Standard data processing method and system based on knowledge graph

Abstract

The invention relates to the field of data processing, in particular to a standard data processing method and system based on a knowledge graph, and the method comprises the steps: determining a graph supervision strategy of a target management graph based on a dynamic entity proportion and a reference connection index; during entity connection analysis, determining a setting mode of an entity connection set of each dynamic supervision entity according to a connection link richness index and a connection link extension parameter, and determining a determination mode of reference data information according to a set dynamic proportion of each entity connection set and a set update response parameter; during time sequence conflict analysis, determining a determination mode of reference data information according to the updated time sequence difference reference value; and determining an auxiliary evaluation strategy of each update supervision entity based on a determination mode of the reference data information under a reference determination completion condition, and determining whether to send an update information early warning to the user according to the information reliability coefficient, thereby improving the timeliness and accuracy of the update result of the knowledge graph.

Description

Technical Field

[0001] This invention relates to the field of data processing, and in particular to a standard data processing method and system based on knowledge graphs. Background Technology

[0002] Constructing knowledge graphs can more intuitively reflect the inherent relationships of standard data information within an industry. This can effectively improve the comprehensiveness and efficiency of analysis results when analyzing industry standard data. However, the process of constructing a knowledge graph requires acquiring standard data information from multiple sources and in various formats. Standard data is characterized by rapid version iterations, traceable sources, and complex cross-system relationships. Therefore, it is necessary to continuously update and supplement the attributes and inherent relationships of each entity within the constructed knowledge graph. However, the update process requires comparing past construction processes with the continuously added standard data information to determine the update results, leading to a continuous increase in the standard data analysis tasks during the update process. Therefore, how to ensure the quality of knowledge graph updates while maintaining the efficiency of processing the acquired standard data information is a problem that urgently needs to be solved by those skilled in the art.

[0003] Chinese patent application CN116108201A discloses a method for dynamically updating a knowledge graph and its terminal, including the following steps: S1, updating the data source and capturing update information, and storing the captured update information in a message queue; S2, extracting the update information, cleaning and filtering the data, and mapping the data to entity relationships; S3, updating the knowledge graph according to the entity relationships, and determining whether the update is successful; S4, if the update fails, storing the failed data in a retry queue, and repeating steps S2 and S3 until the update is successful, thus completing the knowledge graph update. However, the above solution has the following problems: it fails to determine a targeted data information filtering strategy based on the entities that need to be updated within the constructed knowledge graph, resulting in the update process failing to balance the accuracy of the update results and the efficiency of data information processing. Summary of the Invention

[0004] To address this issue, the present invention provides a standard data processing method and system based on knowledge graphs, which overcomes the problem in the prior art that it fails to determine a targeted standard data information filtering strategy based on the entities that need to be updated within the constructed knowledge graph, resulting in the update process being unable to balance the accuracy of the update results and the efficiency of data information processing.

[0005] To achieve the above objectives, this invention provides a standard data processing method based on knowledge graphs, comprising: The target management map is determined based on the dynamic entity ratio and reference connection index. The map supervision strategy is to update the target management map based on entity connection analysis or time series conflict analysis. During entity link analysis, the setting method of entity link set for each dynamic regulatory entity is determined based on the link link richness index and link link extension parameters. The method of determining reference data information is determined based on the dynamic proportion of each entity link set and the set update response parameters. This method involves matching reference data information based on update interval parameters and auxiliary verification parameters, or information association parameters. When performing time-series conflict analysis, the reference data information for updating regulatory entities is determined by either the updated time-series difference reference value and the entity correlation coefficient, or the updated stability index and the entity correlation coefficient. Under the condition of completion, the auxiliary assessment strategy for each updated regulatory entity is determined based on the method of determining the reference data information. The auxiliary assessment strategy is to determine the information reliability coefficient of the relevant information based on the information overlap parameter and update interval parameter, or the abnormal combination ratio. Whether to send an update alert to the user is determined based on the information reliability coefficient.

[0006] Furthermore, the dynamic index of each target regulatory entity is periodically determined based on the data source richness index and the reference update index, and the category of each target regulatory entity is determined based on the dynamic index. Target regulatory entities with a dynamic index greater than the preset dynamic index are recorded as dynamic regulatory entities. The dynamic entity ratio refers to the proportion of the number of dynamic regulatory entities to the number of target regulatory entities within the target management map; The reference connectivity index is the average value of the entity connectivity indices of each dynamic regulatory entity within the target management map.

[0007] Furthermore, if the proportion of dynamic entities in the target management graph is greater than the preset proportion of dynamic entities or the reference connectivity index is greater than the preset reference connectivity index, then the target management graph is updated based on entity connectivity analysis. Entity connectivity analysis includes: Under the condition of entity update, the setting method of entity link set for each updated regulatory entity is determined based on the link richness index and link extension parameters; The method for determining the reference data information is based on the dynamic proportion of each entity's linked set and the set update response parameters. The entity update condition is the existence of a target regulatory entity within the target management graph that responds to the update condition. The target regulatory entity that responds to the update condition is recorded as the updated regulatory entity.

[0008] Furthermore, for a single updated regulatory entity, if the dynamic linkage coefficient is greater than the preset dynamic linkage coefficient, the entity linkage set of the updated regulatory entity is determined based on the update association index and the update radiation index. If the dynamic linkage coefficient is less than or equal to the preset dynamic linkage coefficient, the entity linkage set of the updated regulatory entity is determined based on the relevant construction overlap index and the combined construction relevant index. The dynamic connectivity coefficient is determined based on the connectivity link richness index and the connectivity link extension parameter.

[0009] Furthermore, if the dynamic proportion of a set of entity connections is greater than the preset dynamic proportion of a set or the set update response parameter is greater than the preset set update response parameter, then the reference data information is matched for the updated regulatory entity corresponding to the entity connection set according to the update interval parameter and the auxiliary verification parameter to determine the reference data information. The reference data information refers to the information to be evaluated whose reference coefficient is greater than the preset reference coefficient. The reference coefficient is positively correlated with the update interval parameter and the auxiliary verification parameter.

[0010] Furthermore, if the dynamic proportion of a set of entity connections is less than or equal to the preset dynamic proportion of the set and the set update response parameter is less than or equal to the preset set update response parameter, then the reference data information is matched for the updated regulatory entity corresponding to the entity connection set according to the information association parameter to determine the reference data information. The reference data information is the information to be evaluated whose information association parameter is greater than the preset information association parameter.

[0011] Furthermore, if the proportion of dynamic entities in the target management graph is less than or equal to the preset proportion of dynamic entities and the reference connectivity index is less than or equal to the preset reference connectivity index, then the target management graph is updated based on temporal conflict analysis. Temporal conflict analysis includes: Under the condition of entity update, the reference value of the update time sequence difference for each updated regulatory entity is detected; The method for determining the reference data information for each updated regulatory entity is based on the reference value of the update time sequence difference.

[0012] Furthermore, for a single updated regulatory entity, if the update time series difference reference value of the updated regulatory entity is less than or equal to the preset update time series difference reference value, the reference data information of the updated regulatory entity is determined based on the update time series difference value and the entity correlation coefficient. If the update time series difference reference value of the updated regulatory entity is greater than the preset update time series difference reference value, the reference data information of the updated regulatory entity is determined based on the update stability index and the entity correlation coefficient.

[0013] Furthermore, under the condition that the reference data information of an updated regulatory entity is determined according to the update interval parameter and the auxiliary verification parameter, the information reliability coefficient of the construction-related information of the updated regulatory entity is determined according to the information overlap parameter and the update interval parameter. If the reference data information of an updated regulatory entity is determined based on information association parameters, or determined through time-series conflict analysis, then the information reliability coefficient of the construction-related information of the updated regulatory entity is determined based on the abnormal combination ratio. If the reliability coefficient of the information related to the construction of the updated regulatory entity is less than the preset reliability coefficient, an update information warning will be sent to the user. The condition for determining the reference is that there is a reference standard data information for any updated regulatory entity.

[0014] This invention also provides a standard data processing system based on knowledge graphs, comprising: The graph supervision module is used to determine the graph supervision strategy of the target management graph based on the dynamic entity ratio and reference connection index. The graph supervision strategy is to update the target management graph based on entity connection analysis or time series conflict analysis. The connection analysis module, which is connected to the graph monitoring module, is used to determine the setting method of the entity connection set based on the connection link richness index and the connection link extension parameter, and to determine the reference data information based on the dynamic proportion of the set and the set update response parameter. The determination method is to match the reference data information based on the update interval parameter and the auxiliary verification parameter, or the information association parameter. The time series analysis module, which is connected to the map monitoring module, is used to determine the reference data information based on the updated time series difference reference value. The determination method is to determine the reference data information of the updated monitoring entity based on the updated time series difference value and the entity correlation coefficient, or the updated stability index and the entity correlation coefficient. The auxiliary assessment module, which is connected to the connection analysis module and the time series analysis module, is used to determine the auxiliary assessment strategy for each updated regulatory entity based on the determination method of reference data information. This strategy is based on the information overlap parameter and update interval parameter, or the abnormal combination ratio, to determine the information reliability coefficient of the relevant information.

[0015] Compared with the prior art, the beneficial effects of the present invention are as follows: the technical solution of the present invention determines the graph supervision strategy of the target management graph based on the dynamic entity ratio and reference connection index of the target management graph, obtains standard data information for updating the attributes and internal relationships of each entity according to the targeted graph supervision strategy, and effectively filters the standard data information to ensure the quality of knowledge graph updates while ensuring the processing efficiency of the obtained standard data information. The present invention improves the timeliness and accuracy of knowledge graph update results.

[0016] Furthermore, in this invention, the category of the target regulatory entity is determined based on the data source richness index and the reference update index, thereby determining the dynamic entity ratio and the reference connection index of the target management map to characterize the frequency of update tasks to be performed within the target management map and the data analysis burden. In turn, a targeted map supervision strategy is set to ensure the quality of standard data information screening during the update process while ensuring the data analysis efficiency of the update process.

[0017] Furthermore, in this invention, when the proportion of dynamic entities in the target regulatory graph is greater than the preset proportion of dynamic entities or the reference connection index is greater than the preset reference connection index, the setting method of the entity connection set of each updated regulatory entity is determined according to the dynamic connection coefficient, and a targeted data matching method is determined based on the dynamic proportion of the set and the set update response parameters. This ensures that the correlation between each updated regulatory entity and other target regulatory entities used for matching reference data information is optimal, thereby guaranteeing the reference value of the determined matching reference data information. This invention improves the accuracy of the knowledge graph update results.

[0018] Furthermore, in this invention, when the proportion of dynamic entities in the target management graph is less than or equal to the preset proportion of dynamic entities and the reference connection index is less than or equal to the preset reference connection index, the updated time-series difference reference value is used to characterize whether the updated regulatory entity has completed auxiliary verification data in the near future. This determines the method for determining targeted reference data, making the process of determining reference data information more consistent with the actual situation of updating regulatory entities, ensuring the effectiveness of the obtained reference data information, and thus improving the accuracy of the knowledge graph update results.

[0019] Furthermore, this invention determines the auxiliary evaluation strategy for each updated regulatory entity based on the determination method of reference data information, so as to obtain the update warning coefficient of each updated regulatory entity, and to determine whether it is necessary to send a warning to the user for the update result. By evaluating the reliability of the construction-related information of the updated regulatory entity through reference data information, this invention improves the accuracy of the update results of the knowledge graph. Attached Figure Description

[0020] Figure 1 This is a schematic diagram of the standard data processing method based on knowledge graphs according to the present invention; Figure 2 This is a flowchart of the graph supervision strategy of the present invention, which determines the target management graph based on dynamic entity proportion and reference connection index; Figure 3 This is a flowchart illustrating the method for setting the entity connection set of a dynamically monitored entity based on the dynamic connection coefficient according to the present invention. Figure 4This is a module connection diagram of the standard data processing system based on knowledge graphs according to the present invention. Detailed Implementation

[0021] To make the objectives and advantages of the present invention clearer, the present invention will be further described below with reference to embodiments; it should be understood that the specific embodiments described herein are merely for explaining the present invention and are not intended to limit the present invention.

[0022] Preferred embodiments of the present invention will now be described with reference to the accompanying drawings. Those skilled in the art should understand that these embodiments are merely illustrative of the technical principles of the present invention and are not intended to limit the scope of protection of the present invention.

[0023] It should be noted that in the description of this invention, the terms "upper", "lower", "left", "right", "inner", "outer", etc., which indicate directions or positional relationships, are based on the directions or positional relationships shown in the accompanying drawings. This is only for the convenience of description and is not intended to indicate or imply that the device or element must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, it should not be construed as a limitation of this invention.

[0024] Furthermore, it should be noted that, in the description of this invention, unless otherwise explicitly specified and limited, the terms "installation," "connection," and "linking" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection of two components. Those skilled in the art can understand the specific meaning of the above terms in this invention according to the specific circumstances.

[0025] Please see Figures 1 to 3 As shown, this invention provides a standard data processing method based on knowledge graphs, including: The target management map is determined based on the dynamic entity ratio and reference connection index. The map supervision strategy is to update the target management map based on entity connection analysis or time series conflict analysis. During entity link analysis, the setting method of entity link set for each dynamic regulatory entity is determined based on the link link richness index and link link extension parameters. The method of determining reference data information is determined based on the dynamic proportion of each entity link set and the set update response parameters. This method involves matching reference data information based on update interval parameters and auxiliary verification parameters, or information association parameters. When performing time-series conflict analysis, the reference data information for updating regulatory entities is determined by either the updated time-series difference reference value and the entity correlation coefficient, or the updated stability index and the entity correlation coefficient. Under the condition of completion, the auxiliary assessment strategy for each updated regulatory entity is determined based on the method of determining the reference data information. The auxiliary assessment strategy is to determine the information reliability coefficient of the relevant information based on the information overlap parameter and update interval parameter, or the abnormal combination ratio. Whether to send an update alert to the user is determined based on the information reliability coefficient.

[0026] This invention is used to supervise the updating process of the attributes and internal relationships of entities within a knowledge graph. The knowledge graph currently requiring supervision is designated as the target management graph, and all entities within the target management graph are designated as target supervised entities. The construction of the target management graph in this invention is based on various standard data information, including but not limited to national and industry standards. For a single target supervised entity, the triples existing within the target management graph for that target supervised entity are obtained and designated as the relevant triples for that target supervised entity. The data information upon which the construction and updating of the relevant triples for that target supervised entity are based is designated as the construction-related information of that target supervised entity. Each triple contains two definite values. Data information on the attribute relationships between entities. If a data information contains content that determines the entities and their relationships within the triple, then the triple is constructed or updated based on the data information. The types of data information that can be used to construct or update the triple include, but are not limited to: structured data, unstructured data, and semi-structured data. The forms of data information include, but are not limited to: text data, image data, video data, and XML files. In this invention, each data information is set with a timestamp to represent the last update time of each data information. Some data information is also set with an expiration timestamp to represent the expiration time of each data information. How to determine the entities existing in the target management map is a content that is easy for those skilled in the art to understand, and will not be elaborated here. This invention utilizes several graph monitoring records. Each graph monitoring record documents at least one update process for monitoring the attributes and internal relationships of entities within the knowledge graph. This process includes dynamic indices, dynamic entity proportions, reference connection indices, dynamic connection coefficients, set matching coefficients, set update response parameters, set dynamic proportions, information reference coefficients, update time sequence difference reference values, reference evaluation coefficients, and information reliability coefficients. Each graph monitoring record also has a corresponding qualification mark. The qualification mark indicates whether the timeliness and accuracy of the knowledge graph update results meet user requirements. It is understood that users can determine whether the timeliness and accuracy of the knowledge graph update results meet their needs based on self-defined indicators. For example, self-defined indicators can include, but are not limited to, update execution parameters, which are the average time taken for each target monitored entity from responding to update conditions to completing the update.

[0027] Specifically, the dynamic index of each target regulatory entity is periodically determined based on the data source richness index and the reference update index, and the category of each target regulatory entity is determined based on the dynamic index. Target regulatory entities with a dynamic index greater than the preset dynamic index are recorded as dynamic regulatory entities. The dynamic entity ratio refers to the proportion of the number of dynamic regulatory entities to the number of target regulatory entities within the target management map; The reference connectivity index is the average value of the entity connectivity indices of each dynamic regulatory entity within the target management map.

[0028] In this invention, a cyclic entity supervision cycle is applied. The duration of the entity supervision cycle can be determined by the user. The higher the user's requirements for the timeliness and accuracy of the knowledge graph update results, the shorter the duration of the entity supervision cycle. One entity supervision cycle duration is 5 days. At the end of each entity supervision cycle, the dynamic index of each target supervision entity is determined based on the data source richness index and the reference update index. The category of each target supervision entity is determined based on the dynamic index. For a single target supervision entity, if the dynamic index of the target supervision entity is greater than the preset dynamic index, the target supervision entity is recorded as a dynamic supervision entity. If the dynamic index of the target supervision entity is less than or equal to the preset dynamic index, the target supervision entity is recorded as a steady-state supervision entity. For a single target regulatory entity, the dynamic index is the sum of the data source richness index and the reference update index. The data source richness is the number of construction-related information of the target regulatory entity. The reference update index is the number of times the target regulatory entity completes updates within the preset dynamic evaluation phase. The end time of the preset dynamic evaluation phase is the end time of the current entity regulatory cycle. The duration of the preset dynamic evaluation phase can be determined by the user according to the actual work scenario. For example, the user can set it according to the knowledge graph regulatory records. The higher the user's requirements for the timeliness and accuracy of the knowledge graph update results, the larger the value of the preset dynamic evaluation phase duration. One possible value for the preset dynamic evaluation phase duration is ten times the duration of the entity regulatory cycle. The value of the preset dynamic index can be determined by the user based on the actual work scenario. For example, the user can set it based on the graph supervision record. The higher the user's requirements for the timeliness and accuracy of the knowledge graph update results, the smaller the preset dynamic index. A method for determining the value of the preset dynamic index is provided, which is the minimum value of the dynamic index of each dynamic supervision entity in the graph supervision record that meets the requirements for the timeliness and accuracy of the knowledge graph update results.

[0029] Specifically, when the proportion of dynamic entities in the target management graph is greater than the preset proportion of dynamic entities or the reference connectivity index is greater than the preset reference connectivity index, the target management graph is updated based on entity connectivity analysis. Entity connectivity analysis includes: Under the condition of entity update, the setting method of entity link set for each updated regulatory entity is determined based on the link richness index and link extension parameters; The method for determining the reference data information is based on the dynamic proportion of each entity's linked set and the set update response parameters. The entity update condition is the existence of a target regulatory entity within the target management graph that responds to the update condition. The target regulatory entity that responds to the update condition is recorded as the updated regulatory entity.

[0030] Wherein, the dynamic entity ratio = the number of dynamic regulatory entities within the target management graph / the number of target regulatory entities within the target management graph. For a single dynamic regulatory entity, the entity connection index is the number of related triples of that dynamic regulatory entity. The values ​​of the preset dynamic entity ratio and the preset reference connection index can be determined by the user according to the actual work scenario. For example, the user can set them according to the graph supervision records. The higher the user's requirements for the timeliness and accuracy of the knowledge graph update results, the smaller the value of the preset dynamic entity ratio and the smaller the value of the preset reference connection index. A method for determining the value of the preset dynamic entity ratio is provided, in which the graph supervision records updated based on entity connection analysis for the target management graph are recorded as analysis reference records, and the average value of the dynamic entity ratio in the analysis reference records that meet the user's requirements for the timeliness and accuracy of the knowledge graph update results is recorded as the preset dynamic entity ratio. A method for determining the value of the preset reference connection index is provided, in which the average value of the reference connection index in the analysis reference records that meet the user's requirements for the timeliness and accuracy of the knowledge graph update results is recorded as the preset reference connection index. This invention employs a cyclical update monitoring cycle. The duration of the update monitoring cycle can be determined by the user. The higher the user's requirements for the timeliness and accuracy of the knowledge graph update results, the shorter the update monitoring cycle. One update monitoring cycle duration is provided: 10 hours. At the end of each update monitoring cycle, it is determined whether each target monitoring entity within the target management graph has responded to the update conditions. For a single target monitoring entity, if the timestamp corresponding to the construction-related information of the target monitoring entity changes or the hash value of the relevant triplet of the target monitoring entity changes, then the target monitoring entity is determined to have responded to the update conditions. For a single relevant triplet, each update monitoring cycle... At the end of the current regulatory cycle, the timestamp of the construction-related information corresponding to the relevant triple, the hash value of the relevant triple within the target management graph, and the hash value of the relevant triple within the corresponding construction-related information are obtained. If the hash value of the relevant triple within the target management graph is different from the hash value of the relevant triple within the corresponding construction-related information, it is determined that the hash value of the relevant triple has changed. If the timestamp of the construction-related information obtained at the end of the current regulatory cycle is different from the timestamp of the end of the previous regulatory cycle, it is determined that the timestamp corresponding to the construction-related information of the target regulatory entity has changed. How to determine the hash value of the triple is a topic easily understood by those skilled in the art and will not be elaborated here.

[0031] Specifically, for a single updated regulatory entity, if the dynamic linkage coefficient is greater than the preset dynamic linkage coefficient, the entity linkage set of the updated regulatory entity is determined based on the update association index and the update radiation index. If the dynamic linkage coefficient is less than or equal to the preset dynamic linkage coefficient, the entity linkage set of the updated regulatory entity is determined based on the relevant construction overlap index and the combined construction relevant index. The dynamic connectivity coefficient is determined based on the connectivity link richness index and the connectivity link extension parameter.

[0032] Specifically, for a single updated regulatory entity, the dynamic link coefficient = ln(link richness index × link extension parameter), where the link richness index is the number of link connections within the target management graph for the updated regulatory entity, and the link extension parameter is the average number of target regulatory entities contained in each link of the updated regulatory entity within the target management graph. Each link is a set of several target regulatory entities, and the determined links are not completely identical. For a single link, any target regulatory entity contained in the link must form a triple with at least one other target regulatory entity in the link and at most two other target regulatory entities in the link. The value of the preset dynamic connection coefficient can be determined by the user according to the actual work scenario. For example, the user can set it according to the graph supervision record. The higher the user's requirements for the timeliness and accuracy of the knowledge graph update results, the smaller the value of the preset dynamic connection coefficient. A method for determining the value of the preset dynamic connection coefficient is provided, which records the graph supervision records updated for the target management graph based on entity connection analysis as set reference records, and records the minimum value of the dynamic connection coefficient of the updated supervision entity in the set reference records that meet the user's requirements for the timeliness and accuracy of the knowledge graph update results as the preset dynamic connection coefficient. For a single updated regulatory entity, the target regulatory entity with a link correlation coefficient less than the preset link correlation coefficient is recorded as the link-related entity of the updated regulatory entity. For a single target regulatory entity, the connection link that exists between the target regulatory entity and the updated regulatory entity is recorded as the related link. If there is a related link, the link correlation coefficient is the minimum value of the number of target regulatory entities that exist between the target regulatory entity and the updated regulatory entity in each related link. If there is no related link, the link correlation coefficient of the target regulatory entity is 0. The value of the preset link correlation coefficient can be determined by the user according to the actual working scenario. For example, the user can set it according to the graph regulatory record. The higher the user's requirements for the timeliness and accuracy of the knowledge graph update results, the smaller the value of the preset link correlation coefficient. A method for determining the value of the preset link correlation coefficient is provided, which records the maximum value of the link correlation coefficient of each related link entity in the graph regulatory record that meets the user's requirements for the timeliness and accuracy of the knowledge graph update results as the preset link correlation coefficient. If the dynamic link coefficient of an updated regulatory entity is greater than a preset dynamic link coefficient, then the entity link set of the updated regulatory entity is determined based on the update association index and the update radiation index. The set matching coefficient of each link-related entity of the updated regulatory entity is determined based on the update association index and the update radiation index. The set matching coefficient of any link-related entity within the entity link set of the updated regulatory entity is greater than a first preset set matching coefficient. For a single link-related entity, the set matching coefficient = ln(update association index × update radiation index). The update association index is the number of times the link-related entity and the updated regulatory entity have associated updates within a preset dynamic evaluation phase. For any two target regulatory entities, if the interval between the times when the two target regulatory entities respond to the update condition is less than a preset associated update duration, then it is determined that the two target regulatory entities have an associated update. In this context, the update radiation index is the average number of radiation update entities within the radiation assessment phase after each response update condition time for the related entities in the link. For a single response update condition of the related entity in the link, the radiation update entity is the target regulatory entity responding to the update condition within the radiation assessment phase of the related link between the related entity and the updating regulatory entity. The start time of the radiation assessment phase is the time of the response update condition of the related entity in the link. The duration of the radiation assessment phase is a preset association duration. The value of the preset association update duration can be determined by the user according to the actual work scenario. For example, the user can set it according to the knowledge graph supervision record. The higher the user's requirements for the timeliness and accuracy of the knowledge graph update results, the larger the value of the preset association update duration. One preset association update duration value is provided, which is 10 days. For a single updated regulatory entity, if the dynamic link coefficient is less than or equal to a preset dynamic link coefficient, the entity link set of the updated regulatory entity is determined based on the relevant construction overlap index and the combined construction correlation index. The set matching coefficient of each link-related entity of the updated regulatory entity is determined based on the relevant construction overlap index and the combined construction correlation index. The set matching coefficient of any link-related entity within the entity link set of the updated regulatory entity is greater than a second preset set matching coefficient. For a single link-related entity, the set matching coefficient = ln(relevant construction overlap index × combined construction correlation index), where the relevant construction overlap index is the number of overlapping related links of the link-related entity. For any link link containing the link-related entity, if the link link contains any target regulatory entity involved in the relevant triple of the updated regulatory entity, then the link link is determined to be valid. The path is an overlapping related link. The combined construction related index is the number of change-related information in the related construction information of the related entity of the link and the related construction information of the updated regulatory entity. For any related construction information of the related entity of the link and any related construction information of the updated regulatory entity, if the interval between the timestamps of the two related construction information changes is less than the preset information change related duration, the two related construction information are determined to be change-related information. The value of the preset information change related duration can be determined by the user according to the actual work scenario. For example, the user can set it according to the graph supervision record. The higher the user's requirements for the timeliness and accuracy of the knowledge graph update results, the smaller the value of the preset information change related duration. One preset information change related duration value is provided, which is 20h. The values ​​of the first preset set matching coefficient and the second preset set matching coefficient can be determined by the user according to the actual work scenario. For example, the user can set them according to the graph supervision record. The higher the user's requirements for the timeliness and accuracy of the knowledge graph update results, the larger the value of the first preset set matching coefficient and the second preset set matching coefficient. A method for determining the value of the first preset set matching coefficient is provided, in which the graph supervision record that determines the entity connection set based on the update association index and the update radiation index is recorded as the first partitioning reference record, and the average value of the set matching coefficients of each link-related entity in the entity connection set of the first partitioning reference record that meets the user's requirements for the timeliness and accuracy of the knowledge graph update results is recorded as the first preset set matching coefficient. A method for determining the value of the second preset set matching coefficient is provided, in which the graph supervision record that determines the entity connection set based on the correlation construction overlap index and the combination construction correlation index is recorded as the second partitioning reference record, and the average value of the set matching coefficients of each link-related entity in the entity connection set of the second partitioning reference record that meets the user's requirements for the timeliness and accuracy of the knowledge graph update results is recorded as the second preset set matching coefficient.

[0033] Specifically, if the dynamic proportion of a set of entity connections is greater than the preset dynamic proportion of the set or the set update response parameter is greater than the preset set update response parameter, then the reference data information is matched for the updated regulatory entity corresponding to the entity connection set according to the update interval parameter and the auxiliary verification parameter to determine the reference data information. The reference data information refers to the information to be evaluated whose reference coefficient is greater than the preset reference coefficient. The reference coefficient is positively correlated with the update interval parameter and the auxiliary verification parameter.

[0034] Specifically, for a single entity connection set, the dynamic proportion of the set = the number of dynamically monitored entities in the entity connection set / the number of target monitored entities in the entity connection set. The set update response parameter is the number of updated radiation entities in the entity connection set. For any target monitored entity in the entity connection set, if the time interval between the target monitored entity and the corresponding updated monitored entity in the entity connection set responding to the update condition is less than the preset response interval, then the target monitored entity is determined to be an updated radiation entity. The value of the preset response interval can be determined by the user according to the actual work scenario. For example, the user can set it according to the graph monitoring record. The higher the user's requirements for the timeliness and accuracy of the knowledge graph update results, the smaller the value of the preset information change related time. One preset information change related time value is provided, which is 30h. The user can determine the values ​​of the preset set dynamic ratio and the preset set update response parameter according to the actual work scenario. For example, the user can set them according to the graph supervision record. The higher the user's requirements for the timeliness and accuracy of the knowledge graph update results, the larger the value of the preset set dynamic ratio and the larger the value of the preset set update response parameter. A method for determining the value of the preset set dynamic ratio is provided, in which the graph supervision record that matches the updated supervision entity corresponding to the entity connection set according to the information association parameter is recorded as the second-class matching reference record, and the maximum value of the set dynamic ratio in the second-class matching reference record that meets the user's requirements for the timeliness and accuracy of the knowledge graph update results is recorded as the preset set dynamic ratio. A method for determining the value of the preset set update response parameter is provided, in which the maximum value of the set update response parameter in the second-class matching reference record that meets the user's requirements for the timeliness and accuracy of the knowledge graph update results is recorded as the preset set update response parameter. For a single entity connection set, all relevant construction information of the target regulatory entities within the entity connection set, excluding the updated regulatory entities, is recorded as information to be evaluated. If the dynamic proportion of the entity connection set is greater than the preset dynamic proportion or the set update response parameter is greater than the preset set update response parameter, it indicates that there are many target regulatory entities that have recently completed updates within the entity connection set, meaning there is a lot of trustworthy data information to assist in the judgment. The reference data information of the entity connection set is determined based on the update interval parameter and the auxiliary verification parameter. For a single piece of information to be evaluated, the information reference coefficient of the information to be evaluated is determined according to the update interval parameter and the auxiliary verification parameter. The information reference coefficient is the product of the update interval parameter and the auxiliary influence coefficient. The update interval parameter is the time interval between the time when the information to be evaluated is used to determine the response condition of the target regulatory entity and the time when the updated regulatory entity responds to the update condition. The unit of the update interval parameter is hours. The auxiliary influence coefficient is positively correlated with the auxiliary verification parameter, and the auxiliary influence coefficient is less than 1. The auxiliary verification parameter is the number of times the information to be evaluated is used as the reference data information of the updated regulatory entity within the preset dynamic evaluation stage of the updated regulatory entity. The value of the preset information reference coefficient can be determined by the user according to the actual work scenario. For example, the user can set it according to the graph supervision record. The higher the user's requirements for the timeliness and accuracy of the knowledge graph update results, the larger the value of the preset information reference coefficient. A method for determining the value of the preset information reference coefficient is provided, which records the graph supervision records that match the reference data information of the updated supervision entity corresponding to the entity connection set according to the update interval parameter and the auxiliary verification parameter are recorded as a type of matching reference record. The average value of the information reference coefficients of the reference data information in the type of matching reference records that meet the user's requirements for the timeliness and accuracy of the knowledge graph update results is recorded as the preset information reference coefficient.

[0035] Specifically, if the dynamic proportion of a set of entity associations is less than or equal to the preset dynamic proportion of the set and the set update response parameter is less than or equal to the preset set update response parameter, then the reference data information is matched for the updated regulatory entity corresponding to the entity association set according to the information association parameter to determine the reference data information. The reference data information is the information to be evaluated whose information association parameter is greater than the preset information association parameter.

[0036] For a single entity connection set, if the dynamic proportion of the set is less than or equal to the preset dynamic proportion and the set update response parameter is less than or equal to the preset update response parameter, it indicates that there are few target regulatory entities that have been updated recently in the entity connection set. That is, the relevant construction information of each target regulatory entity has not been verified by other data information for a long time. It is necessary to detect the conflict between the relevant construction information of the target regulatory entities in the entity connection set and the relevant construction information of the updated regulatory entities, in order to determine whether the response update conditions determined by the relevant construction information are reliable. Therefore, the target regulatory entities included in the standard data information used for auxiliary verification need to have a certain overlap with the target regulatory entities included in the relevant construction information of the updated regulatory entities in the entity connection set. For a single piece of information to be evaluated, the information association parameter is the number of target regulatory entities that exist in both the relevant construction information of the updated regulatory entities corresponding to the entity connection set. The value of the preset information association parameter can be determined by the user according to the actual work scenario. For example, the user can set it according to the knowledge graph supervision record. The higher the user's requirements for the timeliness and accuracy of the knowledge graph update results, the larger the value of the preset information association parameter. A method for determining the value of the preset information association parameter is provided, which is the average value of the information association parameter of the reference data information in the second-class matching reference record that meets the user's requirements for the timeliness and accuracy of the knowledge graph update results.

[0037] Specifically, when the proportion of dynamic entities in the target management graph is less than or equal to the preset proportion of dynamic entities and the reference connectivity index is less than or equal to the preset reference connectivity index, the target management graph is updated based on temporal conflict analysis. Temporal conflict analysis includes: Under the condition of entity update, the reference value of the update time sequence difference for each updated regulatory entity is detected; The method for determining the reference data information for each updated regulatory entity is based on the reference value of the update time sequence difference.

[0038] Specifically, for a single updated regulatory entity, if the update time difference reference value of the updated regulatory entity is less than or equal to the preset update time difference reference value, the reference data information of the updated regulatory entity is determined based on the update time difference value and the entity correlation coefficient. If the update time series difference reference value of the updated regulatory entity is greater than the preset update time series difference reference value, the reference data information of the updated regulatory entity is determined based on the update stability index and the entity correlation coefficient.

[0039] Specifically, for a single updated regulatory entity, the update time sequence difference reference value is the interval between the time when the updated regulatory entity responds to the update condition and the time when each link-related entity of the updated regulatory entity most recently responds to the update condition. The value of the preset update time sequence difference reference value can be determined by the user according to the actual working scenario. For example, the user can set it according to the graph regulatory records. The higher the user's requirements for the timeliness and accuracy of the knowledge graph update results, the smaller the value of the preset update time sequence difference reference value. A method for determining the value of the preset update time sequence difference reference value is provided, which records the graph regulatory records that determine the reference data information of the updated regulatory entity based on the update time sequence difference value and the entity correlation coefficient are recorded as a type of difference reference record. The maximum value of the update time sequence difference reference value in the type of difference reference records that meets the user's requirements for the timeliness and accuracy of the knowledge graph update results is recorded as the preset update time sequence difference reference value. For a single updated regulatory entity, the construction-related information of each link-related entity of the updated regulatory entity is recorded as the information to be matched. If the update time sequence difference reference value of the updated regulatory entity is less than or equal to the preset update time sequence difference reference value, the reference data information is the information to be matched with a reference evaluation coefficient greater than the preset first-class evaluation coefficient. The reference evaluation coefficient is determined based on the update time sequence difference value and the entity correlation coefficient. For a single information to be matched, the reference evaluation coefficient is the product of the update time sequence difference value and the entity correlation coefficient. The entity correlation coefficient is the minimum number of target regulatory entities between the target regulatory entity corresponding to the information to be matched and the updated regulatory entity in each common connection link. The update time sequence difference value is the interval between the time when the target regulatory entity corresponding to the information to be matched last responded to the update condition and the time when the updated regulatory entity responded to the update condition. For a single updated regulatory entity, if the update time series difference reference value of the updated regulatory entity is greater than a preset update time series difference reference value, the reference data information is the matching information whose reference evaluation coefficient is greater than a preset second-type evaluation coefficient. The reference evaluation coefficient is determined based on the update stability index and the entity correlation coefficient. For a single matching information, the reference evaluation coefficient is the product of the update stability index and the entity correlation coefficient. , n is the number of times the timestamp of the information to be matched changes within the preset dynamic evaluation phase of the updated regulatory entity, yi is the duration of the change interval of the i-th timestamp change interval of the information to be matched within the preset dynamic evaluation phase of the updated regulatory entity, y0 is the average duration of the change interval of each timestamp change interval of the information to be matched within the preset dynamic evaluation phase of the updated regulatory entity, the timestamp change interval is the interval formed between two adjacent timestamp changes, and the change interval duration is the interval duration between two adjacent timestamp changes; The values ​​of the preset first-class evaluation coefficient and the preset second-class evaluation coefficient can be determined by the user according to the actual work scenario. For example, the user can set them according to the knowledge graph supervision records. The higher the user's requirements for the timeliness and accuracy of the knowledge graph update results, the larger the value of the preset first-class evaluation coefficient and the larger the value of the preset second-class evaluation coefficient. A method for determining the value of the preset first-class evaluation coefficient is provided, in which the minimum value of the reference evaluation coefficient of each reference data information in the first-class difference reference record that meets the user's requirements for the timeliness and accuracy of the knowledge graph update results is recorded as the preset first-class evaluation coefficient. A method for determining the value of the preset second-class evaluation coefficient is provided, in which the graph supervision record that determines the reference data information of the updated supervision entity based on the update stability index and the entity correlation coefficient is recorded as the second-class difference reference record, and the minimum value of the reference evaluation coefficient of each reference data information in the second-class difference reference record that meets the user's requirements for the timeliness and accuracy of the knowledge graph update results is recorded as the preset second-class evaluation coefficient.

[0040] Specifically, under the condition that the reference data information of an updated regulatory entity is determined according to the update interval parameter and the auxiliary verification parameter, the information reliability coefficient of the construction-related information of the updated regulatory entity is determined according to the information overlap parameter and the update interval parameter. If the reference data information of an updated regulatory entity is determined based on information association parameters, or determined through time-series conflict analysis, then the information reliability coefficient of the construction-related information of the updated regulatory entity is determined based on the abnormal combination ratio. If the reliability coefficient of the information related to the construction of the updated regulatory entity is less than the preset reliability coefficient, an update information warning will be sent to the user. The condition for determining the reference is that there is reference data information for any updated regulatory entity.

[0041] Specifically, for a single updated regulatory entity, if the reference data information of the updated regulatory entity is determined based on the update interval parameter and auxiliary verification parameter, the information reliability coefficient is the sum of the products of the information overlap parameter of each reference data information and the corresponding reliability evaluation coefficient. For a single reference data information, the information overlap parameter is the number of target regulatory entities that exist in both the reference data information and the related construction information of the updated regulatory entity. The reliability evaluation coefficient is negatively correlated with the update interval parameter of the reference data information. The reliability evaluation coefficient of each reference data information is less than 1, and the sum of the reliability evaluation coefficients of each reference data information is 1. For a single updated regulatory entity, if the reference data information of the updated regulatory entity is determined according to information association parameters, or determined through time-series conflict analysis, the information reliability coefficient is positively correlated with the abnormal combination ratio of the updated regulatory entity. The abnormal combination ratio = the number of triples determined by the relevant construction information of the updated regulatory entity / the number of abnormal triples. For a single triple, the hash value of the triple in each reference data information involving the triple is the same as the hash value of the triple determined by the relevant construction information. If the number of reference data information that is different from the hash value of the triple determined by the relevant construction information is greater than the preset abnormal assessment number, then the triple is determined to be an abnormal triple. The value of the preset abnormal assessment number can be determined by the user according to the actual working scenario. For example, the user can set it according to the graph regulatory record. The higher the user's requirements for the timeliness and accuracy of the knowledge graph update results, the smaller the value of the preset abnormal assessment number. One preset abnormal assessment number is provided, which is 3. The value of the preset information reliability coefficient can be determined by the user based on the actual working scenario. For example, the user can set it based on the graph monitoring records. The higher the user's requirements for the timeliness and accuracy of the knowledge graph update results, the larger the value of the preset information reliability coefficient. A method for determining the value of the preset information reliability coefficient is provided, in which the graph monitoring record that sends update information warnings to the user is recorded as the warning reference record, and the maximum value of the information reliability coefficient in the warning reference record that meets the user's requirements for the timeliness and accuracy of the knowledge graph update results is recorded as the preset information reliability coefficient.

[0042] Please see Figure 4 The diagram shown is a module connection diagram of the standard data processing system based on knowledge graphs according to the present invention. The present invention also provides a standard data processing system based on knowledge graphs, comprising: The graph supervision module is used to determine the graph supervision strategy of the target management graph based on the dynamic entity ratio and reference connection index. The graph supervision strategy is to update the target management graph based on entity connection analysis or time series conflict analysis. The connection analysis module, which is connected to the graph monitoring module, is used to determine the setting method of the entity connection set based on the connection link richness index and the connection link extension parameter, and to determine the reference data information based on the dynamic proportion of the set and the set update response parameter. The determination method is to match the reference data information based on the update interval parameter and the auxiliary verification parameter, or the information association parameter. The time series analysis module, which is connected to the map monitoring module, is used to determine the reference data information based on the updated time series difference reference value. The determination method is to determine the reference data information of the updated monitoring entity based on the updated time series difference value and the entity correlation coefficient, or the updated stability index and the entity correlation coefficient. The auxiliary assessment module, which is connected to the connection analysis module and the time series analysis module, is used to determine the auxiliary assessment strategy for each updated regulatory entity based on the determination method of reference data information. This strategy is based on the information overlap parameter and update interval parameter, or the abnormal combination ratio, to determine the information reliability coefficient of the relevant information.

[0043] The technical solution of the present invention has been described above with reference to the preferred embodiments shown in the accompanying drawings. However, it will be readily understood by those skilled in the art that the scope of protection of the present invention is obviously not limited to these specific embodiments. Without departing from the principles of the present invention, those skilled in the art can make equivalent changes or substitutions to the relevant technical features, and the technical solutions after these changes or substitutions will all fall within the scope of protection of the present invention.

[0044] The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A knowledge graph-based standard data processing method, characterized in that, The method comprises the following steps: determining a graph management strategy of the target management graph based on the dynamic entity proportion and the reference connection index, wherein the graph management strategy is based on entity connection analysis or time sequence conflict analysis to update the target management graph; in the entity connection analysis, the setting mode of the entity connection set of each dynamic management entity is determined according to the connection link richness index and the connection link extension parameter, and the determination mode of the reference data information is determined according to the set dynamic proportion of each entity connection set and the set update response parameter, that is, the reference data information is matched according to the update interval parameter and the auxiliary verification parameter, or the information correlation parameter; in the time sequence conflict analysis, the determination mode of the reference data information is determined according to the update time sequence difference reference value, that is, the reference data information of the update management entity is determined according to the update time sequence difference value and the entity correlation coefficient, or the update stability index and the entity correlation coefficient; under the reference determination completion condition, the auxiliary evaluation strategy of each update management entity is determined based on the determination mode of the reference data information, that is, the information reliable coefficient of the constructed related information is determined according to the information coincidence parameter and the update interval parameter, or the abnormal combination ratio; whether to send the update information warning to the user is determined according to the information reliable coefficient. 2.The knowledge graph-based standard data processing method according to claim 1, characterized in that, periodically, the dynamic index of each target management entity is determined according to the data source richness index and the reference update index, and the category of each target management entity is determined according to the dynamic index, and the target management entity with a dynamic index greater than a preset dynamic index is recorded as a dynamic management entity; the dynamic entity proportion is the proportion of the number of dynamic management entities in the number of target management entities in the target management graph; the reference connection index is the average value of the entity connection index of each dynamic management entity in the target management graph. 3.The knowledge graph-based standard data processing method according to claim 2, characterized in that, When the dynamic entity proportion of the target management graph is greater than a preset dynamic entity proportion or the reference connection index is greater than a preset reference connection index, the target management graph is updated based on entity connection analysis, and the entity connection analysis comprises the following steps: under the entity update condition, the setting mode of the entity connection set of each update management entity is determined according to the connection link richness index and the connection link extension parameter; the determination mode of the reference data information is determined according to the set dynamic proportion of each entity connection set and the set update response parameter; the entity update condition is that there is a target management entity responding to the update condition in the target management graph, and the target management entity responding to the update condition is recorded as an update management entity. 4.The knowledge graph-based standard data processing method according to claim 3, characterized in that, for a single update management entity, if the dynamic connection coefficient is greater than a preset dynamic connection coefficient, the entity connection set of the update management entity is determined based on the update correlation index and the update radiation index; if the dynamic connection coefficient is less than or equal to the preset dynamic connection coefficient, the entity connection set of the update management entity is determined based on the related construction overlap index and the combination construction correlation index; the dynamic connection coefficient is determined according to the connection link richness index and the connection link extension parameter. 5.The knowledge graph-based standard data processing method according to claim 4, characterized in that, If the set dynamic proportion of an entity connection set is greater than a preset set dynamic proportion or the set update response parameter is greater than a preset set update response parameter, reference data information of an update supervision entity corresponding to the entity connection set is matched according to the update interval parameter and the auxiliary verification parameter, so as to determine the reference data information. The reference data information is to-be-evaluated information with an information reference coefficient greater than a preset information reference coefficient, and the information reference coefficient is in a positive correlation with the update interval parameter and the auxiliary verification parameter. 6.The knowledge graph-based standard data processing method according to claim 5, characterized in that, If the set dynamic proportion of an entity connection set is less than or equal to a preset set dynamic proportion and the set update response parameter is less than or equal to a preset set update response parameter, reference data information of an update supervision entity corresponding to the entity connection set is matched according to the information association parameter, so as to determine the reference data information. The reference data information is to-be-evaluated information with an information association parameter greater than a preset information association parameter. 7.The knowledge graph-based standard data processing method according to claim 6, characterized in that, When the dynamic entity proportion of the target management graph is less than or equal to a preset dynamic entity proportion and the reference connection index is less than or equal to a preset reference connection index, the target management graph is updated based on the time sequence conflict analysis, and the time sequence conflict analysis includes: Under the entity update condition, the update time sequence difference reference value of each update supervision entity is detected. The determination mode of the reference data information of each update supervision entity is determined according to the update time sequence difference reference value. 8.The knowledge graph-based standard data processing method according to claim 7, characterized in that, For a single update supervision entity, if the update time sequence difference reference value of the update supervision entity is less than or equal to a preset update time sequence difference reference value, the reference data information of the update supervision entity is determined according to the update time sequence difference value and the entity correlation coefficient. If the update time sequence difference reference value of the update supervision entity is greater than a preset update time sequence difference reference value, the reference data information of the update supervision entity is determined according to the update stability index and the entity correlation coefficient. 9.The knowledge graph-based standard data processing method according to claim 8, characterized in that, Under the reference determination completion condition, if the reference data information of an update supervision entity is determined according to the update interval parameter and the auxiliary verification parameter, the information reliable coefficient of the construction related information of the update supervision entity is determined according to the information coincidence parameter and the update interval parameter. If the reference data information of an update supervision entity is determined according to the information association parameter or is determined through the time sequence conflict analysis, the information reliable coefficient of the construction related information of the update supervision entity is determined according to the abnormal combination ratio. If the information reliable coefficient of the construction related information of the update supervision entity is less than a preset information reliable coefficient, an update information warning is sent to a user. The reference determination completion condition is that the reference data information of any update supervision entity is determined.

10. A standard data processing system applying the standard data processing method based on the knowledge graph according to any one of claims 1 to 9, characterized in that, The method comprises: The graph supervision module is used to determine the graph supervision strategy of the target management graph based on the dynamic entity proportion and the reference connection index, and the graph supervision strategy is to update the target management graph based on the entity connection analysis or the time sequence conflict analysis. The connection analysis module is connected with the graph supervision module, and is used for determining a setting mode of the entity connection set according to a connection link enrichment index and a connection link extension parameter, and determining a determination mode of the reference data information according to a set dynamic proportion and a set update response parameter, that is, the reference data information is matched according to an update interval parameter and an auxiliary verification parameter, or an information correlation parameter; The time sequence analysis module is connected with the graph supervision module, and is used for determining the determination mode of the reference data information according to an update time sequence difference reference value, that is, the reference data information of the update supervision entity is determined according to an update time sequence difference value and an entity correlation coefficient, or an update stability index and the entity correlation coefficient; The auxiliary evaluation module is connected with the connection analysis module and the time sequence analysis module, and is used for determining an auxiliary evaluation strategy of each update supervision entity according to the determination mode of the reference data information, that is, the information reliable coefficient of the constructed related information is determined according to an information coincidence parameter and an update interval parameter, or an abnormal combination ratio.

Images