t) + w5·IPOt + w6·IRIt·1 + VtVmax, where LSt is the truth score at the current time t, Vt is the current emotional fluctuation amplitude of the player, V maxis the maximum emotional fluctuation value, and w1, w2, w3, w4, w5, and w6 are weight coefficients; specifically, the true word score evaluates the emotional intensity and interaction effect at the current moment by synthesizing various voice and emotional features; w1, w2, w3, w4, w5, and w6 are used to adjust the contributions of the voice sharpness index, voice consistency index, emotional reverberation effect, sound space oscillation index, intention prediction oscillation, and interaction rhythm imbalance index to the true word score, with a value range of [0, 1] and the sum of the weights being 1; in the game, the emotional changes of the character's voice can be dynamically adjusted according to the true word score, so as to give different game plot directions and enhance the immersive interaction experience between the character and the player;

[0076] The melody dynamic score is calculated based on dynamic emotion regulation, pitch dimension expansion, amplitude volatility, melody time series correlation, melody energy growth rate, and note interval isolation; the calculation logic of the melody dynamic score is: OEI(t) = A1·DE(t) + A2·PD(t) + A3·AF(t) + A4·MTC(t) + A5·MEGR(t) + A6·NID(t), where OEI(t) is the melody dynamic score at the current time t, and A1, A2, A3, A4, A5, and A6 are weight coefficients; specifically, the melody dynamic score evaluates the dynamic changes of music or voice through the calculation of multiple melody and emotional features; A1, A2, A3, A4, A5, and A6 are used to determine the contributions of dynamic emotion regulation, pitch dimension expansion, amplitude volatility, melody time series correlation, melody energy growth rate, and note interval isolation to the melody dynamic score, with a value range of [0, 1] and the sum of the weights being 1; by combining various indicators to comprehensively evaluate the player's singing or melody, it plays a role in promoting the triggering of the game plot.

[0077] The present invention is further set as, the calculation logic of the comprehensive score is:

[0078]

[0079] where CS(t) is the comprehensive score at the current time t; S(t) is the silence score at the current time t, LS(t) is the true word score at the current time t, OEI(t) is the melody dynamic score at the current time t, S(t)0 is the silence score reference value, ε0, ε1, and ε2 are weight coefficients, λ s , θ t , α m and δ mis a non - linear adjustment parameter; specifically, the comprehensive score CS(t) measures the overall dynamic performance of music or speech by comprehensively calculating the silence score, truth score, and melody dynamic score; S(t) is used to measure the silence characteristics of the audio; LS(t) is used to measure the emotional coherence and expression strength of the speech; OEI(t) is used to measure the fluency, rhythm, and emotional expression of the audio melody; ε0, ε1, and ε2 are used to determine the contribution of each part to the comprehensive score, and each part includes the silence score adjustment item truth score adjustment item ε1·tanh(θ t ·LS(t)) and the dynamic melody score adjustment item The value range is [0,1], and the sum of the weights is 1; λ s is used to determine the influence intensity of the silence score, and the value range is [0.1,5]; θ t is used to control the non - linear growth rate, and the value range is [0.5,3]; α m is used to adjust the weight of the melody dynamic score, and the value range is [0.1,2]; δ m is used to adjust the non - linear degree of the melody dynamic score, and the value range is [0.5,2]; It can automatically adjust the comprehensive score according to different game scenarios and provide more interactive feedback that conforms to the player experience.

[0080] The present invention is further set as follows: the calculation logic for adjusting the scenario atmosphere is:

[0081] where E is the scenario atmosphere, α e and β e are the reflection coefficients for adjusting the tension of the silence score on the environment, θ s and γ s are the influence coefficients for adjusting the truth score on the scenario atmosphere, δ e and φ m are the influence coefficients for controlling the melody dynamic score on the scenario atmosphere, ω1 is the periodic fluctuation adjustment coefficient of the melody dynamic score on the environment; specifically, the scenario atmosphere reflects the tension, immersion, or rhythm change of the current scene through the comprehensive action of the silence score, truth score, and melody dynamic score. When the value of E is within the range of [0,5], the scenario atmosphere is adjusted as follows: the environment becomes calmer and more soothing, reducing stimulating elements or presenting a more gentle environment, and there is a 10% probability of obtaining a high - value reward. When the value of E is within the range of [5,10], the scenario atmosphere is adjusted as follows: the environment in the game or interaction becomes more tense, intense, or challenging, and there is a 75% probability of obtaining a high - value reward; α e is used to set the influence degree of the silence score on the overall atmosphere, and the value range is [0.1,3]; β eIt is used to control the sensitivity of the atmosphere change when the silence score deviates from the benchmark value S(t)0, and the value range is [0.5, 5]; θ s It is used to set the basic influence intensity of the true word score on the atmosphere, and the value range is [0.1, 2]; γ s It is used to control the non-linear volatility of the true word score, and the value range is [0.5, 3]; δ e It is used to set the influence intensity of the melody dynamic score on the atmosphere, and the value range is [0.1, 2]; φ m It is used to control the non-linear degree of the melody dynamic score, and the value range is [0.5, 3]; ω1 is used to control the periodic fluctuation of the atmosphere, and the value range is [0.1, 2π].

[0082] The present invention is further configured to adjust the calculation logic of the NPC dialogue behavior as follows:

[0083] Among them, R NPC is the NPC dialogue behavior, ω2 is the adjustment factor of the silence score on the NPC response period, α t and β t are the coefficients for controlling the true word score on the NPC response behavior, γ m and δ t are the coefficients for adjusting the emotional response degree of the melody dynamic score on the NPC; specifically, the NPC dialogue behavior is adjusted according to the silence score, the true word score, and the melody dynamic score, so that the dialogue behavior of the NPC in different situations is more intelligent and vivid. When the value of R NPC is in the range of (0, 0.3], the NPC tone is cold or mechanical, the reply is short and lacks emotion, and the interaction response delay to the player increases. There is a 10% probability of triggering a plot mission. When the value of R NPC is in the range of [0.3, 0.7], the NPC dialogue behavior is adjusted to: the tone is natural, the dialogue interaction is normal, using complete sentences to reply and showing certain emotional fluctuations. There is a 30% probability of triggering a plot mission. When the value of R NPC is greater than 0.7, the NPC dialogue behavior is adjusted to: the tone is enthusiastic, and the dialogue is actively extended or additional questions are proposed. There is a 55% probability of triggering a plot mission; ω2 is used to control the adjustment of the silence score on the NPC reaction rhythm, and the value range is [0.1, 5]; α t is used to control the influence degree of the NPC's basic response to voice changes, and the value range is [0.1, 3]; β t is used to adjust the adaptability of the NPC to sharp voice changes, and the value range is [0.5, 5]; γ m is used to set the sensitivity degree of the NPC to melody fluctuations, and the value range is [0.1, 2]; δ tUsed to control the degree of nonlinear response of NPC to melody changes, the value range is [0.5,3].

[0084] The present invention is further configured to adjust the calculation logic of the task progress reward mechanism to: Among them, T reward To adjust the task progress reward mechanism, μ t and t To adjust the bonus effect coefficient of the truth score on the reward, θ m To adjust the influence coefficient of melody score, α a and δ a To adjust the nonlinear influence coefficient of the melody dynamic score, λ a To adjust the influence coefficient of silence score on task rewards; specifically, the task progress reward mechanism dynamically adjusts the reward value according to the changes in the truth score, melody dynamic score and silence score, making the reward mechanism more intelligent and adapting to the player's interactive performance. reward When the value of is in the range of [0,10], the task progress reward mechanism is adjusted as follows: the task is less challenging, and experience value rewards are given. reward When the value of is in the range of [0,50], the task progress reward mechanism is adjusted to: the task is neither too challenging nor too challenging, and game props are given as rewards. reward When the value of exceeds 50, the task progress reward mechanism is adjusted as follows: the task is more challenging, and the game progress reward is given; t It is used to control the basic gain of truth score on rewards, with a value range of [0.1,2]; ζ t The nonlinear growth used to determine the truth score, with a value range of [0.5, 3]; θ m It is used to set the basic influence of the melody score on the reward, and the value range is [0.5,3]; α a It is used to control the influence of melody score, and the value range is [0.1,5]; δ a It is used to control the nonlinear adjustment influence of melody score, and the value range is [0.5,3]; a Used to determine the penalty of silence score on reward, the value range is [0.1,2].

[0085] Embodiment 2

[0086] See also Figure 2 , the exemplary game content customization system based on user behavior includes:

[0087] Data acquisition module: used to obtain audio data and player behavior data and perform preprocessing;

[0088] Feature construction module: used to construct environmental surge features, intention ripple features, and sonic melody features based on the preprocessed audio data and player behavior data;

[0089] First calculation module: used to calculate a silence score based on the environmental surge features, a truth score based on the intention ripple features, and a melody dynamics score based on the sonic melody features;

[0090] Second calculation module: used to fuse the silence score, the truth score, and the melody dynamics score to construct a comprehensive score;

[0091] Adjustment module: used to dynamically adjust the game scenario atmosphere, NPC dialogue behavior, and task progress reward mechanism based on the comprehensive score and the silence score, the truth score, and the melody dynamics score.

[0092] It should be noted that a game content customization system based on user behavior provided in the above embodiments and a game content customization method based on user behavior provided in the above embodiments belong to the same concept. The specific ways in which each module and unit perform operations have been described in detail in the method embodiments, and will not be repeated here. A game content customization system based on user behavior provided in the above embodiments can, in practical applications, allocate the above functions to different functional modules as needed, that is, divide the internal structure of the system into different functional modules to complete all or part of the functions described above. This is not limited here either.

[0093] The above embodiments can be implemented in whole or in part by software, hardware, firmware, or any combination thereof. When implemented using software, the above embodiments can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, the processes or functions described in the embodiments of the present application are generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center by wire (such as infrared, wireless, microwave, etc.). The computer-readable storage medium can be any available medium that the computer can access or a data storage device such as a server or data center that includes one or more collections of available media. The available media can be magnetic media (such as floppy disks, hard disks, magnetic tapes), optical media (such as DVDs), or semiconductor media. The semiconductor media can be a solid-state drive.

[0094] It should be understood that the term "and / or" in this text is merely a description of the association relationship of associated objects, indicating that there can be three relationships. For example, A and / or B can represent three situations: A exists alone, A and B exist simultaneously, and B exists alone. Here, A and B can be singular or plural. Additionally, the character " / " in this text generally represents an "or" relationship between the associated objects before and after, but it may also represent an "and / or" relationship. Specifically, it can be understood by referring to the context before and after.

[0095] In this application, "at least one" means one or more, and "a plurality" means two or more. "At least one of the following items (pieces)" or its similar expressions refer to any combination of these items, including any combination of single items (pieces) or plural items (pieces). For example, at least one of a, b, or c can represent: a, b, c, a - b, a - c, b - c, or a - b - c, where a, b, and c can be single or multiple.

[0096] It should be understood that in various embodiments of this application, the magnitudes of the serial numbers of the above processes do not mean the sequence of execution. The execution sequence of each process should be determined by its function and internal logic, and should not constitute any limitation to the implementation process of the embodiments of this application.

[0097] Those of ordinary skill in the art can realize that the units and algorithm steps of each example described in combination with the embodiments disclosed in this text can be implemented by electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are executed in a hardware or software manner depends on the specific application and design constraints of the technical solution. Professional technicians can use different methods for each specific application to implement the described functions, but such implementation should not be considered to exceed the scope of this application.

[0098] Those skilled in the art can clearly understand that for the convenience and conciseness of description, the specific working processes of the systems, devices, and units described above can refer to the corresponding processes in the foregoing method embodiments, and will not be elaborated herein.

[0099] In several embodiments provided in this application, it should be understood that the disclosed system can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For example, the division of the units is only a logical function division. In actual implementation, there can be other division methods. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the displayed or discussed coupling or direct coupling or communication connection to each other can be through some interfaces. The indirect coupling or communication connection of devices or units can be in an electrical, mechanical, or other form.

[0100] The unit described as a separation component may or may not be physically separated. The component shown as a unit may or may not be a physical unit, that is, it may be located in one place or may be distributed over multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0101] In addition, each functional unit in various embodiments of the present application may be integrated in a processing unit, may exist physically alone for each unit, or two or more units may be integrated in one unit.

[0102] If the above-mentioned function is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on such an understanding, the technical solution of the present application, in essence, or the part that contributes to the prior art or a part of this technical solution can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in various embodiments of the present application. The foregoing storage medium includes: various media such as USB flash drives, mobile hard disks, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical discs that can store program codes.

[0103] As described above, the above are only the specific implementation manners of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present application can easily think of changes or substitutions, which should all be covered by the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claimed rights.

Claims

1. A method for customizing game content based on user behavior, characterized in that: include: Obtain audio data and player behavior data and perform preprocessing; Construct environmental surge features, intention ripple features, and sound wave melody features based on pre-processed audio data and player behavior data; The silence score is calculated based on the environmental surge characteristics, the truth score is calculated based on the intention ripple characteristics, and the melody dynamic score is calculated based on the sound wave melody characteristics; Combine the silence score, mantra score and melody dynamic score to construct a comprehensive score; Dynamically adjust the game situation atmosphere, NPC dialogue behavior and task progress reward mechanism based on the comprehensive score as well as the silence score, truth score and melody dynamic score.

2. A method for customizing game content based on user behavior according to claim 1, characterized in that: Environmental surge characteristics include instantaneous volume, short-term energy change rate, background noise average, spectrum analysis, instantaneous entropy of volume, life rhythm perception, instantaneous sound sparseness and silence stability index; The intention ripple features include voice acuity index, voice consistency index, emotional convolution effect, sound space oscillation index, intention prediction oscillation and interaction rhythm imbalance index; Acoustic melodic features include dynamic affective modulation, pitch dimension extension, amplitude volatility, melodic temporal correlation, melodic energy growth rate, and inter-note isolation; The silence score was calculated based on instantaneous volume, short-term energy change rate, background noise average, spectrum analysis, instantaneous entropy of volume, life rhythm perception, instantaneous sound sparseness and silence stability index; The true speech score was calculated based on the voice acuity index, voice consistency index, emotional convolution effect, sound space oscillation index, intention prediction oscillation and interaction rhythm imbalance index; The melodic dynamics score was calculated based on dynamic emotional regulation, pitch dimension expansion, amplitude volatility, melodic timing correlation, melodic energy growth rate and inter-note isolation.

3. A method for customizing game content based on user behavior according to claim 2, characterized in that: The calculation logic of instantaneous volume is: Where L(t) is the instantaneous volume at time point t, N is the sampling window size, and s i is the sampling value of the ith audio sample. Based on the above calculation logic, L(t) is smoothed to obtain The calculation logic is: in, is the volume after smoothing, α is the smoothing coefficient; The calculation logic of short-time energy change rate is: Among them, EV(t) is the short-term energy change rate at the current time t; The calculation logic of the background noise average is: BNL(t) = β·BNL(t-1) + (1-β)·L env (t), where BNL(t) is the background noise at the current time t, BNL(t-1) is the background noise at the previous time, and L env (t) is the ambient noise measured at the current time t, and β is the weighting factor; The calculation logic of spectrum analysis is: Where HF(t) is the frequency spectrum analysis at time point t, P(f) is the power spectrum density at frequency f, and f max is the maximum frequency; The calculation logic of volume instantaneous entropy is: Among them, LE(t) is the instantaneous entropy of the volume at the current time t, G i is the probability distribution of the i-th sampling volume point, G i The calculation logic is: Among them, L i is the volume sampling value, The calculation logic of life rhythm perception is: Among them, BPI(t) is the life rhythm perception at the current time t; The calculation logic of the silencing stability index is: Among them, SSI(t) is the silence stability index at the current time t, T is the historical time length of the investigation, and κ is the sensitivity adjustment parameter; The calculation logic of silence score is: Among them, S(t) is the silence score at the current time t, α1 is the volume attenuation coefficient, β1 is the energy change weight, γ is the balance factor, η is the noise suppression threshold, and δ is the noise influence coefficient.

4. The method for customizing game content based on user behavior according to claim 2, characterized in that: The calculation logic of speech acuity index is: Among them, SUI(t) is the speech acuteness index at the current time t, P i is the pitch of the i-th sampling point, PD i is the pause time of each sampling point i, SR j is the value of speech rate at the jth moment, U is the number of sampling points of speech data in a given time window, P max For maximum volume, PD max is the maximum pause time, SR max is the maximum speech rate; The calculation logic of the speech consistency index is: Wherein, SCI(t) is the speech consistency index at the current time t; The calculation logic of the emotional convolution effect is: Where ERE(t) is the emotional reverberation effect at the current time t, R is the number of sampling points of emotional intensity or pitch and volume changes within a given time window, and S i is the i-th emotion intensity sample, PD j is the pause time of the jth sampling point; The calculation logic of the sound space oscillation index is: Among them, VSI(t) is the sound space oscillation index at the current time t, V i is the volume of the i-th audio sample, P min is the minimum value of the pitch, V min is the minimum value of the volume; The calculation logic of the intention prediction oscillation is: Where IPO(t) is the intention prediction oscillation at the current time t, O is the number of time segment sampling points in the voice interaction between the player and the NPC, and EI i is the emotional intensity at the i-th moment, P target is the target speech frequency; The calculation logic of the interaction rhythm imbalance index is: Among them, IRI(t) is the interaction rhythm imbalance index at the current time t, SR i For the player's speech speed, SR NPC NPC speech speed, PD NPC Pause time for NPC; The calculation logic of the truth score is: LS(t) = (w1·SUI(t)+w2·SCI(t)+w3·EREt+w4·VSIt+w5·IPOt+w6·IRIt·1+VtVmax, where LSt is the truth score at the current time t, V(t) is the current emotional fluctuation amplitude of the player, and V max is the maximum emotion fluctuation value, w1, w2, w3, w4, w5 and w6 are weight coefficients.

5. The method for customizing game content based on user behavior according to claim 2, characterized in that: The calculation logic of dynamic emotion regulation is: Where DEt is the dynamic emotional adjustment at the current time t, Y is the number of notes in a given time window, Δf i is the pitch change amplitude, Δv i is the volume change amplitude, τ i is the duration of the i-th note, α2, β2, γ1 and δ1 are adjustment parameters; The calculation logic of pitch dimension expansion is: Among them, PD(t) is the pitch dimension expansion at the current time t, is the pitch value of the i-th note in the k-th dimension, m is the number of dimensions of the pitch space, and λ is the nonlinear control factor; The calculation logic of amplitude volatility is: Among them, AF(t) is the amplitude fluctuation at the current time t, σ(f i ) is the nonlinear transformation function of pitch, ∈ and μ are adjustment parameters; The calculation logic of melody timing correlation is: Among them, MTCt is the melody timing correlation at the current time t, υi is the starting time of the i-th note, κ1 and μ1 are parameters for controlling the induction of timing differences; The calculation logic of the melody energy growth rate is: Among them, MEGRt is the melody energy growth rate at the current time t, ζ and η1 are energy growth control parameters; The calculation logic of the isolation between notes is: Among them, NID(t) is the isolation between notes at the current time t, is the nonlinear transformation parameter of the interval; The calculation logic of the melody dynamic score is: OEI(t)=A1·DE(t)+A2·PD(t)+A3·AF(t)+A4·MTC(t)+A5·MEGR(t)+A6·NID(t), where OEI(t) is the melody dynamic score at the current time t, and A1, A2, A3, A4, A5 and A6 are weight coefficients.

6. The method for customizing game content based on user behavior according to claim 1, characterized in that: The calculation logic of the comprehensive score is: Among them, CS(t) is the comprehensive score at the current time t; S(t) is the silence score at the current time t, LS(t) is the truth score at the current time t, OEI(t) is the melody dynamic score at the current time t, S(t)0 is the silence score benchmark value, ε0, ε1 and ε2 are weight coefficients, λ s ,θ t , α m and δ m is the nonlinear adjustment parameter.

7. The method for customizing game content based on user behavior according to claim 1, characterized in that: The calculation logic for adjusting the scene atmosphere is: Among them, E is the scene atmosphere, α e and β e To adjust the silence score's response coefficient to the environment's tension, θ s and γ s To adjust the influence coefficient of truth score on the environmental atmosphere, δ e and φ m is the coefficient for controlling the influence of melody dynamic score on environmental atmosphere, and ω1 is the adjustment coefficient of melody dynamic score on the periodic fluctuation of environment.

8. The method for customizing game content based on user behavior according to claim 1, characterized in that: Adjust the calculation logic of NPC dialogue behavior to: R NPC =(1-cos(ω2·(S(t)-St0·αt·log1+βt·LSt+γm·1-e-δm·OEIt, where RNPC is the NPC dialogue behavior, ω2 is the silence score adjustment factor for the NPC response period, α t and β t To control the coefficient of the truth score on the NPC's behavior, γ m and δ t To adjust the emotional responsiveness coefficient of the NPC to the melody dynamic score.

9. The method for customizing game content based on user behavior according to claim 1, characterized in that: Adjust the calculation logic of the task progress reward mechanism to: Among them, T reward To adjust the task progress reward mechanism, μ t and t To adjust the bonus effect coefficient of the truth score on the reward, θ m To adjust the influence coefficient of melody score, α a and δ a To adjust the nonlinear influence coefficient of the melody dynamic score, λ a To adjust the influence coefficient of silence score on task rewards.

10. A game content customization system based on user behavior, used to implement a game content customization method based on user behavior as claimed in any one of claims 1 to 9, characterized in that: include: Data acquisition module: used to obtain audio data and player behavior data and perform preprocessing; Feature construction module: used to construct environmental surge features, intention ripple features and sound wave melody features based on pre-processed audio data and player behavior data; The first calculation module is used to calculate the silence score according to the environmental surge characteristics, calculate the truth score according to the intention ripple characteristics, and calculate the melody dynamic score according to the sound wave melody characteristics; The second calculation module is used to integrate the silence score, the truth score and the melody dynamic score to construct a comprehensive score; Adjustment module: used to dynamically adjust the game situation atmosphere, NPC dialogue behavior and task progress reward mechanism based on the comprehensive score as well as the silence score, truth score and melody dynamic score.

Images