CN109300059B - Dish recommending method and device - Google Patents
Dish recommending method and device Download PDFInfo
- Publication number
- CN109300059B CN109300059B CN201811069258.1A CN201811069258A CN109300059B CN 109300059 B CN109300059 B CN 109300059B CN 201811069258 A CN201811069258 A CN 201811069258A CN 109300059 B CN109300059 B CN 109300059B
- Authority
- CN
- China
- Prior art keywords
- dish
- sample
- attribute information
- vector
- recommended
- 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.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 60
- 239000013598 vector Substances 0.000 claims abstract description 228
- 238000012549 training Methods 0.000 claims abstract description 123
- 230000006399 behavior Effects 0.000 claims abstract description 79
- 238000013507 mapping Methods 0.000 claims abstract description 34
- 235000013305 food Nutrition 0.000 claims description 36
- 239000011159 matrix material Substances 0.000 claims description 33
- 238000002372 labelling Methods 0.000 claims description 23
- 238000004891 communication Methods 0.000 claims description 15
- 235000012054 meals Nutrition 0.000 description 6
- 238000010586 diagram Methods 0.000 description 5
- 230000002452 interceptive effect Effects 0.000 description 3
- 238000004590 computer program Methods 0.000 description 2
- 230000006870 function Effects 0.000 description 2
- 238000012545 processing Methods 0.000 description 2
- 230000003044 adaptive effect Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 238000011022 operating instruction Methods 0.000 description 1
- 230000011218 segmentation Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/10—Services
- G06Q50/12—Hotels or restaurants
Landscapes
- Business, Economics & Management (AREA)
- Tourism & Hospitality (AREA)
- Health & Medical Sciences (AREA)
- Economics (AREA)
- General Health & Medical Sciences (AREA)
- Human Resources & Organizations (AREA)
- Marketing (AREA)
- Primary Health Care (AREA)
- Strategic Management (AREA)
- Physics & Mathematics (AREA)
- General Business, Economics & Management (AREA)
- General Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a dish recommending method and device, wherein the method comprises the following steps: mapping user attribute information and scene attribute information of a target user into corresponding attribute information vectors; obtaining dish vectors of a plurality of dishes to be recommended; inputting the attribute information vector and the dish vector of each dish to be recommended into the association degree model obtained by training to obtain the association degree value of the attribute information vector and the dish vector of each dish to be recommended; and selecting a recommended dish from the plurality of dishes to be recommended according to the relevance value and recommending the dish to the target user. According to the scheme, dish recommendation can be performed without depending on historical ordering behaviors of the target user, even for the target user with the insufficient ordering behaviors, the relevance degree value of the attribute information vector of the corresponding target user and the dish vector of the dish to be recommended can be obtained by using the relevance degree model, and recommendation is performed according to the relevance degree value, so that the ordering experience of the target user is improved.
Description
Technical Field
The invention relates to the technical field of computers, in particular to a dish recommending method and device.
Background
With the rise of intelligent food and drink technologies, food ordering modes have been developed from traditional paper food ordering to terminal intelligent food ordering, for example, food ordering on a take-away platform or food ordering in a food ordering terminal provided by a restaurant. Meanwhile, in order to improve the ordering rate of dishes and improve the ordering experience of the user, more and more platforms providing ordering service recommend the dishes to the user through the terminal.
In the existing recommendation scheme, preference information of a user is determined according to user history data, and recommendation is performed according to the preference information. For example, in the chinese patent application with application publication No. CN1068157745A, a dish component preference library of a target user is determined according to comment information of the target user, and dishes recommended for the target user are determined according to the dish component preference library. According to the scheme, the dishes recommended for the users are determined according to historical comment information of the target users, and accurate dish recommendation cannot be provided for the target users without comments or with few comments. As another example, in the chinese patent application with application publication No. CN108230009A, a preference prediction result of a target user for a target object is obtained by predicting interaction behavior characteristics (including behavior characteristics such as clicking, purchasing, browsing, and collecting) of the target object based on a pre-trained preference prediction model. The scheme determines the preference degree of the target user to the target object depending on the historical interactive behaviors of the target user, and cannot accurately predict the preference degree of the target user who does not generate the interactive behaviors or has less interactive behaviors with the target object, so that accurate recommendation cannot be performed.
Disclosure of Invention
In view of the above, the present invention has been made to provide a dish recommending method and apparatus that overcomes or at least partially solves the above problems.
According to an aspect of the present invention, there is provided a dish recommending method including:
mapping user attribute information and scene attribute information of a target user into corresponding attribute information vectors;
obtaining dish vectors of a plurality of dishes to be recommended;
inputting the attribute information vector and the dish vector of each dish to be recommended into a trained relevancy model to obtain relevancy values of the attribute information vector and the dish vector of each dish to be recommended;
and selecting a recommended dish from a plurality of dishes to be recommended according to the relevance value and recommending the dish to a target user.
Optionally, before the obtaining the dish vector of the plurality of dishes to be recommended, the method further includes:
constructing a training corpus of a dish vectorization model according to the dish knowledge graph;
acquiring training sample corpora from the training corpus, and inputting stroke characteristic information of the training sample corpora into an initialized dish vectorization model for training; and obtaining a dish vectorization matrix according to the training result.
Optionally, the obtaining the dish vector of the plurality of dishes to be recommended further includes:
and mapping stroke characteristic information corresponding to the dish names of the dishes to be recommended into dish vectors of the dishes to be recommended according to the dish vectorization matrix.
Optionally, the relevance model is obtained by training through the following steps:
obtaining dish sample vectors of a plurality of dish samples; acquiring user sample attribute information of a food ordering user sample corresponding to the multi-time food ordering behavior and scene sample attribute information of a food ordering scene; mapping the user sample attribute information and the scene sample attribute information into corresponding sample attribute information vectors;
according to the multiple ordering behaviors, labeling the relevance between the sample attribute information vector and the dish sample vector to obtain a relevance labeling result;
inputting the dish sample vector and the sample attribute information vector into a relevancy training model to obtain a relevancy output result output by the relevancy training model;
and training a relevance training model by using the loss between the relevance marking result and the relevance output result to obtain the relevance model.
Optionally, the obtaining a dish sample vector of a plurality of dish samples further includes:
and mapping stroke characteristic information corresponding to the dish names of all the dish samples into dish vectors of the dish samples according to the dish vectorization matrix.
Optionally, before the labeling the association degree between the sample attribute information vector and the dish sample vector, the method further includes:
the ordering dishes contained in each ordering behavior correspond to the dish samples to obtain the corresponding relation between the ordering dishes and the dish samples;
the step of labeling the relevance between the sample attribute information vector and the dish sample vector specifically comprises: and marking the association degree of the sample attribute information vector and the dish sample vector according to the corresponding relation.
Optionally, the labeling the association degree of the sample attribute information vector and the dish sample vector according to the correspondence further includes:
according to the corresponding relation, marking the association degree of the dish sample vector of the dish sample contained in the ordering behavior and the sample attribute information vector corresponding to the ordering behavior as a first association degree;
and/or marking the association degree of the dish sample vector of the dish sample not contained in the ordering behavior and the sample attribute information vector corresponding to the ordering behavior as a second association degree.
Optionally, the selecting a recommended dish from a plurality of dishes to be recommended according to the relevance value and recommending the recommended dish to the target user further includes:
and sorting a plurality of dishes to be recommended according to the relevance value from high to low, and selecting a preset number of dishes to be recommended from the sorting result as recommended dishes to be recommended to a target user.
According to another aspect of the present invention, there is provided a dish recommending apparatus including:
the mapping module is suitable for mapping the user attribute information and the scene attribute information of the target user into corresponding attribute information vectors;
the acquisition module is suitable for acquiring dish vectors of a plurality of dishes to be recommended;
the prediction module is suitable for inputting the attribute information vector and the dish vector of each dish to be recommended into a trained relevance model to obtain a relevance value of the attribute information vector and the dish vector of each dish to be recommended;
and the recommending module is suitable for selecting recommended dishes from a plurality of dishes to be recommended according to the relevance value and recommending the recommended dishes to the target user.
Optionally, the apparatus further comprises:
the first training module is suitable for constructing a training corpus of the dish vectorization model according to the dish knowledge graph; acquiring training sample corpora from the training corpus, and inputting stroke characteristic information of the training sample corpora into an initialized dish vectorization model for training; and obtaining a dish vectorization matrix according to the training result.
Optionally, the obtaining module is further adapted to:
and mapping stroke characteristic information corresponding to the dish names of the dishes to be recommended into dish vectors of the dishes to be recommended according to the dish vectorization matrix.
Optionally, the apparatus further comprises:
the second training module is suitable for obtaining dish sample vectors of a plurality of dish samples; acquiring user sample attribute information of a food ordering user sample corresponding to the multi-time food ordering behavior and scene sample attribute information of a food ordering scene; mapping the user sample attribute information and the scene sample attribute information into corresponding sample attribute information vectors; according to the multiple ordering behaviors, labeling the relevance between the sample attribute information vector and the dish sample vector to obtain a relevance labeling result; inputting the dish sample vector and the sample attribute information vector into a relevancy training model to obtain a relevancy output result output by the relevancy training model; and training a relevance training model by using the loss between the relevance marking result and the relevance output result to obtain the relevance model.
Optionally, the second training module is further adapted to:
and mapping stroke characteristic information corresponding to the dish names of all the dish samples into dish vectors of the dish samples according to the dish vectorization matrix.
Optionally, the second training module is further adapted to:
the ordering dishes contained in each ordering behavior correspond to the dish samples to obtain the corresponding relation between the ordering dishes and the dish samples;
and marking the association degree of the sample attribute information vector and the dish sample vector according to the corresponding relation.
Optionally, the second training module is further adapted to:
according to the corresponding relation, marking the association degree of the dish sample vector of the dish sample contained in the ordering behavior and the sample attribute information vector corresponding to the ordering behavior as a first association degree;
and/or marking the association degree of the dish sample vector of the dish sample not contained in the ordering behavior and the sample attribute information vector corresponding to the ordering behavior as a second association degree.
Optionally, the recommendation module is further adapted to: and sorting a plurality of dishes to be recommended according to the relevance value from high to low, and selecting a preset number of dishes to be recommended from the sorting result as recommended dishes to be recommended to a target user.
According to yet another aspect of the present invention, there is provided a computing device comprising: the system comprises a processor, a memory, a communication interface and a communication bus, wherein the processor, the memory and the communication interface complete mutual communication through the communication bus;
the memory is used for storing at least one executable instruction, and the executable instruction enables the processor to execute the operation corresponding to the dish recommending method.
According to still another aspect of the present invention, there is provided a computer storage medium having at least one executable instruction stored therein, the executable instruction causing a processor to perform operations corresponding to the dish recommendation method as described above.
According to the dish recommending method and device, the trained relevance model is used for predicting the relevance value of the attribute information vector of the target user and the dish vector of each dish to be recommended, the preference degree of the target user to each dish to be recommended in the current recommending scene is obtained, the preference degree of the target user to each dish to be recommended is not predicted depending on the historical ordering behavior of the target user, and the preference degree of the target user to each dish to be recommended can be accurately predicted by the method for predicting the target user with the insufficient ordering behavior; and then, dish recommendation is carried out according to the predicted relevance value, so that the matching degree of the recommended dishes recommended to the target user, the user attribute information of the target user and the scene attribute information of the current recommended scene is high, the recommendation efficiency is improved, and the ordering experience of the user is improved.
The foregoing description is only an overview of the technical solutions of the present invention, and the embodiments of the present invention are described below in order to make the technical means of the present invention more clearly understood and to make the above and other objects, features, and advantages of the present invention more clearly understandable.
Drawings
Various other advantages and benefits will become apparent to those of ordinary skill in the art upon reading the following detailed description of the preferred embodiments. The drawings are only for purposes of illustrating the preferred embodiments and are not to be construed as limiting the invention. Also, like reference numerals are used to refer to like parts throughout the drawings. In the drawings:
FIG. 1 illustrates a flow diagram of a dish recommendation method according to one embodiment of the present invention;
FIG. 2 illustrates a flow chart of a dish recommendation method according to another embodiment of the present invention;
FIG. 3 is a flow chart illustrating a process for training a relevance model in an embodiment of the invention;
FIG. 4 shows a functional block diagram of a dish recommendation device according to one embodiment of the present invention;
FIG. 5 illustrates a schematic structural diagram of a computing device according to an embodiment of the invention.
Detailed Description
Exemplary embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. While exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure may be embodied in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the disclosure to those skilled in the art.
Fig. 1 shows a flowchart of a dish recommendation method according to an embodiment of the present invention. As shown in fig. 1, the method includes:
step S101, mapping the user attribute information and the scene attribute information of the target user into corresponding attribute information vectors.
The target user refers to a user for whom dish recommendation needs to be performed currently; the user attribute information includes information related to the user representation, for example, information such as sex, age, and the like; the scene attribute information includes information related to the current recommended scene, for example, information on a recommended period, holidays, and the like.
In this step, the user attribute information and the scene attribute information of the target user are represented by using the vector. In the present invention, the specific way of vectorizing and characterizing the user attribute information and the scene attribute information is not limited, and those skilled in the art should understand that any method that can be used for vectorizing and characterizing the attribute information is included in the scope of the present invention.
And step S102, obtaining dish vectors of a plurality of dishes to be recommended.
The method comprises the following steps that a plurality of dishes to be recommended are matched with a current recommendation scene, and the dish to be recommended can be determined according to the position of a page where a target user is located; optionally, if the target user clicks to enter a specific shop, a plurality of dishes to be recommended are all or part of dishes in the specific shop; and if the target user does not enter any shop, determining a plurality of dishes to be recommended according to the promotion requirement corresponding to the page position of the target user.
Specifically, after a plurality of dishes to be recommended which are matched with the current recommendation scene are determined, dish vectors of the plurality of dishes to be recommended are obtained. The method for obtaining the dish vector is not specifically limited, and optionally, the dish vector can be obtained directly from other channels or generated by using a dish vectorization method.
And step S103, inputting the attribute information vector and the dish vector of each dish to be recommended into the trained association degree model to obtain the association degree value of the attribute information vector and the dish vector of each dish to be recommended.
Specifically, the relevance degree model is used for predicting to obtain the relevance degree value of the attribute information vector and the dish vector of each dish to be recommended, so that the preference degree of the target user for each dish to be recommended in the current recommendation scene is obtained, and the process of predicting and obtaining the preference degree is independent of the historical ordering behavior information of the target user.
And step S104, selecting a recommended dish from the plurality of dishes to be recommended according to the relevance value, and recommending the dish to the target user.
Specifically, dish recommendation is performed according to the relevance value, so that the matching degree of the recommended dishes recommended to the target user, the user attribute information of the target user and the scene attribute information of the current recommended scene is high, the recommendation efficiency is improved, and the ordering experience of the user is improved.
According to the dish recommending method provided by the embodiment, the trained relevance model is used for predicting the relevance value of the attribute information vector of the target user and the dish vector of each dish to be recommended, so that the preference degree of the target user for each dish to be recommended in the current recommending scene is obtained, and the preference degree of the target user for each dish to be recommended is not predicted depending on the historical meal ordering behavior of the target user, so that the preference degree of the target user with the insufficient meal ordering behavior can be accurately predicted; and then, dish recommendation is carried out according to the predicted relevance value, so that the matching degree of the recommended dishes recommended to the target user, the user attribute information of the target user and the scene attribute information of the current recommended scene is high, the recommendation efficiency is improved, and the ordering experience of the user is improved.
Fig. 2 shows a flowchart of a dish recommendation method according to another embodiment of the present invention. As shown in fig. 2, the method includes:
step S201, training to obtain a dish vectorization matrix.
In this embodiment, the dish vectors of a plurality of dishes to be recommended and the dish sample vectors of a plurality of dish samples used for training the association degree model are obtained through the dish vectorization matrix. Before this, a dish vectorization matrix is trained.
The dish vectorization matrix can be obtained by training the text features corresponding to the dish names by using a dish vectorization model; text features include, but are not limited to, radical, character components, and/or stroke features. The following describes the process of training to obtain a dish vectorization matrix by taking only stroke features as an example, but the present invention is not limited to this example:
specifically, a training corpus of a dish vectorization model is constructed according to a dish knowledge graph; acquiring training sample corpora from a training corpus, and inputting stroke characteristic information of the training sample corpora into an initialized dish vectorization model for training; and obtaining a dish vectorization matrix according to the training result. Further, performing word segmentation and other processing on the original corpus in the structured dish knowledge graph to obtain unstructured corpus data, and constructing a training corpus; selecting a preset number of corpora as training sample corpora, taking stroke characteristic information of a central word in each training sample corpus as training input of a dish vectorization model, and taking word vectors of upper and lower words of the central word as training output of the dish vectorization model for training; and obtaining a dish vectorization matrix according to the parameters of the dish vectorization model at the end of the training.
And step S202, mapping the stroke characteristic information corresponding to the dish name of each dish to be recommended into a dish vector of the dish to be recommended according to the dish vectorization matrix.
Multiplying the stroke characteristic information corresponding to the dish name of each dish to be recommended by the dish vectorization matrix, and mapping to obtain a word vector corresponding to the dish name of the dish to be recommended, namely the dish vector of the dish to be recommended.
Step S203, maps the user attribute information and the scene attribute information of the target user into corresponding attribute information vectors.
The user attribute information includes information related to the user portrait, and optionally, the information related to the user portrait data is information affecting the taste and the dish component preference of the user. For example, the user's residence may affect the user's taste preferences to some extent. The scene attribute information includes information related to the current recommended scene, and optionally, the information related to the current recommended scene is information that affects dish selection of the user. For example, the dishes that the user may select are different for different seasons or festivals, or different for weekdays and non-weekdays.
Specifically, user attribute information of a target user is acquired from a platform or a cooperation channel of the platform according to account information of the target user in a platform which currently provides ordering service, or associated information or real-name authentication information corresponding to the account information; determining scene attribute information according to data such as current date, time and weather; and representing the obtained user attribute information and scene attribute information of the target user by using a vector. Further, the user attribute information and the scene attribute information may each include information of a plurality of dimensions, and when mapping into corresponding attribute information vectors, the information of each dimension is mapped into corresponding attribute information vectors.
For example, the information of the target user's gender, age, and residence in three dimensions is selected as the user attribute information, and the information of the current time period, whether the current time period is a workday or not, and the information of the current holiday information in three dimensions is selected as the scene attribute information, and when the information is represented by a vector, the information of the six dimensions is respectively mapped into a vector of a preset dimension.
And step S204, inputting the attribute information vector and the dish vector of each dish to be recommended into the trained association degree model to obtain the association degree value of the attribute information vector and the dish vector of each dish to be recommended.
The relevancy model is obtained by training user attribute information and ordering behavior information of other users (which may also include the target user), and is trained without depending on historical ordering behavior information of the target user. FIG. 3 is a flow chart illustrating a process for training the relevance model in an embodiment of the invention. As shown in fig. 3, the training process includes:
step S301, obtaining dish sample vectors of a plurality of dish samples.
Specifically, the stroke characteristic information corresponding to the dish name of each dish sample is mapped into a dish sample vector of the dish sample according to the dish vectorization matrix. Further, the stroke characteristic information corresponding to the dish name of each dish sample is multiplied by the dish vectorization matrix, and a vector corresponding to the dish name of the dish sample, namely the dish sample vector of the dish sample, is obtained through mapping.
Step S302, collecting user sample attribute information of a food ordering user sample corresponding to a plurality of food ordering behaviors and scene sample attribute information of a food ordering scene; and mapping the user sample attribute information and the scene sample attribute information into corresponding sample attribute information vectors.
Specifically, a meal ordering user and a meal ordering scene are determined according to meal ordering data corresponding to each meal ordering behavior. Optionally, the user sample attribute information is acquired from a corresponding channel according to a food ordering account or a payment account, for example, user portrait data corresponding to an account number is acquired according to a payment account number. And determining scene sample attribute information corresponding to the ordering behavior according to the ordering date, time and weather. And then, representing the user sample attribute information and the scene sample attribute information corresponding to each ordering behavior by using the vector to obtain a sample attribute information vector corresponding to each ordering behavior sample.
And step S303, according to the multiple ordering behaviors, labeling the relevance between the sample attribute information vector and the dish sample vector to obtain a relevance labeling result.
Specifically, the relevance marking means that the selection condition of each dish sample by the ordering user corresponding to the ordering behavior at each time under the ordering scene corresponding to the ordering behavior is marked. And because the acquired ordering behaviors come from different ordering channels or different shops, the dish naming difference may exist, and correspondingly, the order-placed dishes and the dish samples contained in each ordering behavior are corresponding to each other before the relevance between the sample attribute information vector and the dish sample vector is labeled, so that the corresponding relation between the order-placed dishes and the dish samples is obtained. And labeling the association degree of the sample attribute information vector and the dish sample vector according to the corresponding relation.
Further, according to the corresponding relation, the relevance between the dish sample vector of the dish sample contained in the ordering behavior and the sample attribute information vector corresponding to the ordering behavior is marked as a first relevance; and/or marking the association degree of the dish sample vector of the dish sample not contained in the ordering behavior and the sample attribute information vector corresponding to the ordering behavior as a second association degree. For example, the ordering behavior includes ordering dishes a, b, and C, where a, b, and C correspond to dish samples A, B and C, respectively, the relevance between the dish sample vector of dish samples A, B and C and the sample attribute information vector corresponding to the ordering behavior is labeled as 1 (first relevance), and the relevance between the dish sample vector of the remaining dishes in the dish sample and the sample attribute information vector corresponding to the ordering behavior is labeled as 0 (second relevance).
And step S304, inputting the dish sample vector and the sample attribute information vector into the relevancy training model to obtain a relevancy output result output by the relevancy training model.
Specifically, dish sample vectors and sample attribute information vectors are used as training input data of the relevancy training model, and relevancy labeling results of the corresponding sample attribute information vectors and the dish sample vectors are used as training output data of the relevancy training model. In the training process, corresponding to each set of training input data (namely dish sample vectors and sample attribute information vectors), a correlation output result actually output by the correlation training model can be obtained.
And S305, training the relevance training model by using the loss between the relevance marking result and the relevance output result to obtain the relevance model.
Specifically, the loss between the actually output correlation output result and the correlation labeling result is calculated through a loss function, adaptive learning is performed according to the loss, training is stopped until the loss is reduced to a preset loss range, a correlation model is obtained, and training parameters when training is stopped are model parameters of the correlation model.
And S205, sorting a plurality of dishes to be recommended according to the relevance value from high to low, and selecting a preset number of dishes to be recommended from the sorting result as recommended dishes to be recommended to a target user.
Specifically, dish recommendation is performed according to the relevance value, so that the matching degree of the recommended dishes recommended to the target user, the user attribute information of the target user and the scene attribute information of the current recommended scene is high, the recommendation efficiency is improved, and the ordering experience of the user is improved.
In order to facilitate understanding of the implementation process of the present embodiment and the technical effect of the scheme adopting the present embodiment, the following describes a specific implementation process of the present embodiment as a specific application example: applying the trained dish vectorization matrix and the relevance model to a food ordering system, when detecting that a user 1 enters a shop A through the food ordering system to order food, acquiring dish names of 50 dishes in a menu provided by the shop A, and vectorizing the dish names through the dish vectorization matrix to obtain dish vectors corresponding to the 50 dishes; meanwhile, according to a user account input by the user 1 in the ordering system, user attribute information such as the gender, the age, the household address and the like of the user 1 is obtained, scene attribute information such as ordering time, holidays and the like is determined according to the current ordering scene, each information item in the user attribute information and the scene attribute information is mapped into a corresponding attribute information vector, and if 10 attribute information vectors are obtained in total; inputting the 10 attribute information vectors and the dish vector corresponding to each dish into the relevance model for prediction to obtain the relevance value of the dish vector and the attribute information vector; according to the predicted 50 relevance values corresponding to 50 dishes, 5 dish vectors corresponding to the maximum 5 relevance values are determined, dish names of 5 dishes corresponding to the 5 dish vectors are determined, and then the dish names corresponding to the 5 dishes and dish pictures corresponding to the dish names are displayed at preset positions of a page where the shop A is located to serve as selection references when a user orders dishes. Therefore, in the application example, the dish recommending process refers to the personalized factors and the scene factors of the user, so that the dish recommended to the user is more in line with the ordering requirement of the user; and the dish recommending process does not depend on the historical ordering behavior of the user, so that accurate dish recommendation can be performed for the user with little or no ordering data.
According to the dish recommendation method provided by the embodiment, training is performed according to text characteristic information corresponding to the name of the dish, so that a dish vectorization matrix is obtained; the dish vectorization matrix is utilized to obtain dish vectors of dishes to be recommended and dish sample vectors of dish samples for training the association degree model; the method comprises the steps of training food ordering data corresponding to existing food ordering behaviors to obtain a relevance model, predicting relevance values of attribute information vectors of target users and dish vectors of dishes of all dishes to be recommended by using the trained relevance model to obtain the preference degrees of the target users to all dishes to be recommended in a current recommendation scene, and predicting the preference degrees of the target users to all dishes to be recommended instead of relying on historical food ordering behaviors of the target users, wherein the method can accurately predict the preference degrees of any target users (including target users with insufficient food ordering behaviors); and then, dish recommendation is carried out according to the predicted relevance value, so that the matching degree of the recommended dishes recommended to the target user, the user attribute information of the target user and the scene attribute information of the current recommended scene is high, the recommendation efficiency is improved, and the ordering experience of the user is improved.
Fig. 4 shows a functional block diagram of a dish recommending apparatus according to an embodiment of the present invention. As shown in fig. 4, the apparatus includes: a mapping module 401, an obtaining module 402, a predicting module 403 and a recommending module 404; optionally, the apparatus further comprises: a first training module 405, a second training module 406.
A mapping module 401 adapted to map user attribute information and scene attribute information of a target user into corresponding attribute information vectors;
an obtaining module 402, adapted to obtain a dish vector of a plurality of dishes to be recommended;
the prediction module 403 is adapted to input the attribute information vector and the dish vector of each dish to be recommended into a trained relevancy model to obtain a relevancy value between the attribute information vector and the dish vector of each dish to be recommended;
and the recommending module 404 is adapted to select a recommended dish from the plurality of dishes to be recommended according to the relevance value and recommend the recommended dish to the target user.
In an alternative embodiment, the apparatus further comprises:
the first training module 405 is suitable for constructing a training corpus of the dish vectorization model according to the dish knowledge graph; acquiring training sample corpora from the training corpus, and inputting stroke characteristic information of the training sample corpora into an initialized dish vectorization model for training; and obtaining a dish vectorization matrix according to the training result.
In an alternative embodiment, the obtaining module 402 is further adapted to:
and mapping stroke characteristic information corresponding to the dish names of the dishes to be recommended into dish vectors of the dishes to be recommended according to the dish vectorization matrix.
In an alternative embodiment, the apparatus further comprises:
a second training module 406 adapted to obtain a dish sample vector for a plurality of dish samples; acquiring user sample attribute information of a food ordering user sample corresponding to the multi-time food ordering behavior and scene sample attribute information of a food ordering scene; mapping the user sample attribute information and the scene sample attribute information into corresponding sample attribute information vectors; according to the multiple ordering behaviors, labeling the relevance between the sample attribute information vector and the dish sample vector to obtain a relevance labeling result; inputting the dish sample vector and the sample attribute information vector into a relevancy training model to obtain a relevancy output result output by the relevancy training model; and training a relevance training model by using the loss between the relevance marking result and the relevance output result to obtain the relevance model.
In an alternative embodiment, the second training module 406 is further adapted to:
and mapping stroke characteristic information corresponding to the dish names of all the dish samples into dish vectors of the dish samples according to the dish vectorization matrix.
In an alternative embodiment, the second training module 406 is further adapted to:
the ordering dishes contained in each ordering behavior correspond to the dish samples to obtain the corresponding relation between the ordering dishes and the dish samples;
and marking the association degree of the sample attribute information vector and the dish sample vector according to the corresponding relation.
In an alternative embodiment, the second training module 406 is further adapted to:
according to the corresponding relation, marking the association degree of the dish sample vector of the dish sample contained in the ordering behavior and the sample attribute information vector corresponding to the ordering behavior as a first association degree;
and/or marking the association degree of the dish sample vector of the dish sample not contained in the ordering behavior and the sample attribute information vector corresponding to the ordering behavior as a second association degree.
In an alternative embodiment, the recommendation module 404 is further adapted to: and sorting a plurality of dishes to be recommended according to the relevance value from high to low, and selecting a preset number of dishes to be recommended from the sorting result as recommended dishes to be recommended to a target user.
The specific structure and operation principle of each module described above may refer to the description of the corresponding step in the method embodiment, and are not described herein again.
The embodiment of the application provides a nonvolatile computer storage medium, wherein the computer storage medium stores at least one executable instruction, and the computer executable instruction can execute the dish recommending method in any method embodiment.
Fig. 5 is a schematic structural diagram of a computing device according to an embodiment of the present invention, and the specific embodiment of the present invention does not limit the specific implementation of the computing device.
As shown in fig. 5, the computing device may include: a processor (processor)502, a Communications Interface 504, a memory 506, and a communication bus 508.
Wherein:
the processor 502, communication interface 504, and memory 506 communicate with one another via a communication bus 508.
A communication interface 504 for communicating with network elements of other devices, such as clients or other servers.
The processor 502 is configured to execute the program 510, and may specifically execute the relevant steps in the aforementioned dish recommending method embodiment.
In particular, program 510 may include program code that includes computer operating instructions.
The processor 502 may be a central processing unit CPU, or an Application Specific Integrated Circuit (ASIC), or one or more Integrated circuits configured to implement an embodiment of the present invention. The computing device includes one or more processors, which may be the same type of processor, such as one or more CPUs; or may be different types of processors such as one or more CPUs and one or more ASICs.
And a memory 506 for storing a program 510. The memory 506 may comprise high-speed RAM memory, and may also include non-volatile memory (non-volatile memory), such as at least one disk memory.
The program 510 may specifically be used to cause the processor 502 to perform the following operations:
mapping user attribute information and scene attribute information of a target user into corresponding attribute information vectors;
obtaining dish vectors of a plurality of dishes to be recommended;
inputting the attribute information vector and the dish vector of each dish to be recommended into a trained relevancy model to obtain relevancy values of the attribute information vector and the dish vector of each dish to be recommended;
and selecting a recommended dish from a plurality of dishes to be recommended according to the relevance value and recommending the dish to a target user.
In an alternative embodiment, the program 510 may specifically be further configured to cause the processor 502 to perform the following operations:
constructing a training corpus of a dish vectorization model according to the dish knowledge graph;
acquiring training sample corpora from the training corpus, and inputting stroke characteristic information of the training sample corpora into an initialized dish vectorization model for training; and obtaining a dish vectorization matrix according to the training result.
In an alternative embodiment, the program 510 may specifically be further configured to cause the processor 502 to perform the following operations:
and mapping stroke characteristic information corresponding to the dish names of the dishes to be recommended into dish vectors of the dishes to be recommended according to the dish vectorization matrix.
In an alternative embodiment, the program 510 may specifically be further configured to cause the processor 502 to perform the following operations:
obtaining dish sample vectors of a plurality of dish samples; acquiring user sample attribute information of a food ordering user sample corresponding to the multi-time food ordering behavior and scene sample attribute information of a food ordering scene; mapping the user sample attribute information and the scene sample attribute information into corresponding sample attribute information vectors;
according to the multiple ordering behaviors, labeling the relevance between the sample attribute information vector and the dish sample vector to obtain a relevance labeling result;
inputting the dish sample vector and the sample attribute information vector into a relevancy training model to obtain a relevancy output result output by the relevancy training model;
and training a relevance training model by using the loss between the relevance marking result and the relevance output result to obtain the relevance model.
In an alternative embodiment, the program 510 may specifically be further configured to cause the processor 502 to perform the following operations:
and mapping stroke characteristic information corresponding to the dish names of all the dish samples into dish vectors of the dish samples according to the dish vectorization matrix.
In an alternative embodiment, the program 510 may specifically be further configured to cause the processor 502 to perform the following operations:
the ordering dishes contained in each ordering behavior correspond to the dish samples to obtain the corresponding relation between the ordering dishes and the dish samples;
and marking the association degree of the sample attribute information vector and the dish sample vector according to the corresponding relation.
In an alternative embodiment, the program 510 may specifically be further configured to cause the processor 502 to perform the following operations:
according to the corresponding relation, marking the association degree of the dish sample vector of the dish sample contained in the ordering behavior and the sample attribute information vector corresponding to the ordering behavior as a first association degree;
and/or marking the association degree of the dish sample vector of the dish sample not contained in the ordering behavior and the sample attribute information vector corresponding to the ordering behavior as a second association degree.
In an alternative embodiment, the program 510 may specifically be further configured to cause the processor 502 to perform the following operations:
and sorting a plurality of dishes to be recommended according to the relevance value from high to low, and selecting a preset number of dishes to be recommended from the sorting result as recommended dishes to be recommended to a target user.
The algorithms and displays presented herein are not inherently related to any particular computer, virtual machine, or other apparatus. Various general purpose systems may also be used with the teachings herein. The required structure for constructing such a system will be apparent from the description above. Moreover, the present invention is not directed to any particular programming language. It is appreciated that a variety of programming languages may be used to implement the teachings of the present invention as described herein, and any descriptions of specific languages are provided above to disclose the best mode of the invention.
In the description provided herein, numerous specific details are set forth. It is understood, however, that embodiments of the invention may be practiced without these specific details. In some instances, well-known methods, structures and techniques have not been shown in detail in order not to obscure an understanding of this description.
Similarly, it should be appreciated that in the foregoing description of exemplary embodiments of the invention, various features of the invention are sometimes grouped together in a single embodiment, figure, or description thereof for the purpose of streamlining the disclosure and aiding in the understanding of one or more of the various inventive aspects. However, the disclosed method should not be interpreted as reflecting an intention that: that the invention as claimed requires more features than are expressly recited in each claim. Rather, as the following claims reflect, inventive aspects lie in less than all features of a single foregoing disclosed embodiment. Thus, the claims following the detailed description are hereby expressly incorporated into this detailed description, with each claim standing on its own as a separate embodiment of this invention.
Those skilled in the art will appreciate that the modules in the device in an embodiment may be adaptively changed and disposed in one or more devices different from the embodiment. The modules or units or components of the embodiments may be combined into one module or unit or component, and furthermore they may be divided into a plurality of sub-modules or sub-units or sub-components. All of the features disclosed in this specification (including any accompanying claims, abstract and drawings), and all of the processes or elements of any method or apparatus so disclosed, may be combined in any combination, except combinations where at least some of such features and/or processes or elements are mutually exclusive. Each feature disclosed in this specification (including any accompanying claims, abstract and drawings) may be replaced by alternative features serving the same, equivalent or similar purpose, unless expressly stated otherwise.
Furthermore, those skilled in the art will appreciate that while some embodiments described herein include some features included in other embodiments, rather than other features, combinations of features of different embodiments are meant to be within the scope of the invention and form different embodiments. For example, in the following claims, any of the claimed embodiments may be used in any combination.
The various component embodiments of the invention may be implemented in hardware, or in software modules running on one or more processors, or in a combination thereof. It will be appreciated by those skilled in the art that a microprocessor or Digital Signal Processor (DSP) may be used in practice to implement some or all of the functions of some or all of the components of the dish recommending apparatus according to an embodiment of the present invention. The present invention may also be embodied as apparatus or device programs (e.g., computer programs and computer program products) for performing a portion or all of the methods described herein. Such programs implementing the present invention may be stored on computer-readable media or may be in the form of one or more signals. Such a signal may be downloaded from an internet website or provided on a carrier signal or in any other form.
It should be noted that the above-mentioned embodiments illustrate rather than limit the invention, and that those skilled in the art will be able to design alternative embodiments without departing from the scope of the appended claims. In the claims, any reference signs placed between parentheses shall not be construed as limiting the claim. The word "comprising" does not exclude the presence of elements or steps not listed in a claim. The word "a" or "an" preceding an element does not exclude the presence of a plurality of such elements. The invention may be implemented by means of hardware comprising several distinct elements, and by means of a suitably programmed computer. In the unit claims enumerating several means, several of these means may be embodied by one and the same item of hardware. The usage of the words first, second and third, etcetera do not indicate any ordering. These words may be interpreted as names.
Claims (18)
1. A method of dish recommendation, comprising:
mapping user attribute information and scene attribute information of a target user into corresponding attribute information vectors; acquiring user attribute information of a target user according to account information of the target user in a platform for providing ordering service currently or real-name authentication information corresponding to the account information; the user attribute information comprises information influencing the taste and/or the dish preference of the user; determining scene attribute information according to the current date, time and weather data;
obtaining dish vectors of a plurality of dishes to be recommended;
inputting the attribute information vector and the dish vector of each dish to be recommended into a trained relevancy model to obtain relevancy values of the attribute information vector and the dish vector of each dish to be recommended; the dish sample vectors of a plurality of dish samples and sample attribute information vectors corresponding to the multi-time ordering behaviors of other users are used as training input data of a relevancy training model, and the relevancy training model obtains relevancy output results of the dish sample vectors and the sample attribute information vectors through training output;
and selecting a recommended dish from a plurality of dishes to be recommended according to the relevance value and recommending the dish to a target user.
2. The method of claim 1, wherein prior to the obtaining the dish vector for the plurality of dishes to be recommended, the method further comprises:
constructing a training corpus of a dish vectorization model according to the dish knowledge graph;
acquiring training sample corpora from the training corpus, and inputting stroke characteristic information of the training sample corpora into an initialized dish vectorization model for training; and obtaining a dish vectorization matrix according to the training result.
3. The method of claim 2, wherein the obtaining a dish vector for a plurality of dishes to be recommended further comprises:
and mapping stroke characteristic information corresponding to the dish names of the dishes to be recommended into dish vectors of the dishes to be recommended according to the dish vectorization matrix.
4. The method of claim 2, wherein the relevance model is trained by:
obtaining dish sample vectors of a plurality of dish samples; acquiring user sample attribute information of a food ordering user sample corresponding to the multi-time food ordering behavior and scene sample attribute information of a food ordering scene; mapping the user sample attribute information and the scene sample attribute information into corresponding sample attribute information vectors;
according to the multiple ordering behaviors, labeling the relevance between the sample attribute information vector and the dish sample vector to obtain a relevance labeling result;
inputting the dish sample vector and the sample attribute information vector into a relevancy training model to obtain a relevancy output result output by the relevancy training model;
and training a relevance training model by using the loss between the relevance marking result and the relevance output result to obtain the relevance model.
5. The method of claim 4, wherein the obtaining a dish sample vector for a plurality of dish samples further comprises:
and mapping stroke characteristic information corresponding to the dish names of all the dish samples into dish vectors of the dish samples according to the dish vectorization matrix.
6. The method of claim 4, wherein prior to said labeling the degree of association of the sample attribute information vector with the dish sample vector, the method further comprises:
the ordering dishes contained in each ordering behavior correspond to the dish samples to obtain the corresponding relation between the ordering dishes and the dish samples;
the step of labeling the relevance between the sample attribute information vector and the dish sample vector specifically comprises: and marking the association degree of the sample attribute information vector and the dish sample vector according to the corresponding relation.
7. The method of claim 6, wherein said labeling the relevance of the sample attribute information vector to the dish sample vector according to the correspondence further comprises:
according to the corresponding relation, marking the association degree of the dish sample vector of the dish sample contained in the ordering behavior and the sample attribute information vector corresponding to the ordering behavior as a first association degree;
and/or marking the association degree of the dish sample vector of the dish sample not contained in the ordering behavior and the sample attribute information vector corresponding to the ordering behavior as a second association degree.
8. The method of any of claims 1-7, wherein the selecting a recommended dish recommendation from a plurality of dishes to be recommended to a target user according to the relevancy values further comprises:
and sorting a plurality of dishes to be recommended according to the relevance value from high to low, and selecting a preset number of dishes to be recommended from the sorting result as recommended dishes to be recommended to a target user.
9. A dish recommendation device comprising:
the mapping module is suitable for mapping the user attribute information and the scene attribute information of the target user into corresponding attribute information vectors;
the acquisition module is suitable for acquiring dish vectors of a plurality of dishes to be recommended;
the prediction module is suitable for inputting the attribute information vector and the dish vector of each dish to be recommended into a trained relevance model to obtain a relevance value of the attribute information vector and the dish vector of each dish to be recommended;
the recommending module is suitable for selecting recommended dishes from a plurality of dishes to be recommended according to the relevance value and recommending the recommended dishes to a target user;
the apparatus is further adapted to: acquiring user attribute information of a target user according to account information of the target user in a platform for providing ordering service currently or real-name authentication information corresponding to the account information; the user attribute information comprises information influencing the taste and/or the dish preference of the user; determining scene attribute information according to the current date, time and weather data;
the apparatus is further adapted to: and taking dish sample vectors of a plurality of dish samples and sample attribute information vectors corresponding to the multi-time ordering behaviors of other users as training input data of the association degree model, and obtaining an association degree output result of the dish sample vectors and the sample attribute information vectors by training the association degree model to output.
10. The apparatus of claim 9, wherein the apparatus further comprises:
the first training module is suitable for constructing a training corpus of the dish vectorization model according to the dish knowledge graph; acquiring training sample corpora from the training corpus, and inputting stroke characteristic information of the training sample corpora into an initialized dish vectorization model for training; and obtaining a dish vectorization matrix according to the training result.
11. The apparatus of claim 10, wherein the acquisition module is further adapted to:
and mapping stroke characteristic information corresponding to the dish names of the dishes to be recommended into dish vectors of the dishes to be recommended according to the dish vectorization matrix.
12. The apparatus of claim 10, wherein the apparatus further comprises:
the second training module is suitable for obtaining dish sample vectors of a plurality of dish samples; acquiring user sample attribute information of a food ordering user sample corresponding to the multi-time food ordering behavior and scene sample attribute information of a food ordering scene; mapping the user sample attribute information and the scene sample attribute information into corresponding sample attribute information vectors; according to the multiple ordering behaviors, labeling the relevance between the sample attribute information vector and the dish sample vector to obtain a relevance labeling result; inputting the dish sample vector and the sample attribute information vector into a relevancy training model to obtain a relevancy output result output by the relevancy training model; and training a relevance training model by using the loss between the relevance marking result and the relevance output result to obtain the relevance model.
13. The apparatus of claim 12, wherein the second training module is further adapted to:
and mapping stroke characteristic information corresponding to the dish names of all the dish samples into dish vectors of the dish samples according to the dish vectorization matrix.
14. The apparatus of claim 12, wherein the second training module is further adapted to:
the ordering dishes contained in each ordering behavior correspond to the dish samples to obtain the corresponding relation between the ordering dishes and the dish samples;
and marking the association degree of the sample attribute information vector and the dish sample vector according to the corresponding relation.
15. The apparatus of claim 14, wherein the second training module is further adapted to:
according to the corresponding relation, marking the association degree of the dish sample vector of the dish sample contained in the ordering behavior and the sample attribute information vector corresponding to the ordering behavior as a first association degree;
and/or marking the association degree of the dish sample vector of the dish sample not contained in the ordering behavior and the sample attribute information vector corresponding to the ordering behavior as a second association degree.
16. The apparatus of any of claims 9-15, wherein the recommendation module is further adapted to: and sorting a plurality of dishes to be recommended according to the relevance value from high to low, and selecting a preset number of dishes to be recommended from the sorting result as recommended dishes to be recommended to a target user.
17. A computing device, comprising: the system comprises a processor, a memory, a communication interface and a communication bus, wherein the processor, the memory and the communication interface complete mutual communication through the communication bus;
the memory is used for storing at least one executable instruction, and the executable instruction causes the processor to execute the operation corresponding to the dish recommending method of any one of claims 1-8.
18. A computer storage medium having stored therein at least one executable instruction for causing a processor to perform operations corresponding to the dish recommendation method of any one of claims 1-8.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811069258.1A CN109300059B (en) | 2018-09-13 | 2018-09-13 | Dish recommending method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811069258.1A CN109300059B (en) | 2018-09-13 | 2018-09-13 | Dish recommending method and device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109300059A CN109300059A (en) | 2019-02-01 |
| CN109300059B true CN109300059B (en) | 2020-11-17 |
Family
ID=65167000
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811069258.1A Active CN109300059B (en) | 2018-09-13 | 2018-09-13 | Dish recommending method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109300059B (en) |
Families Citing this family (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111695960A (en) * | 2019-03-12 | 2020-09-22 | 阿里巴巴集团控股有限公司 | Object recommendation system, method, electronic device and storage medium |
| CN110162694B (en) * | 2019-04-02 | 2021-01-05 | 莫毓昌 | Recommendation system and method based on paired association rules |
| CN110135646B (en) * | 2019-05-20 | 2021-07-20 | 梁志鹏 | Restaurant pre-estimation quick serving method and device and storage medium |
| CN112016582B (en) * | 2019-05-31 | 2023-11-24 | 口口相传(北京)网络技术有限公司 | Dish recommending method and device |
| CN110782308A (en) * | 2019-09-20 | 2020-02-11 | 北京三快在线科技有限公司 | Pushing method and device for recommended package, electronic equipment and readable storage medium |
| CN110837552B (en) * | 2019-09-30 | 2022-12-09 | 口口相传(北京)网络技术有限公司 | Diet information recommendation method and device |
| CN111563788A (en) * | 2020-03-26 | 2020-08-21 | 口碑(上海)信息技术有限公司 | Object recommendation method and device |
| CN111666418B (en) * | 2020-04-23 | 2024-01-16 | 北京三快在线科技有限公司 | Text regeneration method, device, electronic equipment and computer readable medium |
Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103412936A (en) * | 2013-08-21 | 2013-11-27 | 五八同城信息技术有限公司 | Dish recommendation system based on data mining and cloud computing service |
| CN104133901A (en) * | 2014-08-04 | 2014-11-05 | 上海巨浪信息科技有限公司 | Mobile catering dish ordering and intelligent recommending system |
| CN106548006A (en) * | 2016-10-09 | 2017-03-29 | 浙江大学 | A kind of meals based on user's typical case's taste recommend method |
| CN107590246A (en) * | 2017-09-15 | 2018-01-16 | 百度在线网络技术（北京）有限公司 | Method and apparatus for pushed information |
| CN107665254A (en) * | 2017-09-30 | 2018-02-06 | 济南浪潮高新科技投资发展有限公司 | A kind of menu based on deep learning recommends method |
| CN107862542A (en) * | 2017-09-28 | 2018-03-30 | 北京三快在线科技有限公司 | A kind of vegetable recommends method and apparatus |
| CN107886348A (en) * | 2017-09-30 | 2018-04-06 | 厦门快商通信息技术有限公司 | A kind of artificial intelligence method of ordering and system |
| CN107886400A (en) * | 2017-11-15 | 2018-04-06 | 维沃移动通信有限公司 | A kind of recipe recommendation method, device and mobile terminal |
| CN108492144A (en) * | 2018-03-30 | 2018-09-04 | 百度在线网络技术（北京）有限公司 | Method and system, terminal and the computer readable storage medium that vegetable is recommended |
-
2018
- 2018-09-13 CN CN201811069258.1A patent/CN109300059B/en active Active
Patent Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103412936A (en) * | 2013-08-21 | 2013-11-27 | 五八同城信息技术有限公司 | Dish recommendation system based on data mining and cloud computing service |
| CN104133901A (en) * | 2014-08-04 | 2014-11-05 | 上海巨浪信息科技有限公司 | Mobile catering dish ordering and intelligent recommending system |
| CN106548006A (en) * | 2016-10-09 | 2017-03-29 | 浙江大学 | A kind of meals based on user's typical case's taste recommend method |
| CN107590246A (en) * | 2017-09-15 | 2018-01-16 | 百度在线网络技术（北京）有限公司 | Method and apparatus for pushed information |
| CN107862542A (en) * | 2017-09-28 | 2018-03-30 | 北京三快在线科技有限公司 | A kind of vegetable recommends method and apparatus |
| CN107665254A (en) * | 2017-09-30 | 2018-02-06 | 济南浪潮高新科技投资发展有限公司 | A kind of menu based on deep learning recommends method |
| CN107886348A (en) * | 2017-09-30 | 2018-04-06 | 厦门快商通信息技术有限公司 | A kind of artificial intelligence method of ordering and system |
| CN107886400A (en) * | 2017-11-15 | 2018-04-06 | 维沃移动通信有限公司 | A kind of recipe recommendation method, device and mobile terminal |
| CN108492144A (en) * | 2018-03-30 | 2018-09-04 | 百度在线网络技术（北京）有限公司 | Method and system, terminal and the computer readable storage medium that vegetable is recommended |
Non-Patent Citations (1)
| Title |
|---|
| 个性化菜品推荐系统的研究与实现;曾强;《中国优秀硕士学位论文全文数据库 信息科技辑》;20180115;全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109300059A (en) | 2019-02-01 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109300059B (en) | Dish recommending method and device | |
| CN107730313B (en) | Shop recommendation method and device based on recommendation reason | |
| CN110825957B (en) | Deep learning-based information recommendation method, device, equipment and storage medium | |
| CN110413877A (en) | A kind of resource recommendation method, device and electronic equipment | |
| CN107577807B (en) | Method and device for pushing information | |
| CN108596695B (en) | Entity pushing method and system | |
| CN107590174A (en) | Page access method and device | |
| CN111506849A (en) | Page generation method and device | |
| US20210390365A1 (en) | Aspect Pre-selection using Machine Learning | |
| US20200356781A1 (en) | Method and apparatus for performing categorised matching of videos, and selection engine | |
| CN110647696A (en) | Business object sorting method and device | |
| CN110580278A (en) | personalized search method, system, equipment and storage medium according to user portrait | |
| CN110597965B (en) | Emotion polarity analysis method and device for article, electronic equipment and storage medium | |
| CN111967924A (en) | Commodity recommendation method, commodity recommendation device, computer device, and medium | |
| CN111144974B (en) | Information display method and device | |
| CN110059172B (en) | Method and device for recommending answers based on natural language understanding | |
| CN111680213B (en) | Information recommendation method, data processing method and device | |
| US20190266648A1 (en) | Systems and methods for analyzing consumer products | |
| US11379898B2 (en) | Touch and pressure-based apparel image searching | |
| CN108197288B (en) | House source pushing method, device, equipment and medium according to user characteristic preference | |
| CN110515929A (en) | Books methods of exhibiting calculates equipment and storage medium | |
| CN112445909A (en) | Commodity retrieval method and device | |
| CN115618109A (en) | Content recommendation method and device, electronic equipment and computer-readable storage medium | |
| CN112487277B (en) | Data distribution method and device, readable storage medium and electronic equipment | |
| US11227122B1 (en) | Methods, mediums, and systems for representing a model in a memory of device |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |