CN117764778A - Intelligent ordering method and device, electronic equipment and storage medium - Google Patents
Intelligent ordering method and device, electronic equipment and storage medium Download PDFInfo
- Publication number
- CN117764778A CN117764778A CN202311638480.XA CN202311638480A CN117764778A CN 117764778 A CN117764778 A CN 117764778A CN 202311638480 A CN202311638480 A CN 202311638480A CN 117764778 A CN117764778 A CN 117764778A
- Authority
- CN
- China
- Prior art keywords
- meal
- customer
- dining
- dish
- dishes
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 70
- 235000012054 meals Nutrition 0.000 claims abstract description 216
- 230000000007 visual effect Effects 0.000 claims description 15
- 238000012545 processing Methods 0.000 claims description 14
- 238000004590 computer program Methods 0.000 claims description 11
- 230000003993 interaction Effects 0.000 claims description 3
- 235000019640 taste Nutrition 0.000 description 18
- 235000013305 food Nutrition 0.000 description 10
- 238000003058 natural language processing Methods 0.000 description 10
- 238000000605 extraction Methods 0.000 description 6
- 235000021167 banquet Nutrition 0.000 description 5
- 230000006399 behavior Effects 0.000 description 5
- 238000005516 engineering process Methods 0.000 description 5
- 238000010801 machine learning Methods 0.000 description 5
- 239000000463 material Substances 0.000 description 5
- 238000010411 cooking Methods 0.000 description 4
- 230000001815 facial effect Effects 0.000 description 4
- 235000014102 seafood Nutrition 0.000 description 4
- 239000013598 vector Substances 0.000 description 4
- 238000013473 artificial intelligence Methods 0.000 description 3
- 238000010586 diagram Methods 0.000 description 3
- 235000021158 dinner Nutrition 0.000 description 3
- 235000020095 red wine Nutrition 0.000 description 3
- 238000000611 regression analysis Methods 0.000 description 3
- 238000013528 artificial neural network Methods 0.000 description 2
- 235000015278 beef Nutrition 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 2
- 238000010224 classification analysis Methods 0.000 description 2
- 238000013135 deep learning Methods 0.000 description 2
- 238000013136 deep learning model Methods 0.000 description 2
- 238000001514 detection method Methods 0.000 description 2
- 235000012149 noodles Nutrition 0.000 description 2
- 238000004080 punching Methods 0.000 description 2
- 235000013311 vegetables Nutrition 0.000 description 2
- 239000004278 EU approved seasoning Substances 0.000 description 1
- 241001122315 Polites Species 0.000 description 1
- 230000000172 allergic effect Effects 0.000 description 1
- 208000010668 atopic eczema Diseases 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 238000003066 decision tree Methods 0.000 description 1
- 235000021185 dessert Nutrition 0.000 description 1
- 235000011850 desserts Nutrition 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 235000013410 fast food Nutrition 0.000 description 1
- 238000001914 filtration Methods 0.000 description 1
- 235000011194 food seasoning agent Nutrition 0.000 description 1
- 230000014509 gene expression Effects 0.000 description 1
- 238000007477 logistic regression Methods 0.000 description 1
- 238000000691 measurement method Methods 0.000 description 1
- 238000005065 mining Methods 0.000 description 1
- 230000000050 nutritive effect Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 235000014594 pastries Nutrition 0.000 description 1
- 235000011888 snacks Nutrition 0.000 description 1
- 235000021259 spicy food Nutrition 0.000 description 1
- 230000001502 supplementing effect Effects 0.000 description 1
- 238000012706 support-vector machine Methods 0.000 description 1
- 235000014101 wine Nutrition 0.000 description 1
Abstract
The embodiment of the invention provides an intelligent ordering method, which is used for acquiring multiple rounds of dialogue data of a customer and image data of the customer during ordering; determining dining intention and dining scene of the customer based on the multi-round dialogue data and the image data; determining recommended dishes in a dish library according to the dining intention of a customer, and generating a corresponding recommended menu according to the recommended dishes; pushing the recommended menu to a customer for selection, and receiving the selection menu of the customer; and generating personalized meal recommendation according to the selection menu and the meal scene, and determining the meal menu according to the personalized meal recommendation selected by the customer. Through acquiring the multi-round dialogue data and the image data of the customers during the ordering period, the dining intentions and the dining scenes of the customers can be analyzed more accurately, personalized dining recommendations conforming to the dining intentions and the dining scenes of the customers are generated, and recommended dishes conform to the dining intentions and the dining scenes of the customers more, so that the customer experience is improved.
Description
Technical Field
The invention relates to the technical field of artificial intelligence, in particular to an intelligent ordering method, an intelligent ordering device, electronic equipment and a storage medium.
Background
In recent years, with the rapid development of artificial intelligence technology, many fields have begun to apply artificial intelligence technology to improve efficiency and accuracy. The traditional ordering system usually adopts modes such as manual ordering, paper menus or simple electronic menus, and a great number of dishes are needed to be browsed by customers for selection, especially for new customers, the selection suitable for the customers is difficult to be made in the face of complicated dishes in the menus, the ordering selection cost is increased, the ordering time is long, the ordering efficiency is low, and the customer experience is influenced.
Disclosure of Invention
The embodiment of the invention provides an intelligent ordering method, which aims to solve the problems that in the existing ordering mode, customers are difficult to select proper dishes facing to a menu, the selecting cost of ordering is increased, the ordering time is long, the ordering efficiency is low, and the customer experience is affected. Through acquiring the multi-round dialogue data and the image data of the customers during ordering, the dining intentions and the dining scenes of the customers can be analyzed more accurately, the demands and the preferences of the customers are known in depth, accordingly, more proper recommended dishes are determined in a dish library, personalized dining recommendations conforming to the dining intentions and the dining scenes of the customers are generated, the selection cost of the dishes is reduced, the time of ordering is shortened, the recommended dishes conform to the dining intentions and the dining scenes of the customers more, and the customer experience is improved.
In a first aspect, an embodiment of the present invention provides an intelligent ordering method, where the method includes:
acquiring multiple rounds of dialogue data of a customer during ordering and image data of the customer;
determining dining intention and dining scene of the customer based on the multi-round dialogue data and the image data;
determining recommended dishes in a dish library according to the dining intention of a customer, and generating a corresponding recommended menu according to the recommended dishes;
pushing the recommended menu to a customer for selection, and receiving a selection menu of the customer;
and generating personalized meal recommendation according to the selection menu and the meal scene, and determining a meal menu according to the personalized meal recommendation selected by the customer.
Optionally, the acquiring the multi-round dialogue data of the customer during ordering and the image data of the customer includes:
acquiring audio data of a customer during ordering, and converting the audio data into text data to obtain multi-round dialogue data of the customer during ordering;
and acquiring visual data of the customer, and performing image processing on the visual data to obtain image data of the customer.
Optionally, the determining, based on the multi-round dialogue data and the image data, the dining intention and the dining scene of the customer includes:
Extracting first semantic features of the multi-round dialogue data to obtain first meal semantics corresponding to the multi-round dialogue data;
extracting second semantic features from the image data to obtain second dining semantics corresponding to the image data;
and predicting the first meal semantics and the second meal semantics to obtain the meal intention and the meal scene of the customer.
Optionally, the determining the recommended dishes in the dish library according to the dining intention of the customer, and generating the corresponding recommended menu according to the recommended dishes includes:
encoding all dishes in a dish library and the dining intents into the same feature space to obtain dish features of the dishes and intent features of the dining intents;
performing feature matching on the intention feature and the dish feature to obtain a successfully matched dish feature;
and determining the dishes corresponding to the successfully matched dish characteristics as recommended dishes, and generating a corresponding recommended menu according to the recommended dishes.
Optionally, the selecting menu includes a plurality of target dishes, and the generating personalized meal recommendation according to the selecting menu and the meal scene includes:
Determining attribute information of each target dish;
ordering the meals of the target dishes according to the attribute information to obtain a meal list;
generating a dining flow of the customer based on the dining list and the dining scene, wherein the dining flow comprises personalized dish introduction and personalized dining guidance;
and generating personalized meal recommendation based on the meal process.
Optionally, the generating a dining flow of the customer based on the dining list and the dining scene includes:
determining initial dish introduction of each target dish and initial meal guidance in the meal list;
generating type adjustment is carried out on the initial dish introduction through the dining scene, so that personalized dish introduction corresponding to the dining scene is obtained;
generating type adjustment is carried out on the initial meal instruction through the meal scene, so that personalized meal instruction corresponding to the meal scene is obtained;
and generating a dining flow of the customer based on the personalized dish introduction and the personalized dining guidance.
Optionally, the personalized meal recommendation includes a recommended meal time of each of the target dishes, and the determining the meal menu according to the personalized meal recommendation selected by the customer includes:
Determining the meal-out time of the target dish according to the recommended meal time of the target dish;
and determining a meal menu based on the meal outlet time of the target dishes.
In a second aspect, an embodiment of the present invention further provides an intelligent ordering device, where the intelligent ordering device includes:
the acquisition module is used for acquiring multiple rounds of dialogue data of the customer during ordering and image data of the customer;
the processing module is used for determining the dining intention and the dining scene of the customer based on the multi-round dialogue data and the image data;
the first generation module is used for determining recommended dishes in the dish library according to the dining intention of the customer and generating a corresponding recommendation menu according to the recommended dishes;
the interaction module is used for pushing the recommended menu to a customer for selection and receiving a selection menu of the customer;
and the second generation module is used for generating personalized meal recommendation according to the selection menu and the meal scene, and determining the meal menu according to the personalized meal recommendation selected by the customer.
In a third aspect, an embodiment of the present invention provides an electronic device, including: the intelligent ordering method comprises a memory, a processor and a computer program which is stored in the memory and can run on the processor, wherein the steps in the intelligent ordering method provided by the embodiment of the invention are realized when the processor executes the computer program.
In a fourth aspect, embodiments of the present invention provide a computer readable storage medium having a computer program stored thereon, the computer program when executed by a processor implementing the steps in the intelligent ordering method provided by the embodiments of the present invention.
In the embodiment of the invention, the multi-round dialogue data of the customer and the image data of the customer during the ordering period are obtained; determining dining intention and dining scene of the customer based on the multi-round dialogue data and the image data; determining recommended dishes in a dish library according to the dining intention of a customer, and generating a corresponding recommended menu according to the recommended dishes; pushing the recommended menu to a customer for selection, and receiving a selection menu of the customer; and generating personalized meal recommendation according to the selection menu and the meal scene, and determining a meal menu according to the personalized meal recommendation selected by the customer. Through acquiring the multi-round dialogue data and the image data of the customers during ordering, the dining intentions and the dining scenes of the customers can be analyzed more accurately, the demands and the preferences of the customers are known in depth, accordingly, more proper recommended dishes are determined in a dish library, personalized dining recommendations conforming to the dining intentions and the dining scenes of the customers are generated, the selection cost of the dishes is reduced, the time of ordering is shortened, the recommended dishes conform to the dining intentions and the dining scenes of the customers more, and the customer experience is improved.
Drawings
In order to more clearly illustrate the embodiments of the invention or the technical solutions in the prior art, the drawings that are required in the embodiments or the description of the prior art will be briefly described, it being obvious that the drawings in the following description are only some embodiments of the invention, and that other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.
FIG. 1 is a method flow chart of an intelligent ordering method provided by an embodiment of the invention;
FIG. 2 is a schematic diagram of an intelligent ordering device according to an embodiment of the present invention;
fig. 3 is a schematic structural diagram of an electronic device according to an embodiment of the present invention.
Detailed Description
The following description of the embodiments of the present invention will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present invention, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
As shown in fig. 1, fig. 1 is a method flowchart of an intelligent ordering method according to an embodiment of the present invention. The intelligent ordering method comprises the following steps:
101. multiple rounds of dialogue data of the customer during ordering and image data of the customer are obtained.
In the embodiment of the invention, the intelligent ordering method can be used for a management platform of dining behaviors, the management platform consists of a server or a distributed server, the management platform can comprise ordering service, after a merchant accesses the management platform, the management platform can provide the ordering service as subscription service to the merchant, and the intelligent ordering method of the embodiment of the invention is embedded in the ordering service. After subscribing the ordering service, the merchant needs to provide dish information to construct a corresponding dish library, and provide real-time connection between the ordering terminal and the management platform, so that the ordering service in the management platform can acquire multiple rounds of dialogue data of the customer in the merchant during ordering and image data of the customer.
The ordering terminal can be an ordering robot, a tablet computer or mobile equipment registered with the merchant member, and can acquire multiple rounds of dialogue data of customers during ordering and image data of the customers. The multi-round dialogue data may be multi-round dialogue between a plurality of customers, or multi-round dialogue between a customer and a food ordering terminal. The image data may be the appearance, behavior, etc. of the customer, for example, the image data of the customer such as wearing, hairstyle, wearing, facial movement, and limb movement.
102. Based on the multi-turn dialogue data and the image data, the dining intention of the customer and the dining scene are determined.
In the embodiment of the invention, the dialogue state and the image state of the customer can be determined by carrying out state analysis on the dialogue data and the image data by a rule-based method, a machine learning-based method and a deep learning-based method, wherein the dialogue state can be hidden with the preference and the intention of the customer, the preference can be taste preference, price preference and the like, the intention can be eating, talking, punching cards and the like, the image state can be formal, leisure, tension, elegance and the like, the dining intention of the customer and the dining scene can be determined according to the dialogue state and the image state, the dining intention can comprise the dining intention of taste, price, number of dishes, dining time and the like, and the dining scene can comprise the dining scene of appointment, business banquet, shop card, invitation, affinitive face and the like.
In one possible embodiment, the customer's dining intent, such as taste, price, number of dishes, time of dining, etc., may be determined based on customer preferences and intent in the conversation state, such as taste preferences, price preferences, eating, talking, punching a card, etc. The dining scene such as appointment, business banquet, shop card, doctor's eat, affinity meeting, etc. is determined according to the image state such as formality, leisure, tension, elegance, etc.
The dining intention and the dining scene of the customers in the ordering process are illustrated by the dating among couples, and the following examples are:
customer a says that: "loving, what you want to eat? "
Customer B says: "do i don't call what is you recommended? "
Customer a says that: "I want to eat the mind surface, you try to see. "
Customer B says: "good, that me also tried to try the face bar. "
Customer a says that: "the restaurant looks good and the environment is very ambient. "
Customer B says: the "o" is o, and the "o" is also delicious. "
Customer a says that: "what tastes you also like? "
Customer B says: "I also like Japanese style cuisine. "
During the ordering process, the states of the images of the customers can be obtained by acquiring the speaking and the interrelationships among the customers. For example, customers a and B sit together, talk easily and pleasurably, and have intimate behavior, and can determine that their image status is recreational romantic.
From the dialog states and the avatar states described above, their dining intention may be determined as: the tastes of customers A and B are deliberate noodles and Japanese cuisine, the price is not too much, the number of dishes to be tried is two deliberate noodles and one Japanese cuisine, and the eating time can be 1-2 hours. Meanwhile, according to the dialogue state and the image state, the dining scenes of the people can be determined to be appointment occasions.
Business banquet is used to illustrate the dining intention and the dining scene of customers in the ordering process, and examples are as follows:
customer a says that: "Boss, what is you particularly willing to eat? "
Customer B says: "I don't call, but the customer is somewhat allergic to seafood. "
Customer a says that: "that me first sees what dishes do not contain seafood. "
Customer B says: "can you look at, the price is not too high. "
Customer a says that: good, some bars with high cost performance are selected. "
Customer a says that: the red wine has good taste and moderate price. "
Customer B says: "A client also likes to drink red wine. "
Customer a says that: "that me did several times vegetable, where the vegetable appeared fresh. "
Customer B says: "OK". "
During the ordering process, the state of the customer's image can be obtained by acquiring the language of the customer and the interrelationship. For example, customers A and B sitting together, with polite communication, may determine that their image status is formal.
According to the conversation state and the image state, the dining intention of the customers A and B can be determined to be selecting dishes with high cost performance, wherein seafood is avoided, certain requirements are met for wine, red wine with moderate price is considered, certain consideration is given to price control, the number of dishes to be tried is 3-4, and the dining time can be 1.5-2 hours. Based on the session state and the image state, it is possible to determine that the dining scenes of the customers a and B are business banquet occasions.
103. And determining recommended dishes in the dishes library according to the dining intention of the customer, and generating a corresponding recommended menu according to the recommended dishes.
In the embodiment of the invention, the dining intention can comprise taste, price, quantity of dishes, dining time and the like, and after the dining intention of the customer is determined, the recommended dishes can be determined in the dish library according to the taste, price, quantity of dishes, dining time and the like in the dining intention.
Specifically, recommendation algorithms (e.g., collaborative filtering, content-based recommendation, etc.) may be used to determine recommended dishes based on the customer's dining intent and the information of the dishes in the dish library. For example, if a customer likes spicy food, some spicy dishes may be recommended to the customer. If the customer likes beef, then some dishes containing beef may be recommended to the customer. If the customer likes seafood and is price sensitive, then some moderately priced dishes may be recommended to the customer.
After determining the recommended dishes, a corresponding recommended menu may be generated. The recommended menu may include information of pictures, names, prices, food materials, tastes, etc. of the dishes. For example, recommended dishes may be organized into a table or list, including information on the name, picture, price, food material, taste, etc. of the dishes. The arrangement order of dishes can be adjusted or screened according to the dining time of customers, the number of people eating, and the like.
104. Pushing the recommended menu to the customer for selection and receiving the selection menu of the customer.
In the embodiment of the invention, after the recommended menu is obtained, the recommended menu can be sent to the ordering terminal of the customer, and after the customer receives the recommended menu on the ordering terminal, the recommended menu can be supplemented or deleted to obtain the selected menu which meets the actual requirement of the customer, and the selected menu is uploaded to the ordering service of the management platform.
105. And generating personalized meal recommendation according to the selection menu and the meal scene, and determining the meal menu according to the personalized meal recommendation selected by the customer.
In the embodiment of the present invention, the selection menu is a menu obtained by supplementing or deleting a menu according to a recommendation menu by a customer, and the personalized meal recommendation may include a meal process, such as when to taste which dishes, a description of dishes, which related topics are said when to taste the dishes, and the like.
Personalized multiple meal recommendations can be generated according to the selection menu and the meal scene, the multiple meal recommendations are sent to the ordering terminal of the customer, and the customer can select the meal recommendations after receiving the meal recommendations on the ordering terminal. After the customer selects the personalized meal recommendation, uploading the personalized meal recommendation selected by the customer to a meal ordering service of a management platform, and generating a corresponding meal outlet menu, wherein the meal outlet menu is a meal outlet menu required by a kitchen, the meal outlet menu comprises meal outlet dishes and meal outlet time corresponding to the meal outlet dishes, the kitchen cooks the dishes according to the meal outlet time, and meal outlet is carried out according to the meal outlet time.
Specifically, the personalized meal recommendations may be generated using predefined rules or algorithms based on the information in the menu of choices and in the meal scene. For example, rules may be defined to determine customer preference for peppery taste and dining scenes to recommend corresponding dishes.
In the embodiment of the invention, the multi-round dialogue data of the customer and the image data of the customer during the ordering period are obtained; determining dining intention and dining scene of the customer based on the multi-round dialogue data and the image data; determining recommended dishes in a dish library according to the dining intention of a customer, and generating a corresponding recommended menu according to the recommended dishes; pushing the recommended menu to a customer for selection, and receiving the selection menu of the customer; and generating personalized meal recommendation according to the selection menu and the meal scene, and determining the meal menu according to the personalized meal recommendation selected by the customer. Through acquiring the multi-round dialogue data and the image data of the customers during ordering, the dining intentions and the dining scenes of the customers can be analyzed more accurately, the demands and the preferences of the customers are known in depth, accordingly, more proper recommended dishes are determined in a dish library, personalized dining recommendations conforming to the dining intentions and the dining scenes of the customers are generated, the selection cost of the dishes is reduced, the time of ordering is shortened, the recommended dishes conform to the dining intentions and the dining scenes of the customers more, and the customer experience is improved.
Optionally, in the step of acquiring the multi-round dialogue data of the customer during the ordering and the image data of the customer, the audio data of the customer during the ordering may be acquired, and the audio data is converted into text data, so as to obtain the multi-round dialogue data of the customer during the ordering; and acquiring visual data of the customer, and performing image processing on the visual data to obtain image data of the customer.
In the embodiment of the invention, the audio data of the conversation of the customer during the ordering can be collected through the audio collection equipment arranged on the ordering terminal. The audio data may be stored in the form of raw audio files or transmitted to an ordering service for further processing, and the audio files may be converted to text data using a speech recognition engine. The speech recognition engine may use existing speech recognition techniques, such as deep learning based speech recognition or natural language processing based speech recognition. The converted text data will be used as multiple rounds of dialogue data for the customer during the order.
The visual information of the customer is captured through the camera installed on the ordering terminal, and the visual data can be transmitted to the ordering service for further processing in the form of an original video file. The visual data may be processed using image processing algorithms in particular to extract the customer's image data. The image processing algorithm may include steps of face detection, feature extraction, etc. to identify and extract data of wearing, hairstyle, wearing, facial action, limb action, etc. of the customer as image data of the customer. The avatar data may be structured text data.
Further, the visual data may be processed using computer vision techniques and image processing algorithms. For example, face detection and feature extraction may be performed to identify facial features and expressions of a customer. The appearance characteristics of wearing, hairstyle, accessories and the like of the customer, limb actions, behavior and the like can be analyzed. The extracted avatar data is converted into a structured data description. For example, facial features (e.g., eye size, nose shape, etc.), wearing style (e.g., leisure, formal, etc.), limb movements (e.g., gestures, standing, etc.), etc. of the customer may be extracted and these features are converted into data descriptions, such as using text representations, vector representations, or label representations, etc., to obtain visual data.
Optionally, in the step of determining the dining intention and the dining scene of the customer based on the multi-round dialogue data and the image data, the multi-round dialogue data may be subjected to first semantic feature extraction to obtain first dining semantics corresponding to the multi-round dialogue data; extracting second semantic features from the image data to obtain second dining semantics corresponding to the image data; and predicting the first meal semantics and the second meal semantics to obtain the meal intention and the meal scene of the customer.
In the embodiment of the invention, after the multi-round dialogue data is obtained, the first semantic feature extraction can be realized through a natural language processing technology, so that the first meal semantics corresponding to the multi-round dialogue data are obtained. For example, semantic features of text data are extracted by using a word bag model, TF-IDF (word frequency-inverse document frequency) and other methods, so that first meal semantics corresponding to multi-round dialogue data are obtained.
The image data can be text representation, vector representation or label representation, and the like, and after the image data is obtained, the second semantic feature extraction can be performed through a natural language processing technology, so that the second meal semantics corresponding to the image data are obtained.
After the first meal semantics and the second meal semantics are obtained, the first meal semantics and the second meal semantics are predicted. Specifically, machine learning or deep learning models (such as logistic regression, support vector machines, naive bayes classifier or neural networks) can be used for classifying or performing regression analysis on the extracted semantic features so as to predict and obtain dining intention and dining scene of the customer.
Further, in predicting the dining intention of the customer, classification or regression analysis may be performed according to information such as dishes, dining demands, taste preference, etc. mentioned in the customer session, so as to determine the dining intention of the customer, such as snack, dinner, dessert, etc.
Further, in the process of predicting the dining scene of the customer, classification or regression analysis can be performed according to the image features and behavior of the customer, so as to determine the dining scene of the customer, such as family dinner, business banquet, leisure peeks, and the like.
Optionally, in the step of determining recommended dishes in the dish library according to the dining intention of the customer and generating a corresponding recommended menu according to the recommended dishes, all the dishes in the dish library and the dining intention can be encoded into the same feature space to obtain the dish features of each dish and the intention features of the dining intention; performing feature matching on the intention feature and the dish feature to obtain the dish feature successfully matched; and determining the dishes corresponding to the successfully matched dish characteristics as recommended dishes, and generating a corresponding recommended menu according to the recommended dishes.
In the embodiment of the invention, the file is used for collecting information such as pictures, names, descriptions and the like of all dishes in the dish library, and labels or keywords of dining intentions, such as fast food, dinner, desserts and the like. The menu picture is preprocessed, e.g., background removed, resized, feature extracted, etc., to obtain key features or structured data of the menu. Information such as the name, description, etc. of the dish and the tag or keyword of the dining intention are converted into a machine readable format, for example, the keyword is extracted using a text mining technique or text conversion is performed using a natural language processing technique.
Encoding all dishes and dining intents in the dish library into the same feature space to obtain dish features and dining intention features of all dishes. The dish features may be encoded using natural language processing techniques, such as encoding using natural language processing techniques to obtain dish features of the dish. The intent feature may be encoded using natural language processing techniques, such as using bag of words models, TF-IDF, and the like.
And in the feature space, carrying out feature matching on the intended features and the dish features to obtain the successfully matched dish features. Similarity measurement methods (such as cosine similarity and Euclidean distance) can be used for calculating the similarity between the intention features and the dish features, and the dish features with higher similarity are selected as the dish features successfully matched.
And determining the dishes corresponding to the successfully matched dish features as recommended dishes, determining the recommended dishes according to similarity scores of the successfully matched dish features, generating corresponding recommended menus according to the recommended dishes, and specifically, arranging the recommended dishes according to a certain sequence to generate a menu list to obtain the recommended menus. The recommendation menu can also be customized and adjusted according to the personalized needs and preferences of the customers.
Optionally, the selection menu includes a plurality of target dishes, and in the step of generating the personalized meal recommendation according to the selection menu and the meal scene, attribute information of each target dish can be determined; ordering the meals of each target dish according to the attribute information to obtain a meal list; generating a dining flow of a customer based on the dining list and the dining scene, wherein the dining flow comprises personalized dish introduction and personalized dining guidance; personalized meal recommendations are generated based on the meal flow.
In the embodiment of the invention, the relevant information of each target dish in the menu, such as the dish type, taste, making method, food materials, nutritive value, eating method and other attributes, can be collected. The text information of each target dish in the selection menu can be extracted and classified by using a natural language processing technology so as to obtain the attribute information of the target dish.
And ordering the dining of each target dish according to the extracted attribute information. The attribute information may be classified or regression analyzed using machine learning or deep learning models (e.g., decision trees, neural networks, etc.) to determine the order of eating the individual target dishes. For example, the order of eating the target dishes may be determined in the order of front dishes, staple food, pastry, etc. Of course, the ranking algorithms and rules may also be adjusted to generate personalized meal recommendations based on different meal scenarios and customer preferences.
Based on the meal ordering result and the meal scene, a meal process for the customer is generated, which may include personalized dish introduction and personalized meal instruction. The dish introduction may be customized and adjusted according to the customer's taste preferences and other personalized needs. Personalized meal guidance, such as collocation advice, tableware use, topic advice, and the like, can be provided according to the meal scene and the dish characteristics.
Based on the meal flow, personalized meal recommendations are generated. The recommendation algorithm and rules can be adjusted according to the dining scenes and the demands of the customers to generate dining recommendations more in line with the demands of the customers. Personalized meal recommendations may be presented to customers in the form of text, pictures, or videos, etc., to provide a more intuitive and lively recommendation experience.
Optionally, in the step of generating the dining flow of the customer based on the dining list and the dining scene, an initial dish introduction and an initial dining instruction of each target dish can be determined in the dining list; generating type adjustment is carried out on the initial dish introduction through the dining scene, so that personalized dish introduction corresponding to the dining scene is obtained; generating type adjustment is carried out on the initial meal instruction through the meal scene, so that personalized meal instruction corresponding to the meal scene is obtained; based on the personalized dish introduction and the personalized dining guidance, a dining flow of the customer is generated.
In the embodiment of the invention, the initial dish introduction is a dish introduction of the target dish, the initial meal instruction is also a meal instruction of the target dish, the initial dish introduction may include information such as pictures, names, descriptions, food materials, cooking methods and the like of the dish, and the meal instruction may include information such as collocation suggestions, tableware use, optimal eating time and the like.
After the initial dish introduction is obtained, the initial dish introduction may be feature extracted and classified using natural language processing techniques or machine learning algorithms. The initial menu presentation may then be regenerated or adjusted using a generative model (e.g., generating a countermeasure network, a variational self-encoder, etc.) based on the feature vectors of the dining scene to obtain a personalized menu presentation corresponding to the dining scene. For example, according to different dining scenes, relevant information such as historical background, cultural characteristics, food sources and the like can be added into the dish description.
Similarly, after the initial meal instruction is obtained, the initial meal instruction may be feature extracted and categorized using natural language processing techniques or machine learning algorithms. The initial meal instructions may then be regenerated or adjusted using the generative model based on the feature vectors of the meal scenes to obtain personalized meal instructions corresponding to the meal scenes. For example, according to different dining scenes, food materials or seasonings suitable for local tastes and cultures can be added into the collocation suggestions; recommending tableware suitable for different occasions and etiquette in the use of the tableware; topics related to local culture, history, or current events are provided in topic suggestions.
And arranging the personalized dish introduction and the personalized dining instruction according to the sequence of each target dish to generate a dining flow of a customer. Of course, the dining process can be adjusted and optimized in combination with the personal preference and special requirements of the customers.
Optionally, the personalized meal recommendation includes recommended meal time of each target dish, and in the step of determining the meal menu according to the personalized meal recommendation selected by the customer, the meal out time of the target dish may be determined according to the recommended meal time of the target dish; and determining a meal menu based on the meal outlet time of the target dishes.
In the embodiment of the invention, when a customer selects personalized meal recommendation, the meal outlet time of the target dish is determined according to the recommended meal time of the target dish in the personalized meal recommendation, so that the meal outlet time of the dish is ensured to be in accordance with the meal time of the customer.
According to the meal-out time of the target dish, a corresponding meal-out menu can be determined, for example, the recommended meal time of the target dish A is 19:30, the meal out time of the target dish a may be 19:28, if cooking the target dish a takes 10 minutes, then the kitchen needs to be done at 19:18 begin cooking target dish a.
Of course, the meal output menu can be adjusted and optimized according to personal preference and special requirements of the customer, for example, the customer expects the target dish A to be cooked older, and the cooking time of the kitchen can be advanced by a plurality of minutes so as not to influence the meal time and the meal output time.
As shown in fig. 2, an embodiment of the present invention provides an intelligent ordering device, which includes:
an acquisition module 201, configured to acquire multiple rounds of dialogue data of a customer during ordering and image data of the customer;
a processing module 202, configured to determine a dining intention and a dining scene of the customer based on the multi-round dialogue data and the image data;
the first generation module 203 is configured to determine recommended dishes in a dish library according to a dining intention of a customer, and generate a corresponding recommendation menu according to the recommended dishes;
the interaction module 204 is configured to push the recommended menu to a customer for selection, and receive a selection menu of the customer;
and the second generating module 205 is configured to generate a personalized meal recommendation according to the selection menu and the meal scene, and determine a meal menu according to the personalized meal recommendation selected by the customer.
Optionally, the obtaining module 201 is further configured to obtain audio data of the customer during the ordering process, and convert the audio data into text data, so as to obtain multiple rounds of dialogue data of the customer during the ordering process; and acquiring visual data of the customer, and performing image processing on the visual data to obtain image data of the customer.
Optionally, the processing module 202 is further configured to perform a first semantic feature extraction on the multi-round dialogue data to obtain a first meal semantic corresponding to the multi-round dialogue data; extracting second semantic features from the image data to obtain second dining semantics corresponding to the image data; and predicting the first meal semantics and the second meal semantics to obtain the meal intention and the meal scene of the customer.
Optionally, the first generating module 203 is further configured to encode all dishes in a dish library and the dining intention into a same feature space, so as to obtain a dish feature of each dish and an intention feature of the dining intention; performing feature matching on the intention feature and the dish feature to obtain a successfully matched dish feature; and determining the dishes corresponding to the successfully matched dish characteristics as recommended dishes, and generating a corresponding recommended menu according to the recommended dishes.
Optionally, the selection menu includes a plurality of target dishes, and the second generating module 205 is further configured to determine attribute information of each target dish; ordering the meals of the target dishes according to the attribute information to obtain a meal list; generating a dining flow of the customer based on the dining list and the dining scene, wherein the dining flow comprises personalized dish introduction and personalized dining guidance; and generating personalized meal recommendation based on the meal process.
Optionally, the second generating module 205 is further configured to determine an initial dish introduction and an initial meal instruction of each of the target dishes in the meal list; generating type adjustment is carried out on the initial dish introduction through the dining scene, so that personalized dish introduction corresponding to the dining scene is obtained; generating type adjustment is carried out on the initial meal instruction through the meal scene, so that personalized meal instruction corresponding to the meal scene is obtained; and generating a dining flow of the customer based on the personalized dish introduction and the personalized dining guidance.
Optionally, the second generating module 205 is further configured to determine a meal out time of the target dish according to the recommended meal time of the target dish; and determining a meal menu based on the meal outlet time of the target dishes.
It should be noted that the intelligent ordering device provided by the embodiment of the invention can be applied to equipment such as an ordering robot, an intelligent mobile phone, a computer, a server and the like which can perform intelligent ordering.
The intelligent ordering device provided by the embodiment of the invention can realize each process realized by the intelligent ordering method in the method embodiment, and can achieve the same beneficial effects. In order to avoid repetition, a description thereof is omitted.
Referring to fig. 3, fig. 3 is a schematic structural diagram of an electronic device according to an embodiment of the present invention, as shown in fig. 3, including: memory 302, processor 301, and a computer program stored on memory 302 and executable on processor 301 for an intelligent ordering method, wherein:
the processor 301 is configured to call a computer program stored in the memory 302, and perform the following steps:
acquiring multiple rounds of dialogue data of a customer during ordering and image data of the customer;
determining dining intention and dining scene of the customer based on the multi-round dialogue data and the image data;
determining recommended dishes in a dish library according to the dining intention of a customer, and generating a corresponding recommended menu according to the recommended dishes;
pushing the recommended menu to a customer for selection, and receiving a selection menu of the customer;
And generating personalized meal recommendation according to the selection menu and the meal scene, and determining a meal menu according to the personalized meal recommendation selected by the customer.
Optionally, the acquiring, by the processor 301, multiple rounds of dialogue data of the customer during the ordering and the image data of the customer includes:
acquiring audio data of a customer during ordering, and converting the audio data into text data to obtain multi-round dialogue data of the customer during ordering;
and acquiring visual data of the customer, and performing image processing on the visual data to obtain image data of the customer.
Optionally, the determining, by the processor 301, the dining intention and the dining scene of the customer based on the multi-round dialogue data and the image data includes:
extracting first semantic features of the multi-round dialogue data to obtain first meal semantics corresponding to the multi-round dialogue data;
extracting second semantic features from the image data to obtain second dining semantics corresponding to the image data;
and predicting the first meal semantics and the second meal semantics to obtain the meal intention and the meal scene of the customer.
Optionally, the determining, by the processor 301, a recommended dish in a dish library according to the dining intention of the customer, and generating a corresponding recommended menu according to the recommended dish, includes:
encoding all dishes in a dish library and the dining intents into the same feature space to obtain dish features of the dishes and intent features of the dining intents;
performing feature matching on the intention feature and the dish feature to obtain a successfully matched dish feature;
and determining the dishes corresponding to the successfully matched dish characteristics as recommended dishes, and generating a corresponding recommended menu according to the recommended dishes.
Optionally, the selecting menu includes a plurality of target dishes, and the generating, by the processor 301, a personalized meal recommendation according to the selecting menu and the meal scene includes:
determining attribute information of each target dish;
ordering the meals of the target dishes according to the attribute information to obtain a meal list;
generating a dining flow of the customer based on the dining list and the dining scene, wherein the dining flow comprises personalized dish introduction and personalized dining guidance;
And generating personalized meal recommendation based on the meal process.
Optionally, the generating, by the processor 301, a dining process of the customer based on the dining list and the dining scene includes:
determining initial dish introduction of each target dish and initial meal guidance in the meal list;
generating type adjustment is carried out on the initial dish introduction through the dining scene, so that personalized dish introduction corresponding to the dining scene is obtained;
generating type adjustment is carried out on the initial meal instruction through the meal scene, so that personalized meal instruction corresponding to the meal scene is obtained;
and generating a dining flow of the customer based on the personalized dish introduction and the personalized dining guidance.
Optionally, the personalized meal recommendation executed by the processor 301 includes a recommended meal time of each of the target dishes, and determining a meal menu according to the personalized meal recommendation selected by the customer includes:
determining the meal-out time of the target dish according to the recommended meal time of the target dish;
and determining a meal menu based on the meal outlet time of the target dishes.
It should be noted that, the electronic device provided by the embodiment of the invention can be applied to devices such as a food ordering robot, a smart phone, a computer, a server and the like which can perform the intelligent food ordering method.
The electronic equipment provided by the embodiment of the invention can realize each process realized by the intelligent ordering method in the embodiment of the method, and can achieve the same beneficial effects. In order to avoid repetition, a description thereof is omitted.
The embodiment of the invention also provides a computer readable storage medium, on which a computer program is stored, which when executed by a processor, implements each process of the intelligent ordering method provided by the embodiment of the invention, and can achieve the same technical effect, and in order to avoid repetition, the description is omitted here.
Those skilled in the art will appreciate that implementing all or part of the above-described methods may be accomplished by way of a computer program stored in a computer-readable storage medium, which when executed may comprise the steps of the embodiments of the methods described above. The computer readable storage medium may be a magnetic disk, an optical disk, a Read-Only Memory (ROM), a random access Memory (Random Access Memory, RAM) or the like.
The foregoing disclosure is illustrative of the present invention and is not to be construed as limiting the scope of the invention, which is defined by the appended claims.
Claims (10)
1. An intelligent ordering method is characterized by comprising the following steps:
acquiring multiple rounds of dialogue data of a customer during ordering and image data of the customer;
determining dining intention and dining scene of the customer based on the multi-round dialogue data and the image data;
determining recommended dishes in a dish library according to the dining intention of a customer, and generating a corresponding recommended menu according to the recommended dishes;
pushing the recommended menu to a customer for selection, and receiving a selection menu of the customer;
and generating personalized meal recommendation according to the selection menu and the meal scene, and determining a meal menu according to the personalized meal recommendation selected by the customer.
2. The intelligent ordering method according to claim 1, wherein the acquiring the multi-turn dialogue data of the customer during ordering and the image data of the customer includes:
acquiring audio data of a customer during ordering, and converting the audio data into text data to obtain multi-round dialogue data of the customer during ordering;
and acquiring visual data of the customer, and performing image processing on the visual data to obtain image data of the customer.
3. The intelligent ordering method according to claim 2, wherein the determining the dining intention of the customer and the dining scene based on the multi-turn dialogue data and the image data includes:
extracting first semantic features of the multi-round dialogue data to obtain first meal semantics corresponding to the multi-round dialogue data;
extracting second semantic features from the image data to obtain second dining semantics corresponding to the image data;
and predicting the first meal semantics and the second meal semantics to obtain the meal intention and the meal scene of the customer.
4. The intelligent ordering method according to claim 3, wherein determining recommended dishes in a dish library according to the dining intention of a customer, and generating a corresponding recommended menu according to the recommended dishes, comprises:
encoding all dishes in a dish library and the dining intents into the same feature space to obtain dish features of the dishes and intent features of the dining intents;
performing feature matching on the intention feature and the dish feature to obtain a successfully matched dish feature;
and determining the dishes corresponding to the successfully matched dish characteristics as recommended dishes, and generating a corresponding recommended menu according to the recommended dishes.
5. The intelligent ordering method according to any one of claims 1-4, wherein the selection menu includes a plurality of target dishes, and wherein the generating personalized meal recommendations according to the selection menu and the meal scene includes:
determining attribute information of each target dish;
ordering the meals of the target dishes according to the attribute information to obtain a meal list;
generating a dining flow of the customer based on the dining list and the dining scene, wherein the dining flow comprises personalized dish introduction and personalized dining guidance;
and generating personalized meal recommendation based on the meal process.
6. The intelligent ordering method according to claim 5, wherein the generating a meal process for the customer based on the meal list and the meal scene comprises:
determining initial dish introduction of each target dish and initial meal guidance in the meal list;
generating type adjustment is carried out on the initial dish introduction through the dining scene, so that personalized dish introduction corresponding to the dining scene is obtained;
generating type adjustment is carried out on the initial meal instruction through the meal scene, so that personalized meal instruction corresponding to the meal scene is obtained;
And generating a dining flow of the customer based on the personalized dish introduction and the personalized dining guidance.
7. The intelligent ordering method according to claim 6, wherein the personalized meal recommendation includes a recommended meal time for each of the target dishes, the determining a meal menu based on the personalized meal recommendation selected by the customer comprising:
determining the meal-out time of the target dish according to the recommended meal time of the target dish;
and determining a meal menu based on the meal outlet time of the target dishes.
8. An intelligent ordering device, characterized in that, intelligent ordering device includes:
the acquisition module is used for acquiring multiple rounds of dialogue data of the customer during ordering and image data of the customer;
the processing module is used for determining the dining intention and the dining scene of the customer based on the multi-round dialogue data and the image data;
the first generation module is used for determining recommended dishes in the dish library according to the dining intention of the customer and generating a corresponding recommendation menu according to the recommended dishes;
the interaction module is used for pushing the recommended menu to a customer for selection and receiving a selection menu of the customer;
And the second generation module is used for generating personalized meal recommendation according to the selection menu and the meal scene, and determining the meal menu according to the personalized meal recommendation selected by the customer.
9. An electronic device, comprising: memory, a processor and a computer program stored on the memory and executable on the processor, the processor implementing the steps in the intelligent ordering method according to any one of claims 1 to 7 when the computer program is executed.
10. A computer readable storage medium, characterized in that it has stored thereon a computer program which, when executed by a processor, implements the steps of the intelligent ordering method according to any of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311638480.XA CN117764778A (en) | 2023-12-01 | 2023-12-01 | Intelligent ordering method and device, electronic equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311638480.XA CN117764778A (en) | 2023-12-01 | 2023-12-01 | Intelligent ordering method and device, electronic equipment and storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN117764778A true CN117764778A (en) | 2024-03-26 |
Family
ID=90313440
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311638480.XA Pending CN117764778A (en) | 2023-12-01 | 2023-12-01 | Intelligent ordering method and device, electronic equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117764778A (en) |
-
2023
- 2023-12-01 CN CN202311638480.XA patent/CN117764778A/en active Pending
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Min et al. | A survey on food computing | |
| KR102050334B1 (en) | Automatic suggestion responses to images received in messages, using the language model | |
| KR102119868B1 (en) | System and method for producting promotional media contents | |
| US20170278135A1 (en) | Image recognition artificial intelligence system for ecommerce | |
| US11430347B2 (en) | Providing diet assistance in a session | |
| WO2017210446A1 (en) | Method for providing one or more customized media centric products | |
| US20140164507A1 (en) | Media content portions recommended | |
| WO2014004864A1 (en) | Determining how interested a particular person is in an image | |
| WO2014004062A1 (en) | System for presenting high-interest-level images | |
| CN110554782B (en) | Expression input image synthesis method and system | |
| JP2021525433A (en) | Predicting potentially relevant topics based on searched / created digital media files | |
| US20230316378A1 (en) | System and methods for determining an object property | |
| CN103198089A (en) | Information processing apparatus, information processing method, and program | |
| KR20110035380A (en) | System and method for providing recipe information based on network | |
| CN117350825B (en) | Shopping guide method and system based on large language model | |
| CN115292543A (en) | Data processing method based on voice interaction novel and related product | |
| CN110209777A (en) | The method and electronic equipment of question and answer | |
| Verma et al. | Non-linear consumption of videos using a sequence of personalized multimodal fragments | |
| CN116823408B (en) | Commodity recommendation method, device, terminal and storage medium based on virtual reality | |
| WO2021070641A1 (en) | Management server and product search method | |
| US20230315729A1 (en) | Methods and systems for modifying a search query having a non-character-based input | |
| US8725755B2 (en) | Methods and apparatus or interactive name searching techniques | |
| CN117764778A (en) | Intelligent ordering method and device, electronic equipment and storage medium | |
| Hasan et al. | Eat This, Not That!–a Personalised Restaurant Menu Decoder That Helps You Pick the Right Food | |
| Agius et al. | Experimenting with tagging and context for collaborative MPEG-7 metadata |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |