CN117112915A - Intelligent design method and system based on user characteristics and big data training - Google Patents

Intelligent design method and system based on user characteristics and big data training Download PDF

Info

Publication number
CN117112915A
CN117112915A CN202311377926.8A CN202311377926A CN117112915A CN 117112915 A CN117112915 A CN 117112915A CN 202311377926 A CN202311377926 A CN 202311377926A CN 117112915 A CN117112915 A CN 117112915A
Authority
CN
China
Prior art keywords
user
information
interactive
acquiring
target
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.)
Granted
Application number
CN202311377926.8A
Other languages
Chinese (zh)
Other versions
CN117112915B (en
Inventor
刘再行
赖培源
王昌栋
杨永仕
李奎
廖晓东
赖剑煌
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Guangdong South China Technology Transfer Center Co ltd
Guangzhou Academy of Fine Arts
Sun Yat Sen University
Original Assignee
Guangdong South China Technology Transfer Center Co ltd
Guangzhou Academy of Fine Arts
Sun Yat Sen University
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Guangdong South China Technology Transfer Center Co ltd, Guangzhou Academy of Fine Arts, Sun Yat Sen University filed Critical Guangdong South China Technology Transfer Center Co ltd
Priority to CN202311377926.8A priority Critical patent/CN117112915B/en
Publication of CN117112915A publication Critical patent/CN117112915A/en
Application granted granted Critical
Publication of CN117112915B publication Critical patent/CN117112915B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/90—Details of database functions independent of the retrieved data types
    • G06F16/95—Retrieval from the web
    • G06F16/953—Querying, e.g. by the use of web search engines
    • G06F16/9535—Search customisation based on user profiles and personalisation
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/90—Details of database functions independent of the retrieved data types
    • G06F16/95—Retrieval from the web
    • G06F16/953—Querying, e.g. by the use of web search engines
    • G06F16/9536—Search customisation based on social or collaborative filtering
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/04—Architecture, e.g. interconnection topology
    • G06N3/0464—Convolutional networks [CNN, ConvNet]
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T17/00—Three dimensional [3D] modelling, e.g. data description of 3D objects
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T19/00—Manipulating 3D models or images for computer graphics
    • G06T19/20—Editing of 3D images, e.g. changing shapes or colours, aligning objects or positioning parts
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
    • Y02D10/00—Energy efficient computing, e.g. low power processors, power management or thermal management
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
    • Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
    • Y02P90/30—Computing systems specially adapted for manufacturing

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Databases & Information Systems (AREA)
  • General Physics & Mathematics (AREA)
  • General Engineering & Computer Science (AREA)
  • Software Systems (AREA)
  • Data Mining & Analysis (AREA)
  • Computer Graphics (AREA)
  • Health & Medical Sciences (AREA)
  • Biophysics (AREA)
  • Geometry (AREA)
  • Architecture (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Artificial Intelligence (AREA)
  • Biomedical Technology (AREA)
  • Computer Hardware Design (AREA)
  • Computational Linguistics (AREA)
  • Evolutionary Computation (AREA)
  • General Health & Medical Sciences (AREA)
  • Molecular Biology (AREA)
  • Computing Systems (AREA)
  • Mathematical Physics (AREA)
  • Electrically Operated Instructional Devices (AREA)

Abstract

The invention discloses an intelligent design method and system based on user characteristics and big data training, comprising the following steps: acquiring interactive information of a user to generate characteristic information of the user, and generating user information by combining basic information; acquiring target creation information of a target artwork by a user, preprocessing and formatting, and determining user demand information by combining the user information; constructing an intelligent artwork creation model based on a deep learning method, and constructing three-dimensional information of a target artwork based on user demand information; and feeding back the three-dimensional information of the target artware to the user, acquiring the interactive information of the user on the three-dimensional information, correcting the user demand information, updating the three-dimensional information of the target artware, and enriching the user characteristics by utilizing the corrected user demand information. According to the invention, the user characteristics are constructed through the user preferences, the collaborative recommendation characteristics and the like, so that the intelligent degree of the artwork design is improved, the exclusive artwork creation design of the user is realized, and the user experience is improved.

Description

Intelligent design method and system based on user characteristics and big data training
Technical Field
The invention relates to the technical field of artwork design, in particular to an intelligent design method and system based on user characteristics and big data training.
Background
In recent years, with the rapid development of artificial intelligence technology, the field of artistic creatives is changing over the sky, especially ChatGPT, midjourney, stable Diffusion and other tools appear, so that the threshold for creation is greatly reduced, more and more people are trying to use artificial intelligence technology for artistic creation, and the works not only have unique creative forms and expression forms, but also break through the boundaries of traditional art to a great extent. Meanwhile, the 3D model has unique advantages in the field of intelligent design and the editability of the 3D model in the field of intelligent design, and the upper limit of 2D design is expanded. On the one hand, from the user's point of view, in addition to basic merchandise information, more and more other information is incorporated into the final purchasing decision, for example, when purchasing large pieces of furniture, the user can consider whether the color and style of furniture and the hard clothing in the home match or not, whether the functionality and comfort meet the requirements of family use or not. On the other hand, standing at the seller's point of view, it also starts to discover the effectiveness of 3D model content authoring: for example, in the mobile phone industry, more and more mobile phone manufacturers are willing to invest money to use a 3D model to make short videos, and configure dynamic backgrounds of different colors for specifications and styles of different colors. Content creation based on the 3D model can further improve user experience and improve conversion efficiency to a certain extent.
At present, AI is popular, so that many consumers who do not have painting experience, but are willing to perform secondary creation, experience own special-style craftwork signs. Therefore, how to utilize the requirements and the user characteristics thereof in the process design to generate the customized design result of the artwork is a problem to be solved.
Disclosure of Invention
In order to solve the technical problems, the invention provides an intelligent design method and system based on user characteristics and big data training.
The first aspect of the invention provides an intelligent design method based on user characteristics and big data training, which comprises the following steps:
acquiring interactive information of a user, generating characteristic information of the user according to the interactive information, and generating user information based on the characteristic information and basic information;
acquiring target creation information of a target artwork by a user, preprocessing and formatting the target creation information, and determining user demand information by combining the user information;
constructing an intelligent artwork creation model based on a deep learning method, training by a big data means, and constructing three-dimensional information of a target artwork based on the user demand information through the intelligent artwork creation model;
And feeding back the three-dimensional information of the target artware to the user, acquiring the interactive information of the user on the three-dimensional information, correcting the user demand information, updating the three-dimensional information of the target artware, and enriching the user characteristics by utilizing the corrected user demand information.
In the scheme, the interactive information of the user is obtained, the characteristic information of the user is generated according to the interactive information, and the user information is generated based on the characteristic information and the basic information, specifically:
acquiring interaction information of a user in a historical time period, generating an interaction data set of the user, carrying out embedded representation of the user by utilizing the interaction data set, carrying out similarity calculation by using the embedded representation of the user, and taking the user with the similarity meeting a preset standard as a similar user;
acquiring embedded representations of the similar users, acquiring attention weights by an attention mechanism, characterizing associated features among the users, and acquiring feature information of the similar users by weighting and aggregating the attention weights;
extracting interactive items from interactive information of a user, sequencing the interactive items according to a time sequence to obtain an interactive time sequence, and obtaining initial weights of the interactive items in the interactive time sequence according to the interactive times and the interactive time length of the interactive items by the user;
Learning and representing a plurality of interaction time sequence sequences through a self-attention network, and acquiring characteristic information of the interaction items by combining initial weights and self-attention weights of the interaction items;
feature information of similar users corresponding to the users and feature information of interaction items are subjected to feature fusion to generate feature information of the users, and the feature information of the users is combined with basic information of the users to generate user information.
In the scheme, target creation information of a target artwork is obtained by a user, the target creation information is preprocessed and formatted, and user demand information is determined by combining the user information, specifically:
acquiring the authoring requirement of a user on a target artwork, segmenting the authoring requirement into words, acquiring word vectors of each word, and classifying the authoring requirement according to the word vectors to acquire a requirement category;
clustering the word vectors based on the demand category to generate a sub-data set, setting data tags of the sub-data set by using the demand category, and acquiring target creation information according to the sub-data set of each data tag;
preprocessing and formatting the target creation information according to a preset data format, and matching the target creation information through user information to generate user demand information of a user on a target artwork.
In the scheme, an intelligent artwork creation model is built based on a deep learning method, training is carried out through a big data means, and the method specifically comprises the following steps:
constructing an undirected graph in a low-dimensional space according to a user and an interactive item, taking the expression form of the interactive item as an additional characteristic of a node, and performing representation learning on the undirected graph through a graph convolution neural network;
constructing an intelligent artwork creation model based on a graph roll-up neural network, acquiring training data by using a big data means to perform model training, and importing the user demand information into the graph roll-up neural network to perform feature coding;
the vectorization representation of the user and the vectorization representation of the interactive item are obtained through the combination of a message transmission and neighbor aggregation mechanism of the graph convolution neural network, and the inner product of the vectorization representation of the user and the vectorization representation of the interactive item is calculated;
acquiring inner product sequences according to the inner products, acquiring a preset number of interactive items based on the inner product sequences, and outputting candidate interactive items corresponding to user demand information serving as a graph convolution neural network;
and combining the candidate interaction vector with a basic three-dimensional model of the target artwork to generate three-dimensional information of the target artwork.
In the scheme, three-dimensional information of the target artwork is fed back to a user, interaction information of the three-dimensional information of the user is obtained to correct the user demand information, and the three-dimensional information of the target artwork is updated, specifically:
Three-dimensional information of the target artware is obtained, the three-dimensional information is fed back to a user for visual display, interactive information of the user and the three-dimensional information of the target artware is obtained, and the interactive information is matched with a demand category;
acquiring association features among the user corresponding to the demand categories according to the historical user demands, generating revised recommendation information based on the interaction information of the user through the association features, and judging feedback of the user on the revised recommendation information;
judging the satisfaction degree of the user on the revised recommendation information through feedback of the user, stopping revised recommendation when the satisfaction degree is lower than a preset threshold value, and relearning the associated features;
after the user completes three-dimensional information interaction of the target artware, the final target artware three-dimensional information is sent to 3D printing equipment to generate works, and corrected user demand information is obtained.
In the scheme, the user characteristics are enriched by using the corrected user demand information, and the method specifically comprises the following steps:
acquiring historical user demand information in preset time, constructing a user demand sequence by combining the historical user demand information with a time stamp, and constructing a user personalized database based on the user demand sequence;
Acquiring deviation characteristics of the corrected user demand information and the user demand information, and searching in the user personalized database by utilizing the deviation characteristics to judge whether related data are contained;
if the user demand information does not contain the user demand information, the corrected user demand information is combined with the current time stamp to expand the data in the user personalized database;
if the user personalized database comprises the related data set, acquiring the related data set according to the search result, calculating the average Manhattan distance between the related data set and the related data set by utilizing the deviation feature, and acquiring the membership degree of the deviation feature in the user personalized database;
when the membership is greater than or equal to a preset membership threshold, replacing the related data set according to the deviation feature, and when the membership is less than the preset membership threshold, updating the related data set;
and updating the user characteristics through the real-time updating of the user personalized database, and performing optimization training on the intelligent artwork creation model.
The second aspect of the present invention also provides an intelligent design system based on user characteristics and big data training, the system comprising: the system comprises a user characteristic module, an intelligent creation module, a user interaction module, a user personalized database module, a copyright application module and a transaction management module:
The user characteristic module is used for acquiring characteristic information of a user and generating user information based on the characteristic information and basic information;
the intelligent creation module is used for acquiring user demand information, constructing an intelligent creation model of the artwork based on a deep learning method, constructing three-dimensional information of a target artwork based on the user demand information through the intelligent creation model of the artwork, and returning three-dimensional information of the artwork;
the user interaction module is used for acquiring interaction information of a user on the returned three-dimensional information of the artwork, updating and adjusting the three-dimensional information of the target artwork by using user feedback, and carrying out demand adjustment;
the user personalized database module is used for acquiring user characteristic update data, expanding a training data set of the intelligent creation module, enriching user characteristics and continuously updating and optimizing an intelligent creation model of the artwork;
after a user determines a target artwork, the copyright application module submits three-dimensional information of the target artwork to a blockchain platform to obtain a digital copyright uplink record and lock work creation information;
the transaction management module forms an order based on the price list according to the requirements of the user, and works are created after the payment of the order is received.
The invention discloses an intelligent design method and system based on user characteristics and big data training, comprising the following steps: acquiring interactive information of a user to generate characteristic information of the user, and generating user information by combining basic information; acquiring target creation information of a target artwork by a user, preprocessing and formatting, and determining user demand information by combining the user information; constructing an intelligent artwork creation model based on a deep learning method, training by a big data means, and constructing three-dimensional information of a target artwork based on user demand information; and feeding back the three-dimensional information of the target artware to the user, acquiring the interactive information of the user on the three-dimensional information, correcting the user demand information, updating the three-dimensional information of the target artware, and enriching the user characteristics by utilizing the corrected user demand information. The user characteristics are built by the preference, collaborative recommendation characteristics and the like of the user, the intelligent degree of the artwork design is improved, the exclusive artwork creation design of the user is realized, and the user experience is improved.
Drawings
FIG. 1 shows a flow chart of an intelligent design method based on user features and big data training of the present invention;
FIG. 2 shows a flow chart of a method for constructing an intelligent artwork creation model according to the present application;
FIG. 3 is a flow chart of a method of enriching user features using modified user demand information in accordance with the present application;
FIG. 4 shows a block diagram of an intelligent design system based on user features and big data training of the present application.
Detailed Description
In order that the above-recited objects, features and advantages of the present application will be more clearly understood, a more particular description of the application will be rendered by reference to the appended drawings and appended detailed description. It should be noted that, without conflict, the embodiments of the present application and features in the embodiments may be combined with each other.
In the following description, numerous specific details are set forth in order to provide a thorough understanding of the present application, however, the present application may be practiced in other ways than those described herein, and therefore the scope of the present application is not limited to the specific embodiments disclosed below.
FIG. 1 shows a flow chart of an intelligent design method based on user characteristics and big data training of the present application.
As shown in fig. 1, a first aspect of the present application provides an intelligent design method based on user characteristics and big data training, including:
S102, acquiring interaction information of a user, generating characteristic information of the user according to the interaction information, and generating user information based on the characteristic information and basic information;
s104, acquiring target creation information of a target artwork by a user, preprocessing and formatting the target creation information, and determining user demand information by combining the user information;
s106, constructing an intelligent artwork creation model based on a deep learning method, training by a big data means, and constructing three-dimensional information of a target artwork based on the user demand information through the intelligent artwork creation model;
s108, feeding back the three-dimensional information of the target artwork to the user, acquiring the interactive information of the user on the three-dimensional information, correcting the user demand information, updating the three-dimensional information of the target artwork, and enriching the user characteristics by utilizing the corrected user demand information.
It should be noted that, acquiring interactive information of a user in a process design or related historical time period of a related website of the artwork, including browsing records, transaction records and the like, generating an interactive data set of the user, performing embedded representation of the user by using the interactive data set, performing similarity calculation by using the embedded representation of the user, and taking the user with the similarity meeting a preset standard as a similar user; acquiring an embedded representation of the similar user, the attention-introducing mechanism acquiring an attention weight, the attention weight Calculated as +.>，/>Embedded representation representing user->The method comprises the steps of representing embedded representations of similar users i, n representing the total number of the similar users, T representing matrix transposition, representing association features among users through attention weights, and obtaining feature information of the similar users through weighted aggregation of the attention weights; extracting interactive items from interactive information of a user, sorting the interactive items according to time sequence to obtain an interactive time sequence, and carrying out interaction times on the interactive items according to the userAcquiring initial weights of interactive items in the interactive time sequence by the number and the interactive time length; the self-attention network learns and represents a plurality of interactive time sequence sequences, and the characteristic information of the interactive items is obtained by combining the initial weight of the interactive items and the self-attention weight of the interactive items, wherein the self-attention weight is->The calculation formula of (2) is as follows: />,/>Representing the number of characteristic items>Representing matrix size, +.>Representing a query vector->Representing key vectors +_>Representing a value vector; and carrying out feature fusion on the feature information of the similar user corresponding to the user and the feature information of the interactive item by utilizing a gating unit or an LSTM unit to generate the feature information of the user, and combining the feature information of the user with the basic information of the user to generate the user information.
FIG. 2 shows a flow chart of a method for constructing an intelligent artwork creation model according to the present invention.
According to the embodiment of the invention, an intelligent artwork creation model is constructed based on a deep learning method, and training is performed by a big data means, specifically:
s202, constructing an undirected graph in a low-dimensional space according to a user and an interactive item, and performing representation learning on the undirected graph through a graph convolution neural network by taking the expression form of the interactive item as an additional feature of a node;
s204, an intelligent artwork creation model is built based on the graph roll-up neural network, training data are obtained by using a big data means to carry out model training, and the user demand information is imported into the graph roll-up neural network to carry out feature coding;
s206, combining the vectorization representation of the user and the vectorization representation of the interactive item through a message transmission and neighbor aggregation mechanism of the graph convolution neural network, and calculating the inner product of the vectorization representation of the user and the vectorization representation of the interactive item;
s208, acquiring inner product sequences according to the inner products, and acquiring a preset number of interaction items based on the inner product sequences, wherein the interaction items are used as candidate interaction items corresponding to user demand information and are used as output of a graph convolution neural network;
s210, combining the candidate interaction vector with a basic three-dimensional model of the target artwork to generate three-dimensional information of the target artwork.
It should be noted that, the authoring requirements of the user on the target artwork are obtained, including the form, theme, color, mood, etc. of the target artwork, for example: the cup needs are customized, the main tone is blue, the blue sky and white cloud theme is the mood of a high-altitude bird. Multiple requirements, such as customizing a cup or hanging ornament, may also be submitted simultaneously. Word segmentation is carried out on the creation requirements, word vectors of all words are obtained, the creation requirements are classified according to the word vectors, and requirement categories are obtained, wherein the requirement categories comprise color categories, shape categories, style categories and the like; clustering the word vectors based on the demand category to generate a sub-data set, setting data tags of the sub-data set by using the demand category, and acquiring target creation information according to the sub-data set of each data tag; preprocessing and formatting the target creation information according to a preset data format, and matching the target creation information through user information to generate user demand information of a user on a target artwork. In addition, the demands submitted by the users are formatted and divided, a demand order is generated and submitted to transaction management, and the transaction management returns the user after generating the order and the total price according to the unit price and the demands.
And acquiring mass artwork data from the works of the universities, graduation designs, the artwork on the electronic commerce platform, the digital work of the blockchain copyright platform and the works of the large artistic exhibitions by using a large data means as training data.
The three-dimensional information of the target artware is acquired, the three-dimensional information is fed back to a user for visual display, the interactive information of the user and the three-dimensional information of the target artware is acquired, and the interactive information is matched with the demand category; acquiring associated features, such as corresponding changes of colors or shapes, of the user corresponding to the demand categories according to the historical user demands, generating revised recommendation information based on interaction information of the user through the associated features, and judging feedback of the user on the revised recommendation information; for example, in the process of customizing the cup, if the current interaction information of the user is to change the color of the cup body, the color of the cup handle is revised and recommended according to the associated features of the historical user demands, the satisfaction degree of the user on the revised recommendation information is judged through feedback of the user, and when the satisfaction degree is lower than a preset threshold value, the revised recommendation is stopped, and the associated features are learned again; after the user completes three-dimensional information interaction of the target artware, the final target artware three-dimensional information is sent to 3D printing equipment to generate works, and corrected user demand information is obtained. After the user selects, the creation model submits the works to a blockchain platform to obtain the digital copyright uplink records and lock the creation information of the works.
FIG. 3 is a flow chart of a method of enriching user features using modified user demand information in accordance with the present invention.
According to the embodiment of the invention, the user characteristics are enriched by using the corrected user demand information, and the method specifically comprises the following steps:
s302, acquiring historical user demand information in preset time, constructing a user demand sequence by combining the historical user demand information with a time stamp, and constructing a user personalized database based on the user demand sequence;
s304, obtaining deviation characteristics of the corrected user demand information and the user demand information, and searching in the user personalized database by utilizing the deviation characteristics to judge whether related data are contained;
s306, if the user request information does not contain the user request information, the corrected user request information is combined with the current time stamp to expand the data in the user personalized database;
s308, if the user personalized database comprises the related data set is obtained according to the search result, the average Manhattan distance between the related data set and the deviation feature is calculated, and the membership degree of the deviation feature in the user personalized database is obtained;
s310, when the membership is greater than or equal to a preset membership threshold, replacing the related data set according to the deviation feature, and when the membership is less than the preset membership threshold, updating the related data set;
S312, updating the user characteristics through real-time updating of the user personalized database, and performing optimization training on the intelligent artwork creation model.
The method includes the steps that a craft evaluation model is built, evaluation factors of craft evaluation are obtained in a related database through similarity calculation by a big data method, data statistics is conducted on the evaluation factors, and multi-dimensional evaluation indexes are obtained according to data statistics results; acquiring expert experience of artwork evaluation based on a knowledge graph through the multidimensional evaluation index, training a BP neural network through the expert experience, acquiring data containing an evaluation label, constructing a training set, training the BP neural network to acquire an artwork evaluation model, importing three-dimensional information of a target artwork into the artwork evaluation model, acquiring an evaluation scoring result of the target artwork, optimizing by utilizing a relevant knowledge graph according to the evaluation scoring result, and providing a modification suggestion or providing an optimization suggestion by an expert in the field of systematic distribution of the artwork.
FIG. 4 shows a block diagram of an intelligent design system based on user features and big data training of the present invention.
The second aspect of the present invention also provides an intelligent design system based on user characteristics and big data training, the system comprising: the system comprises a user characteristic module, an intelligent creation module, a user interaction module, a user personalized database module, a copyright application module and a transaction management module:
The user characteristic module is used for acquiring characteristic information of a user and generating user information based on the characteristic information and basic information;
the intelligent creation module is used for acquiring user demand information, constructing an intelligent creation model of the artwork based on a deep learning method, constructing three-dimensional information of a target artwork based on the user demand information through the intelligent creation model of the artwork, and returning three-dimensional information of the artwork;
the user interaction module is used for acquiring interaction information of a user on the returned three-dimensional information of the artwork, updating and adjusting the three-dimensional information of the target artwork by using user feedback, and carrying out demand adjustment;
the user personalized database module is used for acquiring user characteristic update data, expanding a training data set of the intelligent creation module, enriching user characteristics and continuously updating and optimizing an intelligent creation model of the artwork;
after a user determines a target artwork, the copyright application module submits three-dimensional information of the target artwork to a blockchain platform to obtain a digital copyright uplink record and lock work creation information;
the transaction management module forms an order based on the price list according to the requirements of the user, and works are created after the payment of the order is received.
It should be noted that, the user feature module obtains the interactive information of the user in the relevant historical time period of the process design or the related website of the artwork, including browsing records, transaction records, etc., generates the interactive data set of the user, uses the interactive data set to perform the embedded representation of the user, performs the similarity calculation through the embedded representation of the user, and uses the user with the similarity meeting the preset standard as the similar user; acquiring an embedded representation of the similar user, the attention-introducing mechanism acquiring an attention weight, the attention weightCalculated as +.>，/>Embedded representation representing user->The method comprises the steps of representing embedded representations of similar users i, n representing the total number of the similar users, T representing matrix transposition, representing association features among users through attention weights, and obtaining feature information of the similar users through weighted aggregation of the attention weights; extracting interactive items from interactive information of a user, sequencing the interactive items according to a time sequence to obtain an interactive time sequence, and obtaining initial weights of the interactive items in the interactive time sequence according to the interactive times and the interactive time length of the interactive items by the user; the self-attention network learns and represents a plurality of interactive time sequence sequences, and the characteristic information of the interactive items is obtained by combining the initial weight of the interactive items and the self-attention weight of the interactive items, wherein the self-attention weight is- >The calculation formula of (2) is as follows: />,/>Representing the number of characteristic items>Representing matrix size, +.>Representing a query vector->Representing key vectors +_>Representing a value vector; feature information of similar users corresponding to the users and feature information of interaction items are subjected to feature fusion by utilizing a gating unit or an LSTM unit to generate feature information of the users, and the feature information of the users is combined with basic information of the users to generate user informationAnd (5) extinguishing.
It should be noted that, in the intelligent creation module, an intelligent creation model of the artwork is built based on a deep learning method, and training is performed by a big data means, specifically: constructing an undirected graph in a low-dimensional space according to a user and an interactive item, taking the expression form of the interactive item as an additional characteristic of a node, and performing representation learning on the undirected graph through a graph convolution neural network; constructing an intelligent artwork creation model based on a graph roll-up neural network, acquiring training data by using a big data means to perform model training, and importing the user demand information into the graph roll-up neural network to perform feature coding; the vectorization representation of the user and the vectorization representation of the interactive item are obtained through the combination of a message transmission and neighbor aggregation mechanism of the graph convolution neural network, and the inner product of the vectorization representation of the user and the vectorization representation of the interactive item is calculated; acquiring inner product sequences according to the inner products, acquiring a preset number of interactive items based on the inner product sequences, and outputting candidate interactive items corresponding to user demand information serving as a graph convolution neural network; and combining the candidate interaction vector with a basic three-dimensional model of the target artwork to generate three-dimensional information of the target artwork.
It should be noted that, the authoring requirements of the user on the target artwork are obtained, including the form, theme, color, mood, etc. of the target artwork, for example: the cup needs are customized, the main tone is blue, the blue sky and white cloud theme is the mood of a high-altitude bird. Multiple requirements, such as customizing a cup or hanging ornament, may also be submitted simultaneously. Word segmentation is carried out on the creation demands, word vectors of all words are obtained, and the creation demands are classified according to the word vectors to obtain demand categories; clustering the word vectors based on the demand category to generate a sub-data set, setting data tags of the sub-data set by using the demand category, and acquiring target creation information according to the sub-data set of each data tag; preprocessing and formatting the target creation information according to a preset data format, and matching the target creation information through user information to generate user demand information of a user on a target artwork. In addition, the demands submitted by the users are formatted and divided, a demand order is generated and submitted to transaction management, and the transaction management returns the user after generating the order and the total price according to the unit price and the demands.
And acquiring mass artwork data from the works of the universities, graduation designs, the artwork on the electronic commerce platform, the digital work of the blockchain copyright platform and the works of the large artistic exhibitions by using a large data means as training data.
In the user interaction module, three-dimensional information of the target artware is obtained, the three-dimensional information is fed back to a user for visual display, interaction information of the user and the three-dimensional information of the target artware is obtained, and the interaction information is matched with a demand type; acquiring associated features, such as corresponding changes of colors or shapes, of the user corresponding to the demand categories according to the historical user demands, generating revised recommendation information based on interaction information of the user through the associated features, and judging feedback of the user on the revised recommendation information; for example, in the process of customizing the cup, if the current interaction information of the user is to change the color of the cup body, the color of the cup handle is revised and recommended according to the associated features of the historical user demands, the satisfaction degree of the user on the revised recommendation information is judged through feedback of the user, and when the satisfaction degree is lower than a preset threshold value, the revised recommendation is stopped, and the associated features are learned again; after the user completes three-dimensional information interaction of the target artware, the final target artware three-dimensional information is sent to 3D printing equipment to generate works, and corrected user demand information is obtained. After the user selects, the creation model submits the works to a blockchain platform to obtain the digital copyright uplink records and lock the creation information of the works.
It should be noted that, the user characteristics are enriched in the user personalized database by using the corrected user demand information, specifically: acquiring historical user demand information in preset time, constructing a user demand sequence by combining the historical user demand information with a time stamp, and constructing a user personalized database based on the user demand sequence; acquiring deviation characteristics of the corrected user demand information and the user demand information, and searching in the user personalized database by utilizing the deviation characteristics to judge whether related data are contained; if the user demand information does not contain the user demand information, the corrected user demand information is combined with the current time stamp to expand the data in the user personalized database; if the user personalized database comprises the related data set, acquiring the related data set according to the search result, calculating the average Manhattan distance between the related data set and the related data set by utilizing the deviation feature, and acquiring the membership degree of the deviation feature in the user personalized database; when the membership is greater than or equal to a preset membership threshold, replacing the related data set according to the deviation feature, and when the membership is less than the preset membership threshold, updating the related data set; and updating the user characteristics through the real-time updating of the user personalized database, and performing optimization training on the intelligent artwork creation model.
The third aspect of the present application also provides a computer readable storage medium, where the computer readable storage medium includes an intelligent design method program based on user characteristics and big data training, where the intelligent design method program based on user characteristics and big data training implements the steps of the intelligent design method based on user characteristics and big data training as described in any one of the above.
In the several embodiments provided by the present application, it should be understood that the disclosed apparatus and method may be implemented in other ways. The above described device embodiments are only illustrative, e.g. the division of the units is only one logical function division, and there may be other divisions in practice, such as: multiple units or components may be combined or may be integrated into another system, or some features may be omitted, or not performed. In addition, the various components shown or discussed may be coupled or directly coupled or communicatively coupled to each other via some interface, whether indirectly coupled or communicatively coupled to devices or units, whether electrically, mechanically, or otherwise.
The units described above as separate components may or may not be physically separate, and components shown as units may or may not be physical units; can be located in one place or distributed to a plurality of network units; some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.
In addition, each functional unit in each embodiment of the present invention may be integrated in one processing unit, or each unit may be separately used as one unit, or two or more units may be integrated in one unit; the integrated units may be implemented in hardware or in hardware plus software functional units.
Those of ordinary skill in the art will appreciate that: all or part of the steps for implementing the above method embodiments may be implemented by hardware related to program instructions, and the foregoing program may be stored in a computer readable storage medium, where the program, when executed, performs steps including the above method embodiments; and the aforementioned storage medium includes: a mobile storage device, a Read-Only Memory (ROM), a random access Memory (RAM, random Access Memory), a magnetic disk or an optical disk, or the like, which can store program codes.
Alternatively, the above-described integrated units of the present invention may be stored in a computer-readable storage medium if implemented in the form of software functional modules and sold or used as separate products. Based on such understanding, the technical solutions of the embodiments of the present invention may be embodied in essence or a part contributing to the prior art in the form of a software product stored in a storage medium, including several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the methods described in the embodiments of the present invention. And the aforementioned storage medium includes: a removable storage device, ROM, RAM, magnetic or optical disk, or other medium capable of storing program code.
The foregoing is merely illustrative of the present invention, and the present invention is not limited thereto, and any person skilled in the art will readily recognize that variations or substitutions are within the scope of the present invention. Therefore, the protection scope of the present invention shall be subject to the protection scope of the claims.

Claims (10)

1. An intelligent design method based on user characteristics and big data training is characterized by comprising the following steps:
acquiring interactive information of a user, generating characteristic information of the user according to the interactive information, and generating user information based on the characteristic information and basic information;
acquiring target creation information of a target artwork by a user, preprocessing and formatting the target creation information, and determining user demand information by combining the user information;
constructing an intelligent artwork creation model based on a deep learning method, training by a big data means, and constructing three-dimensional information of a target artwork based on the user demand information through the intelligent artwork creation model;
and feeding back the three-dimensional information of the target artware to the user, acquiring the interactive information of the user on the three-dimensional information, correcting the user demand information, updating the three-dimensional information of the target artware, and enriching the user characteristics by utilizing the corrected user demand information.
2. The intelligent design method based on user characteristics and big data training according to claim 1, wherein the method is characterized in that the interactive information of the user is obtained, the characteristic information of the user is generated according to the interactive information, and the user information is generated based on the characteristic information and basic information, specifically:
Acquiring interaction information of a user in a historical time period, generating an interaction data set of the user, carrying out embedded representation of the user by utilizing the interaction data set, carrying out similarity calculation by using the embedded representation of the user, and taking the user with the similarity meeting a preset standard as a similar user;
acquiring embedded representations of the similar users, acquiring attention weights by an attention mechanism, characterizing associated features among the users, and acquiring feature information of the similar users by weighting and aggregating the attention weights;
extracting interactive items from interactive information of a user, sequencing the interactive items according to a time sequence to obtain an interactive time sequence, and obtaining initial weights of the interactive items in the interactive time sequence according to the interactive times and the interactive time length of the interactive items by the user;
learning and representing a plurality of interaction time sequence sequences through a self-attention network, and acquiring characteristic information of the interaction items by combining initial weights and self-attention weights of the interaction items;
feature information of similar users corresponding to the users and feature information of interaction items are subjected to feature fusion to generate feature information of the users, and the feature information of the users is combined with basic information of the users to generate user information.
3. The intelligent design method based on user characteristics and big data training according to claim 1, wherein the method is characterized in that target creation information of a target artwork is obtained by a user, the target creation information is preprocessed and formatted, and user demand information is determined by combining the user information, specifically:
acquiring the authoring requirement of a user on a target artwork, segmenting the authoring requirement into words, acquiring word vectors of each word, and classifying the authoring requirement according to the word vectors to acquire a requirement category;
clustering the word vectors based on the demand category to generate a sub-data set, setting data tags of the sub-data set by using the demand category, and acquiring target creation information according to the sub-data set of each data tag;
preprocessing and formatting the target creation information according to a preset data format, and matching the target creation information through user information to generate user demand information of a user on a target artwork.
4. The intelligent design method based on user characteristics and big data training according to claim 1, wherein the intelligent artwork creation model is constructed based on a deep learning method, and training is performed by big data means, specifically:
Constructing an undirected graph in a low-dimensional space according to a user and an interactive item, taking the expression form of the interactive item as an additional characteristic of a node, and performing representation learning on the undirected graph through a graph convolution neural network;
constructing an intelligent artwork creation model based on a graph roll-up neural network, acquiring training data by using a big data means to perform model training, and importing the user demand information into the graph roll-up neural network to perform feature coding;
the vectorization representation of the user and the vectorization representation of the interactive item are obtained through the combination of a message transmission and neighbor aggregation mechanism of the graph convolution neural network, and the inner product of the vectorization representation of the user and the vectorization representation of the interactive item is calculated;
acquiring inner product sequences according to the inner products, acquiring a preset number of interactive items based on the inner product sequences, and outputting candidate interactive items corresponding to user demand information serving as a graph convolution neural network;
and combining the candidate interaction vector with a basic three-dimensional model of the target artwork to generate three-dimensional information of the target artwork.
5. The intelligent design method based on user characteristics and big data training according to claim 1, wherein three-dimensional information of a target artwork is fed back to a user, interaction information of the user on the three-dimensional information is obtained to correct the user demand information, and the three-dimensional information of the target artwork is updated, specifically:
Three-dimensional information of the target artware is obtained, the three-dimensional information is fed back to a user for visual display, interactive information of the user and the three-dimensional information of the target artware is obtained, and the interactive information is matched with a demand category;
acquiring association features among the user corresponding to the demand categories according to the historical user demands, generating revised recommendation information based on the interaction information of the user through the association features, and judging feedback of the user on the revised recommendation information;
judging the satisfaction degree of the user on the revised recommendation information through feedback of the user, stopping revised recommendation when the satisfaction degree is lower than a preset threshold value, and relearning the associated features;
after the user completes three-dimensional information interaction of the target artware, the final target artware three-dimensional information is sent to 3D printing equipment to generate works, and corrected user demand information is obtained.
6. The intelligent design method based on user characteristics and big data training according to claim 1, wherein the user characteristics are enriched by using the corrected user demand information, specifically:
acquiring historical user demand information in preset time, constructing a user demand sequence by combining the historical user demand information with a time stamp, and constructing a user personalized database based on the user demand sequence;
Acquiring deviation characteristics of the corrected user demand information and the user demand information, and searching in the user personalized database by utilizing the deviation characteristics to judge whether related data are contained;
if the user demand information does not contain the user demand information, the corrected user demand information is combined with the current time stamp to expand the data in the user personalized database;
if the user personalized database comprises the related data set, acquiring the related data set according to the search result, calculating the average Manhattan distance between the related data set and the related data set by utilizing the deviation feature, and acquiring the membership degree of the deviation feature in the user personalized database;
when the membership is greater than or equal to a preset membership threshold, replacing the related data set according to the deviation feature, and when the membership is less than the preset membership threshold, updating the related data set;
and updating the user characteristics through the real-time updating of the user personalized database, and performing optimization training on the intelligent artwork creation model.
7. An intelligent design system based on user characteristics and big data training, which is characterized in that the system comprises: the system comprises a user characteristic module, an intelligent creation module, a user interaction module, a user personalized database module, a copyright application module and a transaction management module:
The user characteristic module is used for acquiring characteristic information of a user and generating user information based on the characteristic information and basic information;
the intelligent creation module is used for acquiring user demand information, constructing an intelligent creation model of the artwork based on a deep learning method, constructing three-dimensional information of a target artwork based on the user demand information through the intelligent creation model of the artwork, and returning three-dimensional information of the artwork;
the user interaction module is used for acquiring interaction information of a user on the returned three-dimensional information of the artwork, updating and adjusting the three-dimensional information of the target artwork by using user feedback, and carrying out demand adjustment;
the user personalized database module is used for acquiring user characteristic update data, expanding a training data set of the intelligent creation module, enriching user characteristics and continuously updating and optimizing an intelligent creation model of the artwork;
after a user determines a target artwork, the copyright application module submits three-dimensional information of the target artwork to a blockchain platform to obtain a digital copyright uplink record and lock work creation information;
the transaction management module forms an order based on the price list according to the requirements of the user, and works are created after the payment of the order is received.
8. The intelligent design system based on user characteristics and big data training according to claim 7, wherein the user characteristics module is configured to obtain characteristic information of a user, and generate user information based on the characteristic information and basic information, specifically:
acquiring interaction information of a user in a historical time period, generating an interaction data set of the user, carrying out embedded representation of the user by utilizing the interaction data set, carrying out similarity calculation by using the embedded representation of the user, and taking the user with the similarity meeting a preset standard as a similar user;
acquiring embedded representations of the similar users, acquiring attention weights by an attention mechanism, characterizing associated features among the users, and acquiring feature information of the similar users by weighting and aggregating the attention weights;
extracting interactive items from interactive information of a user, sequencing the interactive items according to a time sequence to obtain an interactive time sequence, and obtaining initial weights of the interactive items in the interactive time sequence according to the interactive times and the interactive time length of the interactive items by the user;
learning and representing a plurality of interaction time sequence sequences through a self-attention network, and acquiring characteristic information of the interaction items by combining initial weights and self-attention weights of the interaction items;
Feature information of similar users corresponding to the users and feature information of interaction items are subjected to feature fusion to generate feature information of the users, and the feature information of the users is combined with basic information of the users to generate user information.
9. The intelligent design system based on user characteristics and big data training of claim 7, wherein the intelligent creation module is used for obtaining user demand information and constructing an intelligent creation model of an artwork based on a deep learning method, and specifically comprises the following steps:
constructing an undirected graph in a low-dimensional space according to a user and an interactive item, taking the expression form of the interactive item as an additional characteristic of a node, and performing representation learning on the undirected graph through a graph convolution neural network;
constructing an intelligent artwork creation model based on a graph roll-up neural network, acquiring training data by using a big data means to perform model training, and importing the user demand information into the graph roll-up neural network to perform feature coding;
the vectorization representation of the user and the vectorization representation of the interactive item are obtained through the combination of a message transmission and neighbor aggregation mechanism of the graph convolution neural network, and the inner product of the vectorization representation of the user and the vectorization representation of the interactive item is calculated;
Acquiring inner product sequences according to the inner products, acquiring a preset number of interactive items based on the inner product sequences, and outputting candidate interactive items corresponding to user demand information serving as a graph convolution neural network;
and combining the candidate interaction vector with a basic three-dimensional model of the target artwork to generate three-dimensional information of the target artwork.
10. The intelligent design system based on user characteristics and big data training of claim 7, wherein the user interaction module is configured to obtain interaction information of a user on returned three-dimensional information of an artwork, update and adjust the three-dimensional information of the target artwork by using user feedback, and perform requirement adjustment, and specifically comprises:
three-dimensional information of the target artware is obtained, the three-dimensional information is fed back to a user for visual display, interactive information of the user and the three-dimensional information of the target artware is obtained, and the interactive information is matched with a demand category;
acquiring association features among the user corresponding to the demand categories according to the historical user demands, generating revised recommendation information based on the interaction information of the user through the association features, and judging feedback of the user on the revised recommendation information;
Judging the satisfaction degree of the user on the revised recommendation information through feedback of the user, stopping revised recommendation when the satisfaction degree is lower than a preset threshold value, and relearning the associated features;
after the user completes three-dimensional information interaction of the target artware, the final target artware three-dimensional information is sent to 3D printing equipment to generate works, and corrected user demand information is obtained.
CN202311377926.8A 2023-10-24 2023-10-24 Intelligent design method and system based on user characteristics and big data training Active CN117112915B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202311377926.8A CN117112915B (en) 2023-10-24 2023-10-24 Intelligent design method and system based on user characteristics and big data training

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202311377926.8A CN117112915B (en) 2023-10-24 2023-10-24 Intelligent design method and system based on user characteristics and big data training

Publications (2)

Publication Number Publication Date
CN117112915A true CN117112915A (en) 2023-11-24
CN117112915B CN117112915B (en) 2024-02-20

Family

ID=88806064

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202311377926.8A Active CN117112915B (en) 2023-10-24 2023-10-24 Intelligent design method and system based on user characteristics and big data training

Country Status (1)

Country Link
CN (1) CN117112915B (en)

Citations (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113255052A (en) * 2021-07-09 2021-08-13 佛山市陶风互联网络科技有限公司 Home decoration scheme recommendation method and system based on virtual reality and storage medium
CN113297647A (en) * 2021-02-01 2021-08-24 阿里巴巴集团控股有限公司 Auxiliary design method and device for customized product and electronic equipment
CN113536126A (en) * 2021-01-12 2021-10-22 陈漩 Service demand mining method based on big data and cloud server
CN113807422A (en) * 2021-09-07 2021-12-17 南京邮电大学 Weighted graph convolutional neural network score prediction model fusing multi-feature information
CN114722269A (en) * 2022-03-08 2022-07-08 清华大学 Article recommendation method and device based on graph neural network and storage medium
CN116595438A (en) * 2023-05-17 2023-08-15 上海极豆科技有限公司 Picture creation method, device, equipment and storage medium
CN116701669A (en) * 2023-06-09 2023-09-05 科大讯飞股份有限公司 Method, device, equipment and storage medium for generating multimedia content
CN116862454A (en) * 2023-08-31 2023-10-10 北京华邑建设集团有限公司 Indoor building design management method and system

Patent Citations (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113536126A (en) * 2021-01-12 2021-10-22 陈漩 Service demand mining method based on big data and cloud server
CN113297647A (en) * 2021-02-01 2021-08-24 阿里巴巴集团控股有限公司 Auxiliary design method and device for customized product and electronic equipment
CN113255052A (en) * 2021-07-09 2021-08-13 佛山市陶风互联网络科技有限公司 Home decoration scheme recommendation method and system based on virtual reality and storage medium
CN113807422A (en) * 2021-09-07 2021-12-17 南京邮电大学 Weighted graph convolutional neural network score prediction model fusing multi-feature information
CN114722269A (en) * 2022-03-08 2022-07-08 清华大学 Article recommendation method and device based on graph neural network and storage medium
CN116595438A (en) * 2023-05-17 2023-08-15 上海极豆科技有限公司 Picture creation method, device, equipment and storage medium
CN116701669A (en) * 2023-06-09 2023-09-05 科大讯飞股份有限公司 Method, device, equipment and storage medium for generating multimedia content
CN116862454A (en) * 2023-08-31 2023-10-10 北京华邑建设集团有限公司 Indoor building design management method and system

Also Published As

Publication number Publication date
CN117112915B (en) 2024-02-20

Similar Documents

Publication Publication Date Title
CN110909176B (en) Data recommendation method and device, computer equipment and storage medium
WO2019019554A1 (en) Method and apparatus for obtaining recommendation information, and electronic device
JP2017522660A (en) Automatic image-based recommendations using color palettes
CN110728541A (en) Information stream media advertisement creative recommendation method and device
CN110554782B (en) Expression input image synthesis method and system
CN110196930A (en) A kind of multi-modal customer service automatic reply method and system
CN108389113B (en) Collaborative filtering recommendation method and system
CN116894711A (en) Commodity recommendation reason generation method and device and electronic equipment
CN115982463A (en) Resource recommendation method, device, equipment and storage medium
CN116862454B (en) Indoor building design management method and system
CN117112915B (en) Intelligent design method and system based on user characteristics and big data training
CN112989053A (en) Periodical recommendation method and device
CN112508114A (en) Intelligent clothing recommendation system and method
JP7445730B2 (en) Learning device, machine learning model and learning method
CN116934531A (en) Wine information intelligent management method and system based on data analysis
EP4376384A1 (en) Method and apparatus for sharing liquor manufacturing method, and server
CN116911926A (en) Advertisement marketing recommendation method based on data analysis
CN113869960B (en) Poster generation method and device, storage medium and computer equipment
CN107220273B (en) Cartoon character face searching method
CN115375408A (en) Article information recommendation method and device and electronic equipment
CN114283300A (en) Label determining method and device, and model training method and device
CN111105258B (en) Commodity pricing method, device and system
CN114997921A (en) Advertisement case recommendation method and device, equipment, medium and product thereof
CN106528584A (en) An ensemble learning-based group recommendation method
CN113744012A (en) Information processing method, device and storage medium

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