Menu recommendation method, apparatus and device for intelligent kitchen range, and storage medium
By calculating the similarity and matching degree of the ingredients collection and dishes, combined with the confidence of multiple sources of evidence, the accuracy of recipe recommendations is solved when users have incomplete ingredients, and efficient and accurate recipe recommendations are achieved.
Patent Information
- Application Number
- CN202510466162.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-04-15
- Publication Date
- 2025-08-15
AI Technical Summary
The existing recipe recommendation algorithm is not very accurate when the user's ingredients are incomplete and cannot effectively recommend recipes.
By obtaining the user's current food collection, calculate the similarity and matching degree of the food collection and the dishes, and calculate the value and confidence of the functions of the evidence sources such as food matching, taste similarity and cooking complexity to determine the target recommended dishes.
It realizes efficient and accurate recommendation of appropriate recipes when users are incomplete ingredients, and improves the accuracy and efficiency of recipe recommendations.
Smart Images
Figure CN120496739A_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the technical field of digital smart kitchen smart stoves, and in particular to a recipe recommendation method, device, equipment and storage medium for a smart stove. Background Art
[0002] With the continuous progress of society and the continuous development of the economy, people are paying more and more attention to health, and a reasonable diet is closely related to people's health. Therefore, a large number of recipe recommendation methods and systems have appeared on the market.
[0003] Current dish recommendation algorithms are mostly based on content recommendations. These recipe recommendation algorithms have limitations when dealing with situations where users do not have complete ingredients, and the accuracy of recommendations is not high. Summary of the Invention
[0004] The present application provides a recipe recommendation method, device, equipment and storage medium for a smart stove, which can solve the problem of limitations in recipe recommendations when the user does not have complete ingredients.
[0005] In order to achieve the above objectives, this application adopts the following technical solutions:
[0006] In a first aspect of an embodiment of the present application, a method for recommending recipes for a smart stove is provided, the method comprising:
[0007] Obtaining the user's current ingredient set, wherein the ingredient set includes ingredient information of multiple current ingredients;
[0008] Matching a plurality of first dishes including any current ingredient from a preset database;
[0009] Calculating a first similarity and an ingredient matching degree between the ingredient set and each of the first dishes respectively;
[0010] Determining a plurality of second dishes from a plurality of first dishes according to the first similarity and the ingredient matching degree;
[0011] determining a function calculation value and a confidence level for each of a plurality of evidence sources, the evidence sources comprising: ingredient matching, taste similarity, and cooking complexity;
[0012] A target recommended dish is determined from the plurality of second dishes according to the calculated value and confidence level of each evidence source.
[0013] As a possible implementation, the ingredient information of the current ingredient includes ingredient type and ingredient proportion; and obtaining the user's current ingredient set includes:
[0014] Get the user's current food image;
[0015] The current food image is input into the trained target image recognition model to obtain the food set, which includes multiple food types and food proportions of the current food, and the food proportion is the proportion of each food relative to all the food in the food set.
[0016] As a possible implementation, before obtaining the user's current ingredient set, the method further includes:
[0017] Constructing a feature vector for each ingredient, where the feature vector is used to indicate the nutritional composition, taste, and cooking method of the ingredient;
[0018] Construct a feature vector for each dish based on the feature vector of each ingredient and the weight of each ingredient in the dish;
[0019] The characteristic vector of each ingredient and the feature vector of each dish are stored in a preset database.
[0020] As a possible implementation manner, respectively calculating the similarity between the current ingredient set and each first dish and the ingredient matching degree includes:
[0021] Obtaining, from the database, a feature vector of each current ingredient and a first feature vector of each first dish;
[0022] Determining a set feature vector of the set of ingredients based on the feature vector of each current ingredient;
[0023] Calculating the cosine similarity between the set feature vector and each first feature vector respectively to obtain the first similarity;
[0024] Determining matching ingredients between the current ingredient set and each first dish;
[0025] The ingredient matching degree is determined based on the ingredient coverage rate and ingredient importance weight of each matching ingredient and each corresponding first dish.
[0026] As a possible implementation, if the evidence source is food matching, determining the function calculation value and confidence level of each of the multiple evidence sources includes:
[0027] Determining a function calculation value of the matching degree of each of the ingredients according to the ingredient coverage and ingredient weight ratio of the ingredient set to each of the second dishes;
[0028] The confidence level of the ingredient matching degree is determined according to the function calculation value of each ingredient matching degree.
[0029] As a possible implementation, if the evidence source is taste similarity, determining the function calculation value and confidence level of each of the multiple evidence sources includes:
[0030] Determine a function calculation value of the corresponding taste similarity based on the similarity between the user's historical preference information and each second dish;
[0031] The confidence level of each taste similarity is determined according to the function calculation value of each taste similarity.
[0032] As a possible implementation, if the evidence source is cooking complexity, determining the function calculation value and confidence level of each of the multiple evidence sources includes:
[0033] Determining a function-calculated value of the cooking complexity of each second dish by the user based on the user's cooking level and the cooking difficulty of each second dish;
[0034] The confidence level of each cooking complexity is determined according to the function calculation value of the cooking complexity.
[0035] In a second aspect of an embodiment of the present application, a recipe recommendation device for a smart stove is provided, the device comprising:
[0036] An acquisition module is used to acquire a user's current ingredient set, wherein the ingredient set includes ingredient information of multiple current ingredients;
[0037] a matching module, configured to match a plurality of first dishes including any current ingredient from a preset database;
[0038] a calculation module, configured to respectively calculate a first similarity and an ingredient matching degree between the ingredient set and each of the first dishes;
[0039] a first determining module, configured to determine a plurality of second dishes from a plurality of first dishes according to the first similarity and the ingredient matching degree;
[0040] a second determination module for determining a function calculation value and a confidence level for each of a plurality of evidence sources, wherein the evidence sources include: ingredient matching, taste similarity, and cooking complexity;
[0041] The third determination module is used to determine a target recommended dish from the multiple second dishes based on the calculated value and confidence level of each evidence source.
[0042] According to a third aspect of an embodiment of the present application, an electronic device is provided, including a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the recipe recommendation method for the smart stove according to the first aspect of the embodiment of the present application is implemented.
[0043] In a fourth aspect of an embodiment of the present application, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the recipe recommendation method for the smart stove described in the first aspect of the embodiment of the present application is implemented.
[0044] The beneficial effects of the technical solutions provided in the embodiments of the present application include at least:
[0045] The recipe recommendation method for a smart stove provided in an embodiment of the present application obtains the user's current ingredient set, where the ingredient set includes ingredient information of multiple current ingredients; matches multiple first dishes including any current ingredient from a preset database; calculates a first similarity and an ingredient matching degree between the ingredient set and each of the first dishes respectively; determines multiple second dishes from the multiple first dishes based on the first similarity and the ingredient matching degree; determines a function calculation value and confidence level of each evidence source in multiple evidence sources, where the evidence sources include: ingredient matching degree, taste similarity and cooking complexity; determines a target recommended dish from the multiple second dishes based on the calculation value and confidence level of each evidence source, so that suitable recipes can be recommended efficiently and accurately based on the type of ingredients provided by the user. BRIEF DESCRIPTION OF THE DRAWINGS
[0046] Figure 1 A flowchart of a recipe recommendation method for a smart stove provided in an embodiment of the present application;
[0047] Figure 2 A structural diagram of a recipe recommendation device for a smart stove provided in an embodiment of the present application;
[0048] Figure 3 A schematic diagram of the internal structure of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION
[0049] The following will be combined with the drawings in the embodiments of this application to clearly and completely describe the technical solutions in the embodiments of this application. Obviously, the embodiments described are only part of the embodiments of this application, not all of the embodiments. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of this application.
[0050] In the following, the terms "first" and "second" are used for descriptive purposes only and should not be understood to indicate or imply relative importance or implicitly specify the number of the technical features indicated. Therefore, a feature defined as "first" or "second" may explicitly or implicitly include one or more of the features. In the description of the embodiments of the present disclosure, unless otherwise specified, "plurality" means two or more.
[0051] Additionally, the use of “based on” or “according to” is intended to be open and inclusive, in that a process, step, calculation, or other action “based on” or “according to” one or more conditions or values may, in practice, be based on additional conditions or beyond values.
[0052] It should be noted that the executor of the recipe recommendation method for a smart stove provided in an embodiment of the present application is an electronic device, which may be a smart stove, a data processing device or chip in the smart stove, or a server communicating with the smart stove. The embodiment of the present application does not make any specific limitations on this.
[0053] The present application embodiment provides a method for recommending recipes for a smart stove, such as Figure 1 As shown, the method includes:
[0054] Step 101: Obtain the user's current ingredient set, where the ingredient set includes ingredient information of multiple current ingredients.
[0055] Optionally, the ingredient information of the current ingredient includes ingredient type and ingredient proportion; and obtaining the user's current ingredient set includes:
[0056] Get the user's current food image;
[0057] The current food image is input into a trained image recognition model to obtain the food set, which includes multiple food types and food proportions of the current food, and the food proportion is the proportion of each food relative to all the food in the food set.
[0058] It should be noted that, optionally, before obtaining the user's current food image, it is necessary to pre-train the image recognition model. The training process of the image recognition model can be:
[0059] Construct a training dataset containing a large number of food images, covering photos with different lighting, angles, and maturity levels; select an efficient image classifier for the image recognition model, and use the training dataset to train and fine-tune the image recognition model. Use data augmentation technology and expand the training dataset to improve the generalization ability of the image recognition model.
[0060] The process of acquiring the user's current food image can be as follows: after the image acquisition device captures the user's current food image, the image is input into the trained image recognition model to obtain the food type, confidence level, and food percentage of the user's current food. For example: {"tomato": 0.95, 40%; "egg": 0.88, 25%; "green pepper": 0.75, 35%}.
[0061] Optionally, before obtaining the user's current ingredient set, the method also includes: constructing a feature vector for each ingredient, the feature vector of the ingredient being used to indicate the nutritional components, taste and cooking method of the ingredient; constructing a feature vector for each dish based on the feature vector of each ingredient and the weight of each ingredient in the dish; and storing the characteristic vector of each ingredient and the feature vector of each dish in a preset database.
[0062] In the actual implementation process, define the ingredient feature vector Using the organized ingredients and recipe data sets, we can create a i Define a feature vector The feature vector of ingredients includes information on multiple dimensions, such as nutritional components (protein, fat, carbohydrates, etc.), taste (crisp, soft, smooth, etc.), cooking methods (steaming, stir-frying, boiling, etc.); similarly, for each dish R i Also define an eigenvector This vector is the weighted average of all the ingredient feature vectors in the recipe. The specific calculation formula is as follows:
[0063]
[0064] Among them, R ij It is dish R i The jth ingredient in ω j is the weight of the jth ingredient in the recipe (e.g. the main ingredient has a higher weight than the auxiliary ingredient), and m is the total number of ingredients in the dish.
[0065] Step 102: Match multiple first dishes including any current ingredient from a preset database.
[0066] Optionally, during the matching process, the ingredient types in the ingredient set are matched with the ingredient types in multiple dishes in the database. As long as the ingredient types used by the dishes in the database include any ingredient type in the ingredient set, the dish will be used as the first dish.
[0067] Step 103: Calculate the first similarity and ingredient matching degree between the ingredient set and each of the first dishes respectively.
[0068] Optionally, respectively calculating the similarity and ingredient matching between the current ingredient set and each first dish includes:
[0069] Obtaining, from the database, a feature vector of each current ingredient and a first feature vector of each first dish;
[0070] Determining a set feature vector of the set of ingredients based on the feature vector of each current ingredient;
[0071] Calculating the cosine similarity between the set feature vector and each first feature vector respectively to obtain the first similarity;
[0072] Determining matching ingredients between the current ingredient set and each first dish;
[0073] The ingredient matching degree is determined based on the ingredient coverage rate and ingredient importance weight of each matching ingredient and each corresponding first dish.
[0074] In the actual execution process, the above-mentioned characteristic vector V of the ingredient set is determined based on the characteristic vector of each current ingredient. s The calculation process can be:
[0075]
[0076] Among them, V s is the collective feature vector of the food set, is the characteristic vector of the i-th ingredient, ω i is the weight of the i-th ingredient, which can be determined based on the proportion of ingredients, and n is the total number of ingredient types in the ingredient set.
[0077] The process of respectively calculating the cosine similarity between the set feature vector and each first feature vector to obtain the first similarity may be:
[0078]
[0079] in, is the cosine similarity mentioned above, that is, the first similarity, is the dot product of two vectors, ‖V s ‖、 are the modulus lengths of the two vectors respectively. It should be noted that the above cosine similarity calculation formula is the first eigenvector of the first dish.
[0080] The calculation process of the above-mentioned ingredient matching degree can be as follows: a fuzzy membership function is defined by the ingredient coverage and the ingredient importance.
[0081] μ(S,R i )=α*Cov(S,R i )+(1-α)*WCov(S,R i )
[0082]
[0083] Among them, Cov(S, R i) is the calculation company of food coverage, WCov is the calculation formula of food importance, α is the adjustment parameter used to balance importance and coverage, matchω j is the weight of the ingredients in the first dish and the same ingredients in the collection cluster, allω j is the weight of all ingredients in the first dish.
[0084] Step 104: Determine multiple second dishes from multiple first dishes based on the first similarity and the ingredient matching degree.
[0085] Optionally, the calculation process of step 104 may be:
[0086] Combine the ingredient matching degree with the cosine similarity to get the adjusted similarity score:
[0087]
[0088] Among them, β is an adjustment parameter used to balance the weights of cosine similarity and ingredient matching.
[0089] Based on the adjusted similarity score ASim(S, R i ), set a threshold θ to filter out candidate dishes that meet the conditions: CanSet = {R i |ASim(S,R i )≥θ}, that is, ASim(α, R i ) is greater than the threshold θ, the first dish is determined as the second dish. Through this step, dishes that do not match the user's ingredient set can be effectively filtered out, reducing the scale of subsequent calculations.
[0090] Step 105: Determine the function calculation value and confidence level of each of the multiple evidence sources, where the evidence sources include: ingredient matching, taste similarity, and cooking complexity.
[0091] It should be noted that after completing the second dish screening, the DS evidence theory is used to further integrate evidence from multiple dimensions to improve the accuracy and reliability of the dish recommendation. The core of the above step 105 is to construct multiple evidence sources, convert them into BPA, and then fuse them using the Dempster combination rule.
[0092] In this application, the identification framework Θ is defined as a set of all possible recommended dishes, that is, Θ is CanSet, Θ = {Y1, Y2, ..., Y k}, among which, recommended dish Y i ∈CanSet. Define Proposition A i ={Y i}, indicating a recommended dish Y i Definition m(Ai ) indicates that the proposition A i ={Y i}, i.e. recommend dish Y i Among them, the recommended dish Y i It was determined from the second dish.
[0093] Optionally, if the source of evidence is: ingredient matching, determining the function calculation value and confidence of each of the multiple evidence sources includes: determining the corresponding function calculation value of each ingredient matching degree based on the ingredient coverage and ingredient weight ratio of the ingredient set and each of the second dishes; and determining the confidence of the ingredient matching degree based on the function calculation value of each ingredient matching degree.
[0094] In the actual implementation process, based on the ingredient coverage Cov(S, R i ) and weighted food coverage WCov(S, R i ) as the basis for constructing BPA, for each second dish R i , define a proposition A i ={Y i}, indicating "recommended dish Y i "; In addition, based on Cov(S, R i ) and WCov(S, R i ) Construct BPA function m1, which is the function m1 of ingredient matching degree:
[0095] m1(A i )=γ*WCov(S,R i )+(1-γ)*Cov(S,R i )(If A i ={Ri})
[0096]
[0097] m1(other)=0
[0098] Among them, γ is a tuning parameter used to balance the importance of food coverage and weighted food coverage. m1(Θ) represents uncertainty. Specifically, ∑ i m1(A i ) represents all specific propositions A i If the sum is less than 1, it means that part of the confidence has not been assigned to the specific proposition, which indicates uncertainty. m1(other) = 0 means that the confidence of the current evidence source (such as ingredient matching) is completely assigned to the specific dish recommendation proposition A. i and uncertainty Θ, with no remaining portion allocated to other undefined propositions.
[0099] If the source of evidence is: taste similarity, determining the function calculation value and confidence of each of the multiple evidence sources includes: determining the corresponding function calculation value of taste similarity based on the similarity between the user's historical preference information and each second dish; and determining the confidence of the taste similarity based on the function calculation value of each taste similarity.
[0100] In the actual implementation process, based on the user's historical preferences or the common combination of ingredients, the taste similarity between the first dish and the second dish is inferred, and the cosine similarity calculation method in the above step 103 is used to calculate and adjust the taste similarity score in combination with the user's taste preferences. i , define a proposition A i ={Y i}, indicating "recommended dish Y i "; Based on the taste similarity score, construct the BPA function m2, that is, the taste similarity function m2:
[0101] m2(A i )=α·Sim(U,R i )+(1-α)·Sim(S,R i )(If A i ={Ri})
[0102]
[0103] m2(other)=0
[0104] Among them, α∈[0, 1] is a weight parameter used to balance the influence of user historical preferences and ingredient matching rules; U represents the user historical preference vector, which can be obtained by obtaining the user's historical cooking information, and S represents the ingredient set.
[0105] m2(Θ) represents uncertainty, specifically, ∑ i m2(A i ) represents all specific propositions A i If the sum is less than 1, it means that some of the confidence has not been assigned to the specific proposition, which indicates uncertainty. m2*other) = 0 means that the confidence of the current evidence source (such as taste similarity) is completely assigned to the specific dish recommendation proposition A. i and uncertainty Θ, with no remaining portion allocated to other undefined propositions.
[0106] If the evidence source is cooking complexity, determining the function calculation value and confidence level of each of the multiple evidence sources includes:
[0107] According to the user's cooking level and the cooking difficulty of each second dish, a function calculation value of the user's cooking complexity for each second dish is determined; and according to the function calculation value of each cooking complexity, a confidence level of the cooking complexity is determined.
[0108] In the actual implementation process, the cooking difficulty is evaluated based on the complexity of the cooking steps and the time required, and used as a source of evidence. i , define proposition A i ={Y i}, and set different weights according to the user's cooking level (e.g., novice, intermediate, advanced), so that the weight of the user's cooking level and the second dish R i Cooking difficulty score, get the dish R in the following formula i Cooking difficulty score. For example, if the user is a novice, then a simple recipe will receive a higher trust. Construct a BPA function m3 based on the cooking difficulty level and the estimated cooking time:
[0109] m3(A i )=1-(dish R i Cooking difficulty score) / (maximum cooking difficulty score)
[0110]
[0111] m3(other)=0
[0112] Among them, the dish R i The cooking difficulty score can be calculated based on the weight of the user's cooking level and the second dish R i The cooking difficulty score is determined, and the maximum cooking difficulty score is the maximum cooking difficulty score of the dishes in the database. The cooking difficulty score can be comprehensively evaluated based on the number of steps in the recipe, the skills required, and the cooking time.
[0113] m3(Θ) represents uncertainty, specifically, ∑ i m3(A i ) represents all specific propositions A i If the sum is less than 1, it means that some of the confidence has not been allocated to the specific proposition, which indicates uncertainty. m3*other) = 0 means that the confidence of the current evidence source (such as cooking complexity) is fully allocated to the specific dish recommendation proposition A. i and uncertainty Θ, with no remaining portion allocated to other undefined propositions.
[0114] Step 106: Determine a target recommended dish from the plurality of second dishes based on the calculated value and confidence level of each evidence source.
[0115] It is understandable that after a target recommended dish is determined from a plurality of second dishes, the recipe information of the target recommended dish can be obtained at the same time.
[0116] Optionally, the process of step 106 may be to fuse the three evidence sources, and the fused BPA function m(A) is:
[0117]
[0118] K is the normalization factor, and the calculation formula of K is:
[0119] K=1-m1(Θ)*m2(Θ)*m3(Θ)
[0120] According to the fused BPA function m(A), we can calculate the value of each second dish R i The trust level m({R i}) and uncertainty m(Θ). It can be calculated based on m({R i}) to sort the candidate dishes and generate the final recommendation list. A confidence threshold is set to recommend only dishes with a confidence level above the threshold, resulting in the target recommended dishes. Alternatively, the value of the fused BPA function m(A) can be used to recommend only the dishes with the highest values. Furthermore, the size of m(Θ) can be considered. If m(Θ) is too large, indicating insufficient evidence for a reliable recommendation, the user can be prompted to provide more information or choose a different recommendation strategy.
[0121] The recipe recommendation method for a smart stove provided in an embodiment of the present application obtains the user's current ingredient set, where the ingredient set includes ingredient information of multiple current ingredients; matches multiple first dishes including any current ingredient from a preset database; calculates a first similarity and an ingredient matching degree between the ingredient set and each of the first dishes respectively; determines multiple second dishes from the multiple first dishes based on the first similarity and the ingredient matching degree; determines a function calculation value and confidence level of each evidence source in multiple evidence sources, where the evidence sources include: ingredient matching degree, taste similarity and cooking complexity; determines a target recommended dish from the multiple second dishes based on the calculation value and confidence level of each evidence source, so that suitable recipes can be recommended efficiently and accurately based on the type of ingredients provided by the user.
[0122] The recipe recommendation method for the smart stove provided in the embodiment of the present application uses cosine similarity combined with fuzzy rules to perform a rough screening of dishes, and then uses different sources of evidence to give the final dish recommendation results. It is a reasonable and hierarchical recommendation framework. This method processes complex problems in stages, first filtering out candidate recipes that do not meet the requirements to reduce the computational scale, and then sorting and recommending the selected candidate set based on more refined sources of evidence. This can solve the problem of limitations in recipe recommendations when the user's ingredients are incomplete, and at the same time improve the accuracy and efficiency of recipe recommendations.
[0123] The present application also provides a recipe recommendation device for a smart stove, such as Figure 2 As shown, the device includes:
[0124] An acquisition module 11 is used to acquire a user's current ingredient set, where the ingredient set includes ingredient information of multiple current ingredients;
[0125] A matching module 12 is configured to match a plurality of first dishes including any current ingredient from a preset database;
[0126] A calculation module 13, configured to respectively calculate a first similarity and an ingredient matching degree between the ingredient set and each of the first dishes;
[0127] A first determining module 14 is configured to determine a plurality of second dishes from a plurality of first dishes based on the first similarity and the ingredient matching degree;
[0128] a second determination module 15 for determining a function calculation value and a confidence level for each of a plurality of evidence sources, wherein the evidence sources include: ingredient matching, taste similarity, and cooking complexity;
[0129] The third determination module 16 is configured to determine a target recommended dish from the plurality of second dishes based on the calculated value and confidence level of each evidence source.
[0130] In one embodiment, the food information of the current food includes the food type and the food proportion; the acquisition module 11 is specifically used to:
[0131] Get the user's current food image;
[0132] The current food image is input into a trained image recognition model to obtain the food set, which includes multiple food types and food proportions of the current food, and the food proportion is the proportion of each food relative to all the food in the food set.
[0133] In one embodiment, the apparatus further includes a configuration module 17, wherein the configuration module 17 is configured to:
[0134] Constructing a feature vector for each ingredient, where the feature vector is used to indicate the nutritional composition, taste, and cooking method of the ingredient;
[0135] Construct a feature vector for each dish based on the feature vector of each ingredient and the weight of each ingredient in the dish;
[0136] The characteristic vector of each ingredient and the feature vector of each dish are stored in a preset database.
[0137] In one embodiment, the calculation module 13 is specifically configured to:
[0138] Obtaining, from the database, a feature vector of each current ingredient and a first feature vector of each first dish;
[0139] Determining a set feature vector of the set of ingredients based on the feature vector of each current ingredient;
[0140] Calculating the cosine similarity between the set feature vector and each first feature vector respectively to obtain the first similarity;
[0141] Determining matching ingredients between the current ingredient set and each first dish;
[0142] The ingredient matching degree is determined based on the ingredient coverage rate and ingredient importance weight of each matching ingredient and each corresponding first dish.
[0143] In one embodiment, if the evidence source is: food matching, the second determination module 15 is specifically configured to:
[0144] Determining a function calculation value of the matching degree of each of the ingredients according to the ingredient coverage and ingredient weight ratio of the ingredient set to each of the second dishes;
[0145] The confidence level of the ingredient matching degree is determined according to the function calculation value of each ingredient matching degree.
[0146] In one embodiment, if the evidence source is taste similarity, the second determination module 15 is specifically configured to:
[0147] Determine a function calculation value of the corresponding taste similarity based on the similarity between the user's historical preference information and each second dish;
[0148] The confidence level of each taste similarity is determined according to the function calculation value of each taste similarity.
[0149] In one embodiment, if the evidence source is cooking complexity, the second determination module 15 is specifically configured to:
[0150] Determining a function-calculated value of the cooking complexity of each second dish by the user based on the user's cooking level and the cooking difficulty of each second dish;
[0151] The confidence level of each cooking complexity is determined according to the function calculation value of the cooking complexity.
[0152] The recipe recommendation device for the smart stove provided in the embodiment of the present application can execute the above-mentioned recipe recommendation method embodiment of the smart stove. Its implementation principle and technical effects are similar and will not be elaborated here.
[0153] The specific definition of the recipe recommendation device for the smart stove can be found in the definition of the recipe recommendation method for the smart stove above, and will not be repeated here. The various modules in the above-mentioned recipe recommendation device for the smart stove can be implemented in whole or in part through software, hardware, or a combination thereof. The above-mentioned modules can be embedded in or independent of the processor of the electronic device in hardware form, or can be stored in the memory of the electronic device in software form, so that the processor can call and execute the corresponding operations of the above modules.
[0154] The executor of the recipe recommendation method for the smart stove provided in the embodiment of the present application may be an electronic device, which may be a smart stove, a data processing device or chip in the smart stove, or a server or server cluster communicating with the smart stove, etc. The embodiment of the present application does not specifically limit this.
[0155] Figure 3 This is a schematic diagram of the internal structure of an electronic device provided in an embodiment of the present application. Figure 3 As shown, the electronic device includes a processor and memory connected via a system bus. The processor is used to provide computing and control capabilities. The memory may include a non-volatile storage medium and internal memory. The non-volatile storage medium stores an operating system and a computer program. The computer program can be executed by the processor to implement the steps of the recipe recommendation method for the smart stove provided in each of the above embodiments. The internal memory provides a cached operating environment for the operating system and computer program stored in the non-volatile storage medium.
[0156] Those skilled in the art will understand that Figure 3 The internal structure diagram of the electronic device shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the electronic device to which the solution of the present application is applied. The specific electronic device may include more or fewer components than shown in the figure, or combine certain components, or have a different component arrangement.
[0157] In another embodiment of the present application, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, the steps of the recipe recommendation method for the smart stove in the embodiment of the present application are implemented.
[0158] In another embodiment of the present application, a computer program product is also provided, which includes computer instructions. When the computer instructions are run on the recipe recommendation device of the smart stove, the recipe recommendation device of the smart stove executes each step of the recipe recommendation method of the smart stove in the method flow shown in the above method embodiment.
[0159] In the above embodiments, it can be implemented in whole or in part by software, hardware, firmware or any combination thereof. When implemented using a software program, it can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer execution instructions are loaded and executed on a computer, the process or function according to the embodiment of the present application is generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions can be transmitted from a website, computer, server or data center to another website, computer, server or data center by wired (e.g., coaxial cable, optical fiber, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device such as a server or data center that contains one or more media that can be integrated. The available medium may be a magnetic medium (eg, a floppy disk, a hard disk, a magnetic tape), an optical medium (eg, a DVD), or a semiconductor medium (eg, a solid state disk (SSD)).
[0160] The technical features of the above embodiments can be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.
[0161] The above embodiments merely represent several implementation methods of the present application. While the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the present invention. It should be noted that a person skilled in the art could make various modifications and improvements without departing from the spirit of the present application, all of which fall within the scope of protection of the present application. Therefore, the scope of protection of the present patent application shall be determined by the appended claims.
Claims
1. A recipe recommendation method for a smart stove, characterized in that: The method comprises: Obtaining the user's current ingredient set, wherein the ingredient set includes ingredient information of multiple current ingredients; Matching a plurality of first dishes including any current ingredient from a preset database; Calculating a first similarity and an ingredient matching degree between the ingredient set and each of the first dishes respectively; Determining a plurality of second dishes from a plurality of first dishes according to the first similarity and the ingredient matching degree; determining a function calculation value and a confidence level for each of a plurality of evidence sources, the evidence sources comprising: ingredient matching, taste similarity, and cooking complexity; A target recommended dish is determined from the plurality of second dishes according to the calculated value and confidence level of each evidence source.
2. The method according to claim 1, characterized in that The ingredient information of the current ingredient includes the ingredient type and the ingredient ratio; the step of obtaining the user's current ingredient set includes: Get the user's current food image; The current food image is input into a trained image recognition model to obtain the food set, which includes multiple food types and food proportions of the current food, and the food proportion is the proportion of each food relative to all the food in the food set.
3. The method according to claim 1, characterized in that Before obtaining the user's current ingredient set, the method further includes: Constructing a feature vector for each ingredient, where the feature vector is used to indicate the nutritional composition, taste, and cooking method of the ingredient; Construct a feature vector for each dish based on the feature vector of each ingredient and the weight of each ingredient in the dish; The characteristic vector of each ingredient and the feature vector of each dish are stored in a preset database.
4. The method according to claim 3, characterized in that The calculating the similarity between the current ingredient set and each first dish and the ingredient matching degree respectively includes: Obtaining, from the database, a feature vector of each current ingredient and a first feature vector of each first dish; Determining a set feature vector of the set of ingredients based on the feature vector of each current ingredient; Calculating the cosine similarity between the set feature vector and each first feature vector respectively to obtain the first similarity; Determining matching ingredients between the current ingredient set and each first dish; The ingredient matching degree is determined based on the ingredient coverage rate and ingredient importance weight of each matching ingredient and each corresponding first dish.
5. The method according to claim 1, wherein If the evidence source is food matching, determining the function calculation value and confidence level of each of the multiple evidence sources includes: Determining a function calculation value of the matching degree of each of the ingredients according to the ingredient coverage and ingredient weight ratio of the ingredient set to each of the second dishes; The confidence level of the ingredient matching degree is determined according to the function calculation value of each ingredient matching degree.
6. The method according to claim 1, characterized in that If the evidence source is taste similarity, determining the function calculation value and confidence level of each of the multiple evidence sources includes: Determine a function calculation value of the corresponding taste similarity based on the similarity between the user's historical preference information and each second dish; The confidence level of each taste similarity is determined according to the function calculation value of each taste similarity.
7. The method according to claim 1, characterized in that If the evidence source is cooking complexity, determining the function calculation value and confidence level of each of the multiple evidence sources includes: Determining a function-calculated value of the cooking complexity of each second dish by the user based on the user's cooking level and the cooking difficulty of each second dish; The confidence level of each cooking complexity is determined according to the function calculation value of the cooking complexity.
8. A recipe recommendation device for a smart stove, characterized in that: The device comprises: An acquisition module is used to obtain the user's current ingredient set, wherein the ingredient set includes ingredient information of multiple current ingredients; a matching module, configured to match a plurality of first dishes including any current ingredient from a preset database; a calculation module, configured to respectively calculate a first similarity and an ingredient matching degree between the ingredient set and each of the first dishes; a first determining module, configured to determine a plurality of second dishes from a plurality of first dishes according to the first similarity and the ingredient matching degree; a second determination module for determining a function calculation value and a confidence level for each of a plurality of evidence sources, wherein the evidence sources include: ingredient matching, taste similarity, and cooking complexity; The third determination module is used to determine a target recommended dish from the multiple second dishes based on the calculated value and confidence level of each evidence source.
9. An electronic device, characterized in that: The invention comprises a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the recipe recommendation method for the smart stove according to any one of claims 1 to 7 is implemented.
10. A computer-readable storage medium, characterized in that A computer program is stored thereon, and when the computer program is executed by a processor, the recipe recommendation method for the smart stove according to any one of claims 1 to 7 is implemented.