Coffee making parameter decision method and system based on user interaction habits

Abstract

The application discloses a coffee making parameter decision method and system based on user interaction habits, relates to the technical field of data processing, and comprises the following steps: calling implicit interaction behavior data set of a target user; performing user state extraction to establish a dynamic user state vector; querying a historical brewing database, performing similarity matching of the query result, and establishing a preferred brewing parameter set; using the dynamic user state vector to perform general brewing portrait matching to establish a prior reference; adaptively fusing the preferred brewing parameter set and the prior reference to output a target brewing parameter; and configuring a decision display area and performing visual display of the target brewing parameter on a display screen. The application solves the technical problem that, in the prior art, coffee making parameters depend on user explicit setting and personalized brewing parameters cannot be automatically generated according to the current state of the user, and achieves the technical effects of automatically outputting personalized coffee making parameters under the condition that the user does not perform explicit setting, and improving the adaptability of brewing parameters and the convenience of interaction.

Description

Technical Field

[0001] This invention relates to the field of data processing technology, and specifically to a method and system for making coffee-making parameters based on user interaction habits. Background Technology

[0002] As coffee machines become increasingly intelligent, users can typically adjust brewing parameters such as coffee concentration, water volume, temperature, and grind size via displays, buttons, or mobile devices to meet different drinking preferences. However, since selecting brewing parameters often requires manual input or selection from preset modes, and user needs can change depending on time, operating status, behavioral rhythms, and environmental conditions, relying solely on fixed presets or manual adjustments cannot accurately reflect the user's current state, resulting in a mismatch between coffee brewing parameters and the user's immediate preferences. Summary of the Invention

[0003] This application provides a coffee brewing parameter decision-making method and system based on user interaction habits, which is used to address the technical problem in the prior art that coffee brewing parameters rely on user-displayed settings and are difficult to automatically generate personalized brewing parameters based on the user's current state.

[0004] In view of the above problems, this application provides a method and system for making coffee brewing parameters based on user interaction habits.

[0005] The first aspect of this application provides a coffee-making parameter decision-making method based on user interaction habits, the method comprising: In response to a user's non-explicit setting command, the system invokes the target user's implicit interaction behavior dataset, which includes operational dynamic features, digital behavior rhythm records, and environmental features. User state is extracted from the implicit interaction behavior dataset to establish a dynamic user state vector. The system queries the target user's historical brewing database and uses the dynamic user state vector to perform similarity matching of the query results, establishing a preferred brewing parameter set. General brewing profile matching is performed using the dynamic user state vector to establish a priori reference. The preferred brewing parameter set and priori reference are adaptively fused to output at least two target brewing parameters. A decision display area mapped to the priority of the target brewing parameters is configured, and the target brewing parameters are then visualized on the display screen.

[0006] A second aspect of this application provides a coffee-making parameter decision-making system based on user interaction habits, the system comprising: The system comprises the following modules: a data retrieval module, activated in response to a user's non-explicit setting command, which retrieves the target user's implicit interaction behavior dataset, including operational dynamic features, digital behavior rhythm records, and environmental features; a state extraction module, used to extract the user's state from the implicit interaction behavior dataset and establish a dynamic user state vector; a similarity matching module, used to query the target user's historical brewing database and perform similarity matching of the query results using the dynamic user state vector to establish a preferred brewing parameter set; a profile matching module, used to perform general brewing profile matching using the dynamic user state vector to establish a priori reference; a fusion module, used to adaptively fuse the preferred brewing parameter set and the priori reference to output at least two target brewing parameters; and a visualization module, used to configure a decision display area mapped to the priority of the target brewing parameters and to perform a visual display of the target brewing parameters on the screen.

[0007] One or more technical solutions provided in this application have at least the following technical effects or advantages: This application responds to a user's non-explicit setting command by invoking a dataset of implicit interaction behaviors of the target user. This dataset includes dynamic operational features, digital behavior rhythm records, and environmental features. User state is extracted from the dataset to establish a dynamic user state vector. The target user's historical brewing database is queried, and the dynamic user state vector is used to perform similarity matching of the query results to establish a set of preferred brewing parameters. The dynamic user state vector is then used to perform general brewing profile matching to establish a priori reference. The preferred brewing parameter set and the priori reference are adaptively fused to output at least two target brewing parameters. A decision display area mapped to the priority of the target brewing parameters is configured, and the target brewing parameters are then visualized on the display screen. This invention solves the technical problem in existing technologies where coffee brewing parameters rely on explicit user settings, making it difficult to automatically generate personalized brewing parameters based on the user's current state. By extracting a dynamic user state vector based on implicit user interaction behaviors and fusing historical preferences with a general brewing profile for parameter decision-making, it achieves the technical effect of automatically outputting personalized coffee brewing parameters under non-explicit user settings, improving the adaptability of brewing parameters and the convenience of interaction. Attached Figure Description

[0008] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0009] Figure 1A schematic diagram of a coffee-making parameter decision-making method based on user interaction habits provided in an embodiment of this application; Figure 2 A schematic diagram of the compensation analysis results of the initial hit record set of the coffee making parameter decision method based on user interaction habits provided in the embodiments of this application; Figure 3 A schematic diagram of the coffee making parameter decision system based on user interaction habits provided in this application embodiment.

[0010] Figure labeling: Data retrieval module 11, Status extraction module 12, Similarity matching module 13, Profile matching module 14, Fusion module 15, Visualization module 16. Detailed Implementation

[0011] This application provides a coffee brewing parameter decision-making method and system based on user interaction habits. It addresses the technical problem in existing technologies where coffee brewing parameters rely on explicit user settings and it is difficult to automatically generate personalized brewing parameters based on the user's current state. By extracting dynamic user state vectors based on implicit user interaction behavior and integrating historical preferences with general brewing profiles for parameter decision-making, it achieves the technical effect of automatically outputting personalized coffee brewing parameters under conditions where the user does not explicitly set them, thereby improving the adaptability of brewing parameters and the convenience of interaction.

[0012] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application.

[0013] It should be noted that any variation of the terms "comprising" and "having" is intended to cover non-exclusive inclusion, for example, a process, method, system, product, or server that includes a series of steps or units is not necessarily limited to those steps or units that are explicitly listed, but may include other steps or modules that are not explicitly listed or that are inherent to such processes, methods, products, or devices.

[0014] Example 1, as Figure 1 As shown, this application provides a coffee-making parameter decision-making method based on user interaction habits, the method comprising: Step S100: In response to the user's non-explicit setting command, the implicit interaction behavior dataset of the target user is invoked, which includes operation dynamic features, digital behavior rhythm records and environmental features.

[0015] Furthermore, the method provided in the application embodiments also includes: The operational dynamic features include at least one of the following: the user's operating speed, operating force, and operating rhythm during the operation of the controls on the coffee machine. The digital behavior rhythm record includes at least one of the following: the number of times the alarm clock snoozes, the density of daily tasks, and the text input error rate obtained based on the smart terminal. The environmental features include at least one of the following: environmental audio style, lighting conditions, and meteorological parameters.

[0016] In this embodiment, after receiving brewing-related input from the user, the coffee machine reads the corresponding operation content. If the operation only includes starting brewing, automatic brewing, quick brewing, or confirming brewing, but does not include brewing parameters directly set by the user such as water volume, temperature, concentration, grind size, and extraction time, then the operation is determined to be a non-explicit setting instruction from the user. After the coffee machine starts responding to the user's non-explicit setting instruction, it reads the currently logged-in account, the bound smart terminal, near-field communication results, or the user's corresponding record stored in the coffee machine to determine the target user for this brewing. Using the target user as the retrieval object, the coffee machine searches for data records corresponding to the target user in local storage, cloud account data, or smart terminals associated with the target user, and calls the target user's implicit interaction behavior dataset. The implicit interaction behavior dataset is a set of interaction-related data generated by the target user when brewing parameters are not directly set. The implicit interaction behavior dataset includes operation dynamic features, digital behavior rhythm records, and environmental features.

[0017] To obtain operational dynamic characteristics, the coffee machine collects control response data during user operation of the controls on the machine. These controls include at least one of buttons, knobs, and a touchscreen. When a user triggers a control, the coffee machine records the trigger start time, trigger end time, number of consecutive triggers, touch coordinate changes, knob rotation angle, button press depth, pressure detection value, and the time interval between two adjacent control triggers. The coffee machine subtracts the trigger end time from the trigger start time to obtain the duration of a single trigger, and calculates the ratio of touch coordinate changes, knob rotation angle, or press travel to the duration of a single trigger to form the operation speed. The coffee machine reads the pressure detection value, button press depth, touch pressure value, or knob rotation resistance, and categorizes them into the same operation process to form the operation force. The coffee machine arranges consecutive trigger records according to the collection time, calculates the time interval between adjacent triggers, pause duration, number of consecutive triggers, and repetition frequency to form the operation rhythm. Therefore, the operational dynamic characteristics include at least one of the following: operation speed, operation force, and operation rhythm during the user's operation of the controls on the coffee machine.

[0018] When the coffee machine acquires digital behavioral rhythm records, it establishes a data connection with the smart terminal associated with the target user and reads alarm clock records, schedule records, and text input records within the authorized scope. For alarm clock records, the coffee machine counts the number of times the alarm clock delayed closing, repeated reminders, or re-wake-up within a preset time period, forming the alarm snooze count. For schedule records, the coffee machine counts the number of schedule items within a preset time period, the duration of each schedule item, the time interval between adjacent schedule items, and the number of overlapping schedule items, forming the schedule item density. For text input records, the coffee machine counts the number of deletions, modifications, re-entries, or accidental touches during text input and calculates the ratio with the total number of inputs, forming the text input error rate. Therefore, the digital behavioral rhythm record includes at least one of the alarm snooze count, schedule item density, and text input error rate acquired based on the smart terminal. When acquiring environmental characteristics, the coffee machine collects ambient sound through a microphone and records the intensity, frequency distribution, or sound type of the ambient sound, forming the ambient audio style. The coffee machine collects ambient brightness, color temperature, or light change through a light detection component, forming the lighting conditions. The coffee machine reads temperature, humidity, weather type, air pressure or precipitation data through temperature and humidity detection components, air pressure detection components, smart terminals or meteorological data interfaces to form meteorological parameters.

[0019] Therefore, environmental characteristics include at least one of ambient audio style, lighting conditions, and meteorological parameters. The coffee machine stores the operational dynamic characteristics, digital behavioral rhythm records, and environmental characteristics according to the target user and the collection time, and retrieves them when activated in response to a non-explicit setting command from the user.

[0020] Step S200: Extract user states from the implicit interaction behavior dataset and establish a dynamic user state vector.

[0021] In this embodiment, when extracting user states from the implicit interaction behavior dataset, the coffee machine reads the dynamic features of the same target user's operations, digital behavior rhythm records, and environmental features within the time range corresponding to this startup, and performs numerical processing on each data item; operation speed, operation force, and operation rhythm are respectively processed into corresponding operation state components; alarm clock snooze count, schedule item density, and text input error rate are respectively processed into corresponding behavior rhythm state components; and environmental audio style, lighting conditions, and meteorological parameters are respectively processed into corresponding environmental state components; then, each state component is written into the same data sequence in a preset order, and associated with the target user and the collection time to establish a dynamic user state vector.

[0022] Step S300: Query the target user's historical brewing database, use the dynamic user state vector to perform similarity matching of the query results, and establish a set of preferred brewing parameters.

[0023] In this embodiment, when querying the target user's historical brewing database, the coffee machine uses the target user as the retrieval object and reads the historical brewing database corresponding to the target user from local storage or cloud account data. The historical brewing database stores the target user's historical user status, historical brewing parameters, and brewing time during the target user's historical brewing process. When performing similarity matching of query results using dynamic user status vectors, the dynamic user status vectors are used as the overall query conditions, and multi-dimensional overall similarity retrieval is performed in the historical brewing database to establish an initial set of hit records. Then, each independent state component in the dynamic user status vector is used as an independent matching query sub-vector, and the independent matching query sub-vectors are used to evaluate the component criticality and establish a criticality score. Subsequently, the criticality score and the independent matching query sub-vectors are used to perform compensation analysis on the initial set of hit records to establish a compensation order ranking. The selected hit records with the highest matching degree in the compensation order ranking are selected, and the missing identifiers of the independent matching query sub-vectors are extracted from the selected hit records. The missing identifiers are used to perform parameter fusion correction on the selected hit records, and the corrected brewing parameters are combined into a preferred brewing parameter set.

[0024] Furthermore, in the method provided in the application embodiments, the method of performing similarity matching of query results using the dynamic user state vector to establish a preferred brewing parameter set also includes: Using the user state vector as the overall query condition, a multi-dimensional overall similarity retrieval is performed in the historical brewing database to establish an initial hit record set. Each independent state component in the dynamic user state vector is used as an independent matching query sub-vector, and the component keyness is evaluated using the independent matching query sub-vector to establish a keyness score. The keyness score and the independent matching query sub-vector are used to perform compensation analysis on the initial hit record set to establish a compensation order ranking. The selected hit record with the highest matching degree in the compensation order ranking is selected, and the missing identifier of the independent matching query sub-vector is extracted from the selected hit record. The missing identifier is used to perform parameter fusion correction on the selected hit record to establish a preferred brewing parameter set.

[0025] In this embodiment, when using the user state vector as the overall query condition, historical brewing records corresponding to the target user are read from the historical brewing database. Each historical brewing record includes a historical state vector, historical brewing parameters, and recording time. The current dynamic user state vector is arranged in the same dimensional order as the historical state vector, so that each state component corresponds to the same dimension in the historical state vector. After normalizing the current dynamic user state vector and each historical state vector, each state component is transformed to the same numerical range. Then, the Euclidean distance between the current dynamic user state vector and each historical state vector is calculated, and the Euclidean distance is substituted into the similarity calculation formula to obtain the multidimensional overall similarity, where the similarity is equal to 1 divided by 1 plus the Euclidean distance. The historical brewing records are arranged from high to low according to the multidimensional overall similarity, and a preset number of historical brewing records at the top are selected to establish an initial hit record set.

[0026] After establishing the initial hit record set, each independent state component is extracted one by one from the dynamic user state vector and used as an independent matching query subvector. An independent state component is a single state data item divided according to dimension within the dynamic user state vector, and an independent matching query subvector is query data formed by a single independent state component. When evaluating the component criticality of each independent matching query subvector, the historical value sequence of that independent state component is read from the historical brewing database, along with the historical brewing parameter sequence from the same batch of historical brewing records. The absolute value of the Pearson correlation coefficient between the historical value sequence and the historical brewing parameter sequence is calculated, and this absolute value is used as the criticality score for that independent matching query subvector. After calculating all independent matching query subvectors in sequence, a criticality score corresponding one-to-one with each independent matching query subvector is established.

[0027] Next, compensation analysis of the initial hit record set is performed using key score and independent matching query sub-vectors. In this process, each hit record in the initial hit record set is read sequentially, and the historical state vector in each hit record is split into historical state components according to its dimensions. The component similarity between each independent matching query sub-vector and its corresponding historical state component is calculated, where the component similarity equals 1 divided by 1 plus the absolute value of the normalized difference of the corresponding component. Each component similarity is multiplied by its corresponding key score to obtain the component compensation value. The sum of all component compensation values ​​is then divided by the sum of all key scores to obtain the compensated similarity for that hit record. Preset overall weights and preset compensation weights are read. The preset overall weight corresponds to the multidimensional overall similarity, and the preset compensation weight corresponds to the compensation similarity; the sum of the preset overall weight and the preset compensation weight is 1. The multidimensional overall similarity is multiplied by the preset overall weight, and the compensation similarity is multiplied by the preset compensation weight. The two products are then added together to obtain the compensated matching degree. After calculating the compensated matching degree for all hit records in the initial hit record set, the records are sorted from highest to lowest according to their compensated matching degree to establish a compensation order. For example... Figure 2 As shown, Figure 2 This diagram illustrates the compensation analysis results of the initial hit record set. Hit records A, B, C, and D are used as the four hit records in the initial hit record set. The diagram shows the multidimensional overall similarity, compensated similarity, and compensated matching degree for each hit record. Specifically, hit record A has a multidimensional overall similarity of 0.82, a compensated similarity of 0.78, and a compensated matching degree of 0.804; hit record B has a multidimensional overall similarity of 0.79, a compensated similarity of 0.88, and a compensated matching degree of 0.826; hit record C has a multidimensional overall similarity of 0.75, a compensated similarity of 0.72, and a compensated matching degree of 0.738; and hit record D has a multidimensional overall similarity of 0.69, a compensated similarity of 0.81, and a compensated matching degree of 0.738. Based on the compensated matching score, the matching score of the hit record B is the highest after compensation. Therefore, the hit record B is determined as the selected hit record with the highest matching score in the compensation order.

[0028] After establishing the compensation order, the record with the highest matching degree in the compensation order is selected as the chosen hit record. The chosen hit record is the historical brewing record with the highest matching degree after compensation analysis in the initial hit record set. It includes the historical state vector closest to the current dynamic user state vector, the historical brewing parameters corresponding to the historical state vector, and the recording time. The historical state vector in the chosen hit record is read, and the dimension list of the historical state vector is matched item by item with the independent matching query sub-vectors in the dynamic user state vector. For each independent matching query sub-vector, it is checked whether there is a valid record value for the historical state component of the same dimension in the chosen hit record. If there is no valid record value for the same dimension, the missing value of the independent matching query sub-vector is marked as 1. If there is a valid record value for the same dimension, the missing value of the independent matching query sub-vector is marked as 0. After checking all independent matching query sub-vectors, each missing value is saved and matched with the chosen hit record.

[0029] Finally, the missing identifier is used to perform parameter fusion correction on the selected hit records. In this process, firstly, the missing state component corresponding to the selected hit record is determined based on the missing identifier, and the component-parameter influence mapping model corresponding to the missing state component is retrieved from the general brewing profile database. Then, the current actual value of the missing state component is extracted from the dynamic user state vector, and the historical actual value of the missing state component is extracted from the selected hit record, and the component deviation value between the two is calculated. The component deviation value is input into the component-parameter influence mapping model to obtain the parameter correction offset of each target brewing parameter associated with the missing state component. Then, using the original value of the associated target brewing parameter in the selected hit record as a benchmark, the parameter correction offset is superimposed on the corresponding benchmark value to generate the corrected target brewing parameter value. Finally, all target brewing parameter values ​​are combined to establish a preferred brewing parameter set.

[0030] Furthermore, in the method provided in the application embodiments, the parameter fusion correction for selecting hit records using the missing identifier further includes: Based on the missing state component indicated by the missing identifier, a pre-constructed component-parameter influence mapping model is retrieved from the general brewing profile database. This model characterizes the directional influence of changes in the value of the missing state component on each associated target brewing parameter while keeping other state components within a stable range. The current actual value of the missing state component is extracted from the dynamic user state vector, and the historical actual value is extracted from the selected hit record. A component deviation value is calculated. This deviation value is input into the component-parameter influence mapping model to obtain the parameter correction offset for each target brewing parameter associated with the missing state component. Using the original value of the associated target brewing parameter in the selected hit record as a reference, the parameter correction offset is superimposed on the corresponding reference value to generate the corrected target brewing parameter value. All target brewing parameter values ​​are combined to establish the preferred brewing parameter set.

[0031] In this embodiment, firstly, based on the missing state component indicated by the missing identifier, a pre-constructed component-parameter influence mapping model is retrieved from the general brewing profile database. The missing identifier and the corresponding independent matching query sub-vector are saved. Independent matching query sub-vectors with a missing identifier value of one are read, and the state component corresponding to this independent matching query sub-vector is determined as the missing state component. The general brewing profile database pre-saves sample records between multiple state components and target brewing parameters. Each sample record includes the user state component value, the target brewing parameter value, and the sample time. When constructing the component-parameter influence mapping model, sample records corresponding to the same missing state component are selected, and other state components besides the missing state component are restricted to a stable interval. The stable interval is the range of values ​​between a preset minimum and a preset maximum value for the corresponding state component. For the retained sample records, the missing state components are sorted in ascending order of value, and the differences in the missing state component values ​​and the parameter differences of each associated target brewing parameter are calculated between adjacent samples. The parameter differences are divided by the value differences to obtain the unit influence value of the missing state component on the corresponding associated target brewing parameter. Then, the average of multiple unit influence values ​​of the same associated target brewing parameter is calculated to obtain the directional influence magnitude. The resulting component-parameter influence mapping model includes at least the name of the missing state component, the name of the associated target brewing parameter, the stability interval of other state components, and the directional influence magnitude.

[0032] When invoking the component-parameter influence mapping model, the missing state component name is used as the retrieval field. The model with the same missing state component name is retrieved from the general brewing profile database, along with the names of the associated target brewing parameters and their corresponding directional influence magnitudes stored within the model. The current actual value of the missing state component is extracted from the dynamic user state vector; this current actual value represents the value of the missing state component in the current dynamic user state vector. The historical actual value of the missing state component is extracted from the selected hit records; this historical actual value represents the historical value stored in the selected hit records corresponding to the missing state component. If the current and historical actual values ​​use different numerical scales, they are converted according to the same normalization rule stored in the general brewing profile database to ensure that the current and historical actual values ​​fall within the same value range. Once the numerical unification is complete, the component deviation value is obtained by subtracting the historical actual value from the current actual value.

[0033] When inputting component deviation values ​​into the component-parameter influence mapping model, correction calculations are performed item by item according to the associated target brewing parameters stored in the model. For any associated target brewing parameter, the directional influence amplitude corresponding to that parameter is read, and the component deviation value is multiplied by the directional influence amplitude to obtain the parameter correction offset for that parameter. A positive directional influence amplitude indicates that an increase in the value of the missing state component corresponds to an increase in the associated target brewing parameter; a negative directional influence amplitude indicates that an increase in the value of the missing state component corresponds to a decrease in the associated target brewing parameter. Through the above item-by-item calculations, the parameter correction offsets for each target brewing parameter associated with the missing state component are obtained.

[0034] Finally, using the original values ​​of the associated target brewing parameters in the selected hit records as a baseline, the original value of each associated target brewing parameter in the selected hit records is read as the corresponding baseline value. For each associated target brewing parameter, the parameter correction offset of the associated target brewing parameter is added to the corresponding baseline value to generate the corrected target brewing parameter value. If multiple missing state components point to the same associated target brewing parameter, the parameter correction offset corresponding to each missing state component is calculated separately, and the multiple parameter correction offsets are accumulated to the same baseline value to generate the corrected target brewing parameter value of that associated target brewing parameter. When all associated target brewing parameters have completed the correction calculation, all target brewing parameter values ​​are combined according to parameter name and parameter value to establish a preferred brewing parameter set.

[0035] Furthermore, in the method provided in the application embodiments, the method of performing similarity matching of query results using the dynamic user state vector to establish a preferred brewing parameter set also includes: Establish a time-series decay factor for the historical brewing database, which is used to constrain the untrustworthiness of data as time intervals increase; perform similarity matching of query results under the constraint of the time-series decay factor to establish a set of preferred brewing parameters.

[0036] In this embodiment, when establishing the time-series decay factor for the historical brewing database, the record time of each historical brewing record in the database is read, and the current time when the similarity match of the current query result is read. The record time is subtracted from the current time to obtain the time interval corresponding to the historical brewing record, and the time interval is uniformly converted into days. The time-series decay factor is configured according to a preset time segment: the time-series decay factor for historical brewing records with a time interval of no more than 7 days is set to 1; the time-series decay factor for historical brewing records with a time interval of more than 7 days but no more than 30 days is set to 0.8; the time-series decay factor for historical brewing records with a time interval of more than 30 days but no more than 90 days is set to 0.6; the time-series decay factor for historical brewing records with a time interval of more than 90 days but no more than 180 days is set to 0.4; and the time-series decay factor for historical brewing records with a time interval of more than 180 days is set to 0.2. The corresponding time-series decay factor is saved for each historical brewing record, and the result of subtracting the time-series decay factor from 1 is recorded as the data untrustworthiness, thereby establishing the time-series decay factor for the historical brewing database.

[0037] When performing similarity matching of query results under the constraint of time decay factor, the dynamic user state vector and the historical state vector of each historical brewing record in the historical brewing database are read and arranged in the same dimensional order. For numerical state components, the preset lower limit and preset upper limit of the state component are read, the actual value is subtracted from the preset lower limit, and then divided by the difference between the preset upper limit and the preset lower limit to obtain a normalized value in the range of 0 to 1. For typological state components, the preset encoding table is read, the typological value is converted into the corresponding encoding value, and then converted to the range of 0 to 1 according to the same normalization method. For the i-th historical brewing record, the absolute value of the difference between the normalized state component in the dynamic user state vector and the normalized state component in the historical state vector is calculated item by item. The absolute values ​​of all differences are added together and divided by the number of state components involved in the calculation to obtain the overall difference value. Then, 1 is divided by 1 and the overall difference value is added to obtain the multidimensional overall similarity. The time decay factor corresponding to the i-th historical brewing record is read, and the multidimensional overall similarity is multiplied by the time decay factor to obtain the time-constrained overall similarity.

[0038] When obtaining the overall similarity under temporal constraints, each independent state component in the dynamic user state vector is used as an independent matching query sub-vector, and a key score is calculated. For the j-th independent matching query sub-vector, the historical value sequence of this state component in the historical brewing database is read, and the target brewing parameter sequences in the same batch of historical brewing records are also read. For any target brewing parameter sequence, the average value of the historical value sequence of the state component and the average value of the target brewing parameter sequence are calculated. The average value of each corresponding value is subtracted from the average value to obtain the state component deviation sequence and the parameter deviation sequence. The two deviation values ​​at corresponding positions are multiplied and summed to obtain the numerator. The sum of the square roots of the sum of the squares of the state component deviation values ​​and the sum of the squares of the sum of the squares of the parameter deviation values ​​are multiplied to obtain the denominator. The correlation coefficient is obtained by dividing the numerator by the denominator, and the absolute value of the correlation coefficient is taken. The average of the absolute values ​​of each target brewing parameter is used as the key score of the independent matching query sub-vector. For the i-th historical brewing record, calculate the component similarity between each independent matching query sub-vector and the corresponding historical state component. Multiply the component similarity by the corresponding key score and sum them, then divide by the sum of all key scores to obtain the compensated similarity. Read the preset overall weight and preset compensation weight, multiply the temporal constraint overall similarity by the preset overall weight, multiply the compensated similarity by the preset compensation weight, and add the two products to obtain the temporal constraint compensated matching degree. The compensation order is established from high to low according to the temporal constraint compensation matching degree, and the record with the highest matching degree is selected as the selected hit record. The historical state components corresponding to each independent matching query sub-vector in the selected hit record are read, the missing identifier is extracted, and according to the aforementioned parameter fusion correction process, the component-parameter influence mapping model is retrieved from the general brewing profile database according to the missing state component indicated by the missing identifier. The component deviation value and the parameter correction offset of each target brewing parameter are calculated. The parameter correction offset is superimposed on the original value of the corresponding target brewing parameter in the selected hit record to generate the corrected target brewing parameter value. All target brewing parameter values ​​are combined into a preferred brewing parameter set.

[0039] Step S400: Perform general brewing profile matching using the dynamic user state vector to establish a priori reference.

[0040] In this embodiment, when performing general brewing profile matching using dynamic user state vectors, a pre-built general brewing profile database is invoked to read multiple state-parameter mapping clusters and their corresponding cluster center state vectors, intra-cluster group brewing parameter distribution statistics, and component influence weight coefficients. The dynamic user state vectors are then compared with the cluster center state vectors of each state-parameter mapping cluster using weighted multidimensional similarity calculations, with the component influence weight coefficients of each state component on each brewing parameter used as dimensional weights. State-parameter mapping clusters whose weighted multidimensional similarity meets preset conditions are selected as matching profile clusters. Intra-cluster group brewing parameter distribution statistics of the matching profile clusters are then extracted, and weighted multidimensional similarity is used as a fusion weight to perform weighted fusion of the group brewing parameter statistical distribution, generating a priori reference distribution for each target brewing parameter. Representative values ​​are extracted from the priori reference distribution to form a priori reference.

[0041] Furthermore, the method provided in the application embodiment, which utilizes the dynamic user state vector to perform general brewing profile matching and establish a priori reference, further includes: A pre-built general brewing profile database is invoked, wherein the general brewing profile database contains multiple state-parameter mapping clusters. Each state-parameter mapping cluster is associated with and stores a cluster center state vector, statistical information on the distribution of brewing parameters within the cluster, and component influence weight coefficients of each state component on each brewing parameter. The component influence weight coefficients are constructed by statistical analysis of the correlation strength between the state changes and parameter adjustments of users within the cluster. The dynamic user state vector and the cluster center state vector of each mapping cluster are subjected to weighted multidimensional similarity calculation, in which the component influence weight coefficients of each state component are used as the dimensional weights in the similarity calculation. At least one mapping cluster whose weighted multidimensional similarity meets the preset conditions is selected as the matching profile cluster. The statistical information on the distribution of brewing parameters within each matching profile cluster is extracted, and the weighted multidimensional similarity is used as the fusion weight to perform weighted fusion of the corresponding group brewing parameter statistical distribution to generate a priori reference distribution for each target brewing parameter. Representative values ​​are extracted from the priori reference distribution to form the prior reference.

[0042] In this embodiment, when calling the pre-built general brewing profile database, the historical state vectors and corresponding brewing parameter records of the group of users are read. The historical state vectors are arranged in the order of operation speed, operation intensity, operation rhythm, alarm snooze frequency, schedule density, text input error rate, ambient audio style, lighting conditions, and meteorological parameters. For numerical state components, the preset lower limit and preset upper limit of the state component are read, the actual value is subtracted from the preset lower limit, and then divided by the difference between the preset upper limit and the preset lower limit to obtain a normalized value of 0 to 1. For typological state components, a preset encoding table is read, the typological value is converted into an encoded value, and then converted to a numerical range of 0 to 1 using the same normalization method. The number of state-parameter mapping clusters, K, is predetermined. K vectors are selected from the historical state vectors of the group of users as initial center vectors. The squared difference between the state components of each historical state vector and each initial center vector is calculated dimension by dimension. The sum of all squared differences and the square root are obtained to obtain the Euclidean distance. Each historical state vector is assigned to the set corresponding to the initial center vector with the smallest Euclidean distance. The average of each state component in each set is then calculated to form a new center vector. The distance calculation, set partitioning, and center vector update are repeated until the change in the center vector between two consecutive times is less than a preset change threshold, resulting in multiple state-parameter mapping clusters. Each state-parameter mapping cluster is associated with and stores a cluster center state vector, statistical information on the distribution of group brewing parameters within the cluster, and the component influence weight coefficients of each state component on each brewing parameter. The cluster center state vector is a combination of the average values ​​of each state component within the state-parameter mapping cluster. The statistical information on the distribution of group brewing parameters within the cluster is obtained by statistically analyzing the number of values, minimum value, maximum value, average value, median value, number of times the gear appears, and the frequency of gear appearance for each brewing parameter within the state-parameter mapping cluster.

[0043] The component influence weighting coefficients are constructed through statistical analysis of the correlation strength between state changes and parameter adjustments among users within a cluster. For any state-parameter mapping cluster, user records within the cluster are sorted according to recording time. For the j-th state component, the sorted state component value sequence is read, and the state component difference between two adjacent records is calculated to form a state change sequence. For the r-th brewing parameter, the brewing parameter value sequence at the same sorted position is read, and the brewing parameter difference between two adjacent records is calculated to form a parameter adjustment sequence. The average value of the state change sequence and the average value of the parameter adjustment sequence are calculated respectively. The state change deviation value is obtained by subtracting the average state change value from each state change value in the state change sequence, and the parameter adjustment deviation value is obtained by subtracting the average parameter adjustment value from each parameter adjustment value at the corresponding position in the parameter adjustment sequence. Multiply the state change deviation value at the corresponding position by the parameter adjustment deviation value and sum them to obtain the numerator of the correlation strength. Sum the squares of all state change deviation values ​​and take the square root to obtain the first correlation strength denominator. Sum the squares of all parameter adjustment deviation values ​​and take the square root to obtain the second correlation strength denominator. Multiply the first and second correlation strength denominators to obtain the correlation strength denominator. Divide the correlation strength numerator by the correlation strength denominator to obtain the correlation strength value between the j-th state component and the r-th brewing parameter. Take the absolute value of the correlation strength value to obtain the initial influence value. Sum the initial influence values ​​of all state components corresponding to the same brewing parameter, and then divide the initial influence value of each state component by the summation result to obtain the component influence weight coefficient of each state component on the brewing parameter. Perform the above calculations sequentially for each state component and each brewing parameter to obtain the component influence weight coefficient of each state component on each brewing parameter.

[0044] Next, when calculating the weighted multidimensional similarity between the dynamic user state vector and the cluster center state vector of each mapping cluster, the dynamic user state vector is first arranged according to the state component order in the general brewing profile database, and the same numerical transformation rule is used to transform each state component in the dynamic user state vector to a value range of 0 to 1. For any state-parameter mapping cluster, the normalized state component in the dynamic user state vector and the corresponding normalized state component in the cluster center state vector of the state-parameter mapping cluster are read item by item, and the absolute value of the difference between the two is calculated to obtain the difference value of each dimension. Then, the component influence weight coefficient of each state component in the state-parameter mapping cluster is read, and the component influence weight coefficient of each brewing parameter corresponding to the same state component is averaged as the dimension weight of the state component in the similarity calculation. Each dimension difference value is multiplied by the corresponding dimension weight to obtain the weighted difference term. All weighted difference terms are summed and divided by the sum of all dimension weights to obtain the weighted difference value. Then, 1 is divided by 1 and the weighted difference value is added to obtain the weighted multidimensional similarity between the dynamic user state vector and the cluster center state vector of the mapping cluster. The calculation is performed sequentially on each mapping cluster to obtain the weighted multidimensional similarity corresponding to each mapping cluster.

[0045] Then, when selecting at least one mapping cluster whose weighted multidimensional similarity meets the preset conditions as the matching profile cluster, a preset similarity threshold is read, and the weighted multidimensional similarity corresponding to each mapping cluster is compared with the preset similarity threshold. Mapping clusters with weighted multidimensional similarity not lower than the preset similarity threshold are written into the matching profile cluster set. When the matching profile cluster set is empty, all mapping clusters are sorted from high to low according to weighted multidimensional similarity, and the top-ranked mapping cluster is written into the matching profile cluster set. This yields at least one matching profile cluster, which is a state-parameter mapping cluster in the general brewing profile database that meets the preset conditions with the dynamic user state vector.

[0046] When extracting the statistical information of the distribution of group brewing parameters within each matched image cluster, and using weighted multidimensional similarity as the fusion weight, the statistical value of each target brewing parameter in each matched image cluster and the weighted multidimensional similarity corresponding to that matched image cluster are read during the weighted fusion of the statistical distribution of the corresponding group brewing parameters. For continuous numerical target brewing parameters, the average, minimum, and maximum values ​​of the target brewing parameter in the m-th matched image cluster are read, and the weighted multidimensional similarity corresponding to the m-th matched image cluster is used as the fusion weight. The average value of each matched image cluster is multiplied by the corresponding fusion weight to obtain the average weighted term. All average weighted terms are summed to obtain the average weighted sum. All fusion weights are summed to obtain the weighted sum. The average weighted sum is divided by the weighted sum to obtain the weighted average value of the target brewing parameter. In the same way, the minimum value of each matched image cluster is multiplied by the corresponding fusion weight and summed, then divided by the weighted sum to obtain the weighted minimum value. The maximum value of each matched image cluster is multiplied by the corresponding fusion weight and summed, then divided by the weighted sum to obtain the weighted maximum value. The weighted minimum, weighted maximum, and weighted average values ​​are combined to generate the prior reference distribution of the target brewing parameter. For discrete gear-type target brewing parameters, the frequency of occurrence of each gear in each matched profile cluster is read. The frequency of occurrence of the same gear in different matched profile clusters is multiplied by the corresponding weighted multidimensional similarity, and the products are summed and divided by the sum of all weighted multidimensional similarities to obtain the fused occurrence frequency of the gear. The fused occurrence frequencies of all gears together constitute the prior reference distribution of the discrete gear-type target brewing parameter.

[0047] Finally, representative values ​​are extracted from the prior reference distribution to form the prior reference. Specifically, for each target brewing parameter, its corresponding prior reference distribution is read, and the value of the parameter set obtained by fusing the matching profile clusters in the prior reference distribution is taken as the representative value of the target brewing parameter, while retaining the distribution statistics used to characterize the reliability of the representative value. For continuous numerical target brewing parameters, the average value of the target brewing parameter in each matching profile cluster and the corresponding weighted multidimensional similarity are read. Each average value is multiplied by the corresponding weighted multidimensional similarity to obtain a weighted parameter term. The sum of all weighted parameter terms is divided by the sum of all weighted multidimensional similarities to obtain the representative value of the target brewing parameter. Based on the deviation of the value of the target brewing parameter in each matching profile cluster from the representative value, and the matching profile clusters involved in the fusion, the representative value of the target brewing parameter is obtained. The quantity and corresponding sample quantity form the prior distribution statistics of the target brewing parameter. For discrete gear-type target brewing parameters, the occurrence frequency of each gear in each matched image cluster and the corresponding weighted multidimensional similarity are read. The occurrence frequency of the same gear is multiplied by the corresponding weighted multidimensional similarity and summed, then divided by the sum of all weighted multidimensional similarities to obtain the fusion occurrence frequency of that gear. The gear with the highest fusion occurrence frequency is selected as the representative value of the target brewing parameter. At the same time, the fusion occurrence frequency of each gear, the number of matched image clusters participating in the fusion, and the sample quantity are used as the prior distribution statistics of the target brewing parameter. Thus, the parameter name, representative value, and corresponding prior distribution statistics of each target brewing parameter are combined into a prior reference.

[0048] Step S500: Adaptively fuse the preferred brewing parameter set and prior reference to output at least two target brewing parameters.

[0049] In this embodiment, when adaptively fusing the preferred brewing parameter set and the prior reference, for each brewing parameter, the corresponding brewing parameter in the preferred brewing parameter set and the corresponding brewing parameter in the prior reference are read separately, and the individual confidence of the brewing parameter in the preferred brewing parameter set and the prior confidence of the brewing parameter in the prior reference are calculated independently. A fusion weight pair is established based on the individual confidence and the prior confidence, so that the participation ratio of the preferred brewing parameter set and the prior reference on the same brewing parameter can be determined separately. Then, the fusion weight pair is used to perform weighted fusion on the two source values ​​of the same brewing parameter to establish the final output value of the target brewing parameter. The above fusion operation is performed on all brewing parameters one by one, and all the final output values ​​are combined as at least two target brewing parameters for output.

[0050] Furthermore, the method provided in the application embodiment, which adaptively fuses the preferred brewing parameter set and prior reference to output at least two target brewing parameters, also includes: For each brewing parameter, the individual confidence level of the brewing parameter in the preferred brewing parameter set and the prior confidence level of the brewing parameter in the prior reference are calculated independently. A fusion weight pair is established based on the individual confidence level and the prior confidence level. The fusion weight pair is used to perform weighted fusion of the brewing parameters to establish the final output value of the target brewing parameter. The fusion operation is performed on all brewing parameters one by one, and all the final output values ​​are combined as at least two target brewing parameters for output.

[0051] In this embodiment, when independently calculating the individual credibility of the brewing parameters in the preferred brewing parameter set and the prior credibility of the brewing parameters in the prior reference for each brewing parameter, the preferred brewing parameter set and the prior reference are first read according to the brewing parameter name, and multiple brewing parameters that exist in both are selected as fusion objects. Each fusion object corresponds to a brewing parameter value in the preferred brewing parameter set and a representative value of a brewing parameter in the prior reference. For any brewing parameter, the value of the brewing parameter in the preferred brewing parameter set, the matching degree of the selected hit record, the number of state components participating in the matching, and the number of missing identifiers are read; the number of valid state components is obtained by subtracting the number of missing identifiers from the number of state components participating in the matching; the number of valid state components is then divided by the number of state components participating in the matching to obtain the data integrity ratio; the matching degree of the selected hit record is multiplied by the data integrity ratio to obtain the individual credibility of the brewing parameter. Read the representative value of the brewing parameter, the number of matching image clusters involved in generating the representative value, the weighted multidimensional similarity of each matching image cluster, and the number of samples from the prior reference; sum the weighted multidimensional similarities of each matching image cluster and divide by the number of matching image clusters to obtain the average similarity; divide the number of samples by the preset upper limit of the number of samples, and if the result is greater than 1, take the value as 1; then multiply the average similarity by this value to obtain the prior confidence of the brewing parameter.

[0052] When establishing fusion weight pairs based on individual credibility and prior credibility, the individual credibility and prior credibility corresponding to the same brewing parameter are read and added together to obtain the total credibility value. If the total credibility value is not 0, the fusion weight corresponding to the preferred brewing parameter set is obtained by dividing the individual credibility by the total credibility value; the fusion weight corresponding to the prior reference is obtained by dividing the prior credibility by the total credibility value. These two fusion weights constitute the fusion weight pair for that brewing parameter. If the total credibility value is 0, both the fusion weight corresponding to the preferred brewing parameter set and the fusion weight corresponding to the prior reference are set to 0.5, which is used as the fusion weight pair for that brewing parameter. When using fusion weights to weightedly fuse brewing parameters, for continuous numerical brewing parameters, the brewing parameter value from the preferred brewing parameter set is read as one fusion parameter value, and the representative value from the prior reference is read as another fusion parameter value. The brewing parameter value from the preferred brewing parameter set is multiplied by its corresponding fusion weight, and the representative value from the prior reference is multiplied by its corresponding fusion weight. The two products are then added together to obtain the final output value of the brewing parameter. For discrete-level brewing parameters, the fusion weights corresponding to the preferred brewing parameter set and the fusion weights corresponding to the prior reference are compared. If the former is greater than or equal to the latter, the value from the preferred brewing parameter set is used as the final output value of the brewing parameter; if the former is less than the latter, the representative value from the prior reference is used as the final output value of the brewing parameter.

[0053] Finally, when performing the fusion operation on all brewing parameters one by one, the selected fusion objects are processed sequentially according to the brewing parameter name. For each brewing parameter, individual credibility calculation, prior credibility calculation, fusion weight pair establishment, and weighted fusion are performed separately. After obtaining the final output value for each brewing parameter, the executable value range corresponding to that brewing parameter is read; if the final output value is lower than the preset lower limit, it is adjusted to the preset lower limit; if the final output value is higher than the preset upper limit, it is adjusted to the preset upper limit. After all fusion objects have been processed, the final output values ​​of each brewing parameter are combined according to the brewing parameter name, parameter unit, and parameter value. At least two brewing parameters with final output values ​​are output from the combined result as at least two target brewing parameters.

[0054] Step S600: Configure the decision display area mapped to the priority of the target brewing parameters, and perform a visual display of the target brewing parameters on the screen.

[0055] In this embodiment, a decision display area mapped to the priority of the target brewing parameters is configured. When the target brewing parameters are visualized on the display screen, the adaptability of each target brewing parameter is first evaluated, and a corresponding priority value is configured for each target brewing parameter based on the adaptability evaluation result. Then, the display area information of the display screen is obtained, and the display area information is divided into areas according to the priority value of each target brewing parameter to form a decision display area mapped to the priority of the target brewing parameters. Among them, the target brewing parameters with higher priority values ​​correspond to larger display area areas and are configured in the priority selection area of ​​the display screen. Finally, the corresponding target brewing parameters are visualized through each decision display area, so that at least two target brewing parameters are presented on the display screen according to their priority relationship.

[0056] Furthermore, in the method provided in the application embodiments, configuring a decision display area mapped to the priority of the target brewing parameters, and performing a visual display of the target brewing parameters on the screen, further includes: The suitability of the target brewing parameters is evaluated, and priority values ​​are configured based on the evaluation results. The display area information of the display screen is obtained, and the area ratio of the display area information is used to divide the display area information to establish a decision display area. The area with a higher priority value is set with a larger area, and the corresponding decision display area is set in the priority selection area. The display screen is used to perform visualization display through the decision display area.

[0057] In this embodiment, when evaluating the suitability of a target brewing parameter and configuring a priority value using the suitability evaluation result, at least two output target brewing parameters are read, and the parameter name, final output value, parameter unit, executable value range, individual confidence level, and prior confidence level of each target brewing parameter are obtained one by one. For any target brewing parameter, the individual confidence level and prior confidence level are added together and divided by 2 to obtain the confidence evaluation value; the upper limit and lower limit of the executable value range are added together and divided by 2 to obtain the midpoint of the range; the midpoint of the range is subtracted from the final output value and the absolute value is taken to obtain the deviation value; the lower limit of the executable value range is subtracted from the upper limit of the executable value range to obtain the range length; the deviation value is divided by the range length to obtain the deviation ratio; the deviation ratio is subtracted from 1 to obtain the value evaluation value; and the trust evaluation value and the value evaluation value are added together and divided by 2 to obtain the suitability evaluation result of the target brewing parameter. After all target brewing parameters have received their suitability evaluation results, they are ranked from highest to lowest according to the evaluation results, and specific priority values ​​are assigned based on the ranking order. When the number of target brewing parameters displayed is N, the first-ranked target brewing parameter is assigned a priority value of N, the second-ranked target brewing parameter is assigned a priority value of N-1, and so on, with the Nth-ranked target brewing parameter assigned a priority value of 1. Thus, each target brewing parameter obtains a specific numerical priority value, and a mapping relationship is established between the target brewing parameters and their priority values.

[0058] Next, the display area information of the screen is obtained, and the area ratio of the display area information is used to divide the display area information into decision display areas using priority values ​​to establish the display area. In this process, the displayable width, displayable height, starting horizontal coordinate, starting vertical coordinate, edge margin size, and touchable area range of the screen are read. The displayable width is subtracted from the left and right edge margin widths to obtain the divisible width; the displayable height is subtracted from the top and bottom edge margin heights to obtain the divisible height; the divisible width is multiplied by the divisible height to obtain the divisible display area. The priority values ​​of all target brewing parameters are added together to obtain the total priority value. For any target brewing parameter, its priority value is divided by the total priority value to obtain the area ratio corresponding to that target brewing parameter; the divisible display area is multiplied by the area ratio to obtain the area corresponding to that target brewing parameter. For example, when there are 3 target brewing parameters, the corresponding priority values ​​are 3, 2, and 1 respectively, the total priority value is 6, and the area ratios corresponding to the three target brewing parameters are 3 / 6, 2 / 6, and 1 / 6 respectively. Once the area is determined, the decision display area for the target brewing parameter corresponding to the highest priority value is set within the priority selection area. The priority selection area is the pre-defined priority interaction coordinate range in the display layout parameters. For the remaining target brewing parameters, in descending order of priority value, the starting coordinates, area width, and area height are determined sequentially within the remaining display area outside the priority selection area. After determining the corresponding display coordinates, area width, and area height for each target brewing parameter, a decision display area mapped to the priority of the target brewing parameter is established.

[0059] Finally, when the display screen is visualized through the decision display area, each decision display area is bound to the corresponding target brewing parameter, and the parameter name, final output value, and parameter unit of the target brewing parameter are written into the decision display area. For decision display areas whose area reaches the preset full display area threshold, the parameter name, final output value, parameter unit, and confirmation control of the target brewing parameter are displayed in the decision display area; for decision display areas whose area does not reach the preset full display area threshold, the parameter name, final output value, and parameter unit of the target brewing parameter are displayed in the decision display area. The display screen generates a display interface based on the starting coordinates, area width, area height, and display content of each decision display area, so that at least two target brewing parameters are displayed in their respective decision display areas, and the target brewing parameters with priority values ​​are visualized according to the area ratio and the priority selected area.

[0060] Furthermore, in the method provided in the application embodiments, after visually displaying the target brewing parameters on the display screen, it further includes: In response to a user's touch on the decision display area, the brewing control is executed according to the touch result, corresponding to the target brewing parameters.

[0061] In this embodiment, in response to a user touching the decision display area, the coffee machine reads the coordinates of the touch point on the display screen and compares these coordinates with the starting coordinates, area width, and area height of each decision display area to determine the decision display area to which the touch point belongs. It then reads the target brewing parameters bound to the decision display area, including the parameter name, final output value, and parameter unit, and writes the target brewing parameters into the brewing control command. The coffee machine adjusts the corresponding execution components according to the brewing control command; for example, the water supply component operates according to the water volume parameter, the heating component operates according to the temperature parameter, the grinding component operates according to the grinding degree parameter, and the extraction component operates according to the extraction time parameter. Thus, based on the touch result, the coffee machine executes brewing control according to the corresponding target brewing parameters.

[0062] In summary, the embodiments of this application have at least the following technical effects: This application responds to a user's non-explicit setting command by invoking a dataset of implicit interaction behaviors of the target user. This dataset includes dynamic operational features, digital behavior rhythm records, and environmental features. User state is extracted from the dataset to establish a dynamic user state vector. The target user's historical brewing database is queried, and the dynamic user state vector is used to perform similarity matching of the query results to establish a set of preferred brewing parameters. The dynamic user state vector is then used to perform general brewing profile matching to establish a priori reference. The preferred brewing parameter set and the priori reference are adaptively fused to output at least two target brewing parameters. A decision display area mapped to the priority of the target brewing parameters is configured, and the target brewing parameters are then visualized on the display screen. This invention solves the technical problem in existing technologies where coffee brewing parameters rely on explicit user settings, making it difficult to automatically generate personalized brewing parameters based on the user's current state. By extracting a dynamic user state vector based on implicit user interaction behaviors and fusing historical preferences with a general brewing profile for parameter decision-making, it achieves the technical effect of automatically outputting personalized coffee brewing parameters under non-explicit user settings, improving the adaptability of brewing parameters and the convenience of interaction.

[0063] Example 2 is based on the same inventive concept as the coffee-making parameter decision-making method based on user interaction habits in the previous examples, such as... Figure 3 As shown, this application provides a coffee-making parameter decision system based on user interaction habits. The system and method embodiments in this application are based on the same inventive concept. The system includes: The data retrieval module 11 is used to respond to a user's non-explicit setting command and call the implicit interaction behavior dataset of the target user. The implicit interaction behavior dataset includes operation dynamic features, digital behavior rhythm records, and environmental features. The state extraction module 12 is used to extract the user state from the implicit interaction behavior dataset and establish a dynamic user state vector. The similarity matching module 13 is used to query the target user's historical brewing database and use the dynamic user state vector to perform similarity matching of the query results to establish a preferred brewing parameter set. The profile matching module 14 is used to use the dynamic user state vector to perform general brewing profile matching and establish a priori reference. The fusion module 15 is used to adaptively fuse the preferred brewing parameter set and the priori reference to output at least two target brewing parameters. The visualization module 16 is used to configure a decision display area mapped to the priority of the target brewing parameters and perform visualization display of the target brewing parameters on the display screen.

[0064] Furthermore, the system is also used to implement the following functions: Using the user state vector as the overall query condition, a multi-dimensional overall similarity retrieval is performed in the historical brewing database to establish an initial hit record set. Each independent state component in the dynamic user state vector is used as an independent matching query sub-vector, and the component keyness is evaluated using the independent matching query sub-vector to establish a keyness score. The keyness score and the independent matching query sub-vector are used to perform compensation analysis on the initial hit record set to establish a compensation order ranking. The selected hit record with the highest matching degree in the compensation order ranking is selected, and the missing identifier of the independent matching query sub-vector is extracted from the selected hit record. The missing identifier is used to perform parameter fusion correction on the selected hit record to establish a preferred brewing parameter set.

[0065] Furthermore, the system is also used to implement the following functions: Based on the missing state component indicated by the missing identifier, a pre-constructed component-parameter influence mapping model is retrieved from the general brewing profile database. This model characterizes the directional influence of changes in the value of the missing state component on each associated target brewing parameter while keeping other state components within a stable range. The current actual value of the missing state component is extracted from the dynamic user state vector, and the historical actual value is extracted from the selected hit record. A component deviation value is calculated. This deviation value is input into the component-parameter influence mapping model to obtain the parameter correction offset for each target brewing parameter associated with the missing state component. Using the original value of the associated target brewing parameter in the selected hit record as a reference, the parameter correction offset is superimposed on the corresponding reference value to generate the corrected target brewing parameter value. All target brewing parameter values ​​are combined to establish the preferred brewing parameter set.

[0066] Furthermore, the system is also used to implement the following functions: A pre-built general brewing profile database is invoked, wherein the general brewing profile database contains multiple state-parameter mapping clusters. Each state-parameter mapping cluster is associated with and stores a cluster center state vector, statistical information on the distribution of brewing parameters within the cluster, and component influence weight coefficients of each state component on each brewing parameter. The component influence weight coefficients are constructed by statistical analysis of the correlation strength between the state changes and parameter adjustments of users within the cluster. The dynamic user state vector and the cluster center state vector of each mapping cluster are subjected to weighted multidimensional similarity calculation, in which the component influence weight coefficients of each state component are used as the dimensional weights in the similarity calculation. At least one mapping cluster whose weighted multidimensional similarity meets the preset conditions is selected as the matching profile cluster. The statistical information on the distribution of brewing parameters within each matching profile cluster is extracted, and the weighted multidimensional similarity is used as the fusion weight to perform weighted fusion of the corresponding group brewing parameter statistical distribution to generate a priori reference distribution for each target brewing parameter. Representative values ​​are extracted from the priori reference distribution to form the prior reference.

[0067] Furthermore, the system is also used to implement the following functions: For each brewing parameter, the individual confidence level of the brewing parameter in the preferred brewing parameter set and the prior confidence level of the brewing parameter in the prior reference are calculated independently. A fusion weight pair is established based on the individual confidence level and the prior confidence level. The fusion weight pair is used to perform weighted fusion of the brewing parameters to establish the final output value of the target brewing parameter. The fusion operation is performed on all brewing parameters one by one, and all the final output values ​​are combined as at least two target brewing parameters for output.

[0068] Furthermore, the system is also used to implement the following functions: The suitability of the target brewing parameters is evaluated, and priority values ​​are configured based on the evaluation results. The display area information of the display screen is obtained, and the area ratio of the display area information is used to divide the display area information to establish a decision display area. The area with a higher priority value is set with a larger area, and the corresponding decision display area is set in the priority selection area. The display screen is used to perform visualization display through the decision display area.

[0069] Furthermore, the system is also used to implement the following functions: In response to a user's touch on the decision display area, the brewing control is executed according to the touch result, corresponding to the target brewing parameters.

[0070] Furthermore, the system is also used to implement the following functions: The operational dynamic features include at least one of the following: the user's operating speed, operating force, and operating rhythm during the operation of the controls on the coffee machine. The digital behavior rhythm record includes at least one of the following: the number of times the alarm clock snoozes, the density of daily tasks, and the text input error rate obtained based on the smart terminal. The environmental features include at least one of the following: environmental audio style, lighting conditions, and meteorological parameters.

[0071] Furthermore, the system is also used to implement the following functions: Establish a time-series decay factor for the historical brewing database, which is used to constrain the untrustworthiness of data as time intervals increase; perform similarity matching of query results under the constraint of the time-series decay factor to establish a set of preferred brewing parameters.

[0072] It should be noted that the order of the embodiments described above is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. Furthermore, the above description focuses on specific embodiments of this specification. The processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired results. In some implementations, multitasking and parallel processing are possible or may be advantageous.

[0073] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any way. Although the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art can make some modifications or alterations to the above-disclosed technical content to create equivalent embodiments without departing from the scope of the present invention. Any modifications, equivalent changes, and alterations made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the scope of the present invention.

Claims

1. A coffee-making parameter decision-making method based on user interaction habits, characterized in that, The method includes: In response to a user's non-explicit setting command, the implicit interaction behavior dataset of the target user is invoked, which includes operational dynamic features, digital behavior rhythm records, and environmental features. User states are extracted from the implicit interaction behavior dataset to establish a dynamic user state vector; The target user's historical brewing database is queried, and the dynamic user state vector is used to perform similarity matching of the query results to establish a set of preferred brewing parameters; The dynamic user state vector is used to perform general brewing profile matching and establish a priori reference; Adaptive fusion of the preferred brewing parameter set and prior reference is performed to output at least two target brewing parameters; Configure a decision display area that maps to the priority of the target brewing parameters, and then visualize the execution of the target brewing parameters on the display screen. The dynamic user state vector is used to perform similarity matching of query results to establish a preferred brewing parameter set, including: Using the user state vector as the overall query condition, a multi-dimensional overall similarity retrieval is performed in the historical brewing database to establish an initial hit record set. Each independent state component in the dynamic user state vector is used as an independent matching query subvector. The independent matching query subvector is used to evaluate the criticality of the component and establish a criticality score. The criticality score and independent matching query sub-vectors are used to perform compensation analysis on the initial hit record set, and a compensation order ranking is established. Select the selected hit record with the highest matching degree in the compensation order sorting, and extract the missing identifier of the independent matching query sub-vector of the selected hit record; The missing identifier is used to perform parameter fusion correction on the selected hit record to establish a preferred brewing parameter set; Adaptive fusion of the preferred brewing parameter set and prior reference is performed to output at least two target brewing parameters, including: For each brewing parameter, the individual confidence level of the brewing parameter in the preferred brewing parameter set is calculated independently, as well as the prior confidence level of the brewing parameter in the prior reference. A fusion weight pair is established based on individual credibility and prior credibility. The fusion weight pair is used to perform weighted fusion of brewing parameters to establish the final output value of the target brewing parameter. All brewing parameters are blended one by one, and all final output values ​​are combined as at least two target brewing parameters.

2. The coffee-making parameter decision-making method based on user interaction habits as described in claim 1, characterized in that, Performing parameter fusion correction for selecting hit records using the missing identifier includes: Based on the missing state component indicated by the missing identifier, a pre-constructed component-parameter influence mapping model is retrieved from the general brewing profile database. The component-parameter influence mapping model is used to characterize the directional influence of the value change of the missing state component on the brewing parameters of each associated target while keeping other state components in a stable range. Extract the current actual value of the missing state component from the dynamic user state vector, extract the historical actual value of the missing state component from the selected hit record, and calculate the component deviation value. Input the component deviation value into the component-parameter influence mapping model to obtain the parameter correction offset of each target brewing parameter associated with the missing state component; Based on the original value of the target brewing parameter associated in the selected hit record, the parameter correction offset is superimposed on the corresponding reference value to generate the corrected target brewing parameter value. The preferred brewing parameter set is established by combining all target brewing parameter values.

3. The coffee-making parameter decision-making method based on user interaction habits as described in claim 1, characterized in that, The dynamic user state vector is used to perform general brewing profile matching and establish a priori reference, including: The system calls a pre-built general brewing profile database, which contains multiple state-parameter mapping clusters. Each state-parameter mapping cluster stores a cluster center state vector, statistical information on the distribution of brewing parameters within the cluster, and the component influence weight coefficients of each state component on each brewing parameter. The component influence weight coefficients are constructed by statistically analyzing the correlation strength between the state changes and parameter adjustments of users within the cluster. The dynamic user state vector and the cluster center state vector of each mapping cluster are subjected to weighted multidimensional similarity calculation. In the weighted multidimensional similarity calculation, the component influence weight coefficient of each state component is used as the dimension weight in the similarity calculation. At least one mapping cluster whose weighted multidimensional similarity meets the preset conditions is selected as the matching profile cluster; Extract the statistical information of the distribution of group brewing parameters within each matched image cluster, and use the weighted multidimensional similarity as the fusion weight to perform weighted fusion of the corresponding group brewing parameter statistical distribution to generate the prior reference distribution of each target brewing parameter; Representative values ​​are extracted from the prior reference distribution to form the prior reference.

4. The coffee-making parameter decision-making method based on user interaction habits as described in claim 1, characterized in that, Configure a decision display area that maps to the priority of target brewing parameters, and visualize the execution of the target brewing parameters on the display screen, including: The suitability of the target brewing parameters is evaluated, and priority values ​​are configured based on the results of the suitability evaluation. Obtain the display area information of the display screen, use the priority value to perform area ratio segmentation of the display area information, and establish a decision display area. The area with the higher priority value is set with a larger area, and the corresponding decision display area is set in the priority selection area. The decision display area is used to perform the visualization display on the screen.

5. The coffee-making parameter decision-making method based on user interaction habits as described in claim 4, characterized in that, After executing the target brewing parameters and visually displaying them on the screen, the process includes: In response to a user's touch on the decision display area, the brewing control is executed according to the touch result, corresponding to the target brewing parameters.

6. The coffee-making parameter decision-making method based on user interaction habits as described in claim 1, characterized in that, The operational dynamic features include at least one of the following: the user's operating speed, operating force, and operating rhythm during the operation of the controls on the coffee machine. The digital behavior rhythm record includes at least one of the following: the number of times the alarm clock snoozes, the density of daily tasks, and the text input error rate obtained based on the smart terminal. The environmental features include at least one of the following: environmental audio style, lighting conditions, and meteorological parameters.

7. The coffee-making parameter decision-making method based on user interaction habits as described in claim 1, characterized in that, The process of performing similarity matching of query results using the dynamic user state vector to establish a preferred brewing parameter set also includes: Establish a time-series decay factor for the historical brewing database, wherein the time-series decay factor is used to constrain the untrustworthiness of data as time intervals increase; Perform similarity matching of query results under the constraint of time decay factor to establish a set of preferred brewing parameters.

8. A coffee-making parameter decision-making system based on user interaction habits, characterized in that, The system is used to execute the coffee-making parameter decision method based on user interaction habits as described in any one of claims 1-7, the system comprising: The data retrieval module is used to respond to a user's non-display setting command and retrieve the target user's implicit interaction behavior dataset, which includes operation dynamic features, digital behavior rhythm records, and environmental features. The state extraction module is used to extract user states from the implicit interaction behavior dataset and establish a dynamic user state vector. The similarity matching module is used to query the target user's historical brewing database, perform similarity matching of the query results using the dynamic user state vector, and establish a set of preferred brewing parameters. The profile matching module is used to perform general brewing profile matching using the dynamic user state vector and establish a priori reference. The fusion module is used to adaptively fuse the preferred brewing parameter set and prior reference, and output at least two target brewing parameters; The visualization module is used to configure a decision display area that is mapped to the priority of the target brewing parameters, and to visualize the execution of the target brewing parameters on the display screen.

Images