Personalized decoration scheme design auxiliary system and method based on big data

Abstract

The invention relates to the technical field of intelligent matching, and particularly discloses a personalized decoration scheme design auxiliary system and method based on big data, and the system employs a big data technology and an artificial intelligence algorithm to analyze the decoration demands of a user, captures the semantic feature expression of the decoration demands of the user, and carries out the design of a personalized decoration scheme. Meanwhile, the alternative decoration schemes are extracted from the decoration case library, whether the alternative decoration schemes are matched with the user requirements or not is judged by measuring the semantic similarity features between the user decoration requirements and the alternative decoration schemes, and therefore the alternative decoration schemes meeting the user requirements are returned. In this way, the efficiency of decoration design can be improved, meanwhile, the personalized requirements of users are met, and the innovativeness and practicability of decoration design are improved.

Description

Technical Field

[0001] The present application relates to the field of intelligent matching technology, and more specifically, to a personalized decoration scheme design assistance system and method based on big data. Background Art

[0002] With rising living standards and improved living environments, home renovation has become an integral part of daily life. A well-designed home renovation not only creates a comfortable and beautiful space but also showcases the owner's personality and taste. However, different users often have diverse decorating needs, ranging from minimalist modern styles to classical Chinese styles to luxurious European styles. Furthermore, factors such as the home's structure, floor area, lighting, and ventilation must be considered. Therefore, accurately grasping user needs becomes a major challenge in designing renovation plans.

[0003] However, traditional renovation design methods rely primarily on the designer's professional experience and manual drawing, resulting in long design cycles, low efficiency, and difficulty meeting the personalized needs of users. Furthermore, due to a lack of big data support, designers often struggle to access comprehensive renovation case studies, resulting in design solutions that lack innovation and practicality.

[0004] Therefore, we look forward to a personalized decoration scheme design assistance system and method based on big data. Summary of the Invention

[0005] In order to solve the above technical problems, the present application is proposed. The embodiment of the present application provides a personalized decoration scheme design assistance system and method based on big data, which uses big data technology and artificial intelligence algorithms to analyze the user's decoration needs, capture the semantic feature expression of the user's decoration needs, and extract alternative decoration schemes from the decoration case library. By measuring the semantic similarity features between the user's decoration needs and the alternative decoration schemes, it is determined whether the alternative decoration schemes match the user's needs, and thus an alternative decoration scheme that meets the user's needs is returned. In this way, the efficiency of decoration design can be improved, while meeting the personalized needs of users and enhancing the innovation and practicality of decoration design.

[0006] Accordingly, according to one aspect of the present application, a personalized decoration scheme design assistance system based on big data is provided, which includes:

[0007] User demand acquisition module, used to obtain the text description of the user's decoration needs;

[0008] An alternative decoration scheme extraction module is used to extract a first alternative decoration scheme from a decoration case library, wherein the first alternative decoration scheme includes a decoration rendering and a text description of the first alternative decoration scheme;

[0009] A user decoration demand semantic analysis module is used to perform semantic analysis on the text description of the user decoration demand to obtain a user decoration demand semantic feature vector;

[0010] An alternative decoration scheme semantic understanding module is used to perform semantic understanding and fusion analysis on the text description of the first alternative decoration scheme and the decoration rendering to obtain a semantic fusion feature vector of the alternative decoration rendering;

[0011] The alternative decoration scheme matching module is used to determine whether to return the first alternative decoration scheme based on the associated interaction information between the user's decoration demand semantic feature vector and the alternative decoration effect semantic fusion feature vector.

[0012] In the above-mentioned personalized decoration scheme design assistance system based on big data, the user decoration demand semantic analysis module is used to: pass the text description of the user decoration demand through the decoration demand semantic encoder based on the Transformer model to obtain the user decoration demand semantic feature vector.

[0013] In the above-mentioned personalized decoration scheme design assistance system based on big data, the user decoration demand semantic analysis module includes: a word segmentation unit, which is used to perform word segmentation processing on the text description of the user's decoration demand to obtain a sequence of demand description words; a word embedding unit, which is used to pass the sequence of demand description words through the word embedding layer of the decoration demand semantic encoder to obtain a sequence of demand description word embedding vectors; a context semantic encoding unit, which is used to pass the sequence of demand description word embedding vectors through the converter model of the decoration demand semantic encoder to obtain the user decoration demand semantic feature vector.

[0014] In the above-mentioned personalized decoration scheme design assistance system based on big data, the alternative decoration scheme semantic understanding module includes: a decoration scheme text description semantic encoding unit, which is used to pass the text description of the first alternative decoration scheme through a decoration scheme text description semantic encoder based on a Transformer model to obtain an alternative decoration scheme semantic feature vector; a decoration effect semantic feature extraction unit, which is used to pass the decoration effect image through a decoration effect feature extractor based on a ViT model to obtain a decoration effect semantic feature vector; and a fusion unit, which is used to fuse the alternative decoration scheme semantic feature vector and the decoration effect semantic feature vector to obtain the alternative scheme decoration effect semantic fusion feature vector.

[0015] In the above-mentioned personalized decoration scheme design assistance system based on big data, the alternative decoration scheme matching module includes: an association coding unit, which is used to associate and encode the user decoration demand semantic feature vector and the alternative scheme decoration effect semantic fusion feature vector to obtain a corrected user demand-alternative scheme semantic association interaction feature vector; an alternative decoration scheme returning unit, which is used to pass the corrected user demand-alternative scheme semantic association interaction feature vector through a classifier-based alternative scheme matcher to obtain a matching result, and the matching result is used to indicate whether to return the first alternative decoration scheme.

[0016] In the above-mentioned personalized decoration scheme design assistance system based on big data, the association coding unit includes: a semantic interaction fusion subunit, which is used to fuse the user decoration demand semantic feature vector and the alternative scheme decoration effect semantic fusion feature vector to obtain the user demand-alternative scheme semantic association interaction feature vector; a correction subunit, which is used to reconstruct the feature hierarchical internal structure of the user demand-alternative scheme semantic association interaction feature vector based on matrix decomposition to obtain the corrected user demand-alternative scheme semantic association interaction feature vector.

[0017] In the above-mentioned personalized decoration scheme design assistance system based on big data, the correction subunit is used to: calculate the full-channel mutual information association matrix of the user demand-alternative solution semantic association interaction feature vector; perform singular value decomposition on the full-channel mutual information association matrix to obtain a set of user demand-alternative solution semantic association interaction feature hierarchical feature representation vectors; perform feature dimension compression on each user demand-alternative solution semantic association interaction feature hierarchical feature representation vector in the set of user demand-alternative solution semantic association interaction feature hierarchical feature representation vectors to obtain a user demand-alternative solution semantic association interaction feature hierarchical feature representation vector. A set of modulation coding vectors; calculating the structural significance weight coefficient of each user demand-alternative solution semantic association interaction feature hierarchical modulation coding vector in the set of user demand-alternative solution semantic association interaction feature hierarchical modulation coding vectors to obtain a set of structural significance weight coefficients; applying regularization constraints to the set of structural significance weight coefficients to obtain a set of structural significance modulation coefficients; based on the set of structural significance modulation coefficients, hierarchically aggregating the set of user demand-alternative solution semantic association interaction feature hierarchical modulation coding vectors to obtain a corrected user demand-alternative solution semantic association interaction feature vector.

[0018] According to another aspect of the present application, a method for assisting in designing personalized decoration schemes based on big data is provided, which includes:

[0019] Get the text description of the user's decoration requirements;

[0020] Extracting a first alternative decoration plan from a decoration case library, where the first alternative decoration plan includes a decoration rendering and a text description of the first alternative decoration plan;

[0021] Performing semantic analysis on the text description of the user's decoration requirements to obtain a semantic feature vector of the user's decoration requirements;

[0022] Performing semantic understanding and fusion analysis on the text description of the first alternative decoration scheme and the decoration rendering to obtain a semantic fusion feature vector of the alternative decoration effect;

[0023] Based on the correlation interaction information between the user's decoration demand semantic feature vector and the decoration effect semantic fusion feature vector of the alternative solution, it is determined whether to return the first alternative decoration solution.

[0024] Compared to existing technologies, the big data-based personalized renovation plan design assistance system and method provided by this application utilizes big data technology and artificial intelligence algorithms to analyze users' renovation needs, capturing the semantic features of their needs. It also extracts alternative renovation plans from a renovation case library and measures the semantic similarity between the user's needs and the alternative plans to determine whether the alternative plans match the user's needs, thereby returning an alternative renovation plan that meets the user's needs. This improves the efficiency of renovation design while meeting the personalized needs of users and enhancing the innovation and practicality of renovation design. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] The above and other purposes, features, and advantages of the present application will become more apparent through a more detailed description of the embodiments of the present application in conjunction with the accompanying drawings. The accompanying drawings are intended to provide a further understanding of the embodiments of the present application and constitute a part of the specification. Together with the embodiments of the present application, they are used to explain the present application and do not constitute a limitation of the present application. In the drawings, the same reference numerals generally represent the same components or steps.

[0026] Figure 1 This is a block diagram of a personalized decoration scheme design assistance system based on big data according to an embodiment of the present application.

[0027] Figure 2 This is a schematic diagram of the architecture of a personalized decoration scheme design assistance system based on big data according to an embodiment of the present application.

[0028] Figure 3 This is a block diagram of a user decoration demand semantic analysis module in a personalized decoration scheme design assistance system based on big data according to an embodiment of the present application.

[0029] Figure 4This is a block diagram of a semantic understanding module for alternative decoration schemes in a personalized decoration scheme design assistance system based on big data according to an embodiment of the present application.

[0030] Figure 5 This is a block diagram of an alternative decoration scheme matching module in a personalized decoration scheme design assistance system based on big data according to an embodiment of the present application.

[0031] Figure 6 This is a flowchart of a method for assisting in designing personalized decoration plans based on big data according to an embodiment of the present application. DETAILED DESCRIPTION

[0032] Below, the exemplary embodiments according to the present application will be described in detail with reference to the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments of the present application, and it should be understood that the present application is not limited to the exemplary embodiments described herein.

[0033] Figure 1 This is a block diagram of a personalized decoration scheme design assistance system based on big data according to an embodiment of the present application. Figure 2 This is a schematic diagram of the architecture of a personalized decoration scheme design assistance system based on big data according to an embodiment of the present application. Figure 1 and Figure 2 As shown, according to an embodiment of the present application, a personalized decoration scheme design assistance system 100 based on big data includes: a user demand acquisition module 110, which is used to obtain a text description of the user's decoration demand; an alternative decoration scheme extraction module 120, which is used to extract a first alternative decoration scheme from a decoration case library, wherein the first alternative decoration scheme includes a decoration rendering and a text description of the first alternative decoration scheme; a user decoration demand semantic analysis module 130, which is used to perform semantic analysis on the text description of the user's decoration demand to obtain a user decoration demand semantic feature vector; an alternative decoration scheme semantic understanding module 140, which is used to perform semantic understanding and fusion analysis on the text description of the first alternative decoration scheme and the decoration rendering to obtain a semantic fusion feature vector of the alternative decoration effect; an alternative decoration scheme matching module 150, which is used to determine whether to return the first alternative decoration scheme based on the associated interaction information between the user's decoration demand semantic feature vector and the alternative decoration effect semantic fusion feature vector.

[0034] As mentioned in the above background technology, the traditional decoration scheme design method mainly relies on the professional experience and manual drawing of the designer, with a long design cycle, low efficiency, and difficulty in meeting the personalized needs of users. At the same time, due to the lack of big data support, designers often find it difficult to obtain comprehensive decoration cases, resulting in a lack of innovation and practicality in the design scheme. In response to the above technical problems, the technical concept of this application is to use big data technology and artificial intelligence algorithms to analyze the user's decoration needs, capture the semantic feature expression of the user's decoration needs, and extract alternative decoration schemes from the decoration case library. By measuring the semantic similarity features between the user's decoration needs and the alternative decoration schemes, it is determined whether the alternative decoration schemes match the user's needs, thereby returning an alternative decoration scheme that meets the user's needs. In this way, the efficiency of decoration design can be improved, while meeting the personalized needs of users and enhancing the innovation and practicality of decoration design.

[0035] In the above-mentioned big data-based personalized decoration scheme design assistance system 100, the user demand acquisition module 110 is used to obtain a text description of the user's decoration needs. It should be understood that the text description of the user's decoration needs usually includes the user's desired decoration style, budget, required functional areas, specific design requirements, etc. By obtaining the user's decoration demand information, natural language processing technology can be used to semantically understand the user's decoration demand information, capture the user's intentions and preferences, and provide the user with a decoration plan that better meets their actual needs. In specific implementation, users can input their decoration needs through various methods, such as online forms, voice input, or through dialogue with an intelligent assistant.

[0036] In the aforementioned big data-based personalized renovation design assistance system 100, the alternative renovation scheme extraction module 120 is configured to extract a first alternative renovation scheme from a renovation case library. The first alternative renovation scheme includes a renovation rendering and a text description of the first alternative renovation scheme. It should be understood that the renovation case library contains a large number of historical renovation cases. These historical renovation cases can be previous successful renovation projects or virtual cases designed by professional designers. Each renovation case in the renovation case library includes a renovation rendering and a corresponding text description. The renovation rendering can intuitively display the final effect of the renovation. Users can intuitively understand the characteristics of the renovation style and layout design through the rendering, making it easier to determine their preferences and needs. The text description records in detail the design ideas, materials used, and decoration style of the renovation scheme, allowing users to understand more details of the renovation scheme. In the technical solution of the present application, the historical renovation cases in the renovation case library are used as alternative renovation schemes. By analyzing the degree of semantic match between the alternative renovation schemes and the user's renovation requirements, the user is screened for renovation schemes that meet their actual needs. This can greatly improve the efficiency of renovation design and meet the user's personalized needs.

[0037] In the above-mentioned personalized decoration scheme design assistance system 100 based on big data, the user decoration demand semantic analysis module 130 is used to perform semantic analysis on the text description of the user decoration demand to obtain the user decoration demand semantic feature vector. In a specific example of the present application, the encoding method of performing semantic analysis on the text description of the user decoration demand to obtain the user decoration demand semantic feature vector is to pass the text description of the user decoration demand through a decoration demand semantic encoder based on the Transformer model to obtain the user decoration demand semantic feature vector. Those skilled in the art should know that the Transformer model is a deep learning model based on the self-attention mechanism, which has excellent performance in the field of natural language processing. It can learn the semantic information in text data and convert it into a vector representation of fixed dimension. In the technical solution of the present application, the decoration requirement semantic encoder based on the Transformer model can effectively learn the semantic dependency between words at various positions in the text description of the user's decoration requirements through the self-attention mechanism, map the text structure data into the semantic space to capture the key information in the text description of the user's decoration requirements, such as decoration style, budget, functional requirements, etc., and convert the text structure data into a feature vector representation, so that the subsequent matching operation of alternative decoration plans is more convenient and efficient.

[0038] Figure 3This is a block diagram of a user decoration demand semantic analysis module in a personalized decoration scheme design assistance system based on big data according to an embodiment of the present application. Figure 3 As shown, the user decoration demand semantic analysis module 130 includes: a word segmentation unit 131, which is used to perform word segmentation processing on the text description of the user decoration demand to obtain a sequence of demand description words; a word embedding unit 132, which is used to pass the sequence of demand description words through the word embedding layer of the decoration demand semantic encoder to obtain a sequence of demand description word embedding vectors; a context semantic encoding unit 133, which is used to pass the sequence of demand description word embedding vectors through the converter model of the decoration demand semantic encoder to obtain the user decoration demand semantic feature vector.

[0039] In the above-mentioned personalized decoration scheme design assistance system 100 based on big data, the alternative decoration scheme semantic understanding module 140 is used to perform semantic understanding and fusion analysis on the text description of the first alternative decoration scheme and the decoration effect diagram to obtain the semantic fusion feature vector of the alternative decoration effect. Specifically, Figure 4 This is a block diagram of the semantic understanding module of the alternative decoration scheme in the personalized decoration scheme design assistance system based on big data according to the embodiment of the present application. Figure 4 As shown, the alternative decoration scheme semantic understanding module 140 includes: a decoration scheme text description semantic encoding unit 141, which is used to pass the text description of the first alternative decoration scheme through a decoration scheme text description semantic encoder based on a Transformer model to obtain an alternative decoration scheme semantic feature vector; a decoration effect semantic feature extraction unit 142, which is used to pass the decoration effect picture through a decoration effect feature extractor based on a ViT model to obtain a decoration effect semantic feature vector; and a fusion unit 143, which is used to fuse the alternative decoration scheme semantic feature vector and the decoration effect semantic feature vector to obtain the alternative decoration effect semantic fusion feature vector.

[0040] Specifically, the decoration scheme text description semantic encoding unit 141 is used to pass the text description of the first alternative decoration scheme through the decoration scheme text description semantic encoder based on the Transformer model to obtain the alternative decoration scheme semantic feature vector. In a specific example of the present application, the decoration scheme text description semantic encoder and the decoration requirement semantic encoder are the same in model structure. It should be understood that the text description of the first alternative decoration scheme is semantically encoded by the decoration scheme text description semantic encoder based on the Transformer model to convert the text information into a numerical feature vector representation, thereby better characterizing the design style, material selection, spatial layout and other information of the first alternative decoration scheme, providing a data basis for subsequent scheme matching. In addition, by converting the text description of the first alternative decoration scheme into a vector form, it is beneficial to represent and compare the first alternative decoration scheme in a high-dimensional space, more conveniently evaluate the matching degree with user needs, and provide users with personalized recommendations that better meet their needs.

[0041] Specifically, the decoration effect semantic feature extraction unit 142 is used to pass the decoration effect picture through the decoration effect feature extractor based on the ViT model to obtain a decoration effect semantic feature vector. It should be understood that the decoration effect picture is also an important information carrier in the first alternative decoration scheme. Therefore, in the technical solution of the present application, the decoration effect feature extractor based on the ViT model is used to extract image semantic features of the decoration effect picture. Those skilled in the art should know that the ViT model (Vision Transformer) is a visual processing model based on the self-attention mechanism. It has a wide range of applications in the field of image processing and can convert image data into a feature vector of fixed dimension, thereby realizing a numerical representation of the image content. That is, the decoration effect feature extractor based on the ViT model is based on the self-attention mechanism and can effectively extract the visual features in the decoration effect picture, such as color matching, material texture, spatial layout and other information, so as to better characterize the visual characteristics of the decoration scheme and convert the decoration effect picture into a feature vector representation for subsequent matching and recommendation operations.

[0042] In a specific example of the present application, the decoration effect semantic feature extraction unit 142 includes: an image segmentation sub-unit, used to perform image segmentation processing on the decoration effect image to obtain a sequence of decoration effect image blocks; an embedding coding sub-unit, used to use the embedding layer of the ViT model to embed code each decoration effect image block in the sequence of decoration effect image blocks to obtain a sequence of decoration effect image block embedding vectors; a conversion coding sub-unit, used to input the sequence of decoration effect image block embedding vectors into the converter module of the ViT model to obtain the decoration effect semantic feature vector.

[0043] Specifically, the fusion unit 143 is used to fuse the semantic feature vector of the alternative decoration scheme and the semantic feature vector of the decoration effect to obtain the semantic fusion feature vector of the decoration effect of the alternative scheme. It should be understood that the semantic feature vector of the alternative decoration scheme and the semantic feature vector of the decoration effect respectively capture the important features in the text description and visual information of the first alternative decoration scheme. In order to combine visual information with text information, thereby more comprehensively describing the first alternative decoration scheme, the semantic feature vector of the alternative decoration scheme and the semantic feature vector of the decoration effect are further feature fused to achieve multimodal information fusion, so that it can better express the design concept, style characteristics and visual effects of the first alternative decoration scheme, and more comprehensively reflect the characteristics and advantages of the first alternative decoration scheme, thereby providing users with more accurate and personalized decoration scheme recommendations in the subsequent scheme matching and recommendation process.

[0044] In the above-mentioned personalized decoration scheme design assistance system 100 based on big data, the alternative decoration scheme matching module 150 is used to determine whether to return the first alternative decoration scheme based on the correlation interaction information between the user's decoration demand semantic feature vector and the alternative decoration effect semantic fusion feature vector. Specifically, Figure 5 FIG. 1 is a block diagram of an alternative decoration scheme matching module in a personalized decoration scheme design assistance system based on big data according to an embodiment of the present application. Figure 5 As shown, the alternative decoration scheme matching module 150 includes: an association coding unit 151, which is used to perform association coding on the user decoration demand semantic feature vector and the alternative decoration effect semantic fusion feature vector to obtain a corrected user demand-alternative scheme semantic association interaction feature vector; an alternative decoration scheme returning unit 152, which is used to pass the corrected user demand-alternative scheme semantic association interaction feature vector through a classifier-based alternative scheme matcher to obtain a matching result, and the matching result is used to indicate whether to return the first alternative decoration scheme.

[0045] Specifically, the association coding unit 151 includes: a semantic interaction fusion subunit for fusing the user decoration demand semantic feature vector and the alternative solution decoration effect semantic fusion feature vector to obtain the user demand-alternative solution semantic association interaction feature vector; and a correction subunit for performing a feature hierarchical internal structure reconstruction based on matrix decomposition on the user demand-alternative solution semantic association interaction feature vector to obtain a corrected user demand-alternative solution semantic association interaction feature vector. In particular, in the technical solution of the present application, due to the essential heterogeneity of the user decoration demand semantic feature vector and the alternative solution decoration effect semantic fusion feature vector in terms of feature expression dimension, semantic coupling mode and interaction logic, the user demand-alternative solution semantic association interaction feature vector obtained by weighted fusion of the two is prone to decision bias due to noise interference in the high-dimensional feature space and linear superposition of redundant dimensions, making it impossible for the direct matching mechanism to analyze the deep design logic, but instead amplifying the semantic bias, resulting in a decrease in the accuracy of solution recommendation and reduced user satisfaction. To solve this problem, in the technical solution of the present application, the user demand-alternative solution semantic association interaction feature vector is reconstructed based on matrix decomposition to obtain a feature hierarchical internal structure.

[0046] More specifically, the user demand-alternative solution semantic association interaction feature vector is subjected to a feature hierarchical internal structure reconstruction based on matrix decomposition to obtain a corrected user demand-alternative solution semantic association interaction feature vector, including: first, calculating the full-channel mutual information association matrix of the user demand-alternative solution semantic association interaction feature vector, which is expressed as follows:

[0047]

[0048] M=D1⊙D2

[0049] v i ,v j ∈V

[0050] Among them, V represents the initial eigenvector, v i and v j They represent the i-th and j-th eigenvalues of the initial eigenvector, w1, w2, w3 and w4 represent different weight hyperparameters, ⊙ represents matrix dot product, D1 represents the initial feature forward weight matrix, D2 represents the initial feature reverse weight matrix, Represents the value of the (i, j)th position of the initial feature forward weight matrix, represents the value of the (i, j)th position of the initial feature inverse weighting matrix, and M represents the global fine-grained autocorrelation topology matrix.

[0051] That is, through the construction of the full-channel mutual information correlation matrix, the multi-dimensional interactive correlation of the user demand-alternative solution semantic correlation interaction feature vector in the semantic space is systematically quantified. On the basis of retaining the fine-grained structural information of the features, a topological correlation mapping covering the entire feature domain is established, and the discrete semantic features are converted into a measurable correlation strength distribution, thereby achieving high-order semantic alignment between user preferences and design solutions in key dimensions such as style tendencies, spatial layout, and material adaptation.

[0052] Secondly, the full-channel mutual information association matrix is subjected to singular value decomposition to obtain a set of hierarchical feature representation vectors of user demand-alternative solution semantic association interaction features, which can be expressed as follows:

[0053]

[0054] Where Λ represents a diagonal matrix, λ1 and λ m denote the first and mth eigenvalues of the diagonal matrix, respectively, (·) T represents the transpose of the vector, U represents the set of initial feature fine-grained eigencomponent encoding vectors, x1, x2 and x m They respectively represent the first, second and mth initial feature fine-grained eigencomponent encoding vectors in the set of initial feature fine-grained eigencomponent encoding vectors.

[0055] That is, through the rank approximation strategy of the full-channel mutual information correlation matrix, the high-dimensional full-channel mutual information correlation matrix is decoupled into independent components arranged in descending order of singular values, forming a hierarchical feature representation vector of user demand-alternative solution semantic correlation interaction features from global semantic coupling to local detail differences. While retaining the contribution of the main components of the features, the core dimensions and noise interference components that characterize the strength of the semantic correlation are separated, so that the multimodal interaction pattern originally implicit in the mutual information matrix is transformed into a quantifiable hierarchical feature spectrum.

[0056] Then, feature dimension compression is performed on each user demand-alternative solution semantic association interaction feature hierarchical feature representation vector in the set of user demand-alternative solution semantic association interaction feature hierarchical feature representation vectors to obtain a set of user demand-alternative solution semantic association interaction feature hierarchical modulation coding vectors, which is expressed as follows:

[0057]

[0058] Among them, x i represents the i-th initial feature fine-grained intrinsic component encoding vector in the set of initial feature fine-grained intrinsic component encoding vectors, ||·|| represents the Euclidean norm, y i represents the i-th initial feature fine-grained eigencomponent modulation coding vector.

[0059] Specifically, through the synergistic effect of dynamic range adjustment and norm constraints, redundant degrees of freedom are compressed while preserving core structural information, mapping the high-dimensional feature space into a lower-dimensional but more semantically cohesive representational domain. This, while suppressing the interference of noise components and minor details, forms a set of hierarchically discriminative hierarchical modulation coding vectors of user demand-alternative semantic association interaction features. High-dimensional principal components are nonlinearly compressed and transformed into core semantic labels representing stylistic tendencies and spatial topology, while low-rank components are compressed into differentiated feature codes reflecting material texture and decorative elements. This allows style, layout, and material information, originally dispersed across multiple layers of features, to achieve high-level semantic coupling within a compact coding space, providing an optimized foundation for hierarchical aggregation that combines semantic coherence and computational efficiency.

[0060] Next, the structural significance weight coefficient of each user demand-alternative solution semantic association interaction feature hierarchical modulation coding vector in the set of user demand-alternative solution semantic association interaction feature hierarchical modulation coding vectors is calculated to obtain a set of structural significance weight coefficients, which is expressed as follows:

[0061]

[0062] Among them, α and β represent different weight parameters, ||·||1 represents the first norm, ||·||2 represents the second norm, L represents the length of the initial feature fine-grained eigencomponent modulation coding vector, a i represents y i The corresponding internal structure significant modulation factor.

[0063] Specifically, by quantifying the structural contribution of the hierarchical modulation coding vectors of the semantically related interaction features between user needs and alternative solutions, we distinguish between core semantic components and secondary modifying elements, thereby providing differentiated weight coefficients for subsequent feature aggregation. This data-driven weight allocation strategy generates a collection of structural significance weight coefficients that not only strengthens the decision-making dominance of key semantic features but also significantly improves the accuracy and robustness of cross-modal matching by dynamically suppressing the interference of redundant information.

[0064] Then, a regularization constraint is imposed on the set of structural significance weight coefficients to obtain a set of structural significance modulation coefficients, which is expressed as follows:

[0065] w i =Softmax(a i )

[0066] Among them, Softmax(·) represents the regularization function, w i Indicates a i The corresponding internal structure significantly modulates the weight factor.

[0067] Specifically, by introducing a structural significance weight coefficient, we suppress the tendency of extreme weight distribution, preventing gradient instability caused by feature dimensionality differences or local noise interference, while simultaneously optimizing the structural significance weight coefficient's ability to represent the core semantic structure. The resulting set of structural significance modulation coefficients not only provides a robust decision-making basis for subsequent hierarchical aggregation, but also balances the dialectical relationship between core user demands and solution details through dynamic adjustment of constraint strength, ultimately achieving the coordinated optimization of controllability and generalization capabilities of the semantic matching process.

[0068] Finally, based on the set of structural significance modulation coefficients, the set of hierarchical modulation coding vectors of the user demand-alternative solution semantic association interaction features is hierarchically aggregated to obtain a corrected user demand-alternative solution semantic association interaction feature vector, which is expressed as follows:

[0069]

[0070] Where V' represents the corrected user demand-alternative solution semantic association interaction feature vector.

[0071] That is, through hierarchical aggregation guided by the structural saliency modulation coefficient, the synergistic enhancement of cross-scale structural information is achieved while preserving the independent semantic contributions of the hierarchical modulation coding vectors of the semantic association interaction features between each user demand and alternative solution. Specifically, through hierarchical aggregation and combination, the core design elements carried by the high-saliency components form a nonlinear superposition effect during the aggregation process, while the local detail differences carried by the low-saliency components are suppressed as background noise. The resulting corrected user demand-alternative solution semantic association interaction feature vector not only improves the semantic accuracy of cross-modal matching through hierarchical optimization, but also enhances the robustness of feature representation through nonlinear fusion, providing an optimized foundation for subsequent solution generation and ranking that combines semantic coherence and decision discrimination.

[0072] Specifically, the alternative decoration scheme returning unit 152 is used to pass the corrected user demand-alternative scheme semantic association interaction feature vector through a classifier-based alternative scheme matcher to obtain a matching result, and the matching result is used to indicate whether to return the first alternative decoration scheme. It should be understood that a classifier is an algorithm that can perform classification based on the characteristics of input data. In the technical solution of the present application, the classifier-based alternative scheme matcher can determine whether the alternative scheme matches the user demand by learning the semantic similarity features between the user demand and the alternative scheme in the corrected user demand-alternative scheme semantic association interaction feature vector, thereby deciding whether to return the first alternative decoration scheme. In this way, based on the classification learning ability of the classifier, it is possible to screen out the scheme that meets the user's needs from the predefined alternative decoration schemes and return it to the user as a matching result, thereby providing the user with more personalized and accurate decoration scheme recommendations to enhance the user's decoration experience.

[0073] In summary, the big data-based personalized renovation plan design assistance system according to the embodiments of the present application is described. It utilizes big data technology and artificial intelligence algorithms to analyze a user's renovation needs, capturing the semantic features of their needs. It also extracts alternative renovation plans from a library of renovation cases. By measuring the semantic similarity between the user's needs and the alternative plans, it determines whether the alternative plans match the user's needs, thereby returning an alternative plan that meets the user's needs. This improves the efficiency of renovation design while meeting the user's personalized needs and enhancing the innovation and practicality of renovation design.

[0074] Figure 6 Flowchart of the method for assisting in designing personalized decoration schemes based on big data according to an embodiment of the present application. Figure 6 As shown, according to the embodiment of the present application, the personalized decoration scheme design assistance method based on big data includes the following steps: S110, obtaining a text description of the user's decoration needs; S120, extracting a first alternative decoration scheme from a decoration case library, wherein the first alternative decoration scheme includes a decoration rendering and a text description of the first alternative decoration scheme; S130, performing semantic analysis on the text description of the user's decoration needs to obtain a semantic feature vector of the user's decoration needs; S140, performing semantic understanding and fusion analysis on the text description of the first alternative decoration scheme and the decoration rendering to obtain a semantic fusion feature vector of the alternative decoration effect; S150, determining whether to return the first alternative decoration scheme based on the associated interaction information between the user's decoration needs semantic feature vector and the alternative decoration effect semantic fusion feature vector.

[0075] Here, those skilled in the art will understand that the specific operations of each step in the above-mentioned personalized decoration scheme design auxiliary method based on big data have been referred to above. Figures 1 to 5 It has been introduced in detail in the description of the personalized decoration scheme design assistance system based on big data, and therefore, its repeated description will be omitted.

[0076] In the above embodiments, the description of each embodiment has its own emphasis. For parts that are not described or recorded in detail in a certain embodiment, please refer to the relevant description of other embodiments. In the several embodiments provided by the present invention, it should be understood that the disclosed devices, systems and methods can be implemented in other ways. For example, the system embodiments described above are only schematic. For example, the module division is only a logical function division, and there may be other division methods in actual implementation. The modules described as separate components may or may not be physically separated, and the components displayed as modules may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment.

[0077] In addition, the functional modules in various embodiments of the present invention may be integrated into a single processing unit, each unit may exist physically separately, or two or more units may be integrated into a single unit. The aforementioned integrated units may be implemented in the form of hardware or hardware plus software functional modules.

[0078] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above and that the invention can be embodied in other specific forms without departing from the spirit or essential characteristics of the invention. Therefore, the embodiments should be considered in all respects as illustrative and non-restrictive, and the scope of the invention is defined by the appended claims, not the foregoing description, and all variations within the meaning and range of equivalents of the claims are intended to be encompassed therein. Any reference to a figure in a claim should not be construed as limiting the claim to which it relates.

[0079] In addition, it is obvious that the word "comprising" does not exclude other units or steps, and the singular does not exclude the plural. Multiple units stated in the system claims can also be implemented by one unit through software or hardware.

[0080] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not limiting. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. A personalized decoration scheme design auxiliary system based on big data, characterized by: include: User demand acquisition module, used to obtain the text description of the user's decoration needs; An alternative decoration scheme extraction module is used to extract a first alternative decoration scheme from a decoration case library, wherein the first alternative decoration scheme includes a decoration rendering and a text description of the first alternative decoration scheme; A user decoration demand semantic analysis module is used to perform semantic analysis on the text description of the user decoration demand to obtain a user decoration demand semantic feature vector; An alternative decoration scheme semantic understanding module is used to perform semantic understanding and fusion analysis on the text description of the first alternative decoration scheme and the decoration rendering to obtain a semantic fusion feature vector of the alternative decoration rendering; The alternative decoration scheme matching module is used to determine whether to return the first alternative decoration scheme based on the associated interaction information between the user's decoration demand semantic feature vector and the alternative decoration effect semantic fusion feature vector.

2. The personalized decoration scheme design assistance system based on big data according to claim 1 is characterized in that: The user decoration demand semantic analysis module is used to: The text description of the user's decoration requirements is passed through a decoration requirements semantic encoder based on a Transformer model to obtain a semantic feature vector of the user's decoration requirements.

3. The personalized decoration scheme design assistance system based on big data according to claim 2 is characterized in that: The user decoration demand semantic analysis module includes: A word segmentation unit, configured to perform word segmentation processing on the text description of the user's decoration requirements to obtain a sequence of requirement description words; a word embedding unit, configured to pass the sequence of demand description words through the word embedding layer of the renovation demand semantic encoder to obtain a sequence of demand description word embedding vectors; The context semantic encoding unit is used to pass the sequence of the demand description word embedding vectors through the converter model of the decoration demand semantic encoder to obtain the user decoration demand semantic feature vector.

4. The personalized decoration scheme design assistance system based on big data according to claim 3 is characterized in that: The alternative decoration scheme semantic understanding module includes: a decoration plan text description semantic encoding unit, configured to pass the text description of the first alternative decoration plan through a decoration plan text description semantic encoder based on a Transformer model to obtain a semantic feature vector of the alternative decoration plan; A decoration effect semantic feature extraction unit, configured to pass the decoration effect image through a decoration effect feature extractor based on a ViT model to obtain a decoration effect semantic feature vector; A fusion unit is used to fuse the semantic feature vector of the alternative decoration scheme and the semantic feature vector of the decoration effect to obtain the semantic fusion feature vector of the alternative decoration effect.

5. The personalized decoration scheme design assistance system based on big data according to claim 4 is characterized in that: The alternative decoration scheme matching module includes: an association coding unit, configured to perform association coding on the user decoration demand semantic feature vector and the alternative decoration effect semantic fusion feature vector to obtain a corrected user demand-alternative solution semantic association interaction feature vector; The alternative decoration scheme returning unit is used to pass the corrected user demand-alternative scheme semantic association interaction feature vector through a classifier-based alternative scheme matcher to obtain a matching result, and the matching result is used to indicate whether to return the first alternative decoration scheme.

6. The personalized decoration scheme design assistance system based on big data according to claim 5 is characterized in that: The associated coding unit includes: A semantic interaction fusion subunit, configured to fuse the user's decoration demand semantic feature vector and the alternative solution decoration effect semantic fusion feature vector to obtain the user demand-alternative solution semantic association interaction feature vector; The correction subunit is used to reconstruct the feature hierarchical internal structure of the user demand-alternative solution semantic association interaction feature vector based on matrix decomposition to obtain a corrected user demand-alternative solution semantic association interaction feature vector.

7. The personalized decoration scheme design assistance system based on big data according to claim 6 is characterized in that: The syndrome unit is configured to: Calculating the full-channel mutual information association matrix of the user demand-alternative solution semantic association interaction feature vector; Performing singular value decomposition on the full-channel mutual information association matrix to obtain a set of hierarchical feature representation vectors of user demand-alternative solution semantic association interaction features; Performing feature dimension compression on each user demand-alternative solution semantic association interaction feature hierarchical feature representation vector in the set of user demand-alternative solution semantic association interaction feature hierarchical feature representation vectors to obtain a set of user demand-alternative solution semantic association interaction feature hierarchical modulation coding vectors; Calculating the structural significance weight coefficient of each user demand-alternative solution semantic association interaction feature hierarchical modulation coding vector in the set of user demand-alternative solution semantic association interaction feature hierarchical modulation coding vectors to obtain a set of structural significance weight coefficients; Applying regularization constraints to the set of structural significance weight coefficients to obtain a set of structural significance modulation coefficients; Based on the set of structural significance modulation coefficients, the set of user demand-alternative solution semantic association interaction feature hierarchical modulation coding vectors is hierarchically aggregated to obtain a corrected user demand-alternative solution semantic association interaction feature vector.

8. A personalized decoration scheme design auxiliary method based on big data, characterized in that: include: Get the text description of the user's decoration requirements; Extracting a first alternative decoration plan from a decoration case library, where the first alternative decoration plan includes a decoration rendering and a text description of the first alternative decoration plan; Performing semantic analysis on the text description of the user's decoration requirements to obtain a semantic feature vector of the user's decoration requirements; Performing semantic understanding and fusion analysis on the text description of the first alternative decoration scheme and the decoration rendering to obtain a semantic fusion feature vector of the alternative decoration effect; Based on the correlation interaction information between the user's decoration demand semantic feature vector and the decoration effect semantic fusion feature vector of the alternative solution, it is determined whether to return the first alternative decoration solution.

Images