A cloud platform-based home design data management system and method

Abstract

The application discloses a kind of based on cloud platform's home design data management system and method, it is related to data management technical field.The system of the present application includes: historical project data processing module, design template construction module, real-time demand matching module and design adjustment and optimization module;Historical project data processing module analyzes historical project data, and constructs design resource database;Design template construction module creates design template based on design resource database, and generates label feature vector for each design template;Real-time demand matching module constructs demand feature vector according to real-time project demand data, and is matched with the label feature vector of design template, obtains reference design template;Design adjustment and optimization module analyze the deviation between real-time project demand and reference design template, generate optimization prompt information, assist relevant personnel to adjust design template, generate target design scheme.

Description

Technical Field

[0001] This invention relates to the field of data management technology, specifically to a cloud-based home design data management system and method. Background Technology

[0002] With the increasing diversification of modern home design needs, interior designers face a large number of design tasks and workloads, especially in the design of large residential, community, and commercial spaces, where designers often need to repeatedly create a large number of similar design elements. Traditional home design methods mainly rely on designers manually creating each design element, model, or layout, which not only increases repetitive work but also leads to inefficiency and waste of resources in the design process.

[0003] Currently, many home design companies and individual designers still use traditional design tools and methods. While these tools can perform basic design functions, they generally lack flexible resource sharing and reuse mechanisms. For example, in multiple projects, designers need to design living rooms of different sizes or use similar furniture and accessories in multiple room layouts. Traditional design processes require starting from scratch each time. Even if there are existing furniture models or space templates, designers still need to determine which existing models or templates are closest to the current project and which elements of the selected models or templates need adjustment. Because the design requirements and details of each project are different, this approach leads to a significant waste of time and effort. In cases where multiple projects are carried out alternately in large-scale residential, community, and commercial spaces, the problems of inefficiency and repetitive work become even more pronounced. Summary of the Invention

[0004] The purpose of this invention is to provide a cloud-based home design data management system and method to solve the problems mentioned in the background art.

[0005] To solve the above-mentioned technical problems, the present invention provides the following technical solution:

[0006] A cloud-based method for managing home design data includes the following steps:

[0007] Step S100. Obtain historical project data from the home design platform, divide the historical project data according to the preset minimum element to obtain the element set corresponding to the historical project data; classify the elements corresponding to the historical project data according to the corresponding element set to build a design resource database.

[0008] Step S200. Based on the design resource database, create several design templates on the home design platform. By analyzing the preset minimum elements that make up each design template, obtain the tag set corresponding to the design template, and extract keywords from the tag set to construct the corresponding tag feature vector.

[0009] Step S300. Obtain real-time project requirement data, analyze the real-time project requirement data, and construct a real-time project requirement feature vector; analyze the real-time project requirement feature vector and the label feature vector corresponding to each design template, match the design template corresponding to the real-time project requirement data according to the analysis results, and use the design template corresponding to the real-time project requirement data as a reference design template.

[0010] Step S400. Analyze the deviation between the real-time project requirement feature vector and the label feature vector corresponding to the reference design template, and generate corresponding prompts based on the deviation. Relevant personnel then adjust the reference design template based on the prompts to generate the target design scheme.

[0011] Furthermore, step S100 includes:

[0012] S101. Obtain several historical project data from the home design platform, divide each historical project data according to a preset minimum element, and decompose each historical project data into a set of elements Ai, where Ai = {ai1, ai2, ..., ain}, where Ai represents the set of elements corresponding to the i-th historical project data, i represents the data number of the historical project data, which is a positive integer; ai1 represents the first element of the i-th historical project data, ai2 represents the second element of the i-th historical project data, and so on, ain represents the n-th element of the i-th historical project data; the preset minimum element refers to the most basic unit in the home design project that can be independently identified, analyzed, and processed; for example, sofa, bed, table, chair, lamp, wall, window, door, floor, etc.

[0013] S102. Summarize the element set corresponding to all historical project data, and extract the element attribute corresponding to each element in the element set. The element attribute refers to the feature description corresponding to the preset minimum element. Construct the corresponding element attribute feature vector V1 based on the element attribute, and the element attribute feature vector V1=[v11,v12,...,v1u], where u represents the dimension of the element attribute feature vector V1, v11 represents the first dimension feature of the element attribute feature vector V1, v12 represents the second dimension feature of the element attribute feature vector V1, and so on, v1u represents the u-th dimension feature of the element attribute feature vector V1. Calculate the similarity between all element attribute feature vectors V1, and use a clustering algorithm to divide all elements into several categories, and each category corresponds to a category label, thereby constructing a design resource database.

[0014] Furthermore, step S200 includes:

[0015] S201. Based on the design resource database, create several design templates on the home design platform, and each design template represents a complete home design scheme; obtain the design element set Dj corresponding to each design template, represented as: Dj={ej1,ej2,...,ejm}, where Dj represents the design element set corresponding to the j-th design template, j represents the data number of the design template; ej1 represents the first design element of the j-th design template, ej2 represents the second design element of the j-th design template, and so on, ejm represents the m-th design element of the j-th design template; the design elements in the design element set correspond to the preset minimum element in the design resource database;

[0016] S202. For each design element set Dj corresponding to a design template, obtain the category tags of the design elements in the design resource database, and summarize the category tags of all design elements to form the tag set Lj corresponding to the design element set Dj; perform word segmentation and keyword extraction on each category tag in the tag set Lj, and calculate the weight coefficient W of each category tag in the corresponding design template, and the specific calculation formula of the weight coefficient W is as follows:

[0017] W(kc)=[TF(kc,Lj) / ∑kd∈Lj,TF(kd,Lj)]×IDF(kc);

[0018] Where W(kc) represents the weight coefficient of category label kc, and c represents the data number of the category label in the design resource database; TF(kc,Lj) represents the frequency of category label kc in label set Lj; ∑kd∈Lj,TF(kd,Lj) represents the sum of the frequencies of all category labels in label set Lj; IDF(kc) is the inverse document frequency, representing the scarcity of category label kc in all design templates, and the calculation formula is:

[0019] IDF(kc) = log[N / DF(kc)];

[0020] Where N represents the total number of design templates, and DF(kc) represents the number of design templates containing the category label kc;

[0021] S203. Based on the weight coefficient of each category label kc in the label set Lj, and combined with the design resource database, construct the label feature vector V_Lj corresponding to the design template Dj, and V_Lj=[W(k1),W(k2),...,W(kp)], where W(k1) represents the weight coefficient of category label k1, W(k2) represents the weight coefficient of category label k2, and so on, W(kp) represents the weight coefficient of category label kp, p equals the total number of category labels in the design resource database, and p is greater than or equal to m; the order of the category label weight coefficients in the label feature vector V_Lj follows the order of the category labels in the design resource database. If the label set Lj does not contain a certain category label from the design resource database, then the weight coefficient of this category label is assigned to zero.

[0022] Furthermore, step S300 includes:

[0023] S301. Obtain real-time project requirement data. Following the analysis method for historical project data, extract the preset minimum element and its corresponding element attributes from the real-time project requirement data to obtain the real-time requirement element set B and the real-time requirement element attribute feature vector V2. Calculate the similarity between the real-time requirement element attribute feature vector V2 corresponding to each element in the real-time requirement element set B and the element attribute feature vector V1 of each category in the design resource database, and calculate the average similarity of the corresponding categories. Select the category label with the largest average similarity as the category label of the corresponding element in the real-time requirement element set B. Summarize all element category labels in the real-time requirement element set B, and for all element category labels in the real-time requirement element set B, refer to the analysis process of the label feature vector corresponding to the design template to construct the real-time project requirement feature vector V_req, where the dimension of the real-time project requirement feature vector V_req is equal to p.

[0024] S302. Represent the real-time project requirement feature vector and the corresponding label feature vector of each design template using a radar chart. The specific analysis process is as follows:

[0025] In a two-dimensional plane, take any point as the origin of the coordinate system. Use the dimension p of the real-time project requirement feature vector V_req as the number of number axes. Each number axis starts from the origin of the coordinate system, and the angle between each number axis is equal. Mark the feature values ​​corresponding to the real-time project requirement feature vector V_req on the number axes in a clockwise order to obtain p data points. Connect the data points of adjacent number axes to form a closed polygon. Calculate the area S_req of the closed polygon.

[0026] Following the analysis method of the real-time project requirement feature vector V_req, the label feature vector V_Lj corresponding to each design template is represented on a two-dimensional plane, and the area S_Lj of the corresponding closed polygon is calculated. Based on the area S_req and S_Lj of the closed polygon, the matching index R between the real-time project requirement feature vector and the label feature vector corresponding to each design template is calculated sequentially, and the specific calculation formula is as follows:

[0027] R={(S_req·S_Lj) / [S_req+S_Lj-|S_req-S_Lj|]}·α;

[0028] Where α represents the shape similarity between the real-time project requirement feature vector and the closed polygon corresponding to the label feature vector of the design template, and the value ranges from 0 to 1; the matching index R corresponding to each design template is arranged in descending order, and the design template with the largest matching index R is selected as the reference design template.

[0029] Furthermore, step S400 includes:

[0030] The difference between the feature values ​​of the real-time project requirement feature vector V_req and the label feature vector V_Lj corresponding to the reference design template in each dimension is calculated, and the absolute value of the difference calculation result is obtained to obtain the corresponding deviation value Δf. The deviation values ​​Δf corresponding to each dimension are sorted in descending order to generate a deviation sorting list. The deviation sorting list is output as prompt information to relevant personnel, who then adjust the reference design template based on the prompt information. The adjusted reference design template is then used as the target design scheme.

[0031] A cloud-based home design data management system includes: a historical project data processing module, a design template construction module, a real-time demand matching module, and a design adjustment and optimization module;

[0032] The historical project data processing module acquires historical project data from the home design platform, divides the historical project data according to preset minimum elements, and obtains the element set corresponding to the historical project data; based on the corresponding element set, the elements corresponding to the historical project data are classified, thereby constructing a design resource database;

[0033] The design template building module is based on a design resource database. It creates several design templates on the home design platform. By analyzing the preset minimum elements that make up each design template, it obtains the tag set corresponding to the design template, and extracts keywords from the tag set to build the corresponding tag feature vector.

[0034] The real-time requirement matching module acquires and analyzes real-time project requirement data to construct a real-time project requirement feature vector. It then analyzes the real-time project requirement feature vector with the label feature vector corresponding to each design template, matches the design template corresponding to the real-time project requirement data based on the analysis results, and uses the design template corresponding to the real-time project requirement data as a reference design template.

[0035] The design adjustment and optimization module analyzes the deviation between the real-time project requirement feature vector and the corresponding label feature vector of the reference design template, and generates corresponding prompts based on the deviation. Relevant personnel then adjust the reference design template based on the prompts to generate the target design scheme.

[0036] Furthermore, the historical project data processing module includes a historical project data acquisition and segmentation unit, an element attribute extraction and analysis unit, and a design resource database construction unit;

[0037] The historical project data acquisition and segmentation unit acquires multiple historical project data from the home design platform, segments them according to a preset minimum element, and extracts the element set of each historical project data; the element attribute extraction and analysis unit extracts the attributes of each element and constructs an element attribute feature vector based on these attributes; the design resource database construction unit divides all elements into several categories based on the element attribute feature vector, and each category corresponds to a category label, thereby constructing a design resource database.

[0038] Furthermore, the design template construction module includes a design template creation unit, a label extraction and weight calculation unit, and a label feature vector construction unit;

[0039] The design template creation unit creates multiple design templates on the platform based on the design resource database, with each design template representing a complete home design scheme. The tag extraction and weight calculation unit extracts relevant category tags for each design template and calculates the weight coefficient of each category tag through keyword extraction and word segmentation. The tag feature vector construction unit constructs the tag feature vector of the design template based on the weight coefficient of each category tag.

[0040] Furthermore, the real-time demand matching module includes a real-time demand data processing unit, an element attribute and tag matching unit, and a matching design template unit;

[0041] The real-time requirement data processing unit acquires and analyzes real-time project requirement data, extracts the preset minimum elements and their attributes, and constructs a real-time requirement feature vector. The element attribute and label matching unit calculates the similarity between the real-time requirement element attributes and the element attribute feature vectors in the design resource database to obtain the category label of each element in the real-time requirement data, and constructs a real-time project requirement feature vector based on the category labels of the real-time requirement elements. The design template matching unit compares the real-time project requirement feature vector with the label feature vectors of all design templates, calculates the matching index, and selects the design template with the largest matching index as the reference design template.

[0042] Furthermore, the design adjustment and optimization module includes a deviation analysis and prompt generation unit and a target design scheme generation unit;

[0043] The deviation analysis and prompt generation unit analyzes the deviation between the real-time demand feature vector and the label feature vector of the reference design template, generates a deviation ranking list based on the deviation value, and outputs the deviation ranking list as prompt information to relevant personnel; the target design scheme generation unit generates the target design scheme by having relevant personnel adjust the reference design template based on the prompt information.

[0044] Compared with existing technologies, the beneficial effects of this invention are as follows: By constructing a design resource database and a design template library, design elements, furniture, accessories, etc., from historical projects can be categorized and organized. Designers do not need to start from scratch each time and can quickly retrieve and reuse existing design templates and resources. The matching process between real-time project requirement data and design templates can automatically provide designers with the most suitable reference design templates, thereby significantly reducing the preparation work in the early stages of design. In traditional design processes, designers often need to manually determine which existing furniture models or space templates best meet the needs of the current project and also consider model adjustments. In this invention, through the calculation and analysis of tag feature vectors, designers can quickly obtain design templates that match project requirements. The system automatically provides matching scores and optimization suggestions, reducing errors from manual intervention and subjective judgment. This invention, through a tag feature vector construction mechanism based on element sets, enables more accurate matching between real-time project requirements and historical design templates. Through radar chart analysis of the tag feature vectors of design templates and real-time requirement feature vectors, not only is global matching of multi-dimensional design features achieved, but also optimal references can be provided to designers based on the matching index, avoiding the limitations of relying on designers' subjective experience in traditional methods. This invention can generate prompts based on deviation values, guiding designers to make precise adjustments to the reference design template according to project requirements. This data-driven adjustment method can provide designers with specific optimization directions, ensuring that the design scheme can better meet actual needs and that the design results are more reasonable and efficient. Attached Figure Description

[0045] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used in conjunction with embodiments of the invention to explain the invention and do not constitute a limitation thereof. In the drawings:

[0046] Figure 1 This is a schematic diagram of a cloud-based home design data management system module of the present invention. Detailed Implementation

[0047] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0048] Please see Figure 1 The present invention provides the following technical solution:

[0049] A cloud-based home design data management system includes: a historical project data processing module, a design template construction module, a real-time demand matching module, and a design adjustment and optimization module;

[0050] The historical project data processing module acquires historical project data from the home design platform, divides the historical project data according to preset minimum elements, and obtains the element set corresponding to the historical project data; based on the corresponding element set, the elements corresponding to the historical project data are classified, thereby constructing a design resource database;

[0051] The design template building module is based on a design resource database. It creates several design templates on the home design platform. By analyzing the preset minimum elements that make up each design template, it obtains the tag set corresponding to the design template, and extracts keywords from the tag set to build the corresponding tag feature vector.

[0052] The real-time requirement matching module acquires and analyzes real-time project requirement data to construct a real-time project requirement feature vector. It then analyzes the real-time project requirement feature vector with the label feature vector corresponding to each design template, matches the design template corresponding to the real-time project requirement data based on the analysis results, and uses the design template corresponding to the real-time project requirement data as a reference design template.

[0053] The design adjustment and optimization module analyzes the deviation between the real-time project requirement feature vector and the corresponding label feature vector of the reference design template, and generates corresponding prompts based on the deviation. Relevant personnel then adjust the reference design template based on the prompts to generate the target design scheme.

[0054] The historical project data processing module includes a historical project data acquisition and segmentation unit, an element attribute extraction and analysis unit, and a design resource database construction unit;

[0055] The historical project data acquisition and segmentation unit acquires multiple historical project data from the home design platform, segments them according to a preset minimum element, and extracts the element set of each historical project data; the element attribute extraction and analysis unit extracts the attributes of each element and constructs an element attribute feature vector based on these attributes; the design resource database construction unit divides all elements into several categories based on the element attribute feature vector, and each category corresponds to a category label, thereby constructing a design resource database.

[0056] The design template construction module includes a design template creation unit, a label extraction and weight calculation unit, and a label feature vector construction unit;

[0057] The design template creation unit creates multiple design templates on the platform based on the design resource database, with each design template representing a complete home design scheme. The tag extraction and weight calculation unit extracts relevant category tags for each design template and calculates the weight coefficient of each category tag through keyword extraction and word segmentation. The tag feature vector construction unit constructs the tag feature vector of the design template based on the weight coefficient of each category tag.

[0058] The real-time demand matching module includes a real-time demand data processing unit, an element attribute and tag matching unit, and a matching design template unit;

[0059] The real-time requirement data processing unit acquires and analyzes real-time project requirement data, extracts the preset minimum elements and their attributes, and constructs a real-time requirement feature vector. The element attribute and label matching unit calculates the similarity between the real-time requirement element attributes and the element attribute feature vectors in the design resource database to obtain the category label of each element in the real-time requirement data, and constructs a real-time project requirement feature vector based on the category labels of the real-time requirement elements. The design template matching unit compares the real-time project requirement feature vector with the label feature vectors of all design templates, calculates the matching index, and selects the design template with the largest matching index as the reference design template.

[0060] The design adjustment and optimization module includes a deviation analysis and prompt generation unit and a target design scheme generation unit;

[0061] The deviation analysis and prompt generation unit analyzes the deviation between the real-time demand feature vector and the label feature vector of the reference design template, generates a deviation ranking list based on the deviation value, and outputs the deviation ranking list as prompt information to relevant personnel; the target design scheme generation unit generates the target design scheme by having relevant personnel adjust the reference design template based on the prompt information.

[0062] A cloud-based method for managing home design data includes the following steps:

[0063] Step S100. Obtain historical project data from the home design platform, divide the historical project data according to the preset minimum element to obtain the element set corresponding to the historical project data; classify the elements corresponding to the historical project data according to the corresponding element set to build a design resource database.

[0064] Step S200. Based on the design resource database, create several design templates on the home design platform. By analyzing the preset minimum elements that make up each design template, obtain the tag set corresponding to the design template, and extract keywords from the tag set to construct the corresponding tag feature vector.

[0065] Step S300. Obtain real-time project requirement data, analyze the real-time project requirement data, and construct a real-time project requirement feature vector; analyze the real-time project requirement feature vector and the label feature vector corresponding to each design template, match the design template corresponding to the real-time project requirement data according to the analysis results, and use the design template corresponding to the real-time project requirement data as a reference design template.

[0066] Step S400. Analyze the deviation between the real-time project requirement feature vector and the label feature vector corresponding to the reference design template, and generate corresponding prompts based on the deviation. Relevant personnel then adjust the reference design template based on the prompts to generate the target design scheme.

[0067] Step S100 includes:

[0068] S101. Obtain several historical project data from the home design platform, divide each historical project data according to a preset minimum element, and decompose each historical project data into a set of elements Ai, where Ai = {ai1, ai2, ..., ain}, where Ai represents the set of elements corresponding to the i-th historical project data, i represents the data number of the historical project data, which is a positive integer; ai1 represents the first element of the i-th historical project data, ai2 represents the second element of the i-th historical project data, and so on, ain represents the n-th element of the i-th historical project data; the preset minimum element refers to the most basic unit in the home design project that can be independently identified, analyzed, and processed; for example, sofa, bed, table, chair, lamp, wall, window, door, floor, etc.

[0069] S102. Summarize the element set corresponding to all historical project data, and extract the element attribute corresponding to each element in the element set. The element attribute refers to the feature description corresponding to the preset minimum element. Construct the corresponding element attribute feature vector V1 based on the element attribute, and the element attribute feature vector V1=[v11,v12,...,v1u], where u represents the dimension of the element attribute feature vector V1, v11 represents the first dimension feature of the element attribute feature vector V1, v12 represents the second dimension feature of the element attribute feature vector V1, and so on, v1u represents the u-th dimension feature of the element attribute feature vector V1. Calculate the similarity between all element attribute feature vectors V1, and use a clustering algorithm to divide all elements into several categories, and each category corresponds to a category label, thereby constructing a design resource database.

[0070] In this embodiment, a clustering algorithm is used to divide all elements into several categories, and each category corresponds to a category label. The process can be specified through the following steps:

[0071] To cluster based on the similarity between elements, it is necessary to calculate the similarity between all element attribute feature vectors V1. Euclidean distance, cosine similarity, or Manhattan distance can usually be used to measure the similarity between feature vectors.

[0072] After calculating the similarity of all elements, a suitable clustering algorithm can be selected to divide the elements into several categories. Common clustering algorithms include:

[0073] K-means clustering: By selecting K initial cluster centers, the cluster centers are iteratively updated until the clustering results converge. It is suitable for scenarios where the number of clusters is known.

[0074] Hierarchical clustering: It continuously merges or splits clusters by building a clustering tree (hierarchical structure) until a stopping condition is met. It is suitable when the number of clusters is uncertain.

[0075] DBSCAN (Density-based Spatial Clustering Algorithm): It divides clusters based on density and is suitable for situations where the shape of the clusters is irregular and the number of clusters is not fixed.

[0076] When choosing a clustering algorithm, the following factors need to be considered:

[0077] Data set size: If the amount of data is very large, K-means may be a good choice.

[0078] Known number of clusters: If the number of clusters to be divided is known in advance, K-means can be chosen. If it is uncertain, hierarchical clustering or DBSCAN can be considered.

[0079] Data distribution: DBSCAN may perform better if the data is densely distributed and has a clear clustering pattern.

[0080] After clustering is completed, each cluster represents a category; for each cluster, a category label can be assigned (e.g., category 1, category 2, ...). This label can be defined by the characteristics of the elements within the cluster, or it can be set according to actual needs.

[0081] For example, if the K-means algorithm is used to obtain K categories, all elements belonging to the same category will be assigned the same label; each category label can be named by the characteristics of the cluster or the cluster center, or named according to the actual use case (such as the functional modules or styles of home design).

[0082] Step S200 includes:

[0083] S201. Based on the design resource database, create several design templates on the home design platform, and each design template represents a complete home design scheme; obtain the design element set Dj corresponding to each design template, represented as: Dj={ej1,ej2,...,ejm}, where Dj represents the design element set corresponding to the j-th design template, j represents the data number of the design template; ej1 represents the first design element of the j-th design template, ej2 represents the second design element of the j-th design template, and so on, ejm represents the m-th design element of the j-th design template; the design elements in the design element set correspond to the preset minimum element in the design resource database;

[0084] S202. For each design element set Dj corresponding to a design template, obtain the category tags of the design elements in the design resource database, and summarize the category tags of all design elements to form the tag set Lj corresponding to the design element set Dj; perform word segmentation and keyword extraction on each category tag in the tag set Lj, and calculate the weight coefficient W of each category tag in the corresponding design template, and the specific calculation formula of the weight coefficient W is as follows:

[0085] W(kc)=[TF(kc,Lj) / ∑kd∈Lj,TF(kd,Lj)]×IDF(kc);

[0086] Where W(kc) represents the weight coefficient of category label kc, and c represents the data number of the category label in the design resource database; TF(kc,Lj) represents the frequency of category label kc in label set Lj; ∑kd∈Lj,TF(kd,Lj) represents the sum of the frequencies of all category labels in label set Lj; IDF(kc) is the inverse document frequency, representing the scarcity of category label kc in all design templates, and the calculation formula is:

[0087] IDF(kc) = log[N / DF(kc)];

[0088] Where N represents the total number of design templates, and DF(kc) represents the number of design templates containing the category label kc;

[0089] S203. Based on the weight coefficient of each category label kc in the label set Lj, and combined with the design resource database, construct the label feature vector V_Lj corresponding to the design template Dj, and V_Lj=[W(k1),W(k2),...,W(kp)], where W(k1) represents the weight coefficient of category label k1, W(k2) represents the weight coefficient of category label k2, and so on, W(kp) represents the weight coefficient of category label kp, p equals the total number of category labels in the design resource database, and p is greater than or equal to m; the order of the category label weight coefficients in the label feature vector V_Lj follows the order of the category labels in the design resource database. If the label set Lj does not contain a certain category label from the design resource database, then the weight coefficient of this category label is assigned to zero.

[0090] In this embodiment, it is assumed that the design resource database has three category labels: k1, k2, and k3, and the corresponding weight coefficients are W(k1), W(k2), and W(k3), respectively. If the label set Lj contains k1 and k3 but not k2, then the constructed label feature vector VLj is: VLj=[W(k1),0,W(k3)], that is, the weight coefficient corresponding to k2 is assigned to zero.

[0091] Step S300 includes:

[0092] S301. Obtain real-time project requirement data. Following the analysis method for historical project data, extract the preset minimum element and its corresponding element attributes from the real-time project requirement data to obtain the real-time requirement element set B and the real-time requirement element attribute feature vector V2. Calculate the similarity between the real-time requirement element attribute feature vector V2 corresponding to each element in the real-time requirement element set B and the element attribute feature vector V1 of each category in the design resource database, and calculate the average similarity of the corresponding categories. Select the category label with the largest average similarity as the category label of the corresponding element in the real-time requirement element set B. Summarize all element category labels in the real-time requirement element set B, and for all element category labels in the real-time requirement element set B, refer to the analysis process of the label feature vector corresponding to the design template to construct the real-time project requirement feature vector V_req, where the dimension of the real-time project requirement feature vector V_req is equal to p.

[0093] S302. Represent the real-time project requirement feature vector and the corresponding label feature vector of each design template using a radar chart. The specific analysis process is as follows:

[0094] In a two-dimensional plane, take any point as the origin of the coordinate system. Use the dimension p of the real-time project requirement feature vector V_req as the number of number axes. Each number axis starts from the origin of the coordinate system, and the angle between each number axis is equal. Mark the feature values ​​corresponding to the real-time project requirement feature vector V_req on the number axes in a clockwise order to obtain p data points. Connect the data points of adjacent number axes to form a closed polygon. Calculate the area S_req of the closed polygon.

[0095] Following the analysis method of the real-time project requirement feature vector V_req, the label feature vector V_Lj corresponding to each design template is represented on a two-dimensional plane, and the area S_Lj of the corresponding closed polygon is calculated. Based on the area S_req and S_Lj of the closed polygon, the matching index R between the real-time project requirement feature vector and the label feature vector corresponding to each design template is calculated sequentially, and the specific calculation formula is as follows:

[0096] R={(S_req·S_Lj) / [S_req+S_Lj-|S_req-S_Lj|]}·α;

[0097] Where α represents the shape similarity between the real-time project requirement feature vector and the closed polygon corresponding to the label feature vector of the design template, and the value ranges from 0 to 1; the matching index R corresponding to each design template is arranged in descending order, and the design template with the largest matching index R is selected as the reference design template.

[0098] In this embodiment, the following data is available:

[0099] Real-time project requirement feature vector: V_req=(3,5,4,6),

[0100] Design template label feature vector 1: V_L1=(4,5,4,5),

[0101] Design template label feature vector 2: V_L2=(6,5,3,7),

[0102] Design template label feature vector 3: V_L3=(2,6,5,6);

[0103] Calculate the area of ​​a closed polygon

[0104] Suppose the following area is obtained by calculating using the cross product method (example value here):

[0105] S_req=16 (area of ​​the polygon corresponding to the feature vector of real-time project requirements)

[0106] S_L1=15 (Area of ​​the label feature vector of design template 1),

[0107] S_L2=18 (Area of ​​label feature vectors in design template 2)

[0108] S_L3=17 (Area of ​​the label feature vector of design template 3);

[0109] The matching index R is calculated using the formula R={(S_req·S_Lj) / [S_req+S_Lj-|S_req-S_Lj|]}·α. Assuming the calculated matching indices R are ordered from largest to smallest as R1, R2, and R3, then design template 1 is selected as the reference design template.

[0110] Step S400 includes:

[0111] The difference between the feature values ​​of the real-time project requirement feature vector V_req and the label feature vector V_Lj corresponding to the reference design template in each dimension is calculated, and the absolute value of the difference calculation result is obtained to obtain the corresponding deviation value Δf. The deviation values ​​Δf corresponding to each dimension are sorted in descending order to generate a deviation sorting list. The deviation sorting list is output as prompt information to relevant personnel, who then adjust the reference design template based on the prompt information. The adjusted reference design template is then used as the target design scheme.

[0112] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0113] Finally, it should be noted that the above descriptions are merely preferred embodiments of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A cloud platform-based home design data management method, characterized in that: The method comprises the following steps: Step S100. Obtain historical project data of a home design platform, divide the historical project data according to preset minimum elements, thereby obtaining an element set corresponding to the historical project data; according to the corresponding element set, use a clustering algorithm to classify all elements, assign a category label to each category, thereby constructing a design resource database; Step S200. Based on the design resource database, create a plurality of design templates on the home design platform, each design template comprising a plurality of design elements, the design elements corresponding to the preset minimum elements in the design resource database; obtain a design element set corresponding to each design template, according to the category label of the design element in the design resource database, the label set corresponding to each design template is obtained by summarizing; perform keyword extraction and word segmentation processing on the label set, calculate the weight coefficient of each category label, and construct the corresponding label feature vector according to the weight coefficient; Step S300. Obtain real-time project demand data, extract the preset minimum elements and element attributes in the real-time project demand data, and construct a real-time demand element attribute feature vector; calculate the similarity between the real-time demand element attribute feature vector and the element attribute feature vector of each category in the design resource database, to obtain the category label corresponding to each real-time demand element; summarize the category labels of all real-time demand elements, and refer to the method of constructing the label feature vector in step S200 to construct a real-time project demand feature vector; analyze the real-time project demand feature vector and the label feature vector corresponding to each design template, and match the design template corresponding to the real-time project demand data according to the analysis result, and take the design template corresponding to the real-time project demand data as a reference design template; Step S400. Analyze the deviation value between the real-time project demand feature vector and the label feature vector corresponding to the reference design template, and generate corresponding prompt information based on the deviation value, and adjust the reference design template based on the prompt information, thereby generating a target design scheme. 2.The cloud platform-based home design data management method of claim 1, wherein: The step S100 comprises: S101. Obtain a plurality of historical project data from the home design platform, divide each historical project data according to the preset minimum elements, and each historical project data is decomposed into a set of element sets Ai, and Ai={ai1, ai2,..., ain}, wherein Ai represents the element set corresponding to the i-th historical project data, i represents the data number of the historical project data, and takes a positive integer; ai1 represents the first element of the i-th historical project data, ai2 represents the second element of the i-th historical project data, and so on, and ain represents the n-th element of the i-th historical project data; the preset minimum element refers to the most basic unit that can be independently recognized, analyzed and processed in the home design project; S102. Aggregate the element set corresponding to all historical project data, and extract the element attribute corresponding to each element in the element set, wherein the element attribute refers to the characteristic description corresponding to a preset minimum element; construct the corresponding element attribute feature vector V1 based on the element attribute, and the element attribute feature vector V1=[v11, v12,..., v1u], wherein u represents the dimension of the element attribute feature vector V1, v11 represents the first dimension feature of the element attribute feature vector V1, v12 represents the second dimension feature of the element attribute feature vector V1, and so on, and v1u represents the u-th dimension feature of the element attribute feature vector V1; calculate the similarity between all element attribute feature vectors V1, and divide all elements into several categories by using a clustering algorithm, and each category corresponds to a category label, thereby constructing a design resource database. 3.The cloud platform-based home design data management method of claim 2, wherein: The step S200 comprises: S201. Based on the design resource database, create several design templates on the home design platform, and each design template represents a complete home design scheme; obtain the design element set Dj corresponding to each design template, which is represented as: Dj={ej1, ej2,..., ejm}, wherein Dj represents the design element set corresponding to the j-th design template, j represents the data number of the design template; ej1 represents the first design element of the j-th design template, ej2 represents the second design element of the j-th design template, and so on, and ejm represents the m-th design element of the j-th design template; the design elements in the design element set correspond to the preset minimum elements in the design resource database; S202. For the design element set Dj corresponding to each design template, obtain the category label of the design element in the design resource database, and aggregate the category labels of all design elements to form the label set Lj corresponding to the design element set Dj; perform word segmentation processing and keyword extraction on each category label in the label set Lj, and calculate the weight coefficient W of each category label in the corresponding design template, and the specific calculation formula of the weight coefficient W is: W(kc)=[TF(kc, Lj) / ∑kd∈Lj, TF(kd, Lj)]xIDF(kc); Wherein, W(kc) represents the weight coefficient of the category label kc, c represents the data number of the category label in the design resource database; TF(kc, Lj) represents the frequency of the category label kc in the label set Lj; ∑kd∈Lj, TF(kd, Lj) represents the total sum of the frequencies of all category labels in the label set Lj; IDF(kc) is the inverse document frequency, which represents the scarcity of the category label kc in all design templates, and the calculation formula is: IDF(kc)=log[N / DF(kc)]; Wherein, N represents the total number of design templates, and DF(kc) represents the number of design templates containing the category label kc; S203. According to the weight coefficient of each category label kc in the label set Lj, the design resource database is combined to construct the label feature vector V_Lj corresponding to the design template Dj, and V_Lj = [W(k1), W(k2), …, W(kp)], wherein W(k1) represents the weight coefficient of the category label k1, W(k2) represents the weight coefficient of the category label k2, and so on, W(kp) represents the weight coefficient of the category label kp, p is equal to the total number of category labels in the design resource database, and p is greater than or equal to m; the order of the category label weight coefficients in the label feature vector V_Lj is according to the order of the category labels in the design resource database, and if the label set Lj does not contain a certain category label in the design resource database, the weight coefficient of this category label is assigned as zero.

4. The cloud platform-based home design data management method of claim 3, wherein: The step S300 comprises: S301. Obtain real-time project demand data, extract the preset minimum element and the element attribute corresponding to the preset minimum element from the real-time project demand data according to the analysis method of the historical project data, so as to obtain the real-time demand element set B corresponding to the real-time project demand data and the real-time demand element attribute feature vector V2; perform similarity calculation on the real-time demand element attribute feature vector V2 corresponding to each element in the real-time demand element set B and the element attribute feature vector V1 of each category in the design resource database, and calculate the average value of the similarity of the corresponding category, select the category label with the maximum average value of the similarity as the category label of the corresponding element in the real-time demand element set B; aggregate all element category labels in the real-time demand element set B, and refer to the analysis process of the label feature vector corresponding to the design template for all element category labels in the real-time demand element set B, so as to construct the real-time project demand feature vector V_req, and the dimension of the real-time project demand feature vector V_req is equal to p; S302. Perform radar chart representation on the real-time project demand feature vector and the label feature vector corresponding to each design template, and the specific analysis process is as follows: Take any point in the two-dimensional plane as the coordinate origin, take the dimension p of the real-time project demand feature vector V_req as the number of number axes, and each number axis takes the coordinate origin as the starting point, and the included angle between each number axis is equal; mark the feature values corresponding to the real-time project demand feature vector V_req in sequence on the number axes in clockwise order, so as to obtain p data points, and connect the data points of adjacent number axes in sequence, so as to form a closed polygon, and calculate the area S_req of the closed polygon; According to the analysis method of the real-time project demand feature vector V_req, the label feature vector V_Lj corresponding to each design template is represented on the two-dimensional plane, the area S_Lj of the corresponding closed polygon is calculated, and the matching index R corresponding to the real-time project demand feature vector and the label feature vector corresponding to each design template is calculated in sequence according to the areas S_req and S_Lj of the closed polygons, and the specific calculation formula is: R={(S_req·S_Lj) / [S_req+S_Lj-|S_req-S_Lj|]}·α; Wherein, α represents the shape similarity of the closed polygon corresponding to the label feature vector of the real-time project demand feature vector and the design template, and the value is 0 to 1; arrange the matching index R corresponding to each design template in descending order, and select the design template with the largest matching index R as the reference design template.

5. The cloud platform-based home design data management method of claim 4, wherein: The step S400 comprises: Difference calculation is performed on the feature values of the real-time project demand feature vector V_req and the label feature vector V_Lj corresponding to the reference design template in each dimension, and the absolute value of the difference calculation result is calculated to obtain the corresponding deviation value Δf; the deviation value Δf corresponding to each dimension is sorted in descending order, and a deviation sorting list is generated, and the deviation sorting list is output to the relevant personnel as prompt information, and the reference design template is adjusted based on the prompt information by the relevant personnel, and the adjusted reference design template is used as the target design scheme.

6. A cloud platform-based home design data management system applied to the cloud platform-based home design data management method of any one of claims 1-5, characterized in that: The system comprises: a historical project data processing module, a design template construction module, a real-time demand matching module, and a design adjustment and optimization module; The historical project data processing module obtains historical project data of a home design platform, divides the historical project data according to a preset minimum element, thereby obtaining an element set corresponding to the historical project data; according to the corresponding element set, the elements corresponding to the historical project data are classified, thereby constructing a design resource database; The design template construction module creates a plurality of design templates on the home design platform based on the design resource database, obtains a label set corresponding to the design template by analyzing the preset minimum elements constituting each design template, and extracts keywords from the label set, thereby constructing a corresponding label feature vector; The real-time demand matching module obtains real-time project demand data, analyzes the real-time project demand data, thereby constructing a real-time project demand feature vector; the real-time project demand feature vector and the label feature vector corresponding to each design template are analyzed, and the design template corresponding to the real-time project demand data is matched according to the analysis result, and the design template corresponding to the real-time project demand data is used as the reference design template; The design adjustment and optimization module analyzes the deviation value between the real-time project demand feature vector and the label feature vector corresponding to the reference design template, and generates corresponding prompt information based on the deviation value, and the reference design template is adjusted by the relevant personnel based on the prompt information, thereby generating a target design scheme.

7. The cloud platform-based home design data management system of claim 6, wherein: The historical project data processing module comprises a historical project data acquisition and division unit, an element attribute extraction and analysis unit, and a design resource database construction unit; The historical project data processing module comprises a historical project data acquisition and division unit, an element attribute extraction and analysis unit, and a design resource database construction unit; The historical project data acquisition and division unit acquires a plurality of historical project data from a home design platform, divides the historical project data according to preset minimum elements, and extracts an element set of each historical project data; the element attribute extraction and analysis unit extracts attributes of each element and constructs an element attribute feature vector based on the attributes; and the design resource database construction unit divides all elements into a plurality of categories based on the element attribute feature vector, each category corresponds to a category label, and a design resource database is constructed.

8. The cloud platform-based home design data management system of claim 6, wherein: The design template construction module includes a design template creation unit, a label extraction and weight calculation unit, and a label feature vector construction unit. The design template creation unit creates a plurality of design templates on the platform based on the design resource database, each design template representing a complete home design scheme; the label extraction and weight calculation unit extracts category labels related to each design template and calculates weight coefficients of each category label through keyword extraction and word segmentation processing; and the label feature vector construction unit constructs a label feature vector of the design template according to the weight coefficients of each category label. 9.The cloud platform-based home design data management system of claim 6, wherein: The real-time demand matching module includes a real-time demand data processing unit, an element attribute and label matching unit, and a matching design template unit. The real-time demand data processing unit acquires and analyzes real-time project demand data, extracts preset minimum elements and their attributes, and constructs a real-time demand feature vector; the element attribute and label matching unit obtains the category label of each element in the real-time demand data by calculating the similarity between the real-time demand element attributes and the element attribute feature vectors in the design resource database, and constructs a real-time project demand feature vector based on the category label of the real-time demand element; and the matching design template unit compares the real-time project demand feature vector with the label feature vectors of all design templates, calculates a matching index, and selects the design template with the largest matching index as a reference design template.

10. The cloud platform-based home design data management system of claim 6, wherein: The design adjustment and optimization module includes a deviation analysis and prompt generation unit and a target design scheme generation unit. The deviation analysis and prompt generation unit analyzes the deviation between the real-time demand feature vector and the label feature vector of the reference design template, generates a deviation sorting list based on the deviation value, and outputs the deviation sorting list as prompt information to relevant personnel; and the target design scheme generation unit generates a target design scheme by adjusting the reference design template based on the prompt information.

Images