A cultural relic data cross-platform transmission method and related device

Abstract

This application discloses a method and related apparatus for cross-platform transmission of cultural relic data, relating to the field of data transmission. The method includes: performing cross-platform alignment mapping on the metadata items described by each platform in a platform set to obtain a metadata alignment mapping table; generating a data transmission rule set for future operating indicators based on a first operating indicator under each platform and the metadata alignment mapping table; transmitting the cultural relic data across platforms according to the data transmission rule set; and monitoring a second operating indicator in real time during transmission. If the change in the second operating indicator compared to the future operating indicator exceeds a change threshold, the second operating indicator is used as the first operating indicator, and a new data transmission rule set is generated. This application improves the transmission efficiency of cultural relic data and reduces the risk of packet loss by monitoring platform operating indicators in real time and adaptively adjusting the data transmission rule set.

Description

Technical Field

[0001] This application relates to the field of data transmission technology, and in particular to a method and related apparatus for cross-platform transmission of cultural relic data. Background Technology

[0002] Currently, before cultural relic data can be transmitted across platforms, a data transmission rule set needs to be generated in advance based on the platform's current operating metrics. The data is then transmitted according to this predetermined rule set. However, this rule set is determined under the assumption that the platform's operating metrics remain constant. In actual transmission, however, the platform's operating metrics may change, potentially leading to low transmission efficiency and packet loss risks when transmitting data across platforms according to the pre-generated rule set. Summary of the Invention

[0003] In view of the above problems, this application provides a method and related apparatus for cross-platform transmission of cultural relic data to solve the problems of low transmission efficiency and packet loss risk in existing technologies that transmit data across platforms according to a pre-generated data transmission rule set. The specific solution is as follows:

[0004] The first aspect of this application provides a method for cross-platform transmission of cultural relic data, including:

[0005] In one possible implementation, metadata description information and first operating indicators are obtained for each platform in the platform set. Each platform is configured with multiple metadata items of cultural relic data to be transmitted, and the metadata description information is the description information of the multiple metadata items.

[0006] Based on the metadata description information of each platform in the platform set, cross-platform alignment mapping is performed on the metadata items described by the metadata description information to obtain a metadata alignment mapping table.

[0007] Generate a data transmission rule set for future operating metrics based on the first operating metric of each platform in the platform set and the metadata alignment mapping table;

[0008] The cultural relic data is transmitted across platforms according to the data transmission rule set, and the second operating indicator is monitored in real time during the transmission process. If the change of the second operating indicator compared with the future operating indicator is greater than the preset change threshold, the second operating indicator is used as the first operating indicator, and the data transmission rule set under the future operating indicator generated according to the first operating indicator under each platform in the platform set and the metadata alignment mapping table is returned.

[0009] In one possible implementation, the step of performing cross-platform alignment mapping on the metadata items described by the metadata description information under each platform in the platform set to obtain a metadata alignment mapping table includes:

[0010] Based on the metadata description information under each platform, determine the characteristics of each metadata item under each platform;

[0011] For every two platforms in the aforementioned platform set:

[0012] Cross-platform pairing of metadata items under the two platforms yields multiple candidate metadata pairs.

[0013] The correlation degree of the candidate metadata pair is determined based on the characteristics of the two metadata items in the candidate metadata pair;

[0014] Based on the relevance of each of the multiple candidate metadata pairs, candidate metadata pairs that maximize the global relevance of the two platforms are selected from the multiple candidate metadata pairs, and the selected candidate metadata pairs are added to the metadata alignment mapping table.

[0015] In one possible implementation, determining the characteristics of each metadata item under each platform based on the metadata description information under each platform includes:

[0016] Based on the name and annotation of each metadata item contained in the metadata description information of each platform, determine the semantic embedding vector of each metadata item under each platform;

[0017] Based on the structural attribute values ​​of each metadata item contained in the metadata description information of each platform, determine the structural feature vector of each metadata item under each platform;

[0018] Based on the compliance attribute value of each metadata item contained in the metadata description information under each platform, determine the compliance feature vector of each metadata item under each platform;

[0019] The features of each metadata item under each platform are determined based on the semantic embedding vector, structural feature vector, and compliance feature vector of each metadata item under each platform.

[0020] In one possible implementation, determining the relevance of the candidate metadata pair based on the characteristics of the two metadata items in the candidate metadata pair includes:

[0021] Calculate the cosine similarity of the features of the two metadata items in the candidate metadata pair;

[0022] Obtain the correlation coefficient between the two metadata items in the candidate metadata pair. The correlation coefficient is determined based on the compliance attribute values ​​of the two metadata items in the candidate metadata pair. The correlation coefficient is used to characterize the degree of alignment between the two metadata items in the candidate metadata pair.

[0023] The cosine similarity is calibrated using the correlation coefficient to obtain the correlation degree of the candidate metadata pair.

[0024] In one possible implementation, adding the selected candidate metadata pairs to the metadata alignment mapping table includes:

[0025] Each selected candidate metadata pair will be used as the target candidate metadata pair;

[0026] Using a pre-constructed domain knowledge graph, the semantic difference between two metadata items in the target candidate metadata pair is determined as the semantic difference between the target candidate metadata pair. The domain knowledge graph is a knowledge graph constructed based on concepts and relationships in the cultural heritage domain.

[0027] Determine the structural-functional equivalence results of the two metadata items in the target candidate metadata pair in their respective metadata systems, and generate the structural difference degree of the target candidate metadata pair based on the structural-functional equivalence results;

[0028] Based on the semantic and structural differences of the target candidate metadata pairs, determine whether the two metadata items in the target candidate metadata pairs can be mapped to each other; to obtain all candidate metadata pairs that can be mapped to each other.

[0029] The candidate metadata pairs that can be mapped to each other are added to the metadata alignment mapping table.

[0030] In one possible implementation, generating a data transmission rule set for future operating metrics based on the first operating metrics for each platform in the platform set and the metadata alignment mapping table includes:

[0031] Based on the first operating metric of each platform in the platform set and the metadata alignment mapping table, generate the profile features of each platform.

[0032] Based on the individual profile characteristics of each platform, multiple clustering algorithms are used to cluster each platform, resulting in clustering results corresponding to each of the multiple clustering algorithms.

[0033] Based on the clustering results corresponding to the various clustering algorithms, the final clustering results of each platform are determined by combining the consensus algorithm.

[0034] The data transmission rule set is generated based on the first operational metric, the final clustering result, and the metadata alignment mapping table.

[0035] In one possible implementation, the metadata description information includes structural attribute values ​​for metadata items, and the structural attribute values ​​include data types.

[0036] The step of generating profile features for each platform based on the first operating metric of each platform in the platform set and the metadata alignment mapping table includes:

[0037] Obtain the policy risk level of each platform;

[0038] Calculate the information entropy corresponding to each platform based on the data types of the multiple metadata items under each platform;

[0039] Based on the information entropy corresponding to each platform, the strategy risk level of each platform, the first operating indicator of each platform, and the metadata alignment mapping table, the profile features of each platform are generated.

[0040] In one possible implementation, the step of performing cross-platform alignment mapping on the metadata items described by the metadata description information under each platform in the platform set to obtain a metadata alignment mapping table includes:

[0041] Using a pre-trained alignment mapping model, cross-platform alignment mapping is performed on the metadata items described by the metadata description information under each platform in the platform set to obtain the metadata alignment mapping table;

[0042] The process involves generating profile features for each platform based on its first operational metrics and the metadata alignment mapping table within the platform set; clustering each platform using multiple clustering algorithms based on these profile features; obtaining clustering results corresponding to each of the multiple clustering algorithms; and determining the final clustering result for each platform based on the clustering results corresponding to each of the multiple clustering algorithms, combined with a consensus algorithm. This process includes:

[0043] The final clustering result of each platform is determined by a pre-trained population partitioning model based on the first operating metric of each platform in the platform set and the metadata alignment mapping table.

[0044] One possible implementation also includes:

[0045] The process of acquiring the data transmission rule set adjustment records and the data transmission efficiency evaluation data of the cultural relics data are also included.

[0046] Based on the adjustment records and the transmission performance evaluation data, and combining the deviations between the prediction results of the alignment mapping model and the group partitioning model and the true labels, the mapping error loss corresponding to the alignment mapping model and the group partitioning model is generated.

[0047] The alignment mapping model and the group partitioning model are trained respectively based on their respective mapping error losses.

[0048] A second aspect of this application provides an electronic device, including at least one processor and a memory connected to the processor, wherein:

[0049] The memory is used to store computer programs;

[0050] The processor is used to execute the computer program so that the electronic device can implement the cross-platform data transmission method for cultural relics as described in the first aspect or any implementation thereof.

[0051] By employing the above technical solution, the cross-platform transmission method for cultural relic data provided in this application obtains metadata description information and a first operational indicator for each platform in a platform set. Based on the metadata description information for each platform in the platform set, the metadata items described by the metadata description information are aligned and mapped to obtain a metadata alignment mapping table. Considering that the operational indicators of each platform are likely to change, in order to adapt the generated data transmission rule set to the future changes in the operational indicators, this application can generate a data transmission rule set under future operational indicators based on the first operational indicators and the metadata alignment mapping table for each platform in the platform set. Since this data transmission rule set is based on the rule set of predicted future operational indicators, and the future operational indicators are the operational indicators that the predicted cultural relic data may rely on during future transmission, transmitting cultural relic data across platforms according to this data transmission rule set enables the cultural relic data to be transmitted to the target platform more quickly and with as little packet loss as possible, improving the transmission efficiency of cultural relic data and reducing the risk of packet loss.

[0052] Furthermore, to avoid inaccurate predictions of future operational indicators, which could lead to reduced data transmission efficiency and packet loss, this application can also monitor a second operational indicator in real time during the transmission of cultural relic data. If the change in the second operational indicator compared to the future operational indicator exceeds a preset threshold, the second operational indicator is used as the first operational indicator. A data transmission rule set for the future operational indicator is then generated based on the alignment mapping table of the first operational indicators and metadata for each platform in the platform set. Because this application can adjust the data transmission rule set in a timely manner based on real-time monitored platform operational indicators, the data transmission rule set is more consistent with the actual platform operational indicators, further improving the transmission efficiency of cultural relic data and reducing the risk of packet loss. Attached Figure Description

[0053] The above and other features, advantages, and aspects of the embodiments of this disclosure will become more apparent from the accompanying drawings and the following detailed description. Throughout the drawings, the same or similar reference numerals denote the same or similar elements. It should be understood that the drawings are schematic, and the originals and elements are not necessarily drawn to scale.

[0054] Figure 1 A schematic diagram of a system architecture provided for this application;

[0055] Figure 2 A flowchart illustrating a cross-platform data transmission method for cultural relics provided in this application;

[0056] Figure 3 A schematic diagram of the structure of a cross-platform data transmission device for cultural relics provided in this application;

[0057] Figure 4 This is a schematic diagram of the structure of an electronic device provided in this application. Detailed Implementation

[0058] The embodiments of this application are described below with reference to the accompanying drawings. The terminology used in the implementation section of this application is for explaining specific embodiments only and is not intended to limit the scope of this application.

[0059] The embodiments of this application will now be described with reference to the accompanying drawings. Those skilled in the art will recognize that, with technological advancements and the emergence of new scenarios, the technical solutions provided in the embodiments of this application are equally applicable to similar technical problems.

[0060] The terms "first," "second," etc., used in the specification, claims, and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such terms are interchangeable where appropriate; this is merely a way of distinguishing objects with the same attributes in the embodiments of this application. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion, so that a process, method, system, product, or apparatus that comprises a series of elements is not necessarily limited to those elements, but may include other elements not explicitly listed or inherent to those processes, methods, products, or apparatuses.

[0061] Optionally, the cross-platform data transmission method for cultural relics provided in this application can be applied to, for example... Figure 1 The system architecture shown includes a terminal 100 and a server 200. The server 200 may include one or more servers (…). Figure 1 (This example uses a server as an illustration).

[0062] Either terminal 100 or server 200 can be used independently to execute the cross-platform data transmission method for cultural relics provided in this application embodiment. Alternatively, terminal 100 and server 200 can also be used collaboratively to execute the cross-platform data transmission method for cultural relics provided in this application embodiment.

[0063] The following description Figure 1 The product form of the mid-terminal 100;

[0064] The terminal 100 in this application embodiment can be a mobile phone, tablet computer, wearable device, vehicle device, augmented reality (AR) / virtual reality (VR) device, laptop computer, ultra-mobile personal computer (UMPC), netbook, personal digital assistant (PDA), etc., and this application embodiment does not impose any restrictions on it.

[0065] To enable those skilled in the art to better understand this application, the cross-platform data transmission method for cultural relics according to embodiments of this application will be described in detail below with reference to the accompanying drawings.

[0066] Reference Figure 2 , Figure 2 This application provides a flowchart illustrating a method for cross-platform transmission of cultural relic data, as shown in the embodiments below. Figure 2 As shown, the cross-platform data transmission method for cultural relics may include:

[0067] Step S201: Obtain the metadata description information and the first running metrics for each platform in the platform set.

[0068] Here, the platform set is a set composed of platforms that need to transmit and store the to-be-transmitted cultural relic data. For example, heterogeneous data management platforms and cloud environments in the cultural relics and archaeology fields, etc.; different cultural relic data may correspond to different platform sets.

[0069] In this embodiment, for the to-be-transmitted cultural relic data, multiple metadata items of this cultural relic data are configured in each platform to facilitate the management and recording of this cultural relic data. Here, the metadata item refers to cultural relic metadata, that is, structured data for describing, recording, and managing cultural relic-related information, such as the name of the cultural relic, the age of the cultural relic, the material of the cultural relic, the place where the cultural relic was unearthed, the preservation status, the encryption situation of cultural relic transmission, etc.

[0070] The above to-be-transmitted cultural relic data includes the content of the cultural relic itself and / or the metadata values of the cultural relic; the metadata description information is the description information of multiple metadata items corresponding to the to-be-transmitted cultural relic data, that is, the information for describing the data type, value range, transmission compliance, etc. of each of the multiple metadata items, and the metadata value is the field value of the metadata item.

[0071] For example, for an image cultural relic, the content of the cultural relic itself refers to the pixel values of the image cultural relic. Taking the metadata item "age of the cultural relic" as an example, the metadata description information may include "field name: age; data type: string; value range: Shang, Zhou, Qin, Han, etc.; annotation: refers to the production period of the cultural relic", and the metadata value may be "age = Shang Dynasty".

[0072] In this embodiment, the metadata description information for each platform includes the attribute values of each of the multiple metadata items. Optionally, the attribute value of each metadata item includes: the name and annotation of the metadata item, the structural attribute value, and the compliance attribute value.

[0073] Among them, the name of the metadata item refers to the name displayed by the metadata item in the user interface or schema definition, such as the field name artiface_name or its Chinese display "name of the cultural relic", which is usually short and used for identification.

[0074] The annotation is a detailed text description of the meaning, usage, or filling specification of the metadata item, such as the standard naming of the cultural relic in the official archives.

[0075] The structural attribute values include but are not limited to: data type, value range, and whether it is required.

[0076] Compliance attribute values ​​include, but are not limited to: the data sensitivity of metadata items (such as ownership information, discovery coordinates), and the compliance control policies that metadata items are subject to (such as security classification, coding standards, permission management mechanisms, access control rules, etc.).

[0077] Data sensitivity refers to the sensitivity level of the metadata value corresponding to the metadata item stored and processed by the platform. For example, if the metadata value corresponding to the metadata item contains precise excavation coordinates, then the metadata item is at the "core sensitive" level.

[0078] Security classification is a level of confidentiality assessed through methods such as national information security level protection evaluation, such as level one and level two.

[0079] Encoding specifications are used to constrain the encoding format of metadata items. For example, dates and times must follow the ISO 8601 international standard format, i.e., YYYY-MM-DD. If it is a time period, it is separated by slashes (e.g., 1926-02-24 / 2026-02-24).

[0080] Access control mechanisms are used to restrict the transmission permissions of metadata items. For example, metadata values ​​marked as "core sensitive" must not be transmitted to servers or cloud platforms outside of China.

[0081] Access control rules include access and use authorization rules and privacy protection rules. For example, the access and use authorization rule is that "cultural relic image data can only be used for academic research and curation, and is strictly prohibited from being used for commercial artificial intelligence model training." The privacy protection rule is that "when metadata items involve personal information of archaeological donors, they must be anonymized before being provided to the public."

[0082] Of course, the aforementioned structural attribute values ​​and compliance attribute values ​​may include others, which are not specifically limited in this application.

[0083] Considering that the transmission of cultural relic data may be limited by environmental changes, leading to problems such as transmission interruption, packet loss, and reduced transmission efficiency, in order to minimize the impact of environmental changes, this embodiment can obtain the first operating index of each platform while obtaining the metadata description information. Here, the first operating index includes the platform operating index at the current moment. Optionally, the first operating index may also include the platform operating index at other moments before the current moment. The current moment refers to the moment when the metadata description information is obtained.

[0084] The aforementioned platform operation metrics refer to a series of measurable values ​​used to quantitatively evaluate the platform's health status, performance, and resource utilization efficiency, including but not limited to: network throughput, response time, bandwidth utilization, resource utilization, and platform availability.

[0085] Optionally, after obtaining the metadata description information of each platform in the platform set, the metadata description information can also be standardized, such as format unification, value domain normalization, and structure flattening / normalization, so that the metadata description information of different platforms can be compared and calculated at the same semantic and structural level, and the accuracy of subsequent alignment mapping can be improved.

[0086] Step S202: Based on the metadata description information of each platform in the platform set, perform cross-platform alignment mapping on the metadata items described by the metadata description information to obtain the metadata alignment mapping table.

[0087] It is understandable that different platforms may configure different metadata items for the same cultural relic. For example, for the above-mentioned image cultural relic, the metadata item "cultural relic age" may be defined as "image age" on platform a, "age" on platform b, "period" on platform c, and undefined on platform d, etc.

[0088] In order to determine the mapping relationship between metadata items of the source platform and metadata items of the target platform, so that when cultural relic data is transmitted from the source platform to the target platform, it is clear where the metadata value of the source platform is transmitted to the target platform, this embodiment can perform cross-platform alignment mapping on the metadata items described by the metadata description information under each platform in the platform set, and obtain a metadata alignment mapping table.

[0089] In other words, the purpose of alignment mapping is to eliminate the barriers to understanding between heterogeneous platforms by establishing semantic equivalence relationships between metadata items on different platforms, and ultimately achieve the interconnection and interoperability of cultural relic data.

[0090] Step S203: Generate a data transmission rule set for future operating metrics based on the alignment mapping table of the first operating metrics and metadata of each platform in the platform set.

[0091] The data transmission rule set is a set of rules for the compliant transformation and transmission of cultural relic data based on the metadata alignment mapping table under the predicted future operating indicators. The future operating indicators are the operating indicators that the predicted cultural relic data may rely on in the future transmission, and are determined according to the first operating indicator.

[0092] It is understandable that the changing trends of platform operation indicators are somewhat predictable. In this embodiment, the future operation indicators (i.e., the predicted values ​​of platform operation indicators) can be predicted based on the first operation indicator, and then a data transmission rule set consistent with the future operation indicators can be generated by combining the metadata alignment mapping table.

[0093] Optionally, the determination of future operating indicators can be based on the following formula (1):

[0094] Formula (1);

[0095] in, This represents the future performance indicators at time t. This represents the actual operating performance at time t-1. This represents the adjustment coefficient.

[0096] Optionally, the data transmission rule set includes, but is not limited to, rules from the following perspectives:

[0097] 1. Transmission path and routing decision rules, used to define the direction of the cultural relic data flow and the nodes it passes through;

[0098] 2. Data encoding / decoding and format conversion strategy rules, which are used to specify the format conversion rules of cultural relic data before, during and after transmission, to ensure technical interoperability and compliance;

[0099] 3. Compliance verification and policy execution process rules, used to transform laws, regulations and internal policies into checkpoints that can be automatically executed in the transmission process;

[0100] 4. Access control and permission inheritance rules are used to ensure that the access permissions of cultural relic data are consistent with or equivalent to the policies of the source platform after the data arrives at the target platform.

[0101] Of course, the above rules are merely examples. In addition, the data transmission rule set may include rules from other perspectives, which are not specifically limited in this application. For example, the data transmission rule set may also include: audit, log, and traceability record specifications to create an undeniable and traceable chain of evidence for each cultural relic data transmission, meeting regulatory and post-review requirements; performance and resource adaptation strategy rules to optimize transmission efficiency and ensure service stability based on the platform's real-time operating indicators; and exception handling and emergency response process rules to predefine handling plans for unexpected events (such as policy conflicts, verification failures, and network interruptions) during transmission.

[0102] Optionally, the transmission path and routing decision rules are as follows: That is, satisfying The transmission path is the actual path used for transmitting cultural relic data; among which, This represents an edge in the transmission path from the source platform to the target platform, that is, a link between two adjacent nodes in the transmission path. Indicates passing through the edge The cost.

[0103] Optionally, the data encoding / decoding and format conversion strategy rules can be... That is, satisfying a set of predefined encoding and decoding schemes. The encoding and decoding scheme is the one actually used for transmitting cultural relics data; among them, This represents the information entropy corresponding to the source platform and / or target platform for cultural relic data transmission. Indicates the length of the encoded data. This indicates the available bandwidth for transmission.

[0104] Step S204: Transmit cultural relic data across platforms according to the data transmission rule set, and monitor the second operating indicator in real time during the transmission process. If the change of the second operating indicator compared with the future operating indicator is greater than the preset change threshold, then the second operating indicator is used as the first operating indicator, and the data transmission rule set under the future operating indicator is generated based on the first operating indicator and metadata alignment mapping table of each platform in the platform set.

[0105] Although the generation process of the data transmission rule set takes into account changes in platform operating metrics, the predicted values ​​may deviate from the actual observed values. Understandably, if the deviation is small, cross-platform transmission according to the original data transmission rule set will have almost no impact on the transmission process. However, if the deviation is large, it may lead to problems such as transmission interruption, packet loss, and decreased efficiency. To address these issues, this embodiment can monitor platform operating metrics in real time during transmission. Here, the platform operating metrics monitored in real time are defined as the second operating metric (i.e., the observed value).

[0106] If the change of the second operating indicator compared to the future operating indicator at the same time... If the change exceeds the preset threshold, it indicates that the data transmission rule set generated earlier deviates from the current real network environment. In this case, the second operating indicator can be used as the first operating indicator, and the process returns to step S203 to regenerate a data transmission rule set that adapts to the current real network environment before continuing to transmit the cultural relic data.

[0107] The cross-platform data transmission method for cultural relics provided in this application obtains metadata description information and a first operational indicator for each platform in a platform set. Based on the metadata description information for each platform in the platform set, the metadata items described by the metadata description information are aligned and mapped to obtain a metadata alignment mapping table. Considering that the operational indicators of each platform are likely to change, in order to adapt the generated data transmission rule set to the future changes in the operational indicators, this application can generate a data transmission rule set under future operational indicators based on the first operational indicators and the metadata alignment mapping table for each platform in the platform set. Since this data transmission rule set is based on the rule set of predicted future operational indicators, and the future operational indicators are the operational indicators that the predicted cultural relics data may rely on during future transmission, transmitting cultural relics data across platforms according to this data transmission rule set enables the cultural relics data to be transmitted to the target platform more quickly and with as little packet loss as possible, improving the transmission efficiency of cultural relics data and reducing the risk of packet loss.

[0108] Furthermore, to avoid inaccurate predictions of future operational indicators, which could lead to reduced data transmission efficiency and packet loss, this application can also monitor a second operational indicator in real time during the transmission of cultural relic data. If the change in the second operational indicator compared to the future operational indicator exceeds a preset threshold, the second operational indicator is used as the first operational indicator. A data transmission rule set for the future operational indicator is then generated based on the alignment mapping table of the first operational indicators and metadata for each platform in the platform set. Because this application can adjust the data transmission rule set in a timely manner based on real-time monitored platform operational indicators, the data transmission rule set is more consistent with the actual platform operational indicators, further improving the transmission efficiency of cultural relic data and reducing the risk of packet loss.

[0109] In some embodiments of this application, the process of step S202, "according to the metadata description information of each platform in the platform set, performing cross-platform alignment mapping on the metadata items described by the metadata description information to obtain a metadata alignment mapping table," is described in detail.

[0110] It is understood that each platform in the platform set can be either a source platform or a target platform for the transmission of cultural relics data. Therefore, in order to determine the mapping relationship between the metadata items of the source platform and the target platform, this embodiment can generate the features of each metadata item under each platform based on the metadata description information of each platform in the platform set. That is, the features of each metadata item are generated based on the attribute values ​​of each metadata item under each platform in the platform set. Then, for every two platforms in the platform set, the metadata items between the two platforms are aligned and mapped in combination with the features of the metadata items.

[0111] Optionally, as mentioned above, the metadata description information under each platform includes the names, annotations, structural attribute values, and compliance attribute values ​​of multiple metadata items under that platform. Then, the process of "determining the characteristics of each metadata item under each platform based on the metadata description information under each platform" includes: determining the semantic embedding vector of each metadata item based on the name and annotation of each metadata item under each platform in the platform set; determining the structural feature vector of each metadata item based on the structural attribute values ​​of each metadata item under each platform in the platform set; determining the compliance feature vector of each metadata item based on the compliance attribute values ​​of each metadata item under each platform in the platform set; and determining the characteristics of each metadata item under each platform based on the semantic embedding vector, structural feature vector, and compliance feature vector of each metadata item under each platform.

[0112] Optionally, the process of "determining the semantic embedding vector of each metadata item based on the name and annotation of each metadata item under each platform in the platform set" may include: inputting the name and annotation of the metadata item into a pre-trained domain-specific language model to obtain a fixed-dimensional vector output by the model, which serves as the semantic embedding vector of the metadata item.

[0113] Optionally, the process of "determining the structural feature vector of each metadata item based on the structural attribute value of each metadata item under each platform in the platform set" may include: converting each structural numerical type of the metadata item into a numerical feature through one-hot encoding, numerical encoding and other techniques, and then concatenating them into a structural feature vector.

[0114] Optionally, the process of "determining the features of each metadata item under each platform based on the semantic embedding vector, structural feature vector, and compliance feature vector of each metadata item under each platform" may include: concatenating the semantic embedding vector, structural feature vector, and compliance feature vector of each metadata item under each platform together (or weighted fusion, etc., which are not specifically limited in this application) to form a fused vector, which serves as the feature of each metadata item under each platform.

[0115] Furthermore, for each pair of platforms in the platform set, the process of aligning and mapping the metadata items between the two platforms may include: pairing the metadata items under the two platforms across platforms to obtain multiple candidate metadata pairs, where the two metadata items in each candidate metadata pair come from the two platforms; determining the relevance of the candidate metadata pair based on the characteristics of the two metadata items in each candidate metadata pair to obtain the relevance of each of the multiple candidate metadata pairs; and then, based on the relevance of each of the multiple candidate metadata pairs, selecting the candidate metadata pairs that maximize the global relevance between the two platforms from the multiple candidate metadata pairs, and adding the selected candidate metadata pairs to the metadata alignment mapping table.

[0116] For ease of explanation below, we will define the two platforms as Platform A and Platform B, and define the feature of the i-th metadata item under Platform A as... Define the feature of the j-th metadata item under platform B as follows: Then you can press ( , The metadata items under the two platforms are paired up in a way that yields multiple candidate metadata pairs. For example, if platforms A and B each have 3 metadata items, then 9 candidate metadata pairs can be obtained.

[0117] In this embodiment, the characteristics of the two metadata items in each candidate metadata pair can be used as a basis ( , The correlation between the two metadata items in the candidate metadata pair is calculated. For ease of description, this is referred to as the correlation between the candidate metadata pair below.

[0118] In one possible implementation, the cosine similarity of the features of the two metadata items in the candidate metadata pair can be calculated as the relevance of the candidate metadata pair.

[0119] Considering that candidate metadata pairs may be determined by the compliance attribute values ​​of the two metadata items to be subject to mandatory alignment, mandatory differentiation, or fundamental type conflicts, the cosine similarity method mentioned above may force the degree of association between the two metadata items to be closer or further apart due to the similarity relationship between semantic embedding vectors or structural feature vectors, resulting in a serious inconsistency between the final determined degree of association and the alignment relationship constrained by the compliance attribute values.

[0120] In another possible implementation, in order to comprehensively consider the alignment relationship between two metadata items constrained by compliance attribute values, this embodiment can use rule engines, lightweight classification models, manual methods, etc., to pre-label the correlation coefficient for each candidate metadata. Then, the cosine similarity is calibrated using the correlation coefficient to obtain the correlation degree of the candidate metadata pairs. The correlation coefficient is... The compliance attribute value of each of the two metadata items in the candidate metadata pair is used to determine the compliance attribute value. Used to characterize the alignment of two metadata items in a candidate metadata pair.

[0121] Optionally, in this embodiment, the correlation degree of candidate metadata pairs can be obtained using the following formula (2).

[0122] Formula (2);

[0123] in, Indicates candidate metadata pairs ( , The degree of correlation between ) This represents the correlation coefficient.

[0124] In this embodiment, when the compliance attribute values ​​of the two metadata items in a candidate metadata pair indicate that the two metadata items must be forcibly aligned (e.g., the metadata item "data owner" must be forcibly aligned across all platforms), Assigning a higher value (e.g., 0.95) significantly improves the relevance of candidate metadata pairs, ensuring a high degree of matching between the two metadata items. Conversely, when the compliance attribute values ​​of the two metadata items in a candidate metadata pair indicate that they involve sensitive information that must be distinguished (such as the "display level" of a public system versus the "confidentiality level" of an internal system), or when there is a fundamental type conflict, Assigning extremely low values ​​(such as 0.1) can significantly reduce the correlation between candidate metadata pairs, suppressing the matching of two metadata items.

[0125] After obtaining the relevance of multiple candidate metadata pairs using the above method, the candidate metadata pairs that maximize the global relevance between the two platforms can be selected and added to the metadata alignment mapping table. Here, maximizing the global relevance means that, for the platform with fewer metadata items, the sum of the relevance of each candidate metadata pair formed after matching each metadata item of the platform with a metadata item of the other platform is maximized.

[0126] For example, suppose the characteristics of the three metadata items under platform A are as follows: , , The characteristics of the two metadata items under platform B are as follows: , Then the candidate metadata pairs include: ( , ), ( , ), ( , ), ( , ), ( , ), ( , Among the six candidate metadata pairs mentioned above, the combinations that allow platform A and platform B to match include:

[0127] Combination 1: ( , ), ( , );

[0128] Combination 2: ( , ), ( , );

[0129] Combination 3: ( , ), ( , );

[0130] Combination 4: ( , ), ( , );

[0131] Combination 5: ( , ), ( , );

[0132] Combination 6: ( , ), ( , );

[0133] Therefore, in this embodiment, the sum of the correlation of each of the above combinations can be calculated, and the combination with the largest sum can be found. The candidate metadata pair in this combination is the candidate metadata pair that maximizes the global correlation between the two platforms.

[0134] It is understandable that the number of metadata items under each platform may be much greater than 2 or 3 items. Using the exhaustive method described above may be time-consuming and laborious. In order to improve the screening efficiency, this embodiment can also adopt the following method: First, construct a benefit matrix based on the correlation of multiple candidate metadata pairs, and then use the Hungarian algorithm to screen and obtain candidate metadata pairs that maximize the global correlation between the two platforms.

[0135] Similarly, all paired platform metadata items (i.e., the selected candidate metadata pairs) can be added to the metadata alignment mapping table.

[0136] Optionally, in addition to the selected candidate metadata pairs, the metadata alignment mapping table also includes the confidence level and compliance attribute value of each candidate metadata pair.

[0137] In this embodiment, a correlation coefficient based on compliance attribute values ​​is introduced. This can effectively improve the matching accuracy of metadata items across platforms, ensure that cultural relic data complies with security rules during cross-platform transmission, and reduce the risk of violations.

[0138] In another possible implementation, considering the above-mentioned correlation-based approach... The candidate metadata pair selection method, because the characteristics of metadata items are vector fusion features of three dimensions—semantics, structure, and compliance—leads to a high degree of relevance. It can only reflect the overall matching degree of three dimensions, which may lead to incorrect screening of candidate metadata pairs.

[0139] To further verify whether there are semantic or structural differences in the candidate metadata pairs selected above, this embodiment can also verify the selected candidate metadata pairs before adding them to the metadata alignment mapping table.

[0140] Specifically, each candidate metadata pair selected from the screening is taken as a target candidate metadata pair. Using a pre-built professional domain knowledge graph, the semantic difference between the two metadata items in the target candidate metadata pair is determined as the semantic difference of the target candidate metadata pair. Here, the professional domain knowledge graph is a knowledge graph constructed based on the concepts and relationships in the cultural heritage domain. The structural and functional equivalence of the two metadata items in the target candidate metadata pair in their respective metadata systems is determined, and the structural difference of the target candidate metadata pair is generated based on the structural and functional equivalence results. Based on the semantic and structural difference of the target candidate metadata pair, it is determined whether the two metadata items in the target candidate metadata pair can be mapped to each other, so as to obtain all candidate metadata pairs that can be mapped to each other.

[0141] Therefore, in this embodiment, only candidate metadata pairs that can be mapped to each other can be added to the metadata alignment mapping table.

[0142] For example, in the candidate metadata pairs selected above, the two metadata items in a certain candidate metadata pair are "producer" and "creator". However, in the professional domain knowledge graph, "producer" is an organization and "creator" is an individual. This indicates that there is a semantic difference between "producer" and "creator". Based on this, the semantic difference degree of this candidate metadata pair can be generated. For example, "organization" and "individual" can be converted into vectors and then the vector similarity can be calculated as the semantic difference degree.

[0143] For example, in the candidate metadata pairs selected earlier, the two metadata items in a certain candidate metadata pair are "Generation Date" and "Registration Time". In platform A where "Generation Date" is located, "Generation Date" is a root-level metadata item, while in platform B where "Registration Time" is located, "Registration Time" is a leaf-level metadata item (i.e., a subordinate attribute of the root-level metadata item). Therefore, it can be determined that there is a structural asymmetry problem between "Generation Date" and "Registration Time". Thus, the structural difference degree of this candidate metadata pair can be generated. For example, the greater the difference in the structural levels of "Generation Date" and "Registration Time", the greater the difference degree of the result.

[0144] Optionally, for each candidate metadata pair (i.e. target candidate metadata pair) among the selected candidate metadata pairs, this embodiment can perform a weighted summation of the semantic difference and structural difference of the target candidate metadata pair to obtain a weighted summation value.

[0145] For example, ,in, This represents the semantic difference between candidate metadata and m. This indicates the degree of structural difference of the candidate metadata relative to m. and These represent two weighting coefficients used to adjust the relative importance of semantic difference and outcome difference in the overall optimization objective, and can be adjusted according to the scenario.

[0146] It is understandable that the smaller the weighted sum, the smaller the semantic and structural differences between the two metadata items in the target candidate metadata pair. Therefore, the two metadata items in the target candidate metadata pair are more likely to map to each other. Thus, this embodiment can further determine whether the two metadata items in the target candidate metadata pair can map to each other based on the weighted sum.

[0147] For example, sort all the selected candidate metadata pairs in ascending order of their weighted sum values, and take the top M candidate metadata pairs from the sorting results. It is assumed that the two metadata items in these M candidate metadata pairs can be mapped to each other, that is, the M candidate metadata pairs can be mapped to each other.

[0148] For example, the weighted sum of each selected candidate metadata pair is compared with a preset mapping threshold. If the weighted sum is greater than the mapping threshold, then the M candidate metadata pairs are determined to be mutually mappable.

[0149] Of course, there are other ways, which will not be specifically limited here.

[0150] By using the above method, candidate metadata pairs that are significantly inconsistent in both semantic and structural dimensions can be eliminated, thereby improving the accuracy of the metadata alignment mapping table.

[0151] In some other embodiments of this application, the process of "generating a data transmission rule set under future operating indicators based on the alignment mapping table of the first operating indicators and metadata of each platform in the platform set" in step S203 above is described.

[0152] As mentioned earlier, each platform in the platform set can be either a source platform or a target platform for the transmission of cultural relics data. Therefore, in one possible implementation, the data transmission rule set generated in step S203 includes: a set of data transmission rule sets (hereinafter referred to as data circulation guidelines) generated for every two platforms. Then, p platforms need to maintain 2p(p-1) sets of data circulation guidelines, which is a very large number and has a very high maintenance cost.

[0153] Considering that some platforms in the platform set may have similar metadata item configurations and operational indicators for cultural relics data, the data transmission rule sets generated for these platforms are also similar. Therefore, this embodiment can cluster these similar platforms together, thereby reducing the number of data circulation guidelines and reducing dimensional costs.

[0154] Specifically, in order to facilitate clustering, this embodiment can first generate the profile features of each platform based on the first operating indicators and metadata alignment mapping table of each platform in the platform set, and then perform clustering based on the profile features of each platform to obtain multiple platform clusters. Finally, a data transmission rule set is generated based on the multiple platform clusters and the metadata alignment mapping table.

[0155] Optionally, the process of "generating profile features for each platform based on the first operational indicators and metadata alignment mapping table for each platform in the platform set" may include: obtaining the policy risk level of each platform, calculating the information entropy of each platform based on the data type of each of the multiple metadata items under each platform, and generating profile features for each platform based on the information entropy of each platform, the policy risk level of each platform, the first operational indicators and metadata alignment mapping table under each platform.

[0156] The policy risk level of any platform is determined by the compliance attribute values ​​of each metadata item under that platform, the platform's compliance, and its historical security records.

[0157] The aforementioned compliance refers to whether the platform meets specific regulations (such as restrictions on data export) or industry standards; historical security records refer to whether the platform has experienced incidents such as data breaches or violations of regulations.

[0158] Optionally, the information entropy corresponding to each platform can be calculated using the following formula (3).

[0159] Formula (3);

[0160] in, This represents the information entropy corresponding to platform X. This represents the number of all possible values ​​for the data type of multiple metadata items under platform X. This represents the i-th possible value. This represents the frequency of the i-th value.

[0161] For example, platform X includes 10 metadata items, each with the following data types: integer, decimal, string, string, decimal, date and time, integer, string, boolean, boolean. Then n=5, when... When the type is integer, decimal, or boolean, When it is 0.2, When it is a string, When it is 0.3, When the data type is date and time, It is 0.1.

[0162] This embodiment can effectively measure the heterogeneity of the metadata system within the platform by calculating the information entropy corresponding to the platform, and can be used to assess the complexity and difficulty of cross-platform data transmission.

[0163] Considering that different clustering algorithms may produce different platform clustering results, to avoid discrepancies among the multiple platform clusters due to unsuitable clustering algorithms, which could affect the accuracy of the data transmission rule set, this embodiment employs multiple clustering algorithms to cluster each platform based on its unique profile characteristics. This yields clustering results corresponding to each algorithm. Then, based on these results and a consensus algorithm, the final clustering result for each platform is determined. Finally, a data transmission rule set is generated based on the first operational metric, the final clustering result, and the metadata alignment mapping table. Different clustering techniques (such as K-means clustering and density clustering) belong to different clustering algorithms, and different parameter configurations under the same clustering technique belong to different clustering algorithms.

[0164] For example, if three or more of the five clustering algorithms cluster platform A and platform B together, then the final clustering result will cluster platform A and platform B into one platform cluster.

[0165] By combining multiple clustering algorithms with consensus algorithms, a platform clustering method with higher consensus can be obtained, which improves the accuracy of platform clustering and thus enhances the effectiveness and accuracy of data transmission rule sets.

[0166] It is understandable that, over time or in different scenarios, the compliance attribute values ​​of metadata items may change for each platform in the platform set. For example, changes in data specifications or policies may lead to changes in policy risk levels or metadata alignment mapping tables, which may affect the platform clustering results.

[0167] In order to re-evaluate the final clustering result above when the compliance attribute value of any metadata item of the platform (hereinafter referred to as the change platform) changes, this embodiment can use the following formula (4) to update the cluster center of the platform cluster where the change platform is located, and then adjust the cluster of the change platform based on the cluster center, or re-cluster the platform cluster where the change platform is located and its neighboring clusters.

[0168] Formula (4);

[0169] in, This indicates the platform cluster where the changed platform belongs (i.e., the kth platform cluster in the final clustering result). express The updated cluster centers (represented by time t+1), express A profile feature of a platform within the platform. express The corresponding weighting coefficient is determined by the magnitude of the platform's changes and its current state.

[0170] The magnitude of the change is determined by the norm of the difference between x before and after the update, and the current state is determined by... The corresponding platform's real-time operating metrics and business importance are determined at the time of update. Business importance indicates whether the platform is a core business system, and is marked by the administrator.

[0171] In this embodiment, by updating the cluster center of the cluster where the platform is located, the affected platform cluster can be locally clustered based on the cluster center, avoiding global recalculation and improving clustering efficiency.

[0172] Of course, in one possible scenario, when the compliance attribute value does not change much, it may not affect the final clustering result. Therefore, in one possible implementation, this embodiment can determine whether the changed compliance attribute value is continued or conflicted after mapping before updating the clustering center of the platform cluster where the changed platform is located according to formula (4). If it is continued, it means that the final clustering result meets the requirements, so the local clustering based on formula (4) can be omitted. If a conflict occurs, it means that the final clustering result needs to be adjusted, so the clustering center of the platform cluster where the changed platform is located is updated according to formula (4).

[0173] Here, "being able to continue" means that during the metadata alignment and mapping process, the compliance attribute values ​​of metadata items under the source platform (i.e., the changing platform) must be equally guaranteed or reflected in the corresponding metadata items under the target platform.

[0174] For example, if metadata item 'a' under source platform A and metadata item 'b' under target platform B form a candidate metadata pair and have been added to the metadata alignment mapping table, then if the compliance attribute value of metadata item 'a' is {Sensitivity level: Core secret, Encryption requirement: Yes}, but the compliance attribute value of metadata item 'b' is {Sensitivity level: General secret, Encryption requirement: No} (this may be because target platform B has no encryption capability, resulting in no encryption requirement), it indicates that this mapping may lead to a compliance policy conflict, i.e., it has not been continued.

[0175] Through the continuity verification in this embodiment, blind local clustering can be avoided, thus improving the transmission efficiency of cultural relic data.

[0176] In one possible implementation, the process of step S202, "based on the metadata description information of each platform in the platform set, performing cross-platform alignment mapping on the metadata items described by the metadata description information to obtain a metadata alignment mapping table," can be implemented using a pre-trained alignment mapping model. That is, in this embodiment, a pre-trained alignment mapping model can be used to perform cross-platform alignment mapping on the metadata items described by the metadata description information of each platform in the platform set to obtain a metadata alignment mapping table. The alignment mapping model is trained using training metadata description information labeled with real metadata alignment mapping table tags as training samples.

[0177] Optionally, the process in step S203 above, which involves "generating profile features for each platform based on the first operational metrics and metadata alignment mapping table for each platform in the platform set, clustering each platform using multiple clustering algorithms based on these profile features, obtaining clustering results corresponding to each algorithm, and determining the final clustering result for each platform based on the clustering results and a consensus algorithm," can be implemented using a pre-trained group partitioning model. That is, the pre-trained group partitioning model determines the final clustering result for each platform based on the first operational metrics and metadata alignment mapping table for each platform in the platform set. The group partitioning model is trained using training operational metrics labeled with real clustering result tags and a training metadata alignment mapping table as training samples.

[0178] Optionally, in order to train the above-mentioned alignment mapping model and group partitioning model to output more accurate prediction results, this embodiment can obtain the adjustment records of the data transmission rule set during the transmission of cultural relics data and the transmission efficiency evaluation data of cultural relics data. Then, based on the adjustment records and transmission efficiency evaluation data, and combined with the deviation between the prediction results of the alignment mapping model and the group partitioning model and the true labels, the corresponding mapping error loss of the alignment mapping model and the group partitioning model is generated. Finally, the alignment mapping model and the group partitioning model are trained respectively based on the corresponding mapping error loss of the alignment mapping model and the group partitioning model.

[0179] Specifically, as mentioned above, during the transmission of cultural relic data, the data transmission rule set may be updated and adjusted due to changes in platform operation indicators, compliance attribute values, etc. In this embodiment, the adjustment record of the data transmission rule set can be obtained.

[0180] In this embodiment, transmission efficiency evaluation data of cultural relics data can also be recorded. This transmission efficiency evaluation data includes, but is not limited to, the actual efficiency of the transmission process of cultural relics data and the compliance verification results.

[0181] Furthermore, this embodiment can generate the mapping error loss corresponding to the alignment mapping model based on the above-mentioned adjustment records, transmission performance evaluation data, and the deviation between the metadata alignment mapping table output by the alignment mapping model and the actual metadata alignment mapping table labels. Then, based on the mapping error loss corresponding to the alignment mapping model... Train the alignment mapping model.

[0182] Optionally, the parameters of the alignment mapping model can be trained using stochastic gradient descent, as shown in the following formula:

[0183] Formula (5);

[0184] in, This represents the model parameters of the alignment mapping model after training in the (t+1)th training batch (or the (t+1)th training session). The learning rate controls the step size for each parameter update. Represents the mapping error loss At the current location The gradient.

[0185] Similarly, in this embodiment, based on the aforementioned adjustment records, transmission performance evaluation data, and the deviation between the final clustering results output by the group partitioning model and the actual clustering results labels, a mapping error loss corresponding to the group partitioning model can be generated. Then, the group partitioning model is trained based on this mapping error loss. The detailed training process is similar to that described above and will not be repeated here.

[0186] In one possible implementation, this embodiment can perform model training once each time the data transmission rule set is updated, i.e., according to the above process. Of course, model training can also be performed after the cultural relic data transmission is completed, based on the adjustment record of the data transmission rule set during the entire transmission process and the transmission performance evaluation data. This application does not impose any specific limitations.

[0187] As mentioned earlier, when the compliance attribute value of any metadata item of the change platform changes, the cluster center of the platform cluster to which the change platform belongs can be passively updated, so as to achieve the update of the affected platform cluster through a local clustering strategy.

[0188] In some scenarios, even if the platform's profile features remain unchanged, the engine of the group segmentation model may find from historical transmission performance evaluation data that the currently clustered platform clusters are not the optimal solution. Therefore, the cluster centers of each platform cluster can be updated using the following formula and used as the initial cluster centers to re-cluster the platforms.

[0189] Formula (6);

[0190] in, This represents the updated cluster center of the k-th platform cluster in the current final clustering result. This represents the pre-update cluster center of the k-th platform cluster in the current final clustering result. Indicates the learning rate. The amount of update or adjustment can be calculated using the following method.

[0191] The negative direction value obtained by calculating the gradient of the mapping error loss in the previous text (- As Of course, there are other ways to determine this. For example, it can be determined through preset business rules, etc., but this application does not make any specific limitations.

[0192] The above describes a method for cross-platform transmission of cultural relic data provided by embodiments of this application. The following describes the apparatus for performing the above-described method for cross-platform transmission of cultural relic data.

[0193] Please see Figure 3 , Figure 3 This is a schematic diagram of the structure of a cross-platform data transmission device for cultural relics provided in an embodiment of this application. Figure 3 As shown, the cross-platform data transmission device for cultural relics may include:

[0194] The data acquisition unit 301 is used to acquire metadata description information and first operating indicators for each platform in the platform set. Each platform is configured with multiple metadata items of cultural relic data to be transmitted, and the metadata description information is the description information of the multiple metadata items.

[0195] Alignment mapping unit 302 is used to perform cross-platform alignment mapping on the metadata items described by the metadata description information according to the metadata description information of each platform in the platform set, so as to obtain a metadata alignment mapping table;

[0196] The guide generation unit 303 is used to generate a data transmission rule set for future operating indicators based on the first operating indicators of each platform in the platform set and the metadata alignment mapping table.

[0197] The data transmission unit 304 is used to transmit the cultural relic data across platforms according to the data transmission rule set, and to monitor the second operating indicator in real time during the transmission process. If the change of the second operating indicator compared with the future operating indicator is greater than the preset change threshold, the second operating indicator is used as the first operating indicator, and the data transmission rule set under the future operating indicator generated according to the first operating indicator under each platform in the platform set and the metadata alignment mapping table is returned.

[0198] Each module in the aforementioned cross-platform data transmission device for cultural relics can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device, or stored in the memory of a computer device as software, so that the processor can call and execute the corresponding operations of each module.

[0199] This application also provides an electronic device, which may include at least one processor and a memory connected to the processor, wherein:

[0200] Memory is used to store computer programs;

[0201] The processor is used to execute computer programs to enable electronic devices to implement any of the cross-platform data transmission methods for cultural relics provided in the embodiments of this application.

[0202] refer to Figure 4 The diagram illustrates a structural schematic suitable for implementing the electronic device in the embodiments of this application. The electronic device in the embodiments of this application may include, but is not limited to, fixed terminals such as mobile phones, laptops, PDAs (personal digital assistants), PADs (tablet computers), desktop computers, etc. Figure 4 The electronic device shown is merely an example and should not impose any limitation on the functionality and scope of use of the embodiments of this application.

[0203] like Figure 4 As shown, the electronic device may include a processing unit (e.g., a central processing unit, a graphics processing unit, etc.) 601, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 602 or a program loaded from a storage device 608 into a random access memory (RAM) 603. When the electronic device is powered on, the RAM 603 also stores various programs and data required for the operation of the electronic device. The processing unit 601, the ROM 602, and the RAM 603 are interconnected via a bus 604. An input / output (I / O) interface 605 is also connected to the bus 604.

[0204] Typically, the following devices can be connected to the input / output interface 605: input devices 606 including, for example, a touchscreen, touchpad, keyboard, mouse, camera, microphone, accelerometer, gyroscope, etc.; output devices 607 including, for example, a liquid crystal display (LCD), speaker, vibrator, etc.; storage devices 608 including, for example, memory card, hard disk, etc.; and communication devices 609. Communication device 609 allows electronic devices to communicate wirelessly or wiredly with other devices to exchange data. Although Figure 4Electronic devices with various devices are shown, but it should be understood that it is not required to implement or have all of the devices shown. More or fewer devices may be implemented or have alternatively.

[0205] This application also provides a computer program product including computer-readable instructions, which, when executed on an electronic device, cause the electronic device to implement any of the cross-platform data transmission methods for cultural relics provided in this application.

[0206] This application also provides a computer-readable storage medium that carries one or more computer programs. When the one or more computer programs are executed by an electronic device, the electronic device can implement any of the cross-platform data transmission methods for cultural relics provided in this application.

[0207] It should also be noted that the device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. In addition, in the device embodiment drawings provided in this application, the connection relationship between modules indicates that they have a communication connection, which can be implemented as one or more communication buses or signal lines.

[0208] Through the above description of the embodiments, those skilled in the art can clearly understand that this application can be implemented by means of software plus necessary general-purpose hardware, or it can be implemented by special-purpose hardware including application-specific integrated circuits, special-purpose CPUs, special-purpose memory, special-purpose components, etc. Generally, any function performed by a computer program can be easily implemented by corresponding hardware, and the specific hardware structure used to implement the same function can also be diverse, such as analog circuits, digital circuits, or special-purpose circuits. However, for this application, software program implementation is more often the preferred implementation method. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a readable storage medium, such as a computer floppy disk, USB flash drive, mobile hard disk, ROM, RAM, magnetic disk, or optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, training equipment, or network device, etc.) to execute the methods described in the various embodiments of this application.

[0209] In the above embodiments, the implementation can be achieved, in whole or in part, through software, hardware, firmware, or any combination thereof. When implemented in software, it can be implemented, in whole or in part, in the form of a computer program product.

[0210] The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, all or part of the processes or functions described in the embodiments of this application are generated. The computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions may be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions may be transmitted from one website, computer, training device, or data center to another website, computer, training device, or data center via wired (e.g., coaxial cable, fiber optic, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium may be any available medium that a computer can store or a data storage device such as a training device or data center that integrates one or more available media. The available media may be magnetic media (e.g., floppy disks, hard disks, magnetic tapes), optical media (e.g., DVDs), or semiconductor media (e.g., solid-state drives (SSDs)).

Claims

1. A method for cross-platform transmission of cultural relic data, characterized in that, include: Obtain metadata description information and first operating indicators for each platform in the platform set. Each platform is configured with multiple metadata items of cultural relic data to be transmitted, and the metadata description information is the description information of the multiple metadata items. Based on the metadata description information of each platform in the platform set, the characteristics of the metadata items described by the metadata description information are determined, and cross-platform alignment mapping is performed on the metadata items according to the characteristics of the metadata items to obtain a metadata alignment mapping table; the characteristics of the metadata items are obtained based on the semantic embedding vector, structural feature vector and compliance feature vector of the metadata items. Based on the first operating indicators of each platform in the platform set and the metadata alignment mapping table, as well as the information entropy and the policy risk level of each platform obtained in advance, the profile features of each platform are generated, and the platforms are clustered according to their respective profile features to obtain the final clustering results of each platform; the information entropy of each platform is calculated based on the data type of each of the multiple metadata items under that platform. Based on the first operating metric under each platform, the final clustering result, and the metadata alignment mapping table, a data transmission rule set under the future operating metric is generated. The cultural relic data is transmitted across platforms according to the data transmission rule set, and the second operating indicator is monitored in real time during the transmission process. If the change of the second operating indicator compared with the future operating indicator is greater than the preset change threshold, the second operating indicator is used as the first operating indicator, and the process returns to the step of generating the portrait features of each platform.

2. The method for cross-platform transmission of cultural relic data according to claim 1, characterized in that, The step of determining the characteristics of the metadata items described by the metadata description information based on the metadata description information of each platform in the platform set includes: Based on the metadata description information under each platform, determine the characteristics of each metadata item under each platform; The step of performing cross-platform alignment mapping on the metadata items based on their characteristics to obtain a metadata alignment mapping table includes: For every two platforms in the aforementioned platform set: Cross-platform pairing of metadata items under the two platforms yields multiple candidate metadata pairs. The correlation degree of the candidate metadata pair is determined based on the characteristics of the two metadata items in the candidate metadata pair; Based on the relevance of each of the multiple candidate metadata pairs, candidate metadata pairs that maximize the global relevance of the two platforms are selected from the multiple candidate metadata pairs, and the selected candidate metadata pairs are added to the metadata alignment mapping table.

3. The method for cross-platform transmission of cultural relic data according to claim 2, characterized in that, The process of determining the semantic embedding vector, structural feature vector, and compliance feature vector of the metadata item includes: Based on the name and annotation of each metadata item contained in the metadata description information of each platform, determine the semantic embedding vector of each metadata item under each platform; Based on the structural attribute values ​​of each metadata item contained in the metadata description information under each platform, determine the structural feature vector of each metadata item under each platform; Based on the compliance attribute value of each metadata item contained in the metadata description information of each platform, determine the compliance feature vector of each metadata item under each platform.

4. The method for cross-platform transmission of cultural relic data according to claim 2, characterized in that, The step of determining the relevance of the candidate metadata pair based on the characteristics of the two metadata items in the candidate metadata pair includes: Calculate the cosine similarity of the features of the two metadata items in the candidate metadata pair; Obtain the correlation coefficient between the two metadata items in the candidate metadata pair. The correlation coefficient is determined based on the compliance attribute values ​​of the two metadata items in the candidate metadata pair. The correlation coefficient is used to characterize the degree of alignment between the two metadata items in the candidate metadata pair. The cosine similarity is calibrated using the correlation coefficient to obtain the correlation degree of the candidate metadata pair.

5. The method for cross-platform transmission of cultural relic data according to any one of claims 2-4, characterized in that, The step of adding the selected candidate metadata pairs to the metadata alignment mapping table includes: Each selected candidate metadata pair will be used as the target candidate metadata pair; Using a pre-constructed domain knowledge graph, the semantic difference between two metadata items in the target candidate metadata pair is determined as the semantic difference of the target candidate metadata pair. The domain knowledge graph is a knowledge graph constructed based on concepts and relationships in the field of cultural heritage. Determine the structural-functional equivalence of the two metadata items in the target candidate metadata pair in their respective metadata systems, and generate the structural difference degree of the target candidate metadata pair based on the structural-functional equivalence result. Based on the semantic and structural differences of the target candidate metadata pairs, determine whether the two metadata items in the target candidate metadata pairs can be mapped to each other; to obtain all candidate metadata pairs that can be mapped to each other. The candidate metadata pairs that can be mapped to each other are added to the metadata alignment mapping table.

6. The method for cross-platform transmission of cultural relic data according to claim 1, characterized in that, The step of clustering the platforms based on their profile features to obtain the final clustering results for each platform includes: Based on the individual profile characteristics of each platform, multiple clustering algorithms are used to cluster each platform, resulting in clustering results corresponding to each of the multiple clustering algorithms. Based on the clustering results corresponding to the various clustering algorithms, the final clustering results of each platform are determined by combining the consensus algorithm.

7. The method for cross-platform transmission of cultural relic data according to claim 6, characterized in that, The step of determining the characteristics of the metadata items described by the metadata description information based on the metadata description information of each platform in the platform set, and performing cross-platform alignment mapping on the metadata items based on the characteristics of the metadata items to obtain a metadata alignment mapping table includes: Using a pre-trained alignment mapping model, the features of the metadata items described by the metadata description information are determined based on the metadata description information of each platform in the platform set. Based on the features of the metadata items, cross-platform alignment mapping is performed on the metadata items to obtain the metadata alignment mapping table. The process involves generating profile features for each platform based on its first operational metrics and metadata alignment mapping table, as well as pre-obtained information entropy and strategy risk level. Then, based on these profile features, multiple clustering algorithms are used to cluster each platform, yielding clustering results for each algorithm. Finally, a consensus algorithm is used to determine the final clustering result for each platform, based on these clustering results. The final clustering result of each platform is determined by a pre-trained group partitioning model based on the first operating metric of each platform in the platform set, the metadata alignment mapping table, and the information entropy and policy risk level of each platform obtained in advance.

8. The method for cross-platform transmission of cultural relic data according to claim 7, characterized in that, Also includes: The process of acquiring the data transmission rule set adjustment records and the data transmission efficiency evaluation data of the cultural relics data are also included. Based on the adjustment records and the transmission performance evaluation data, and combining the deviations between the prediction results of the alignment mapping model and the group partitioning model and the true labels, the mapping error loss corresponding to the alignment mapping model and the group partitioning model is generated. The alignment mapping model and the group partitioning model are trained respectively based on their respective mapping error losses.

9. An electronic device, characterized in that, It includes at least one processor and a memory connected to the processor, wherein: The memory is used to store computer programs; The processor is used to execute the computer program to enable the electronic device to implement the cross-platform data transmission method for cultural relics as described in any one of claims 1 to 8.

Images