Cartoon design material library management method based on big data

Abstract

The invention discloses an animation design material library management method based on big data, and relates to the technical field of animation design, and the method comprises the following steps: obtaining all materials in a material library and attribute information of the materials, the attribute information specifically comprising resolution, format, style and use scene; classifying all the materials in the material library according to the attribute information of the materials, and outputting a classification result of the materials in the material library; acquiring historical user behavior data of the material library, wherein the historical user behavior data specifically comprises search history, download records, use scenes and recompilation behaviors; according to the steps 1-3, the problems that an existing material library is not clear in classification and insufficient in personalized recommendation are solved, efficient retrieval and personalized recommendation of materials are achieved through refined classification and labeling processing in combination with precise calculation of the user preference degree, the screening cost of a designer is reduced in the process, and the screening efficiency of the designer is improved. And the convenience and the use efficiency of material searching are obviously improved.

Description

Technical Field

[0001] The present invention relates to the technical field of animation design, and more specifically, to a method for managing an animation design material library based on big data. Background Art

[0002] With the booming development of the animation industry, animation design has become an important creative field. Animation design forms a unique visual art creation mode through the combination of comics and animations with storylines, using expression techniques such as two-dimensional plane, three-dimensional animation, and animation special effects. An animation design material library is a resource library specifically used for storing, managing, and providing animation design-related materials. Its main function is to provide rich material support for animation designers, production teams, and related practitioners to improve creative efficiency, stimulate creativity, and ensure the quality and diversity of design works. However, in the actual use process, designers need to quickly find suitable materials according to project requirements and collaborate with other team members. Although some systems attempt to push materials through user behavior analysis, existing algorithms are mostly based on simple download or usage frequency, making it difficult to accurately match the creative needs of designers. There is a lack of an effective personalized recommendation mechanism. In multi-person collaboration projects, designers need to share and synchronize the material usage situation in real time, but existing systems lack an efficient collaborative design support function, resulting in low design efficiency. In addition, animation design involves various styles and application scenarios (such as two-dimensional, three-dimensional, special effects, etc.), and existing material libraries fail to provide flexible material adaptation and classification according to these scenarios, increasing the screening cost for designers and further affecting the creative efficiency and quality. With the rapid development of the animation industry, the material library needs to continuously update and integrate new resources to meet the needs of designers, but existing systems face many challenges in this regard. First, the update frequency of some material libraries is low, and they cannot introduce the latest design trends and technological achievements in a timely manner, making it difficult for designers to obtain cutting-edge materials. Second, materials from different sources have differences in formats, resolutions, styles, etc., and existing systems lack an effective resource integration mechanism, making it difficult to achieve seamless docking and unified management of materials. Summary of the Invention

[0003] To solve the above problems, the present invention provides a method for managing an animation design material library based on big data.

[0004] The present invention provides a method for managing an animation design material library based on big data, including the following steps: Step 1: Obtain all materials in the material library and the attribute information of the materials. The attribute information specifically includes resolution, format, style, and usage scenario. Classify all the materials in the material library according to the attribute information of the materials, and output the classification results of the materials in the material library; Step 2: Obtain the historical user behavior data of the material library, specifically including search history, download records, usage scenarios, and adaptation behaviors; Receive the classification results of the materials in the material library, calculate the preference degree of each historical user for each type of material in the material library according to the historical user behavior data, and transmit it; Step 3: Receive the preference degree of each historical user for each type of material in the material library. According to the preference degree of the user for each type of material in the material library, calculate the matching degree of each user with the materials in the material library. According to the matching degree, generate a personalized material recommendation list for each user; Step 4: Every other period, record the project requirement information of the user and the usage information of the existing materials in the material library. According to the project requirement information of the user and the usage information of the existing materials in the material library, calculate the update priority of the materials in the material library, and update the materials in the material library according to the update priority of the materials in the material library.

[0005] Preferably, the specific working steps of the above Step 1 are as follows: Extract the attribute information of the fine-tuning materials from the material library, including resolution, format, style, and usage scenario, and store it in a structured data format to form a material attribute data table. The attribute information in different formats is unified into a standard format; According to the attribute information of the materials, allocate the materials to different categories and add corresponding tags. The classification rules are as follows: Classify by style: 2D, 3D, special effects, Japanese anime, American anime, Chinese anime. Classify by usage scenario: character, background, special effects, props. The specific classification types are 24.

[0006] Add tags to each material. The tag content includes style, usage scenario, resolution, and format in sequence. Use the tags of each material as the classification results of the materials in the material library and output them.

[0007] Preferably, the specific working steps of the above Step 2 are as follows: Obtain the user historical behavior data of the material library, including search history, download records, usage scenarios, and adaptation behaviors; Extract the user behavior data from the database, record the user ID, behavior type, behavior object, and behavior time, and clean the behavior data to remove invalid or duplicate records; Weights are assigned to each behavior according to the importance of different behavior types. The weights for search history are 0.2, download records are 0.5, usage scenario weights are 0.3, and adaptation behavior weights are 0.6. Weights are assigned to each user behavior record according to the behavior type to obtain weighted user behavior data; Calculate the preference degree value of each user for each type of material , and use the preference degree value to measure the preference degree of each historical user for each type of material in the material library, and transmit the preference degree value of each user for each type of material .

[0008] Preferably, the second step is also used to consider the time sensitivity of user behavior, and calculate the updated updated behavior weight. The specific steps are as follows: First, define the time decay coefficient formula, specifically , where W is the decay coefficient, is the decay rate, with a value of 0.1, is the difference between the time when the user's historical behavior was sent and the current time; Multiply the weight assigned to each behavior by the corresponding time decay coefficient of each behavior record to obtain the updated updated behavior weight.

[0009] Preferably, the calculation of the preference degree value of each user for each type of material is calculated as follows: According to the formula Calculate and obtain the preference degree value of each user for each type of material , where represents the user, represents the material type, represents the updated updated behavior weight of user P for material type T.

[0010] Preferably, the specific working method of the third step is as follows: Receive the preference degree value of each user for each type of material , and organize the preference degree value along with the user ID and material type into a structured table; For each piece of material, extract the feature weight according to its classification result; For each user, calculate the matching degree H between the user and each piece of material according to their preference weights and the feature weights of the material; For each user, sort the materials according to the matching degree H, select the top N materials with the highest matching degree as the recommendation results, remove the materials that the user has already used from the recommendation results to avoid repeated recommendations, and finally generate a personalized recommendation list, which includes the user ID and the list of recommended material IDs.

[0011] Preferably, the specific working steps for extracting the feature weight according to the classification result for each material are as follows: According to the classification result of the material, assign a feature weight to each material, define the weight value of each type of feature, and extract the corresponding weight according to the classification information of the material to obtain the feature weight of each type of material. 。

[0012] Preferably, the specific working steps for calculating the matching degree H between each user and each material according to the user's preference weight and the material's feature weight further include: According to the formula , calculate and obtain the matching degree H between each user and each material.

[0013] Preferably, the specific solution for step four is as follows: Extract information from the project requirements submitted by the user, including the project type, the required material style, and the usage scenario, and extract the usage information of each material from the material library, including the download times, usage frequency, and user evaluation; Obtain the importance value G of the material in the current project requirements and the popularity value V of the material in actual use; According to the formula , calculate and obtain the comprehensive priority Z of each material, sort the materials according to the comprehensive priority, and preferentially update the material with the highest comprehensive priority.

[0014] Preferably, the specific calculation steps for the importance value G of the material in the current project requirements and the popularity value V of the material in actual use are as follows: For the importance value G, first, assign weights to the project requirements according to the project type, style, and usage scenario, then calculate the matching degree of the material according to the matching degree between the classification information of the material and the project requirements. If the classification information of the material matches the project requirements, get one point, and if it does not match, get zero points. Finally, multiply the project requirement weight by the material matching degree and sum to obtain the importance value G of each material; For the popularity value V, first assign weight coefficients to the download times, usage frequency, and user evaluation. Secondly, normalize these data to the range of 0 to 1. Finally, multiply the normalized download times, usage frequency, and user evaluation by the corresponding weight coefficients and sum to obtain the popularity value V of each material.

[0015] Beneficial effects: According to Steps 1 to 3, the problems of unclear classification and insufficient personalized recommendation in the existing material library are solved. Through refined classification and tagging, combined with the accurate calculation of user preference levels, efficient retrieval and personalized recommendation of materials are achieved. This process not only reduces the screening cost of designers, but also significantly improves the convenience and usage efficiency of material search. According to Step 4, the problems of untimely update of the material library and lag in real-time monitoring are solved. Brief Description of the Drawings

[0016] Figure 1 is the flowchart of the method of the present invention. Detailed Embodiments

[0017] Application scenarios: In the actual usage process, designers need to quickly find suitable materials according to project requirements and collaborate with other team members. Although some systems try to push materials through user behavior analysis, existing algorithms are mostly based on simple download or usage frequency, making it difficult to accurately match the creative needs of designers and lacking an effective personalized recommendation mechanism. In multi-person collaborative projects, designers need to share and synchronize the usage of materials in real time, but existing systems lack efficient collaborative design support functions, resulting in low design efficiency. In addition, anime design involves multiple styles and application scenarios (such as 2D, 3D, special effects, etc.), and the existing material library fails to provide flexible material adaptation and classification according to these scenarios, increasing the screening cost of designers and further affecting the creation efficiency and quality.

[0018] As Figure 1 shown: A method for managing an anime design material library based on big data includes the following steps: Step 1: Obtain all materials in the material library and the attribute information of the materials. The attribute information specifically includes resolution, format, style, and usage scenario; Classify all materials in the material library according to the attribute information of the materials, and output the classification results of the materials in the material library. It should be noted that to improve the material retrieval efficiency, first collect and organize all materials in the material library and their attribute information (including resolution, format, style, usage scenario, etc.). Through refined classification and tagging, the materials are classified according to style (such as 2D, 3D, special effects, Japanese anime, American anime, Chinese anime, etc.) and application scenario, and the classified material library is output. This process effectively solves the problems of unclear material classification and poor adaptability, significantly reduces the screening cost of designers, and improves the convenience of material search; Step 2: Obtain the historical user behavior data of the material library, specifically including search history, download records, usage scenarios, and adaptation behaviors; Receive the classification results of the materials in the material library, calculate the preference degree of each historical user for each type of material in the material library according to the historical user behavior data, and transmit it; it should be noted that based on the historical user behavior data of the material library (such as search history, download records, usage scenarios, adaptation behaviors, etc.), further analyze the user's preferences, and by calculating the preference degree of the user for different classified materials, provide accurate data support for subsequent personalized recommendations. This step solves the problem of single-dimensional analysis of user behavior in the existing system, can capture the actual needs and usage habits of users more comprehensively, and lays a foundation for the personalized recommendation mechanism. Step 3: Receive the preference degree of each historical user for each type of material in the material library, calculate the matching degree between each user and the materials in the material library according to the preference degree of the user for each type of material in the material library, and generate a personalized material recommendation list for each user according to the matching degree; it should be noted that by combining the preference degree of the user for different classified materials, calculate the matching degree between each user and the materials in the material library, and generate a personalized material recommendation list accordingly. By accurately matching the creative needs of designers, the usage efficiency of the materials is improved, the problem of lack of personalized recommendation mechanism in the existing system is solved, and the user experience and creation efficiency are further improved; Step 4: Every other period, record the project requirement information of the user and the usage information of the existing materials in the material library, calculate the update priority of the materials in the material library according to the project requirement information of the user and the usage information of the existing materials in the material library, and update the materials in the material library according to the update priority of the materials in the material library; it should be noted that to ensure the timeliness and diversity of the content of the material library, regularly analyze the project requirement information of the user and the usage of the existing materials in the material library, calculate the update priority of the materials, and dynamically update and resource integrate the material library according to the priority, introduce the latest design trends and technical achievements, solve the problems of untimely update of the material library and insufficient resource integration, and meet the needs of designers for cutting-edge materials; It should also be noted that according to Steps 1 to 3, the problems of unclear classification and insufficient personalized recommendation in the existing material library are solved. Through refined classification and tagging processing, combined with the accurate calculation of the user preference degree, efficient retrieval and personalized recommendation of materials are realized. This process not only reduces the screening cost of designers, but also significantly improves the convenience and usage efficiency of material search. According to Step 4, the problems of untimely update of the material library and lag in real-time monitoring are solved.

[0019] As an optional embodiment: The specific working steps of the said Step 1 are as follows: Extract the attribute information of unadjusted materials from the material library, including resolution, format, style, and usage scenarios, and store it in a structured data format to form a material attribute data table. The attribute information in different formats is unified into a standard format. It should be noted that scripts or programs (such as Python scripts) are used to traverse all material files in the material library and read the metadata of the files (such as file format, resolution). For unstructured information such as style and usage scenarios, extraction can be carried out through predefined tags or metadata fields. If there is no clear metadata in the material file, preliminary classification can be performed through manual annotation or the use of image recognition technologies (such as deep learning models). For attribute information that cannot be automatically extracted (such as style and usage scenarios), professional personnel can be organized to manually annotate the materials to ensure the accuracy of the information. It should also be noted that the standardization steps are to unify the resolution into a standard pixel format and the file format into a common extension name. According to the attribute information of the materials, the materials are assigned to different categories and corresponding tags are added. The classification rules are as follows: Classify according to style: 2D, 3D, special effects, Japanese comics, American comics, Chinese comics, and classify according to usage scenarios: characters, backgrounds, special effects, props. The specific classification types are 24. Comprehensive classification will be carried out according to style classification: 2D, 3D, special effects, Japanese comics, American comics, Chinese comics, and usage scenario classification: characters, backgrounds, special effects, props. Specifically, it includes four types of 2D characters, backgrounds, special effects, and props, four types of 3D characters, backgrounds, special effects, and props, and so on in sequence. Add tags to each material. The tag content includes style, usage scenario, resolution, and format in sequence. The tags of each material are used as the classification result of the materials in the material library and output.

[0020] As an optional embodiment: The specific working steps of the second step are as follows: Obtain the user's historical behavior data of the material library, including search history, download records, usage scenarios, and adaptation behaviors. Extract the user behavior data from the database, record the user ID, behavior type, behavior object, and behavior time, and clean the behavior data to remove invalid or duplicate records. Weigh each behavior according to the importance of different behavior types. Search history weight: 0.2, download record: 0.5, usage scenario weight: 0.3, adaptation behavior weight: 0.6. Assign weights to each user behavior record according to the behavior type to obtain weighted user behavior data. It should be noted that the core purpose of behavior weight assignment is to assign reasonable weights to each behavior according to the importance of different behavior types, so as to more accurately reflect the user's interest and preference for different material types. By assigning weights to each user behavior record, a weighted user behavior data table is generated. These weighted data are the basis for calculating the user's preference degree for different material types; Calculate the preference degree value of each user for each material type , and use the preference degree value to measure the preference degree of each historical user for each type of material in the material library, and transmit the preference degree value of each user for each material type .

[0021] As an optional embodiment: The second step is also used to consider the time sensitivity of user behavior and calculate the updated behavior weight. The specific steps are as follows: First, define the time decay coefficient formula, specifically , where W is the decay coefficient, is the decay rate, with a value of 0.1, is the difference between the time when the user's historical behavior was sent and the current time; Multiply the weight assigned to each behavior by the corresponding time decay coefficient of each behavior record to obtain the updated behavior weight.

[0022] 5. As an optional embodiment: The calculation of the preference degree value of each user for each material type is calculated as follows: According to the formula calculate and obtain the preference degree value of each user for each material type , where represents the user, represents the material type, represents the updated behavior weight of user P for material type T.

[0023] As an optional embodiment: The specific working mode of the third step is as follows: Receive the preference degree value of each user for each type of material , and organize the preference degree value together with the user id and the material type into a structured table; For each piece of material, extract the feature weight according to its classification result; For each user, calculate the matching degree H between the user and each piece of material according to the user's preference weight and the characteristic weight of the material; For each user, sort the materials according to the matching degree H, select the top N materials with the highest matching degree as the recommended results, remove the materials that the user has already used from the recommended results to avoid repeated recommendations, and finally generate a personalized recommendation list, which includes the user ID and the list of recommended material IDs.

[0024] As an optional embodiment: The specific working steps for extracting the characteristic weight according to the classification result for each piece of material are as follows: According to the classification result of the material, assign characteristic weights to each piece of material, define the weight values of each type of characteristic, and extract the corresponding weights according to the classification information of the material to obtain the characteristic weights of each type of material 。

[0025] As an optional embodiment: The specific working steps for calculating the matching degree H between each user and each piece of material according to the user's preference weight and the characteristic weight of the material further include: According to the formula , calculate and obtain the matching degree H between each user and each piece of material. It should be noted that by combining the user's preference weights for different material types and the characteristic weights of the materials, the matching degree between the user and each piece of material is calculated by means of weighted summation. The matching degree reflects the user's interest in the material and provides a quantitative basis for personalized recommendation. Finally, output the matching degree matrix between the user and the material to support the generation of the personalized recommendation list.

[0026] As an optional embodiment: The specific solution for step four is as follows: Extract information from the project requirements submitted by the user, specifically including the project type, the required material style, and the usage scenario, and extract the usage information of each piece of material from the material library, including the download times, usage frequency, and user evaluations; Obtain the importance value G of the material in the current project requirements and the popularity value V of the material in actual use; According to the formula , calculate and obtain the comprehensive priority Z of each piece of material, sort the materials according to the comprehensive priority, and preferentially update the material with the highest comprehensive priority.

[0027] As an optional embodiment: In a method for managing an animation design material library based on big data as described in claim 9, the specific calculation steps for the importance value G of the material in the current project requirements and the popularity value V of the material in actual use are as follows: For the importance value G, first, weights are assigned to project requirements according to the project type, style, and usage scenario. Then, the matching degree of the material is calculated based on the matching degree between the classification information of the material and the project requirements. If the classification information of the material matches the project requirements, 1 point is obtained; if not, 0 points are obtained. Finally, the product of the project requirement weight and the material matching degree is multiplied and summed to obtain the importance value G of each material. The higher the importance value G, the higher the requirement priority, indicating the higher the importance of the material in the current project requirements. It should also be noted that factors such as project type, style, and usage scenario are used to assign weights to project requirements. The specific weights are set by project personnel. In this embodiment, the initial value is one-third. For the popularity value V, first, weight coefficients are assigned to the download times, usage frequency, and user evaluation. Secondly, these data are normalized to the range of 0 to 1. Finally, the normalized download times, usage frequency, and user evaluation are multiplied by the corresponding weight coefficients and summed to obtain the popularity value V of each material. The higher the popularity value V, the more popular the material is in actual use and the higher the value to users. It should be noted that the weight coefficients of the download times, usage frequency, and user evaluation are preset values and are one-third in this embodiment.

[0028] Working principle: Step 1: Obtain all materials and their attribute information in the material library. The attribute information specifically includes resolution, format, style, and usage scenario. According to the attribute information of the materials, all materials in the material library are classified, and the classification results of the materials in the material library are output. It should be noted that to improve the material retrieval efficiency, first, all materials and their attribute information (including resolution, format, style, usage scenario, etc.) in the material library are collected and sorted. Through refined classification and tagging, the materials are classified by style (such as 2D, 3D, special effects, Japanese comics, American comics, Chinese comics, etc.) and application scenarios, and the classified material library is output. This process effectively solves the problems of unclear material classification and poor adaptability, significantly reduces the screening cost of designers, and improves the convenience of material search. Step 2: Obtain the historical user behavior data of the material library, specifically including search history, download records, usage scenarios, and adaptation behaviors. Receive the classification results of the materials in the material library, calculate the preference degree of each historical user for each type of material in the material library according to the historical user behavior data, and transmit it; it should be noted that based on the historical user behavior data of the material library (such as search history, download records, usage scenarios, adaptation behaviors, etc.), further analyze the user's preferences, and by calculating the preference degree of the user for different classified materials, provide accurate data support for subsequent personalized recommendations. This step solves the problem of the single dimension of user behavior analysis in the existing system, can capture the actual needs and usage habits of users more comprehensively, and lays a foundation for the personalized recommendation mechanism. Step 3: Receive the preference degree of each historical user for each type of material in the material library, calculate the matching degree between each user and the materials in the material library according to the preference degree of the user for each type of material in the material library, and generate a personalized material recommendation list for each user according to the matching degree; it should be noted that by combining the preference degree of the user for different classified materials, calculate the matching degree between each user and the materials in the material library, and generate a personalized material recommendation list accordingly. By accurately matching the creative needs of designers, the usage efficiency of materials is improved, the problem of the lack of a personalized recommendation mechanism in the existing system is solved, and the user experience and creative efficiency are further improved; Step 4: Every other period, record the project requirement information of the user and the usage information of the existing materials in the material library, calculate the update priority of the materials in the material library according to the project requirement information of the user and the usage information of the existing materials in the material library, and update the materials in the material library according to the update priority of the materials in the material library; it should be noted that to ensure the timeliness and diversity of the content of the material library, regularly analyze the project requirement information of the user and the usage of the existing materials in the material library, calculate the update priority of the materials, and dynamically update and resource integrate the material library according to the priority, introduce the latest design trends and technical achievements, solve the problems of untimely update and insufficient resource integration of the material library, and meet the needs of designers for cutting-edge materials; It should also be noted that according to Steps 1 to 3, the problems of unclear classification and insufficient personalized recommendation in the existing material library are solved. Through refined classification and tagging processing, combined with the accurate calculation of the user preference degree, efficient retrieval and personalized recommendation of materials are realized. This process not only reduces the screening cost of designers, but also significantly improves the convenience and usage efficiency of material search; According to Steps 4 and 5, the problems of untimely update and lagging real-time monitoring of the material library are solved. Through the dynamic update mechanism and real-time behavior monitoring, the timeliness of the content of the material library is ensured, and at the same time, abnormal behaviors are effectively identified and warned, protecting the legitimate rights and interests of the copyright holders and improving the overall efficiency and security of the material library management.

[0029] The above are only the preferred embodiments of the present invention. The protection scope of the present invention is not limited to the above embodiments. Any technical solutions falling within the concept of the present invention belong to the protection scope of the present invention. It should be noted that for those of ordinary skill in the art, without departing from the principle of the present invention, several improvements and refinements should also be regarded as within the protection scope of this template.

Claims

1. A method for managing an animation design material library based on big data, characterized in that: The following steps are involved: Step 1: Get all the materials in the material library and their attribute information, including resolution, format, style and usage scenario; Classify all materials in the material library according to the material attribute information, and output the classification results of the materials in the material library; Step 2: Obtain historical user behavior data of the material library, including search history, download records, usage scenarios, and adaptation behaviors; Receive the classification results of the materials in the material library, calculate the preference of each historical user for each type of material in the material library based on the historical user behavior data, and transmit it; Step 3: Receive the preference of each historical user for each type of material in the material library, calculate the matching degree between each user and the material in the material library according to the preference degree of each user for each type of material in the material library, and generate a personalized material recommendation list for each user according to the matching degree; Step 4: Every once in a while, record the user's project requirement information and the usage information of the existing materials in the material library. Based on the user's project requirement information and the usage information of the existing materials in the material library, calculate the priority of updating the materials in the material library, and update the materials in the material library according to the priority of updating the materials in the material library.

2. A method for managing an animation design material library based on big data according to claim 1, characterized in that: The specific working steps of step one are as follows: Extract the attribute information of the fine-tuned material from the material library, including resolution, format, style and usage scenario, and store it in a structured data format to form a material attribute data table. The attribute information in different formats is unified into a standard format. According to the attribute information of the material, the material is assigned to different categories and corresponding tags are added. The classification rules are as follows: According to the style classification: 2D, 3D, special effects, Japanese comics, American comics, Chinese comics; according to the usage scenario classification: characters, backgrounds, special effects, props, there are 24 specific classification types; Add a label to each piece of material. The label content includes style, usage scenario, resolution, and format in order. The label of each piece of material is used as the classification result of the material in the material library and output.

3. The method for managing an animation design material library based on big data according to claim 1, characterized in that: The specific working steps of step 2 are as follows: Obtain the user's historical behavior data of the material library, including search history, download records, usage scenarios and adaptation behavior; Extract user behavior data from the database, record user ID, behavior type, behavior object, and behavior time, clean the behavior data, and remove invalid or duplicate records; According to the importance of different types of behaviors, weights are assigned to each behavior. Search history weight: 0.2, download history: 0.5, usage scenario weight: 0.3 Adaptation behavior weight: 0.6, assign a weight to each user behavior record according to the behavior type to obtain weighted user behavior data; Calculate each user's preference for each material type , the preference value Used to measure each historical user's preference for each type of material in the material library, and transmit each user's preference value for each type of material .

4. The method for managing an animation design material library based on big data according to claim 3, characterized in that: The step 2 is also used to consider the time sensitivity of the user behavior and calculate the updated update behavior weight. The specific steps are: First, define the time attenuation coefficient formula, specifically: , where W is the attenuation coefficient, is the attenuation rate, the value is 0.1, The difference between the time when the user's historical behavior was sent and the current time; The weight assigned to each behavior is multiplied by the time decay coefficient of each corresponding behavior record to obtain the updated update behavior weight.

5. The method for managing an animation design material library based on big data according to claim 3, characterized in that: The calculation obtains the preference value of each user for each material type The calculation steps are: According to the formula Calculate and obtain each user's preference value for each material type ,in Represents the user, Indicates the material type. Indicates the update behavior weight of user P after updating material type T.

6. The method for managing an animation design material library based on big data according to claim 1, characterized in that: The specific working method of step three is: Receive each user's preference value for each type of material , the preference value and user ID and material type, organized into a structured table; For each piece of material, extract feature weights based on its classification results; For each user, the matching degree H between the user and each piece of material is calculated based on the user's preference weight and the feature weight of the material; For each user, the materials are sorted according to the matching degree H, and the top N materials with the highest matching degree are selected as the recommendation results. The materials that the user has used are removed from the recommendation results to avoid repeated recommendations. Finally, a personalized recommendation list is generated, including the user ID and the recommended material ID list.

7. The method for managing an animation design material library based on big data according to claim 6, characterized in that: For each piece of material, the specific steps of extracting feature weights according to its classification results are as follows: According to the classification results of the materials, feature weights are assigned to each material, the weight value of each type of feature is defined, and the corresponding weights are extracted according to the classification information of the material to obtain the feature weights of each type of material. .

8. The method for managing an animation design material library based on big data according to claim 6, characterized in that: The specific working steps of calculating the matching degree H between each user and each piece of material according to the preference weight of the user and the feature weight of the material also include: According to the formula , the matching degree H between each user and each material is obtained by calculation.

9. The method for visual inspection of images in a tin bath according to claim 1, characterized in that: The specific scheme of step 4 is: Extract information from project requirements submitted by users, including project type, required material style, and usage scenarios; and extract usage information of each material from the material library, including download count, usage frequency, and user reviews; Obtain the importance value G of the material in the current project requirements, and obtain the popularity value V of the material in actual use; According to the formula , calculate and obtain the comprehensive priority Z of each material, sort the materials according to the comprehensive priority, and update the material with the highest comprehensive priority first.

10. A method for managing an animation design material library based on big data according to claim 9, characterized in that: The specific calculation steps for the importance value G of the material in the current project requirements and the popularity value V of the material in actual use are: For the importance value G, first, assign weights to project requirements based on project type, style, and usage scenarios. Then, calculate the matching degree of the material based on the matching degree between the material classification information and the project requirements. If the material classification information matches the project requirements, one point is awarded, and if it does not match, 0 point is awarded. Finally, multiply the project requirement weight by the material matching degree and sum them up to get the importance value G of each material. The popularity value V is first assigned weight coefficients for the number of downloads, usage frequency and user evaluation. Secondly, these data are normalized to the range of 0 to 1. Finally, the normalized number of downloads, usage frequency and user evaluation are multiplied by the corresponding weight coefficients and summed to obtain the popularity value V of each material.

Images