CN117635275B - Intelligent electronic commerce operation commodity management platform and method based on big data - Google Patents
Intelligent electronic commerce operation commodity management platform and method based on big data Download PDFInfo
- Publication number
- CN117635275B CN117635275B CN202311750792.XA CN202311750792A CN117635275B CN 117635275 B CN117635275 B CN 117635275B CN 202311750792 A CN202311750792 A CN 202311750792A CN 117635275 B CN117635275 B CN 117635275B
- Authority
- CN
- China
- Prior art keywords
- commodity
- display image
- semantic feature
- semantic
- feature vector
- 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 abstract description 17
- 239000013598 vector Substances 0.000 claims abstract description 196
- 238000005457 optimization Methods 0.000 claims abstract description 31
- 238000003062 neural network model Methods 0.000 claims abstract description 13
- 238000007726 management method Methods 0.000 claims description 37
- 238000012937 correction Methods 0.000 claims description 32
- 230000004927 fusion Effects 0.000 claims description 21
- 230000011218 segmentation Effects 0.000 claims description 17
- 238000012545 processing Methods 0.000 claims description 16
- 238000004458 analytical method Methods 0.000 claims description 15
- 238000013527 convolutional neural network Methods 0.000 claims description 14
- 238000013507 mapping Methods 0.000 claims description 7
- 238000012546 transfer Methods 0.000 claims description 7
- 239000011159 matrix material Substances 0.000 claims description 6
- 238000006243 chemical reaction Methods 0.000 abstract description 8
- 238000000605 extraction Methods 0.000 abstract description 8
- 238000010586 diagram Methods 0.000 description 10
- 230000000694 effects Effects 0.000 description 8
- 230000006870 function Effects 0.000 description 8
- 230000014509 gene expression Effects 0.000 description 6
- 238000013473 artificial intelligence Methods 0.000 description 5
- 230000008901 benefit Effects 0.000 description 4
- 238000003058 natural language processing Methods 0.000 description 4
- 230000008569 process Effects 0.000 description 4
- 238000013519 translation Methods 0.000 description 4
- 238000005516 engineering process Methods 0.000 description 3
- 230000003993 interaction Effects 0.000 description 3
- 238000013135 deep learning Methods 0.000 description 2
- 238000011161 development Methods 0.000 description 2
- 230000018109 developmental process Effects 0.000 description 2
- 238000010191 image analysis Methods 0.000 description 2
- 230000006872 improvement Effects 0.000 description 2
- 238000005065 mining Methods 0.000 description 2
- 238000011176 pooling Methods 0.000 description 2
- 238000007781 pre-processing Methods 0.000 description 2
- 230000000007 visual effect Effects 0.000 description 2
- 241000282414 Homo sapiens Species 0.000 description 1
- 230000008485 antagonism Effects 0.000 description 1
- 238000013528 artificial neural network Methods 0.000 description 1
- 238000004590 computer program Methods 0.000 description 1
- 238000013136 deep learning model Methods 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 230000008451 emotion Effects 0.000 description 1
- 238000003709 image segmentation Methods 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 230000003252 repetitive effect Effects 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
- 238000012549 training Methods 0.000 description 1
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
- G06Q30/00—Commerce
- G06Q30/06—Buying, selling or leasing transactions
- G06Q30/0601—Electronic shopping [e-shopping]
- G06Q30/0641—Shopping interfaces
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/30—Semantic analysis
-
- 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
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Computational Linguistics (AREA)
- General Health & Medical Sciences (AREA)
- Business, Economics & Management (AREA)
- Artificial Intelligence (AREA)
- Health & Medical Sciences (AREA)
- General Engineering & Computer Science (AREA)
- Molecular Biology (AREA)
- Biophysics (AREA)
- Computing Systems (AREA)
- Evolutionary Computation (AREA)
- Life Sciences & Earth Sciences (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Data Mining & Analysis (AREA)
- Finance (AREA)
- Biomedical Technology (AREA)
- Accounting & Taxation (AREA)
- Audiology, Speech & Language Pathology (AREA)
- Development Economics (AREA)
- Economics (AREA)
- Marketing (AREA)
- Strategic Management (AREA)
- General Business, Economics & Management (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
An intelligent electronic commerce operation commodity management platform and method based on big data are disclosed. Firstly, acquiring text description of a commodity to be optimized, then, carrying out semantic coding on the text description of the commodity to be optimized to obtain a sequence of semantic feature vectors of granularity of a commodity text description word, then, acquiring an initial display image of the commodity to be optimized, then, carrying out feature extraction on the initial display image through an image style feature extractor based on a deep neural network model to obtain an initial display image style semantic feature vector, then, carrying out optimization on the initial display image style semantic feature vector based on the sequence of semantic feature vectors of granularity of the commodity text description word to obtain updated display image style semantic features, and finally, generating an optimized display image based on the updated display image style semantic features. Therefore, the exposure rate and conversion rate of the commodity can be increased, and the attraction and competitiveness of the commodity are improved.
Description
Technical Field
The application relates to the field of intelligent electronic commerce operation, in particular to an intelligent electronic commerce operation commodity management platform and method based on big data.
Background
With the development of the e-commerce platform, the number and variety of commodities are increased, and users face massive commodity selection. Under the condition, how to effectively display the commodities attracts the attention of consumers, and the improvement of the exposure rate and the conversion rate of the commodities becomes an important problem for the operation of the electronic commerce.
Conventional e-commerce platforms typically draw the attention of users by displaying pictures and textual descriptions of the merchandise. However, the commodity display pictures are generally uploaded by a merchant or unified and standardized by a platform, and often only the appearance of the commodity can be displayed by the pictures, the information contained in the text description of the commodity cannot be fully reflected, the characteristics and advantages of the commodity cannot be completely conveyed, the display effect of the commodity is poor, and the purchase will of consumers is influenced. While textual descriptions of the merchandise may provide more information, users often have insufficient tolerance to carefully read the descriptions of each merchandise, resulting in lower exposure and conversion of the merchandise.
Therefore, an intelligent electronic commerce operation commodity management platform based on big data is desired.
Disclosure of Invention
In view of the above, the application provides an intelligent electronic commerce operation commodity management platform and method based on big data, which can acquire text description and display images of commodities, and introduce semantic understanding and image processing algorithms at the rear end to analyze the text description and display images of the commodities so as to optimize display pictures of the commodities based on the text description of the commodities, so that the display pictures are more in line with the attribute, title and feature description of the commodities, thereby increasing the exposure rate and conversion rate of the commodities, and improving the attractiveness and competitiveness of the commodities.
According to an aspect of the present application, there is provided an intelligent electronic commerce operation commodity management platform based on big data, which includes:
the commodity text description acquisition module is used for acquiring text description of the commodity to be optimized, wherein the text description comprises commodity attributes, titles and feature description;
the text semantic coding module is used for carrying out semantic coding on the text description of the commodity to be optimized to obtain a sequence of granularity semantic feature vectors of the commodity text description words;
the commodity display image acquisition module is used for acquiring an initial display image of the commodity to be optimized;
the image feature analysis module is used for extracting features of the initial display image through an image style feature extractor based on a deep neural network model so as to obtain an initial display image style semantic feature vector;
the display image feature optimization updating module is used for optimizing the initial display image style semantic feature vector based on the sequence of the commodity text descriptor granularity semantic feature vector to obtain updated display image style semantic features; and
And the optimized display image generation module is used for generating an optimized display image based on the updated display image style semantic features.
According to another aspect of the present application, there is provided an intelligent electronic commerce operation commodity management method based on big data, including:
acquiring text description of a commodity to be optimized, wherein the text description comprises commodity attributes, titles and feature descriptions;
Carrying out semantic coding on the text description of the commodity to be optimized to obtain a sequence of granularity semantic feature vectors of the commodity text description words;
Acquiring an initial display image of the commodity to be optimized;
extracting features of the initial display image through an image style feature extractor based on a deep neural network model to obtain an initial display image style semantic feature vector;
Optimizing the initial display image style semantic feature vector based on the sequence of the commodity text descriptor granularity semantic feature vector to obtain updated display image style semantic features; and
And generating an optimized display image based on the updated display image style semantic features.
According to the embodiment of the application, firstly, text description of a commodity to be optimized is obtained, then, semantic coding is carried out on the text description of the commodity to be optimized to obtain a sequence of semantic feature vectors of granularity of commodity text descriptors, then, an initial display image of the commodity to be optimized is obtained, then, feature extraction is carried out on the initial display image through an image style feature extractor based on a deep neural network model to obtain an initial display image style semantic feature vector, then, optimization is carried out on the initial display image style semantic feature vector based on the sequence of semantic feature vectors of granularity of commodity text descriptors to obtain updated display image style semantic features, and finally, an optimized display image is generated based on the updated display image style semantic features. Therefore, the exposure rate and conversion rate of the commodity can be increased, and the attraction and competitiveness of the commodity are improved.
Other features and aspects of the present application will become apparent from the following detailed description of exemplary embodiments, which proceeds with reference to the accompanying drawings.
Drawings
The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate exemplary embodiments, features and aspects of the application and together with the description, serve to explain the principles of the application.
Fig. 1 shows a block diagram of a big data based intelligent e-commerce operation commodity management platform according to an embodiment of the present application.
Fig. 2 shows a flowchart of a big data based intelligent e-commerce operation commodity management method according to an embodiment of the present application.
Fig. 3 shows an architecture diagram of a big data based intelligent e-commerce operation commodity management method according to an embodiment of the present application.
Fig. 4 shows an application scenario diagram of a big data based intelligent e-commerce operation commodity management platform according to an embodiment of the present application.
Detailed Description
The following description of the embodiments of the present application will be made clearly and fully with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some, but not all embodiments of the application. All other embodiments, which can be made by those skilled in the art based on the embodiments of the application without making any inventive effort, are also within the scope of the application.
As used in the specification and in the claims, the terms "a," "an," "the," and/or "the" are not specific to a singular, but may include a plurality, unless the context clearly dictates otherwise. In general, the terms "comprises" and "comprising" merely indicate that the steps and elements are explicitly identified, and they do not constitute an exclusive list, as other steps or elements may be included in a method or apparatus.
Various exemplary embodiments, features and aspects of the application will be described in detail below with reference to the drawings. In the drawings, like reference numbers indicate identical or functionally similar elements. Although various aspects of the embodiments are illustrated in the accompanying drawings, the drawings are not necessarily drawn to scale unless specifically indicated.
In addition, numerous specific details are set forth in the following description in order to provide a better illustration of the application. It will be understood by those skilled in the art that the present application may be practiced without some of these specific details. In some instances, well known methods, procedures, components, and circuits have not been described in detail so as not to obscure the present application.
Aiming at the technical problems, the technical concept of the application is to acquire the text description and the display image of the commodity, introduce semantic understanding and image processing algorithm at the rear end to analyze the text description and the display image of the commodity, so as to optimize the display image of the commodity based on the text description of the commodity, and enable the display image to be more in line with the attribute, title and feature description of the commodity, thereby increasing the exposure rate and conversion rate of the commodity and improving the attraction and competitiveness of the commodity.
Fig. 1 shows a block diagram schematic of a big data based intelligent e-commerce operation commodity management platform according to an embodiment of the present application. As shown in fig. 1, the intelligent electronic commerce operation commodity management platform 100 based on big data according to the embodiment of the present application includes: the commodity text description acquisition module 110 is configured to acquire a text description of a commodity to be optimized, where the text description includes a commodity attribute, a title and a feature description; the text semantic coding module 120 is configured to perform semantic coding on the text description of the commodity to be optimized to obtain a sequence of granularity semantic feature vectors of the text description word of the commodity; the commodity display image acquisition module 130 is used for acquiring an initial display image of the commodity to be optimized; the image feature analysis module 140 is configured to perform feature extraction on the initial display image through an image style feature extractor based on a deep neural network model to obtain an initial display image style semantic feature vector; a display image feature optimization updating module 150, configured to optimize the initial display image style semantic feature vector based on the sequence of the commodity text descriptor granularity semantic feature vector to obtain updated display image style semantic features; and an optimized presentation image generation module 160 for generating an optimized presentation image based on the updated presentation image style semantic features.
It should be appreciated that the merchandise text description gathering module 110 is responsible for gathering text information of the merchandise and providing input data for subsequent processing. The text semantic coding module 120 converts the commodity text description into semantic feature vectors that can be used for subsequent commodity analysis and optimization. The merchandise display image acquisition module 130 is responsible for acquiring the display image of the merchandise and providing input data for subsequent feature extraction and optimization. The image feature analysis module 140 may extract features of the image for subsequent image analysis and optimization. The presentation image feature optimization update module 150 optimizes using semantic features and image features of the textual description to improve the quality and style of the presentation image. The optimized display image generation module 160 module generates a new display image by using the optimized image characteristics so as to improve the display effect and the attraction of the commodity. The modules form a core function of an intelligent electronic commerce operation commodity management platform based on big data, and the description and the display effect of commodities are improved through text and image processing and optimization, so that the operation efficiency and the user experience of the electronic commerce are improved.
Specifically, in the technical scheme of the application, firstly, a text description of a commodity to be optimized is obtained, wherein the text description comprises commodity attributes, titles and feature descriptions. Next, considering that the text description of the commodity generally includes information such as attributes, titles, and feature descriptions of the commodity, the text description includes rich semantic information, and the semantic information is based on word granularity. Therefore, in order to perform semantic analysis and understanding on the text description, in the technical scheme of the application, the text description of the commodity to be optimized is further subjected to semantic coding so as to extract semantic understanding characteristic information based on word granularity in the text description of the commodity to be optimized, and thus a sequence of semantic characteristic vectors of the commodity text description word granularity is obtained.
Accordingly, the text semantic coding module 120 includes: the word segmentation unit is used for carrying out word segmentation processing on the text description of the commodity to be optimized so as to convert the text description of the commodity to be optimized into a word sequence consisting of a plurality of words; an embedded encoding unit for mapping each word in the word sequence to a word vector using an embedded layer of a context encoder to obtain a sequence of word vectors; and a context semantic coding unit, configured to perform global-based context semantic coding on the sequence of word vectors using the context encoder to obtain a sequence of the commodity text descriptor granularity semantic feature vectors.
It should be appreciated that word segmentation is a basic task in Natural Language Processing (NLP) that segments continuous natural language text into words or tag sequences of a certain meaning. The purpose of word segmentation is to divide text into discrete word units so that a computer can better understand and process the text. The following are some uses of the word segmentation process: 1. language understanding and semantic analysis: word segmentation is a fundamental step of language processing, which converts text into discrete semantic units that help understand the structure and semantics of sentences. In natural language processing tasks, such as text classification, named entity recognition, emotion analysis, etc., word segmentation can provide more accurate semantic representations, helping to calculate the meaning of the mechanism solution text. 2. Information retrieval and search engine: in information retrieval and search engines, the terms may represent the query terms and documents in a unified manner for matching and retrieval. The query words and the documents are divided into independent words, so that the accuracy and recall rate of searching can be improved. 3. Machine translation: in a machine translation task, the segmentation may segment sentences in the source language and the target language into corresponding word sequences for translation and alignment. Word segmentation is critical to the accuracy and fluency of translation. 4. Text preprocessing: in text analysis and text mining tasks, word segmentation is an important preprocessing step. By segmenting the text into word sequences, word frequency statistics, word vector representation, feature extraction and other operations can be performed, and meaningful input is provided for subsequent analysis and modeling. 5. Information extraction and knowledge graph: the word segmentation can help extract information such as entities, relations, events and the like from the text, and construct a knowledge graph or graph database. Through word segmentation, key information such as entity nouns, verbs and the like in the text can be better identified and extracted. In other words, word segmentation processes play an important role in natural language processing and text analysis, which is capable of converting continuous natural language text into discrete word sequences, providing a basis for subsequent processing and analysis.
The embedding layer (Embedding Layer) is a common technique in text processing for mapping discrete word representations (e.g., words) into a continuous vector space, i.e., converting each word into a word vector. The embedding layer encodes semantic information of the words into a vector representation by learning associations between the words. The embedded layer is used for converting text data into a form which can be understood and processed by a computer, and certain semantic information is reserved. By mapping words into a continuous vector space, the embedding layer is able to capture semantic similarity and relevance between words. Thus, similar words are closer together in the embedding space, thereby facilitating subsequent semantic analysis and processing. In the text semantic coding module, an embedded coding unit maps each word in the word sequence to a word vector, i.e., converts each word in the text description into a vector representation, using the embedded layer of the context encoder. This may translate the text description into a sequence of word vectors for subsequent context semantic coding. In general, the embedding layer functions to convert words in text into vector representations so that a computer can understand and process text data and provide a way to capture semantic associations between words. In the text semantic coding module, an embedding layer is used for converting text description of goods to be optimized into a sequence of word vectors and providing input for subsequent context semantic coding.
And then, acquiring an initial display image of the commodity to be optimized, and performing feature mining on the initial display image by using an image style feature extractor with excellent performance in terms of implicit feature extraction of the image and based on a convolutional neural network model so as to extract style semantic feature information about the commodity in the initial display image, thereby obtaining a style semantic feature vector of the initial display image.
Accordingly, in the image feature analysis module 140, the deep neural network model is a convolutional neural network model. It should be noted that the convolutional neural network (Convolutional Neural Network, abbreviated as CNN) is a deep learning model, and is particularly suitable for processing image and visual data, and is a neural network structure composed of a plurality of convolutional layers, a pooling layer and a full-connection layer. Convolutional neural networks extract features in an image through a convolutional operation, and progressively extract higher level abstract features through a stack of multiple convolutional layers and pooled layers. The convolution layer convolves the input image with a set of learnable filters (also referred to as convolution kernels) to capture local spatial information in the image. The pooling layer is then used to reduce the size of the feature map and preserve the most salient features. Convolutional neural networks have wide applications in the field of image processing, including tasks such as image classification, object detection, image segmentation, and the like. Since the convolutional neural network can automatically learn the feature representation in the image, it can effectively extract the visual features of the image and achieve excellent performance in various image tasks. In the image feature analysis module 140, a convolutional neural network based on a deep neural network model is used for image feature extraction. Through the trained convolutional neural network model, feature vectors with semantic information can be extracted from an initial display image of the commodity to be optimized. These feature vectors may be used for subsequent image analysis, optimization and generation to improve the presentation and style of the merchandise. Therefore, in the operation of the electronic commerce, the convolutional neural network model plays an important role in image processing.
It should be understood that the sequence of the article text descriptor granularity semantic feature vectors and the initial presentation image style semantic feature vector represent semantic understanding features and presentation image style semantic features based on word granularity of the text description related to the article to be optimized, respectively. But consider that the presentation picture style semantics of the merchandise do not fully convey the features and advantages of the merchandise. While the text description semantics of the merchandise may provide more information including attributes, titles, and feature descriptions of the merchandise, etc. Therefore, in order to realize information interaction and fusion between text description semantics and image semantics of the commodity so as to obtain more comprehensive and accurate presentation image style semantic feature representation, in the technical scheme of the application, a cross-mode information transfer fusion module is further used to fuse the sequence of the commodity text description word granularity semantic feature vector into the initial presentation image style semantic feature vector so as to obtain an updated presentation image style semantic feature vector. It should be appreciated that by fusing the sequence of article text descriptor granularity semantic feature vectors into the initial presentation image style semantic feature vector, style semantic optimization expression and updating of article images can be performed with semantic features based on word granularity with respect to the article text description, thereby generating a richer, more accurate image semantic representation. Specifically, the cross-modal information transfer fusion module can adjust the feature representation of the commodity display image according to the importance and the relativity of the commodity text description semantics, so that the feature representation better reflects the characteristics and advantages of the commodity.
Accordingly, the display image feature optimization updating module 150 is configured to: and using a cross-modal information transfer fusion module to fuse the sequence of the commodity text descriptor granularity semantic feature vector into the initial display image style semantic feature vector to obtain an updated display image style semantic feature vector as the updated display image style semantic feature.
Specifically, the display image feature optimization updating module 150 is configured to: using a cross-modal information transfer fusion module, and fusing the sequence of the commodity text descriptor granularity semantic feature vectors into the initial display image style semantic feature vector by using a fusion formula to obtain the updated display image style semantic feature vector; wherein, the fusion formula is:
，
wherein, Representing the semantic feature vector of the style of the initial display image,/>Representation/>Matrix of/>Equal to the scale of the initial presentation image style semantic feature vector,/>Is/>Matrix of/>The number of the commodity text descriptor granularity semantic feature vectors in the sequence equal to the commodity text descriptor granularity semantic feature vectors,/>Representing each item text descriptor granularity semantic feature vector in the sequence of item text descriptor granularity semantic feature vectors,/>Representing the scale of each commodity text descriptor granularity semantic feature vector in the sequence of commodity text descriptor granularity semantic feature vectors,/>Is a Sigmoid function,/>Is a weight coefficient,/>And/>Convolution operation representing a 1 x1 convolution kernel,/>Representing the updated presentation image style semantic feature vector.
And then, the updated presentation image style semantic feature vector is passed through a AIGC-based presentation image optimization generator to obtain an optimized presentation image. That is, the commodity image after the commodity is optimized and expressed based on the semantic features of the word granularity is utilized to update the semantic feature information to perform the optimized generation of the commodity display picture, so that the optimized display image with more attractive and expressive force is obtained.
Accordingly, the optimized presentation image generation module 160 includes: the updating display image style semantic feature correction unit is used for carrying out feature distribution correction on the updating display image style semantic feature vector to obtain a corrected updating display image style semantic feature vector; and the picture optimization generating unit is used for enabling the corrected updated display image style semantic feature vector to pass through a AIGC-based display picture optimization generator to obtain the optimized display image.
It should be understood that the update display image style semantic feature correction unit is configured to perform feature distribution correction on the update display image style semantic feature vector to obtain a corrected update display image style semantic feature vector, and the function of this unit is to adjust the distribution of the feature vector so as to better conform to the desired feature distribution, and through feature distribution correction, the expression capability of the feature vector can be improved so as to be more suitable for subsequent image optimization generation. The picture optimization generating unit uses the corrected updated display image style semantic feature vector, generates an optimized display image through a display picture optimization generator based on AIGC (ARTIFICIAL INTELLIGENCE GENERATED Content), inputs the corrected feature vector into the optimization generator, generates a display image with an optimization effect through the operation and the optimization algorithm of the generator, and can generate a display image which is more in line with expectations according to semantic information and style features in the feature vector so as to improve the display effect and attractive force of commodities. In summary, the update presentation image style semantic feature correction unit is used for adjusting the distribution of feature vectors to obtain better feature representation; and the picture optimization generating unit generates an optimized display image by using the corrected feature vector through an optimization generator. The two units play different roles in optimizing the display image generation module and cooperate together to realize the optimization and generation of the display image.
Notably AIGC is an abbreviation for ARTIFICIAL INTELLIGENCE GENERATED Content, meaning artificial intelligence generated Content. AIGC refers to generating various forms of content such as images, audio, video, text and the like by using artificial intelligence technology, and can enable a computer system to learn and imitate the capability of human beings to create the content by using machine learning, deep learning, model generation and other technologies to generate the content with a certain creativity and expressive force. AIGC can be applied in a number of fields, and the development of AIGC benefits from the advancement of deep learning and generation models, particularly the advent of generation of model of antagonism network (GAN) and variational self-encoder (VAE). These models can learn the distribution of real data and generate new data similar thereto. By training the models, various and creative contents can be generated, and the application of artificial intelligence in the field of content generation is expanded.
In particular, in the above technical solution, the sequence of the article text descriptor granularity semantic feature vectors expresses the text semantic feature of the text description of the article to be optimized, and the initial display image style semantic feature vector expresses the image semantic feature of the initial display image for representing the image style, and in consideration of the feature mode difference between the sequence of the article text descriptor granularity semantic feature vectors and the initial display image style semantic feature vector, when the sequence of the article text descriptor granularity semantic feature vectors is fused into the initial display image style semantic feature vector by using the cross-mode information transfer fusion module, there is significant cross-mode information correspondence sparsity between the sequence of the article text descriptor granularity semantic feature vectors and the initial display image style semantic feature vector, so as to affect the expression effect of the update display image style semantic feature vector, so that cross-mode information correspondence optimization needs to be performed based on the semantic significance and the semantic significance of the respective feature expressions of the sequence of the article text descriptor granularity semantic feature vectors and the initial display image style semantic feature vector, thereby improving the expression effect of the update display image style semantic feature vector.
Based on the above, the applicant of the present application corrects the sequence of the commodity text descriptor granularity semantic feature vectors and the initial presentation image style semantic feature vectors.
In one example, the updating the presentation image style semantic feature correction unit includes: the feature correction subunit is used for correcting the sequence of the commodity text descriptor granularity semantic feature vectors and the initial display image style semantic feature vectors to obtain corrected feature vectors; and the correction feature fusion subunit is used for fusing the correction feature vector with the updated display image style semantic feature vector to obtain the corrected updated display image style semantic feature vector.
It should be understood that the feature correction subunit is configured to correct the sequence of the semantic feature vectors of the granularity of the commodity text descriptor and the semantic feature vectors of the style of the initially displayed image, so as to obtain corrected feature vectors, which are used to improve the expressive power and semantic consistency of the feature vectors by correcting the text description and the initial image features. Specifically, the feature correction subunit may effectively fuse and correct the text description and the image features by using some correction techniques, such as feature mapping, feature fusion, feature alignment, and the like, to generate a correction feature vector with more accuracy and more expressive force. The correction feature fusion subunit is used for fusing the correction feature vector with the updated display image style semantic feature vector to obtain a corrected updated display image style semantic feature vector. The function of this subunit is to fuse the feature corrected text description features with the original image features to obtain a more comprehensive and richer representation of the features. The fusion mode can be a simple vector splicing method, a weighted summation method and the like, and two feature vectors are fused into a comprehensive feature vector. The fused feature vector contains the text description and the image feature information, so that the style and semantic features of the commodity can be better described. In summary, the feature correction subunit is configured to correct the text description and the initial image feature, and generate a corrected feature vector; and the correction feature fusion subunit fuses the correction feature vector with the updated display image style semantic feature vector to obtain a corrected updated display image style semantic feature vector. The two subunits play different roles in updating the display image style semantic feature correction unit and cooperate together to realize correction and fusion of feature vectors.
Wherein the feature correction subunit is configured to: correcting the sequence of the commodity text descriptor granularity semantic feature vectors and the initial display image style semantic feature vectors by using the following correction formula to obtain corrected feature vectors; wherein, the correction formula is:
，
wherein, Is a cascade feature vector after the sequence cascade of the commodity text descriptor granularity semantic feature vector, and/>Is the semantic feature vector of the style of the initial display image,/>Representing the position-wise evolution of the feature vector,And/>Feature vector/>, respectivelyAnd/>Reciprocal of maximum eigenvalue,/>And/>Is a weight superparameter,/>Representing multiplication by location,/>Representing vector subtraction,/>Is the correction feature vector.
Here, by the cascade feature vectorAnd the initial presentation image style semantic feature vector/>To obtain a pre-segmented local set of feature value sets and regressing the concatenated feature vector/>, therefromAnd the initial presentation image style semantic feature vector/>In this way, the saliency distribution by position of the feature values can be promoted based on the idea of the furthest point sampling, so that sparse interaction control among feature vectors is performed through key features with the saliency distribution, and the corrected feature vector/>Cascading feature vectors/>, for the sequence of article text descriptor granularity semantic feature vectorsAnd the initial presentation image style semantic feature vector/>Is a reduction of the original manifold geometry of (a). In this way, the correction feature vector/>And fusing the semantic feature vector with the style semantic feature vector of the updated display image to improve the expression effect of the semantic feature vector of the style semantic feature vector of the updated display image, so that the image quality of the optimized display image obtained by a display image optimization generator based on AIGC is improved. Therefore, the display picture of the commodity can be optimized based on the text description of the commodity, so that the display picture of the commodity is more in line with the attribute, title and feature description of the commodity, the exposure rate and conversion rate of the commodity are increased, and the attraction and competitiveness of the commodity are improved.
In summary, the intelligent electronic commerce operation commodity management platform 100 based on big data according to the embodiment of the present application is illustrated, which can increase the exposure rate and conversion rate of the commodity and improve the attraction and competitiveness of the commodity.
As described above, the smart e-commerce operation merchandise management platform 100 based on big data according to the embodiment of the present application may be implemented in various terminal devices, for example, a server or the like having a smart e-commerce operation merchandise management algorithm based on big data. In one example, the big data based intelligent electronic commerce operator operations commodity management platform 100 may be integrated into the terminal device as one software module and/or hardware module. For example, the big data based intelligent e-commerce operation merchandise management platform 100 may be a software module in the operating system of the terminal device or may be an application developed for the terminal device; of course, the intelligent electronic commerce operation merchandise management platform 100 based on big data may also be one of a plurality of hardware modules of the terminal device.
Alternatively, in another example, the big data based smart e-commerce operation merchandise management platform 100 and the terminal device may be separate devices, and the big data based smart e-commerce operation merchandise management platform 100 may be connected to the terminal device through a wired and/or wireless network and transmit the interaction information according to the agreed data format.
Fig. 2 shows a flowchart of a big data based intelligent e-commerce operation commodity management method according to an embodiment of the present application. Fig. 3 shows a schematic diagram of a system architecture of a big data based intelligent e-commerce operation commodity management method according to an embodiment of the present application. As shown in fig. 2 and 3, the intelligent electronic commerce operation commodity management method based on big data according to the embodiment of the present application includes: s110, acquiring text description of a commodity to be optimized, wherein the text description comprises commodity attributes, titles and feature descriptions; s120, carrying out semantic coding on the text description of the commodity to be optimized to obtain a sequence of granularity semantic feature vectors of the commodity text description words; s130, acquiring an initial display image of the commodity to be optimized; s140, extracting features of the initial display image through an image style feature extractor based on a deep neural network model to obtain an initial display image style semantic feature vector; s150, optimizing the initial display image style semantic feature vector based on the sequence of the commodity text descriptor granularity semantic feature vector to obtain updated display image style semantic features; and S160, generating an optimized display image based on the updated display image style semantic features.
In one possible implementation manner, performing semantic coding on the text description of the commodity to be optimized to obtain a sequence of granularity semantic feature vectors of the text description word of the commodity, including: word segmentation processing is carried out on the text description of the commodity to be optimized so as to convert the text description of the commodity to be optimized into a word sequence consisting of a plurality of words; mapping each word in the word sequence to a word vector using an embedding layer of a context encoder to obtain a sequence of word vectors; and performing global-based context semantic coding on the sequence of word vectors using the context encoder to obtain the sequence of commodity text descriptor granularity semantic feature vectors.
In one possible implementation, the deep neural network model is a convolutional neural network model.
Here, it will be understood by those skilled in the art that the specific operations of the respective steps in the above-described big data based intelligent electronic commerce operation commodity management method have been described in detail in the above description with reference to the big data based intelligent electronic commerce operation commodity management platform of fig. 1, and thus, repetitive descriptions thereof will be omitted.
Fig. 4 shows an application scenario diagram of a big data based intelligent e-commerce operation commodity management platform according to an embodiment of the present application. As shown in fig. 4, in this application scenario, first, a text description of a commodity to be optimized (e.g., D1 illustrated in fig. 4) and an initial presentation image of the commodity to be optimized (e.g., D2 illustrated in fig. 4) are acquired, wherein the text description includes commodity attributes, titles, and feature descriptions, and then the text description of the commodity to be optimized and the initial presentation image are input to a server (e.g., S illustrated in fig. 4) deployed with a big data-based smart e-commerce operation commodity management algorithm, wherein the server is capable of processing the text description of the commodity to be optimized and the initial presentation image using the big data-based smart e-commerce operation commodity management algorithm to obtain an optimized presentation image.
The flowcharts and block diagrams in the figures illustrate the architecture, functionality, and operation of possible implementations of systems, methods and computer program products according to various embodiments of the present application. In this regard, each block in the flowchart or block diagrams may represent a module, segment, or portion of instructions, which comprises one or more executable instructions for implementing the specified logical function(s). In some alternative implementations, the functions noted in the block may occur out of the order noted in the figures. For example, two blocks shown in succession may, in fact, be executed substantially concurrently, or the blocks may sometimes be executed in the reverse order, depending upon the functionality involved. It will also be noted that each block of the block diagrams and/or flowchart illustration, and combinations of blocks in the block diagrams and/or flowchart illustration, can be implemented by special purpose hardware-based systems which perform the specified functions or acts, or combinations of special purpose hardware and computer instructions.
The foregoing description of embodiments of the application has been presented for purposes of illustration and description, and is not intended to be exhaustive or limited to the embodiments disclosed. Many modifications and variations will be apparent to those of ordinary skill in the art without departing from the scope and spirit of the various embodiments described. The terminology used herein was chosen in order to best explain the principles of the embodiments, the practical application, or the improvement of technology in the marketplace, or to enable others of ordinary skill in the art to understand the embodiments disclosed herein.
Claims (8)
1. An intelligent electronic commerce operation commodity management platform based on big data, which is characterized by comprising:
the commodity text description acquisition module is used for acquiring text description of the commodity to be optimized, wherein the text description comprises commodity attributes, titles and feature description;
the text semantic coding module is used for carrying out semantic coding on the text description of the commodity to be optimized to obtain a sequence of granularity semantic feature vectors of the commodity text description words;
the commodity display image acquisition module is used for acquiring an initial display image of the commodity to be optimized;
the image feature analysis module is used for extracting features of the initial display image through an image style feature extractor based on a deep neural network model so as to obtain an initial display image style semantic feature vector;
the display image feature optimization updating module is used for optimizing the initial display image style semantic feature vector based on the sequence of the commodity text descriptor granularity semantic feature vector to obtain updated display image style semantic features; and
The optimized display image generation module is used for generating an optimized display image based on the updated display image style semantic features;
the display image feature optimization updating module comprises:
Using a cross-modal information transfer fusion module, and fusing the sequence of the commodity text descriptor granularity semantic feature vectors into the initial display image style semantic feature vector by using a fusion formula to obtain an update display image style semantic feature vector as the update display image style semantic feature; wherein, the fusion formula is:
，
wherein, Representing the semantic feature vector of the style of the initial display image,/>Representation/>Matrix of/>Equal to the scale of the initial presentation image style semantic feature vector,/>Is/>Matrix of/>The number of the commodity text descriptor granularity semantic feature vectors in the sequence equal to the commodity text descriptor granularity semantic feature vectors,/>Representing each item text descriptor granularity semantic feature vector in the sequence of item text descriptor granularity semantic feature vectors,/>Representing the scale of each commodity text descriptor granularity semantic feature vector in the sequence of commodity text descriptor granularity semantic feature vectors,/>Is a Sigmoid function,/>Is a weight coefficient,/>And/>Convolution operation representing a 1 x1 convolution kernel,/>Representing the updated presentation image style semantic feature vector.
2. The big data based intelligent electronic commerce operation commodity management platform according to claim 1, wherein the text semantic coding module comprises:
the word segmentation unit is used for carrying out word segmentation processing on the text description of the commodity to be optimized so as to convert the text description of the commodity to be optimized into a word sequence consisting of a plurality of words;
An embedded encoding unit for mapping each word in the word sequence to a word vector using an embedded layer of a context encoder to obtain a sequence of word vectors; and
And the context semantic coding unit is used for carrying out global-based context semantic coding on the sequence of the word vectors by using the context encoder so as to obtain the sequence of the commodity text descriptor granularity semantic feature vectors.
3. The big data based intelligent electronic commerce operation commodity management platform of claim 2, wherein the deep neural network model is a convolutional neural network model.
4. The big data based intelligent electronic commerce operation commodity management platform according to claim 3, wherein said optimized presentation image generation module comprises:
the updating display image style semantic feature correction unit is used for carrying out feature distribution correction on the updating display image style semantic feature vector to obtain a corrected updating display image style semantic feature vector; and
And the picture optimization generating unit is used for enabling the corrected updated display image style semantic feature vector to pass through a display picture optimization generator based on AIGC so as to obtain the optimized display image.
5. The big data based intelligent electronic commerce operation commodity management platform according to claim 4, wherein the update presentation image style semantic feature correction unit comprises:
the feature correction subunit is used for correcting the sequence of the commodity text descriptor granularity semantic feature vectors and the initial display image style semantic feature vectors to obtain corrected feature vectors; and
And the correction feature fusion subunit is used for fusing the correction feature vector with the updated display image style semantic feature vector to obtain the corrected updated display image style semantic feature vector.
6. The intelligent electronic commerce operation commodity management method based on big data is characterized by comprising the following steps of:
acquiring text description of a commodity to be optimized, wherein the text description comprises commodity attributes, titles and feature descriptions;
Carrying out semantic coding on the text description of the commodity to be optimized to obtain a sequence of granularity semantic feature vectors of the commodity text description words;
Acquiring an initial display image of the commodity to be optimized;
extracting features of the initial display image through an image style feature extractor based on a deep neural network model to obtain an initial display image style semantic feature vector;
Optimizing the initial display image style semantic feature vector based on the sequence of the commodity text descriptor granularity semantic feature vector to obtain updated display image style semantic features; and
Generating an optimized display image based on the updated display image style semantic features;
the optimizing the initial display image style semantic feature vector based on the sequence of the commodity text descriptor granularity semantic feature vector to obtain updated display image style semantic features comprises the following steps:
Using a cross-modal information transfer fusion module, and fusing the sequence of the commodity text descriptor granularity semantic feature vectors into the initial display image style semantic feature vector by using a fusion formula to obtain an update display image style semantic feature vector as the update display image style semantic feature; wherein, the fusion formula is:
，
wherein, Representing the semantic feature vector of the style of the initial display image,/>Representation/>Matrix of/>Equal to the scale of the initial presentation image style semantic feature vector,/>Is/>Matrix of/>The number of the commodity text descriptor granularity semantic feature vectors in the sequence equal to the commodity text descriptor granularity semantic feature vectors,/>Representing each item text descriptor granularity semantic feature vector in the sequence of item text descriptor granularity semantic feature vectors,/>Representing the scale of each commodity text descriptor granularity semantic feature vector in the sequence of commodity text descriptor granularity semantic feature vectors,/>Is a Sigmoid function,/>Is a weight coefficient,/>And/>Convolution operation representing a 1 x1 convolution kernel,/>Representing the updated presentation image style semantic feature vector.
7. The big data-based intelligent electronic commerce operation commodity management method according to claim 6, wherein the semantic coding of the text description of the commodity to be optimized to obtain the sequence of the commodity text descriptor granularity semantic feature vector comprises the following steps:
word segmentation processing is carried out on the text description of the commodity to be optimized so as to convert the text description of the commodity to be optimized into a word sequence consisting of a plurality of words;
mapping each word in the word sequence to a word vector using an embedding layer of a context encoder to obtain a sequence of word vectors; and
And performing global-based context semantic coding on the sequence of word vectors by using the context encoder to obtain the sequence of commodity text descriptor granularity semantic feature vectors.
8. The big data based intelligent e-commerce operation commodity management method according to claim 7, wherein the deep neural network model is a convolutional neural network model.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311750792.XA CN117635275B (en) | 2023-12-19 | 2023-12-19 | Intelligent electronic commerce operation commodity management platform and method based on big data |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311750792.XA CN117635275B (en) | 2023-12-19 | 2023-12-19 | Intelligent electronic commerce operation commodity management platform and method based on big data |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN117635275A CN117635275A (en) | 2024-03-01 |
| CN117635275B true CN117635275B (en) | 2024-05-24 |
Family
ID=90021416
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311750792.XA Active CN117635275B (en) | 2023-12-19 | 2023-12-19 | Intelligent electronic commerce operation commodity management platform and method based on big data |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117635275B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117611245B (en) * | 2023-12-14 | 2024-05-31 | 浙江博观瑞思科技有限公司 | Data analysis management system and method for planning E-business operation activities |
Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109886330A (en) * | 2019-02-18 | 2019-06-14 | 腾讯科技（深圳）有限公司 | Method for text detection, device, computer readable storage medium and computer equipment |
| CN111402365A (en) * | 2020-03-17 | 2020-07-10 | 湖南大学 | Method for generating picture from characters based on bidirectional architecture confrontation generation network |
| CN114266840A (en) * | 2021-12-21 | 2022-04-01 | 北京达佳互联信息技术有限公司 | Image processing method, image processing device, electronic equipment and storage medium |
| CN114329025A (en) * | 2021-10-29 | 2022-04-12 | 南京南邮信息产业技术研究院有限公司 | Cross-modal text-to-image generation method based on countermeasure generation network |
| CN114898349A (en) * | 2022-05-25 | 2022-08-12 | 广州欢聚时代信息科技有限公司 | Target commodity identification method and device, equipment, medium and product thereof |
| CN115760646A (en) * | 2022-12-09 | 2023-03-07 | 中山大学·深圳 | Multi-modal face image restoration method and system for irregular holes |
| CN115760607A (en) * | 2022-11-11 | 2023-03-07 | 抖音视界有限公司 | Image restoration method, device, readable medium and electronic equipment |
| CN116597039A (en) * | 2023-05-22 | 2023-08-15 | 阿里巴巴（中国）有限公司 | Image generation method and server |
| CN117058271A (en) * | 2023-07-11 | 2023-11-14 | 浙江天猫技术有限公司 | Method and computing device for generating commodity main graph background |
| CN117151826A (en) * | 2023-09-13 | 2023-12-01 | 广州数说故事信息科技有限公司 | Multi-mode electronic commerce commodity alignment method and device, electronic equipment and storage medium |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20220351251A1 (en) * | 2021-04-30 | 2022-11-03 | Zeta Global Corp. | Generating accompanying text creative |
-
2023
- 2023-12-19 CN CN202311750792.XA patent/CN117635275B/en active Active
Patent Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109886330A (en) * | 2019-02-18 | 2019-06-14 | 腾讯科技（深圳）有限公司 | Method for text detection, device, computer readable storage medium and computer equipment |
| CN111402365A (en) * | 2020-03-17 | 2020-07-10 | 湖南大学 | Method for generating picture from characters based on bidirectional architecture confrontation generation network |
| CN114329025A (en) * | 2021-10-29 | 2022-04-12 | 南京南邮信息产业技术研究院有限公司 | Cross-modal text-to-image generation method based on countermeasure generation network |
| CN114266840A (en) * | 2021-12-21 | 2022-04-01 | 北京达佳互联信息技术有限公司 | Image processing method, image processing device, electronic equipment and storage medium |
| CN114898349A (en) * | 2022-05-25 | 2022-08-12 | 广州欢聚时代信息科技有限公司 | Target commodity identification method and device, equipment, medium and product thereof |
| CN115760607A (en) * | 2022-11-11 | 2023-03-07 | 抖音视界有限公司 | Image restoration method, device, readable medium and electronic equipment |
| CN115760646A (en) * | 2022-12-09 | 2023-03-07 | 中山大学·深圳 | Multi-modal face image restoration method and system for irregular holes |
| CN116597039A (en) * | 2023-05-22 | 2023-08-15 | 阿里巴巴（中国）有限公司 | Image generation method and server |
| CN117058271A (en) * | 2023-07-11 | 2023-11-14 | 浙江天猫技术有限公司 | Method and computing device for generating commodity main graph background |
| CN117151826A (en) * | 2023-09-13 | 2023-12-01 | 广州数说故事信息科技有限公司 | Multi-mode electronic commerce commodity alignment method and device, electronic equipment and storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN117635275A (en) | 2024-03-01 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111858954A (en) | Task-oriented text-generated image network model | |
| CN113627447B (en) | Label identification method, label identification device, computer equipment, storage medium and program product | |
| CN111079532A (en) | Video content description method based on text self-encoder | |
| CN110083729B (en) | Image searching method and system | |
| CN107145485B (en) | Method and apparatus for compressing topic models | |
| US11158349B2 (en) | Methods and systems of automatically generating video content from scripts/text | |
| CN117635275B (en) | Intelligent electronic commerce operation commodity management platform and method based on big data | |
| CN113705313A (en) | Text recognition method, device, equipment and medium | |
| CN117114063A (en) | Method for training a generative large language model and for processing image tasks | |
| CN105989067A (en) | Method for generating text abstract from image, user equipment and training server | |
| CN115964560B (en) | Information recommendation method and equipment based on multi-mode pre-training model | |
| CN115994990A (en) | Three-dimensional model automatic modeling method based on text information guidance | |
| CN113641797A (en) | Data processing method, device, equipment, storage medium and computer program product | |
| CN116955591A (en) | Recommendation language generation method, related device and medium for content recommendation | |
| CN114281982B (en) | Book propaganda abstract generation method and system adopting multi-mode fusion technology | |
| CN116662495A (en) | Question-answering processing method, and method and device for training question-answering processing model | |
| CN111988668B (en) | Video recommendation method and device, computer equipment and storage medium | |
| CN118113901A (en) | Multi-mode large language model training method, correlation calculation and label generation method | |
| CN117131222A (en) | Semi-automatic labeling method and device based on open world large model | |
| CN113806536B (en) | Text classification method and device, equipment, medium and product thereof | |
| CN113297485B (en) | Method for generating cross-modal representation vector and cross-modal recommendation method | |
| CN113688938A (en) | Method for determining object emotion and method and device for training emotion classification model | |
| CN114692715A (en) | Sample labeling method and device | |
| CN115269901A (en) | Method, device and equipment for generating extended image | |
| CN113239215A (en) | Multimedia resource classification method and device, electronic equipment 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 |