A knowledge graph construction method, device, electronic device and storage medium

Abstract

The present disclosure relates to a knowledge graph construction method, device, electronic device and storage medium, which can improve the efficiency of knowledge graph construction and reduce the impact on downstream data tasks. The specific scheme includes: obtaining incremental data and a first historical database; the incremental data is data updated after the target timestamp, and the first historical database includes historical data before the target timestamp; obtaining a first data group from the first historical database, and fusing the first data group and the incremental data to obtain a target fused data group; the first data group is historical data in the first historical database that is associated with the incremental data; fusing the target fused data group and the first historical database to obtain a target fused database, and generating a target knowledge graph based on the target fused database.

Description

Technical Field

[0001] The present disclosure relates to the field of network technology, and in particular to a knowledge graph construction method, device, electronic device and storage medium. Background Art

[0002] The construction of knowledge graph can usually be divided into two stages: pattern layer construction and data layer construction. Among them, the pattern layer is the core of the knowledge graph, and the pattern layer stores refined knowledge content; the data layer stores specific data information. In the process of building a knowledge graph, a large amount of entity (knowledge) data needs to be processed in multiple steps. Specifically, first, a large amount of source data is subjected to data cleaning, pattern mapping, entity classification and other operations to obtain a cleaned vertical entity warehouse, and then bucketing is performed in the vertical entity warehouse according to specific attributes, and multiple entities in the same bucket are made into candidate entity pairs, and each pair of candidate entity pairs is scored. The aligned fusion entity warehouse is obtained based on the scoring results, and finally the attributes of the fusionable entity groups are selected to obtain the corresponding vertical entity graph.

[0003] However, from the overall construction process, it can be seen that because a large amount of entity data needs to be processed in multiple steps, the amount of entity data is very large, so it takes a long time to process the entity data and the efficiency is low; therefore, the efficiency of knowledge graph construction is low, which has a greater impact on downstream data tasks. Summary of the invention

[0004] The present disclosure provides a knowledge graph construction method, device, electronic device and storage medium, which can improve the efficiency of knowledge graph construction and reduce the impact on downstream data tasks. The technical solution of the present disclosure is as follows:

[0005] According to a first aspect of the present disclosure, a knowledge graph construction method is provided, the method comprising: obtaining incremental data and a first historical database; the incremental data is data updated after a target timestamp, and the first historical database includes historical data before the target timestamp; obtaining a first data group from the first historical database, and fusing the first data group and the incremental data to obtain a target fused data group; the first data group is historical data in the first historical database associated with the incremental data; fusing the target fused data group and the first historical database to obtain a target fused database, and generating a target knowledge graph based on the target fused database.

[0006] From the above, it can be seen that the electronic device can obtain the incremental data updated after the target timestamp and obtain the first historical data corresponding to the target timestamp, so as to obtain the first data group associated with the incremental data from the first historical database, and fuse the first data group and the incremental data to obtain the target fused data group; further, the electronic device can also fuse the target fused data group and the first historical database to obtain the target fused database, and generate the target knowledge graph according to the target fused database. In this case, the electronic device can only process the incremental data and the first data group associated with the incremental data in the first historical database to obtain the latest knowledge graph, without processing all the data to obtain the latest knowledge graph, thereby effectively reducing the amount of data processed by the electronic device, improving the efficiency of knowledge graph construction, and thus reducing the impact on downstream data tasks.

[0007] Optionally, before the above-mentioned "obtaining a first data group from a first historical database, and fusing the first data group and incremental data to obtain a target fused data group", the method specifically also includes: obtaining a second historical database; wherein the data in the first historical database is data obtained by performing a first data processing on the data in the second historical database; performing a second data processing on the incremental data to obtain target update data, and performing a data fusion processing on the target update data and the second historical database to obtain a target database, and the second data processing is different from the first data processing.

[0008] From the above, it can be seen that the electronic device can also obtain a second historical database, and after performing a second data processing on the incremental data to obtain the target update data, the target update data is fused with the second historical database to obtain the target database, so that when the target fusion data group is obtained, the target fusion data group can also be determined based on the data in the target database. A more specific data processing process is further given, and a specific implementation method for the electronic device to obtain the target fusion data group is given, so that the electronic device can quickly and efficiently perform data fusion processing to obtain the target fusion data group.

[0009] Optionally, the above method of "obtaining the first data group from the first historical database, and fusing the first data group with the incremental data to obtain a target fused data group" specifically includes: obtaining the first data group from the first historical database, and obtaining the second data group from the target database; the second data group is data associated with the incremental data in the target database; and determining the target fused data group based on the comparison result of the data in the first data group and the second data group. The above method of "fusing the target fused data group with the first historical database to obtain a target fused database" specifically includes: fusing the target fused data group with the data in the first historical database other than the first data group to obtain the target fused database.

[0010] As can be seen from the above, when fusing data to obtain a target fusion database, the electronic device can obtain a second data group associated with incremental data from the target database after obtaining a first data group from the first historical database, so that the target fusion data group can be determined based on the comparison result of the data in the first data group and the second data group, and further, the determined target fusion data group is fused with the data in the first historical database except the first data group to obtain the target fusion database. A specific implementation method for an electronic device to obtain a target fusion database is given, through which the electronic device can specifically identify the data that needs to be fused, so as to more accurately perform data fusion processing, obtain target fusion data, and improve the efficiency of data fusion.

[0011] Optionally, the above-mentioned method of "obtaining a first data group from a first historical database and obtaining a second data group from a target database" specifically includes: obtaining a first data group corresponding to the incremental data from the first historical database according to attribute information of the incremental data, and obtaining a second data group from the target database according to the attribute information; the attribute information includes at least one of the following: a name identifier of the incremental data, an alias identifier of the incremental data.

[0012] As can be seen from the above, when the electronic device obtains the first data group from the first historical database and obtains the second data group from the target database, it can obtain the first data group corresponding to the incremental data from the first historical database according to at least one of the name identifier and the alias identifier of the incremental data, and obtain the second data group from the target database according to at least one of the name identifier and the alias identifier of the incremental data. A specific implementation method for an electronic device to obtain a data group is given, through which the electronic device can accurately obtain the first data group and the second data group according to the attribute information of the incremental data, thereby improving the accuracy of the electronic device in obtaining the first data group and the second data group.

[0013] Optionally, the above-mentioned first historical database is a historical fusion database, which is used to store historical data after the first data processing, and the first data group includes the third data group; the above-mentioned method of "obtaining the first data group corresponding to the incremental data from the first historical database according to the attribute information of the incremental data" specifically includes: obtaining the third data group corresponding to the incremental data from the historical fusion database according to the attribute information of the incremental data.

[0014] As can be seen from the above, when the first historical database is a historical fusion database for storing historical data after the first data processing, the electronic device can obtain the third data group corresponding to the incremental data from the historical fusion database according to the attribute information of the incremental data. A specific implementation method for the electronic device to obtain the data group is given, whereby the electronic device can specifically obtain the data corresponding to the incremental data from the historical fusion database according to the attribute information of the incremental data, thereby improving the accuracy of the electronic device in obtaining the data group.

[0015] Optionally, the above-mentioned first historical database is a historical knowledge graph, the historical knowledge graph is a knowledge graph generated according to a historical fusion database, the historical fusion database is used to store historical data after the first data processing, and the first data group includes a fourth data group; the above-mentioned method of "obtaining the first data group from the first historical database" specifically includes: according to the attribute information of the incremental data, obtaining the fourth data group corresponding to the incremental data from the historical knowledge graph, the data contained in the fourth data group is the data obtained after the data in the third data group is fused, and the third data group includes the historical data obtained after the first data processing in the historical fusion database.

[0016] From the above, it can be seen that when the first historical database is a historical knowledge graph generated according to the historical fusion database, the electronic device can obtain the fourth data group corresponding to the incremental data from the historical knowledge graph according to the attribute information of the incremental data. A specific implementation method for an electronic device to obtain a data group is given, whereby the electronic device can obtain the data corresponding to the incremental data from the historical knowledge graph according to the attribute information of the incremental data, which can effectively reduce the acquired data groups, thereby improving the efficiency of the electronic device in obtaining the data group.

[0017] Optionally, the above-mentioned method of "performing a second data processing on the incremental data to obtain target updated data" specifically includes: determining the data to be processed according to the update time corresponding to each data in the incremental data, and the update time corresponding to the data to be processed is the time after the target timestamp; performing target processing on the data to be processed to obtain target updated data, and the target processing includes at least one of the following: data processing, pattern mapping processing, and data classification processing.

[0018] As can be seen from the above, when the electronic device performs the second data processing on the incremental data to obtain the target update data, the electronic device can first determine the data to be processed within the time range after the target timestamp according to the update time corresponding to each data in the incremental data, and further, perform at least one of data processing, pattern mapping, and data classification on the data to be processed to obtain the target update data. A specific implementation method for an electronic device to obtain the target update data is given, and the data to be processed is specifically determined according to the update time corresponding to each data, so that the electronic device can quickly and accurately determine the required data to be processed, and then obtain the corresponding target update data.

[0019] Optionally, the above-mentioned method of "determining the target fusion data group based on the comparison results of the data in the first data group and the second data group" specifically includes: scoring the data included in the first data group and the second data group, and determining the correlation between the data included in the first data group and the second data group; determining the data in the first data group and the second data group whose scoring results are greater than or equal to the preset score as the target fusion data group.

[0020] From the above, it can be seen that when the electronic device determines the target fusion data group based on the comparison result of the data in the first data group and the second data group, the data included in the first data group and the second data group can be scored first to determine the correlation between the data included in the first data group and the second data group, so as to determine the data in the first data group and the second data group whose scoring results are greater than or equal to the preset score as the target fusion data group. A specific implementation method for an electronic device to determine the fusion data group is given. By scoring the data included in the first data group and the second data group, the corresponding target fusion data group can be accurately determined according to the scoring results, thereby improving the accuracy of determining the target fusion data group.

[0021] According to a second aspect of the present disclosure, a knowledge graph construction device is provided, which includes: an acquisition unit, a processing unit and a generation unit; the acquisition unit is configured to execute acquisition of incremental data and a first historical database; the incremental data is data updated after a target timestamp, and the first historical database includes historical data before the target timestamp; the acquisition unit is also configured to execute acquisition of a first data group from the first historical database; the processing unit is configured to execute fusion processing of the first data group and the incremental data to obtain a target fused data group; the first data group is historical data in the first historical database associated with the incremental data; the processing unit is also configured to execute fusion processing of the target fused data group and the first historical database to obtain a target fused database; the generation unit is configured to execute generation of a target knowledge graph according to the target fused database.

[0022] Optionally, the acquisition unit is further configured to execute acquisition of a second historical database; wherein the data in the first historical database is data obtained by performing a first data processing on the data in the second historical database; the processing unit is further configured to execute a second data processing on the incremental data to obtain target update data, and perform data fusion processing on the target update data and the second historical database to obtain the target database, and the second data processing is different from the first data processing.

[0023] Optionally, the acquisition unit is specifically configured to execute acquisition of a first data group from a first historical database and a second data group from a target database; the second data group is historical data associated with incremental data in the target database; the processing unit is specifically configured to execute determination of a target fusion data group based on a comparison result of data in the first data group and the second data group, and to fuse the target fusion data group with data in the first historical database except the first data group to obtain a target fusion database.

[0024] Optionally, the acquisition unit is specifically configured to execute, based on the attribute information of the incremental data, acquiring a first data group corresponding to the incremental data from a first historical database, and acquiring a second data group from a target database based on the attribute information; the attribute information includes at least one of the following: a name identifier of the incremental data, an alias identifier of the incremental data.

[0025] Optionally, the first historical database is a historical fusion database, which is used to store historical data after the first data processing, and the first data group includes the third data group; the acquisition unit is specifically configured to execute the acquisition of the third data group corresponding to the incremental data from the historical fusion database according to the attribute information of the incremental data.

[0026] Optionally, the first historical database is a historical knowledge graph, the historical knowledge graph is a knowledge graph generated based on a historical fusion database, the historical fusion database is used to store historical data after the first data processing, and the first data group includes a fourth data group; the acquisition unit is specifically configured to execute, based on the attribute information of the incremental data, acquiring the fourth data group corresponding to the incremental data from the historical knowledge graph, the data contained in the fourth data group is the data obtained after the data in the third data group is fused, and the third data group includes the historical data obtained after the first data processing in the historical fusion database.

[0027] Optionally, the processing unit is specifically configured to determine the data to be processed according to the update time corresponding to each data in the incremental data, where the update time corresponding to the data to be processed is the time after the target timestamp; and perform target processing on the data to be processed to obtain target updated data, where the target processing includes at least one of the following: data processing, pattern mapping processing, and data classification processing.

[0028] Optionally, the processing unit is specifically configured to perform scoring on the data included in the first data group and the second data group, determine the correlation between the data included in the first data group and the second data group; and determine the data in the first data group and the second data group whose scoring results are greater than or equal to a preset score as the target fusion data group.

[0029] According to a third aspect of the present disclosure, an electronic device is provided, the electronic device comprising:

[0030] A processor. A memory for storing processor executable instructions. The processor is configured to execute instructions to implement any optional knowledge graph construction method in the first aspect above.

[0031] According to a fourth aspect of the present disclosure, a computer-readable storage medium is provided, on which instructions are stored. When the instructions in the computer-readable storage medium are executed by a processor of an electronic device, the electronic device is enabled to execute any one of the optional knowledge graph construction methods in the first aspect described above.

[0032] According to a fifth aspect of the present disclosure, there is provided a computer program product comprising instructions which, when executed on a computer, enable the computer to execute any optional knowledge graph construction method as described in the first aspect.

[0033] According to the sixth aspect of the present disclosure, the present disclosure provides a chip, the chip includes a processor and a communication interface, the communication interface and the processor are coupled, and the processor is used to run a computer program or instructions to implement the knowledge graph construction method described in the first aspect and any possible implementation method of the first aspect.

[0034] The technical solution provided by the present disclosure brings at least the following beneficial effects:

[0035] Based on any of the above aspects, in the present disclosure, the electronic device can obtain the incremental data updated after the target timestamp and obtain the first historical data corresponding to the target timestamp, so as to obtain the first data group associated with the incremental data from the first historical database, and fuse the first data group and the incremental data to obtain the target fused data group; further, the electronic device can also fuse the target fused data group and the first historical database to obtain the target fused database, and generate the target knowledge graph according to the target fused database. In this case, the electronic device can only process the incremental data and the first data group associated with the incremental data in the first historical database to obtain the latest knowledge graph, without processing all the data to obtain the latest knowledge graph, thereby effectively reducing the amount of data processed by the electronic device, improving the efficiency of knowledge graph construction, and reducing the impact on downstream data tasks.

[0036] It is to be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the present disclosure. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] The drawings herein are incorporated into and constitute a part of the specification, illustrate embodiments consistent with the present disclosure, and together with the description are used to explain the principles of the present disclosure, and do not constitute improper limitations on the present disclosure.

[0038] Figure 1 It is a flowchart of the existing knowledge graph construction system;

[0039] Figure 2 is a schematic diagram of a knowledge graph construction system according to an embodiment of the present disclosure;

[0040] Figure 3 It is a flowchart of a knowledge graph construction method according to an embodiment of the present disclosure;

[0041] Figure 4 It is a flowchart of another knowledge graph construction method according to an embodiment of the present disclosure;

[0042] Figure 5 is a flowchart of another method for constructing a knowledge graph according to an embodiment of the present disclosure;

[0043] Figure 6 is a flowchart of another method for constructing a knowledge graph according to an embodiment of the present disclosure;

[0044] Figure 7 is a flowchart of another method for constructing a knowledge graph according to an embodiment of the present disclosure;

[0045] Figure 8 It is a schematic diagram of a process of constructing a knowledge graph system according to an embodiment of the present disclosure;

[0046] Fig. 9 is a flowchart of another method for constructing a knowledge graph according to an embodiment of the present disclosure;

[0047] Fig.10 It is a flowchart of another knowledge graph construction system according to an embodiment of the present disclosure;

[0048] Fig.11 is a flowchart of another method for constructing a knowledge graph according to an embodiment of the present disclosure;

[0049] Fig.12 is a flowchart of another method for constructing a knowledge graph according to an embodiment of the present disclosure;

[0050] Fig.13 It is a structural schematic diagram of a knowledge graph construction device according to an embodiment of the present disclosure;

[0051] Fig.14 It is a structural diagram of another knowledge graph construction device according to an embodiment of the present disclosure. DETAILED DESCRIPTION

[0052] In order to enable ordinary persons in the art to better understand the technical solutions of the present disclosure, the technical solutions in the embodiments of the present disclosure will be clearly and completely described below in conjunction with the accompanying drawings.

[0053] It should be noted that the terms "first", "second", etc. in the specification and claims of the present disclosure and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments of the present disclosure described herein can be implemented in an order other than those illustrated or described herein. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present disclosure. Instead, they are merely examples of devices and methods consistent with some aspects of the present disclosure as detailed in the appended claims.

[0054] First, the application scenarios of the embodiments of the present disclosure are introduced.

[0055] The knowledge graph construction method of the disclosed embodiment is applied to the scenario of constructing a knowledge graph. In the related art, the knowledge graph aims to describe various entities or concepts and their relationships existing in the real world. The knowledge graph can be represented by a huge semantic network diagram, where nodes represent entities or concepts or attribute values, and edges represent attributes or relationships. Knowledge graphs are used to refer to various large-scale knowledge bases. The construction of knowledge graphs can generally be divided into two stages: pattern layer construction and data layer construction. Among them, the pattern layer is the core of the knowledge graph, and the pattern layer stores refined knowledge; the data layer stores specific data information. In the process of building a vertical entity graph that supports the search business, it is necessary to first observe a large amount of entity data of this category, and at the same time refer to the general category entity pattern definition Chinese site to design the vertical entity pattern layer, and then the pattern layer guides the completion of subsequent data cleaning, entity comparison and entity fusion processes.

[0056] like Figure 1The figure shows a schematic diagram of the existing knowledge graph construction process. In the existing entity graph construction scheme, tens of millions of entity data need to be processed every day. First, the vertical source data 01 needs to be cleaned, schema mapped, data classified and other knowledge processing operations (data processing) to obtain the cleaned vertical database 02 (vertical entity warehouse). Then, the name and alias attributes of the entity are used to perform bucketing operations in the vertical database 02. The entities in the same bucket are paired into candidate entity pairs to form candidate entity pair groups. Then, a set of entity comparison rules and models are used to score each pair of candidate entity pairs. According to the scoring results, the aligned fusion database 03 (fusion entity warehouse) is obtained. Finally, the attributes of the fusion entity group in the fusion database 03 can be selected to obtain the vertical knowledge graph 04 of the day. Figure 1 The difference between the fusion database 03 and the vertical database 02 is that after comparison, it has been determined which entities should be merged, so the fusion entity warehouse has one more identifier. The entity groups that should be merged share the same identifier, and the identifiers of different groups are different. The vertical knowledge graph 04 is the final result after the attribute selection of the entity groups that should be merged is optimized.

[0057] From the overall process, we can see that because the amount of entity data is very large, tens of millions of data processing operations need to be performed every day, and the number of entity comparisons (number of candidate entity pairs) also reaches tens of millions. The main time consumption of existing solutions is also concentrated in the data processing and entity comparison process. In addition, in order to ensure that the latest data can be obtained every day, it is necessary to update the knowledge graph construction on a daily basis. Therefore, how to improve the efficiency of knowledge graph construction (inclusion) is a very critical issue, which has a great impact on downstream tasks.

[0058] In order to solve the above problems, the disclosed embodiment provides a knowledge graph construction method, whereby the electronic device can obtain the incremental data updated after the target timestamp and obtain the first historical data corresponding to the target timestamp, thereby obtaining the first data group associated with the incremental data from the first historical database, and fuse the first data group and the incremental data to obtain the target fused data group; further, the electronic device can also fuse the target fused data group and the first historical database to obtain the target fused database, and generate the target knowledge graph according to the target fused database. In this case, the electronic device can obtain the latest knowledge graph by processing only the incremental data and the first data group associated with the incremental data in the first historical database, without processing all the data to obtain the latest knowledge graph, thereby effectively reducing the amount of data processed by the electronic device, improving the efficiency of knowledge graph construction, and thereby reducing the impact on downstream data tasks.

[0059] The following is an exemplary description of the content display method provided by the embodiment of the present disclosure in conjunction with the accompanying drawings:

[0060] Figure 2 is a schematic diagram of a knowledge graph construction system provided by an embodiment of the present disclosure, such as Figure 2 As shown, the knowledge graph construction system may include a server 11 and a client 12 ( Figure 2 Only one client 12 is given as an example, and more clients may be provided in a specific implementation). Among them, the server 11 may establish a communication connection with the client 12. The server 11 and the client 12 may be connected in a wired manner or in a wireless manner, which is not limited in the embodiments of the present disclosure.

[0061] The server 11 is used to receive and save the real-time data information sent by the client 12. For example, the server 11 receives the account information, name information, attribute information, etc. sent by the client 12, and saves these data information for building a knowledge graph.

[0062] The client 12 is used to generate real-time data information and send the real-time data information to the server 11. For example, the client 12 generates account information, name information, attribute information, etc. in real time according to the user's operation, and sends this data information to the server 11 to be stored in the server 11 for building a knowledge graph.

[0063] In an practicable manner, the server 11 may be a single server, or a server cluster consisting of multiple servers, or a cloud computing service center. The server 11 may include a processor, a memory, a network interface, and the like.

[0064] In an operative manner, the client 12 is used to provide voice and / or data connectivity services to the user. The client 12 may have different names, such as UE, terminal unit, terminal station, mobile station, mobile station, remote station, remote terminal, mobile device, wireless communication device, vehicle user equipment, terminal agent or terminal device, etc.

[0065] Optionally, the client 12 may be any handheld device, vehicle-mounted device, wearable device, or computer with communication functions, which is not limited in the embodiments of the present disclosure. For example, the handheld device may be a smart phone. The vehicle-mounted device may be a vehicle-mounted navigation system. The wearable device may be a smart bracelet. The computer may be a personal digital assistant (PDA), a tablet computer, or a laptop computer.

[0066] The knowledge graph construction method provided in the present disclosure embodiment can be applied to the aforementioned Figure 2 The server 11 and the client 12 in the knowledge graph construction system shown. The electronic device involved in the present disclosure may be the server 11 or the client 12. Taking the application of the knowledge graph construction method of the present disclosure on the server in the process of constructing the knowledge graph as an example, the knowledge graph construction method provided by the embodiment of the present disclosure is introduced in detail.

[0067] After introducing the application scenarios and knowledge graph construction system of the embodiment of the present disclosure, Figure 2 The knowledge graph construction system shown provides a detailed introduction to the knowledge graph construction method provided in the embodiment of the present disclosure.

[0068] like Figure 3 As shown, a flow chart of a knowledge graph construction method according to an exemplary embodiment is shown. The knowledge graph construction method may include S201-S203.

[0069] S201. Acquire incremental data and a first historical database.

[0070] Specifically, the incremental data is data updated after the target timestamp, and the first historical database includes historical data before the target timestamp.

[0071] In the disclosed embodiment, the electronic device can construct a knowledge graph at a preset interval to ensure the integrity of the data information included in the knowledge graph and to update (include) the knowledge graph in a timely manner. In the process of constructing the knowledge graph by the electronic device, it is necessary to obtain the target source data including the latest data, and the target source data also includes all the historical data, so that the electronic device can construct the knowledge graph based on these data. It can be understood that the incremental data and the first historical database are both data obtained from the target source data.

[0072] Optionally, the electronic device can also obtain a first historical database constructed before the current moment, which can be a historical vertical database obtained by performing full knowledge processing on historical data to perform data cleaning, pattern mapping, and data classification.

[0073] Optionally, the target timestamp can be a time point corresponding to the current moment, which can be understood as the time point when the knowledge graph was last constructed before the current moment. Therefore, the incremental data can be understood as: the updated data after the last knowledge graph was constructed; the first historical database can be understood as the historical database corresponding to the last knowledge graph constructed.

[0074] Exemplarily, the electronic device needs to build a knowledge graph every day. For the vertical entity source data of the day (i.e., the target source data), the data after the target timestamp of the data update (update means new data is added or the data field content is updated) can be intercepted, thereby obtaining incremental data. Furthermore, the incremental data is merged with the vertical database (i.e., the second historical database) obtained the previous day (merging here means that only the latest data of the day is retained when data is repeated) to obtain the vertical entity warehouse of the day (i.e., the target database), thereby reducing the number of data cleaning and processing.

[0075] S202: Acquire a first data group from a first historical database, and fuse the first data group and incremental data to obtain a target fused data group.

[0076] Specifically, the first data group is historical data associated with the incremental data in the first historical database.

[0077] Optionally, the electronic device can also obtain a first historical database constructed before the current moment, and obtain a first data group (also referred to as data to be updated) that has an associated relationship with the incremental data from the first historical database, so that the incremental data in the target source data and the first data group can be fused to obtain a target fused data group.

[0078] Exemplarily, the electronic device can obtain the fusion database (i.e., the first historical database) constructed the day before, and obtain the first data group affected by the current incremental data therefrom, so that the affected data group can be updated in the fusion database constructed the day before to achieve the effect of updating the database.

[0079] It should be noted that the fusion database constructed the day before is obtained based on the vertical database constructed the day before. The corresponding fusion database is obtained by comparing, aligning, and grouping the data in the vertical database (i.e. fusing and associating related data).

[0080] S203: Fuse the target fusion data group and the first historical database to obtain a target fusion database, and generate a target knowledge graph based on the target fusion database.

[0081] Optionally, the electronic device can fuse the target fusion data group and the first historical database to obtain the latest fusion database corresponding to the current moment, thereby obtaining the latest knowledge graph. That is, it can be understood that the latest fusion database is obtained by updating the data to be updated in the first historical database.

[0082] Optionally, the electronic device can further perform data fusion and optimization on the data in the target fusion database based on the obtained target fusion database to obtain a corresponding knowledge graph.

[0083] Exemplarily, the electronic device can obtain the latest knowledge graph of the day based on the latest fusion database (i.e., target fusion database) obtained by current data processing and based on existing attribute fusion, optimization strategies and solutions.

[0084] The technical problem solved by the present disclosure is how to reduce the number of data processing operations such as knowledge processing and data comparison as much as possible while ensuring that the vertical entity knowledge graph is constructed (included) on a daily basis, thereby improving the efficiency of knowledge graph inclusion and allowing the daily routine results to include the latest (including new and updated) data to support downstream business channels.

[0085] The technical solution provided by the above embodiment brings at least the following beneficial effects: the electronic device can obtain the incremental data updated after the target timestamp and obtain the first historical data corresponding to the target timestamp, so as to obtain the first data group associated with the incremental data from the first historical database, and fuse the first data group and the incremental data to obtain the target fused data group; further, the electronic device can also fuse the target fused data group and the first historical database to obtain the target fused database, and generate the target knowledge graph according to the target fused database. In this case, the electronic device can only process the incremental data and the first data group associated with the incremental data in the first historical database to obtain the latest knowledge graph, without processing all the data to obtain the latest knowledge graph, thereby effectively reducing the amount of data processed by the electronic device, improving the efficiency of knowledge graph construction, and reducing the impact on downstream data tasks.

[0086] In one practicable manner, combining Figure 3 ,like Figure 4 As shown, before the above S202, S301-S302 may also be specifically included.

[0087] S301, obtaining a second historical database.

[0088] The data in the first historical database is data obtained by performing a first data processing on the data in the second historical database.

[0089] S302: Perform a second data processing on the incremental data to obtain target update data, and perform data fusion processing on the target update data and the second historical database to obtain a target database.

[0090] The second data processing is different from the first data processing in terms of processing method.

[0091] Optionally, the second data processing is to perform incremental data processing on the incremental data to perform data cleaning, pattern mapping, data classification and other processing operations on the incremental data, so that when the target database is obtained based on the incremental data and the second historical database, the target update data can be obtained by further processing the incremental data, so as to facilitate data fusion processing of the target update data with the second historical database to obtain the target database.

[0092] Exemplarily, after intercepting the incremental data after the timestamp of the data update, the data updated on that day can be filtered out for data cleaning, pattern mapping and other operations, thereby obtaining the changed entity data of the day (i.e., the target update data), and then the target update data can be merged with the vertical database of the previous day to obtain the vertical database of the day (i.e., the target database).

[0093] The technical solution provided by the above embodiment brings at least the following beneficial effects: the electronic device can also obtain a second historical database, and after performing a second data processing on the incremental data to obtain the target update data, perform data fusion processing on the target update data and the second historical database to obtain the target database, so that when the target fusion data group is obtained, the target fusion data group can also be determined based on the data in the target database, and a more specific data processing process is further given, and a specific implementation method for the electronic device to obtain the target fusion data group is given, so that the electronic device can quickly and efficiently perform data fusion processing to obtain the target fusion data group.

[0094] In one practicable manner, combining Figure 4 ,like Figure 5 As shown, the method in the above S202 may specifically include S401-S402, and the method in the above S203 may specifically include S403.

[0095] S401: Acquire a first data group from a first historical database, and acquire a second data group from a target database.

[0096] The second data group is historical data associated with the incremental data in the target database.

[0097] Optionally, in one implementation, while the electronic device acquires the first data group from the first historical database, it may acquire a second data group associated with the incremental data from the obtained target database, where the second data group is data in the target database affected by the incremental data.

[0098] S402: Determine a target fusion data group according to a comparison result of data in the first data group and the second data group.

[0099] S403: Fusing the target fused data group with the data in the first historical database except the first data group to obtain a target fused database.

[0100] Optionally, after obtaining the first data group and the second data group, the electronic device may compare the data included in the first data group and the second data group to determine whether there is associated data between the data included in the first data group and the second data group, thereby obtaining a fused data group.

[0101] Optionally, after obtaining the fused data group, the fused data group may be fused with the first historical database. Specifically, the fused data group is fused with the first historical database after deleting the first data group, so as to obtain a target fused database.

[0102] Exemplarily, the candidate entity groups affected by the incremental data can be obtained from the fusion database of the previous day to obtain all the entity pairs affected in the graph of the previous day (i.e., the first data group); at the same time, the candidate entity groups that may need to be fused and compared on the day of the recall can be obtained from the vertical database of the day using the incremental data of the day to obtain the entity pairs to be compared (possibly affected) on the day (the second data group). Furthermore, using the existing entity comparison strategies and models, the fused data group that changes on the day is obtained based on the first data group and the second data group, and the fused data group that changes on the day is merged with the unaffected fused entity group in the fusion database of the previous day to obtain the latest fused database of the day.

[0103] The technical solution provided by the above embodiment brings at least the following beneficial effects: when fusing data to obtain a target fusion database, the electronic device can obtain a second data group associated with incremental data from the target database after obtaining a first data group from the first historical database, so that the target fusion data group can be determined according to the comparison result of the data in the first data group and the second data group, and further, the determined target fusion data group is fused with the data in the first historical database except the first data group to obtain the target fusion database. A specific implementation method for an electronic device to obtain a target fusion database is given, through which the electronic device can specifically identify the data that needs to be fused, so as to more accurately perform data fusion processing, obtain target fusion data, and improve the efficiency of data fusion.

[0104] In one practicable manner, combining Figure 5 ,like Figure 6 As shown, the method in the above S401 may specifically include S501.

[0105] S501. Acquire a first data group corresponding to the incremental data from a first historical database according to attribute information of the incremental data, and acquire a second data group from a target database according to the attribute information.

[0106] The attribute information includes at least one of the following: a name identifier of the incremental data and an alias identifier of the incremental data.

[0107] Optionally, the electronic device can determine the corresponding first data group from the first historical database based on the name identifier and alias identifier (i.e., attribute information) of the incremental data. Similarly, the electronic device can also determine the corresponding second data group from the target database based on the name identifier and alias identifier of the incremental data.

[0108] Exemplarily, the goal of the present disclosure is to reduce the number of data processing operations such as entity data cleaning and processing, entity comparison and alignment, so as to reduce the time spent in the knowledge graph construction process and improve the efficiency of data collection. After actual testing, using the knowledge graph incremental construction process provided by the present disclosure, data processing operations such as knowledge processing (data cleaning, pattern mapping) can be reduced to one percent of the existing solutions, and the number of entity comparisons is also reduced to one third of the existing solutions. Data processing is reduced from tens of millions to millions, the time spent is greatly reduced, and the efficiency of the entity knowledge graph construction process is also improved.

[0109] The technical solution provided by the above embodiment brings at least the following beneficial effects: when the electronic device obtains the first data group from the first historical database and obtains the second data group from the target database, it can obtain the first data group corresponding to the incremental data from the first historical database according to at least one of the name identifier and the alias identifier of the incremental data, and obtain the second data group from the target database according to at least one of the name identifier and the alias identifier of the incremental data. A specific implementation method for an electronic device to obtain a data group is provided, through which the electronic device can accurately obtain the first data group and the second data group according to the attribute information of the incremental data, thereby improving the accuracy of the electronic device in obtaining the first data group and the second data group.

[0110] In an practicable manner, the first historical database is a historical fusion database, which is used to store historical data after the first data processing, and the first data group includes the third data group. Figure 6 ,like Figure 7 As shown, the above-mentioned "acquiring a first data group corresponding to the incremental data from a first historical database according to the attribute information of the incremental data" in S501 may specifically include S5011.

[0111] S5011. Acquire a third data group corresponding to the incremental data from the historical fusion database according to the attribute information of the incremental data, and acquire a second data group from the target database according to the attribute information.

[0112] Optionally, the third data processing is to perform entity data comparison and alignment processing on the historical data before the target timestamp to obtain a historical fusion database.

[0113] Optionally, when the first historical database is a historical fusion database, the electronic device obtains a data group (ie, a third data group) corresponding to the incremental data from the historical fusion database according to a name identifier and an alias identifier (ie, attribute information) of the incremental data.

[0114] For example, Figure 8 As shown, it is a schematic diagram of the specific knowledge graph construction process when the first historical database is a historical fusion database. First, after obtaining the target source data 21 of the day, the electronic device can determine the time point corresponding to the updated incremental data, that is, the time point corresponding to the target timestamp. For example, the target timestamp is 00:00 every day, and the incremental data is the current change data. Then, the incremental data (new data or updated data) corresponding to the time point corresponding to the updated incremental data to the current moment is intercepted, and the incremental data is processed (incremental knowledge processing) to obtain the change data 22 of the day (that is, the incremental data). Furthermore, the vertical category database 23 of the previous day (that is, the second historical database) is obtained, and the vertical category database 24 of the day (that is, the target database) is obtained by fusing the change data 22 of the day and the vertical category database 23 of the previous day. The electronic device then obtains the first data group 26 from the previous day's fusion database 25 (i.e., the first historical database) based on the attribute information of the day's change data 22, and obtains the second data group 27 from the day's vertical database 24, and finally performs data comparison and alignment processing based on the fusion data of the first data group 26 and the second data group 27, as well as the previous day's fusion database 25, to obtain the day's fusion database 28 (i.e., the target fusion database). Then, the electronic device performs data fusion on the day's fusion database 28 and selects the best to obtain the day's knowledge graph 29.

[0115] The technical solution provided by the above embodiment brings at least the following beneficial effects: when the first historical database is a historical fusion database for storing historical data after the first data processing, the electronic device can obtain the third data group corresponding to the incremental data from the historical fusion database according to the attribute information of the incremental data. A specific implementation method for the electronic device to obtain the data group is provided, whereby the electronic device can specifically obtain the data corresponding to the incremental data from the historical fusion database according to the attribute information of the incremental data, thereby improving the accuracy of the electronic device to obtain the data group.

[0116] In an implementable manner, the first historical database is a historical knowledge graph, the historical knowledge graph is a knowledge graph generated based on a historical fusion database, the historical fusion database is used to store historical data after the first data processing, and the first data group includes the fourth data group. Figure 3 ,like Fig. 9 As shown, the “obtaining a first data group from a first historical database” in the above S202 may specifically include S2021.

[0117] S2021. According to the attribute information of the incremental data, obtain the fourth data group corresponding to the incremental data from the historical knowledge graph, and fuse the fourth data group and the incremental data to obtain the target fused data group.

[0118] The data included in the fourth data group is data obtained after fusion of the data in the third data group, and the third data group includes historical data obtained after the first data processing is performed in the historical fusion database.

[0119] Optionally, the above historical knowledge graph is a knowledge graph obtained by performing data fusion and preferential processing on the historical fusion database.

[0120] Optionally, when the first historical database is a historical knowledge graph, the electronic device obtains a data group (ie, the fourth data group) corresponding to the incremental data from the historical knowledge graph based on the name identifier and alias identifier (ie, attribute information) of the incremental data.

[0121] Optionally, the amount of data included in the fourth data group corresponding to the incremental data obtained by the electronic device from the historical knowledge graph is less than the amount of data included in the third data group corresponding to the incremental data obtained from the historical fusion database.

[0122] For example, when obtaining and recalling candidate data groups affected by entities that changed on the previous day from the knowledge graph of the previous day, you can consider obtaining candidate entities directly from the knowledge graph of the previous day (using the name and alias attributes of the incremental data). This will reduce the number of candidate data groups obtained, which can theoretically further reduce the computational time and improve efficiency. Finally, the final knowledge graph of the day can be obtained based on the comparison results of the candidate data groups.

[0123] Another example is Fig.10As shown, when the first historical database is a historical knowledge graph, a specific knowledge graph construction process diagram is shown. First, after obtaining the target source data 31 of the day, the electronic device can determine the time point corresponding to the updated incremental data, and intercept the incremental data (new data or updated data) corresponding to the time point corresponding to the updated incremental data to the current moment, and perform data processing (incremental knowledge processing) on ​​the incremental data to obtain the day's change data 32 (i.e., incremental data). Furthermore, the previous day's vertical category database 33 (i.e., the second historical database) is obtained, and the day's vertical category database 34 (i.e., the target database) is obtained by fusing the day's change data 32 and the previous day's vertical category database 33. The electronic device then obtains the first data group 36 from the previous day's knowledge graph 35 (i.e., the first historical database) according to the attribute information of the day's change data 32, and finally performs data comparison and alignment processing on the first data group 36 and the previous day's knowledge graph 35 to obtain a graph comparison result 37 (i.e., the target fusion database), and directly obtains the day's knowledge graph 38.

[0124] The technical solution provided by the above embodiment brings at least the following beneficial effects: when the first historical database is a historical knowledge graph generated according to the historical fusion database, the electronic device can obtain the fourth data group corresponding to the incremental data from the historical knowledge graph according to the attribute information of the incremental data. A specific implementation method for an electronic device to obtain a data group is given, whereby the electronic device can obtain the data corresponding to the incremental data from the historical knowledge graph according to the attribute information of the incremental data, which can effectively reduce the acquired data groups, thereby improving the efficiency of the electronic device in obtaining the data group.

[0125] In one practicable manner, combining Figure 4 ,like Fig.11 As shown, the "performing a second data processing on the incremental data to obtain target update data" in the above S302 may specifically include S3021-S3022.

[0126] S3021. Determine the data to be processed according to the update time corresponding to each data in the incremental data.

[0127] The update time corresponding to the data to be processed is the time after the target timestamp.

[0128] Optionally, after the electronic device determines the incremental data, the electronic device may obtain the update time corresponding to each data in the incremental data, thereby determining the data whose data update time is within a time range after the target timestamp as the data to be processed.

[0129] S3022. Perform target processing on the data to be processed to obtain target update data, and perform data fusion processing on the target update data and the second historical database to obtain a target database.

[0130] Among them, the target processing includes at least one of the following: data processing, pattern mapping processing, and data classification processing.

[0131] Optionally, during the process of processing the incremental data, the electronic device may perform at least one of data processing, pattern mapping, and data classification on the determined data to be processed to obtain corresponding target update data.

[0132] Optionally, the above-mentioned data processing can be understood as further analysis or correction processing of the data; the above-mentioned pattern mapping processing can be understood as mapping the data to obtain mapping data; the above-mentioned data classification processing can be understood as classifying the data according to specific characteristics.

[0133] It can be understood that the electronic device can perform data processing on the data to be processed according to actual processing requirements; that is, the electronic device can perform data processing, pattern mapping processing, and data classification processing on the data to be processed in accordance with the processing order; or the electronic device can only perform one or two of the above three types of data processing on the data to be processed.

[0134] The technical solution provided by the above embodiment brings at least the following beneficial effects: when the electronic device performs the second data processing on the incremental data to obtain the target update data, the electronic device can first determine the data to be processed within the time range after the target timestamp according to the update time corresponding to each data in the incremental data, and further, perform at least one of data processing, pattern mapping, and data classification on the data to be processed to obtain the target update data. A specific implementation method for an electronic device to obtain the target update data is provided, and the data to be processed is specifically determined according to the update time corresponding to each data, so that the electronic device can quickly and accurately determine the required data to be processed, and then obtain the corresponding target update data.

[0135] In one practicable manner, combining Figure 5 ,like Fig.12 As shown, the "determining the target fusion data group according to the comparison result of the data in the first data group and the second data group" in the above S402 may specifically include S4021-S4022.

[0136] S4021. Score the data included in the first data group and the second data group to determine the degree of association between the data included in the first data group and the second data group.

[0137] S4022: Determine the data in the first data group and the second data group whose scoring results are greater than or equal to a preset score as the target fusion data group.

[0138] Optionally, the electronic device can score the degree of association between each data included in the first data group and the second data group, thereby determining the data whose scoring result is greater than or equal to a preset score as data with an associated relationship, thereby determining the target fusion data group through the obtained scoring result and the associated relationship between the data.

[0139] Optionally, the target fused data group includes all data in the first data group and the second data group. The processing of this step is to fuse the data having a correlation relationship in the first data group and the second data group to obtain a total fused data group.

[0140] The technical solution provided by the above embodiment brings at least the following beneficial effects: when the electronic device determines the target fusion data group based on the comparison result of the data in the first data group and the second data group, the data included in the first data group and the second data group can be scored first to determine the correlation between the data included in the first data group and the second data group, thereby determining the data in the first data group and the second data group whose scoring results are greater than or equal to the preset score as the target fusion data group. A specific implementation method for an electronic device to determine the fusion data group is provided. By scoring the data included in the first data group and the second data group, the corresponding target fusion data group can be accurately determined according to the scoring results, thereby improving the accuracy of determining the target fusion data group.

[0141] Combined with the above examples, it can be seen that in the existing scheme, in the process of building a knowledge graph, the number of data processing operations such as knowledge processing that need to be performed every day reaches the order of tens of millions, and the number of entity comparisons also reaches the order of tens of millions. These are the main sources of time consumption in the current knowledge graph full-volume construction flow, which is easy to affect downstream tasks. The goal of this disclosure is to minimize the impact of this problem. We found that although the data changes every day, the amount of change is limited compared to the total amount of data. Therefore, the entry point proposed in this scheme is to start with the incremental data of daily data. The incremental data here includes not only the newly added data of the day, but also the data that has been in the past but the content has changed. Therefore, the current incremental scheme cannot be fully applicable to the incremental construction process of the knowledge graph of this scheme. This scheme starts from the incremental data of the day. While reducing data processing operations, by obtaining and recalling the graph entities (data) of the previous day and the entities of the day affected by the entity data of the day, the number of candidate entity pairs that need to be compared is as low as possible, thereby further achieving the purpose of reducing the number of entity comparisons, thereby shortening the time consumption of knowledge graph construction and improving efficiency.

[0142] It is understandable that the above method can be implemented by a knowledge graph construction device. In order to realize the above functions, the knowledge graph construction device includes hardware structures and / or software modules corresponding to the execution of each function. Those skilled in the art should easily realize that, in combination with the modules and algorithm steps of each example described in the embodiments disclosed herein, the embodiments of the present disclosure can be implemented in the form of hardware or a combination of hardware and computer software. Whether a function is executed in the form of hardware or computer software driving hardware depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to exceed the scope of the embodiments of the present disclosure.

[0143] The embodiments of the present disclosure may divide the functional modules of the above-mentioned knowledge graph construction device etc. according to the above-mentioned method examples. For example, each functional module may be divided corresponding to each function, or two or more functions may be integrated into one processing module. The above-mentioned integrated modules may be implemented in the form of hardware or in the form of software functional modules. It should be noted that the division of modules in the embodiments of the present disclosure is schematic and is only a logical functional division. There may be other division methods in actual implementation.

[0144] Fig.13 FIG. 1 is a schematic diagram of a knowledge graph construction device according to an exemplary embodiment. Fig.13 , the knowledge graph construction device 130 may include: an acquisition unit 1301, a processing unit 1302 and a generation unit 1303.

[0145] The acquisition unit 1301 is configured to execute acquisition of incremental data and a first historical database; the incremental data is data updated after the target timestamp, and the first historical database includes historical data before the target timestamp; for example, the acquisition unit 1301 can be used to execute Figure 3 The steps in step 201 in .

[0146] The acquisition unit 1301 is further configured to acquire the first data group from the first historical database; for example, the acquisition unit 1301 can be used to execute Figure 3 The steps in step 201 in .

[0147] The processing unit 1302 is configured to perform a fusion process on the first data group and the incremental data to obtain a target fused data group; the first data group is the historical data associated with the incremental data in the first historical database; for example, the processing unit 1302 can be used to perform Figure 3 The steps in step 201 in .

[0148] The processing unit 1302 is further configured to perform a fusion process on the target fusion data set and the first historical database to obtain a target fusion database; for example, the processing unit 1302 can be used to perform Figure 3 The steps in step 202 in .

[0149] The generating unit 1303 is configured to generate a target knowledge graph according to the target fusion database; for example, the generating unit 1303 can be used to perform Figure 3 The steps in step 203 in .

[0150] Optionally, the acquisition unit 1301 is further configured to execute acquisition of a second historical database; wherein the data in the first historical database is data obtained by performing a first data processing on the data in the second historical database; for example, the processing unit 1302 may be used to execute Figure 4 Step 301 in .

[0151] The processing unit 1302 is further configured to perform a second data processing on the incremental data to obtain target update data, and perform data fusion processing on the target update data and the second historical database to obtain the target database. The second data processing is different from the first data processing. For example, the processing unit 1302 can be used to perform Figure 4 Step 301 in .

[0152] Optionally, the acquisition unit 1301 is specifically configured to execute acquisition of a first data group from a first historical database and acquisition of a second data group from a target database; the second data group is historical data associated with the incremental data in the target database; for example, the acquisition unit 1301 may be used to execute Figure 5 Step 401 in .

[0153] The processing unit 1302 is specifically configured to determine a target fusion data group according to a comparison result of data in the first data group and the second data group, and fuse the target fusion data group with data in the first historical database other than the first data group to obtain a target fusion database; for example, the processing unit 1302 can be used to perform Figure 5 Step 402 in .

[0154] Optionally, the acquisition unit 1301 is specifically configured to acquire a first data group corresponding to the incremental data from the first historical database according to the attribute information of the incremental data, and acquire a second data group from the target database according to the attribute information; the attribute information includes at least one of the following: a name identifier of the incremental data, an alias identifier of the incremental data; for example, the acquisition unit 1301 can be used to execute Figure 6 Step 501 in .

[0155] Optionally, the first historical database is a historical fusion database, which is used to store historical data after the first data processing, and the first data group includes the third data group; the acquisition unit 1301 is specifically configured to execute, according to the attribute information of the incremental data, to acquire the third data group corresponding to the incremental data from the historical fusion database; for example, the acquisition unit 1301 can be used to execute Figure 7 Step 5011 in .

[0156] Optionally, the first historical database is a historical knowledge graph, the historical knowledge graph is a knowledge graph generated according to a historical fusion database, the historical fusion database is used to store historical data after the first data processing, and the first data group includes a fourth data group; the acquisition unit 1301 is specifically configured to execute, according to the attribute information of the incremental data, to obtain a fourth data group corresponding to the incremental data from the historical knowledge graph, the data contained in the fourth data group is the data obtained after the data in the third data group is fused, and the third data group includes the historical data obtained after the first data processing in the historical fusion database; for example, the acquisition unit 1301 can be used to execute Fig. 9 Step 2011.

[0157] Optionally, the processing unit 1302 is specifically configured to determine the data to be processed according to the update time corresponding to each data in the incremental data, and the update time corresponding to the data to be processed is the time after the target timestamp; and perform target processing on the data to be processed to obtain target updated data, and the target processing includes at least one of the following: data processing, pattern mapping processing, and data classification processing; for example, the processing unit 1302 can be used to perform Fig.11 Step 3011 and step 3012 in .

[0158] Optionally, the processing unit 1302 is specifically configured to score the data included in the first data group and the second data group, determine the correlation between the data included in the first data group and the second data group; and determine the data in the first data group and the second data group whose scoring results are greater than or equal to the preset score as the target fusion data group; for example, the processing unit 1302 can be used to perform Fig.12 Step 4021 in .

[0159] As described above, the embodiments of the present disclosure can divide the functional modules of the electronic device according to the above method examples. Among them, the above integrated modules can be implemented in the form of hardware or in the form of software functional modules. In addition, it should be noted that the division of modules in the embodiments of the present disclosure is schematic and is only a logical function division. There may be other division methods in actual implementation. For example, each functional module can be divided according to each function, or two or more functions can be integrated into one processing module.

[0160] Regarding the knowledge graph construction device in the above embodiment, the specific manner in which each module performs operations has been described in detail in the embodiment of the method and will not be elaborated here.

[0161] Fig.14 60 is a schematic diagram of the structure of a knowledge graph construction device 60 provided by the present disclosure. Fig.14 , the knowledge graph construction device 60 may include at least one processor 601 and a memory 603 for storing executable instructions of the processor 601. The processor 601 is configured to execute the instructions in the memory 603 to implement the knowledge graph construction method in the above embodiment.

[0162] In addition, the knowledge graph construction device 60 may also include a communication bus 602 and at least one communication interface 604 .

[0163] The processor 601 may be a GPU, a microprocessing unit, an ASIC, or one or more integrated circuits for controlling the execution of the program of the disclosed solution.

[0164] Communication bus 602 may include a pathway for transmitting information between the above-mentioned components.

[0165] The communication interface 604 uses any transceiver or other device for communicating with other devices or communication networks, such as Ethernet, radio access network (RAN), wireless local area networks (WLAN), etc.

[0166] The memory 603 may be a read-only memory (ROM) or other types of static storage devices that can store static information and instructions, a random access memory (RAM) or other types of dynamic storage devices that can store information and instructions, or an electrically erasable programmable read-only memory (EEPROM), a compact disc read-only memory (CD-ROM) or other optical disc storage, optical disc storage (including compressed optical disc, laser disc, optical disc, digital versatile disc, Blu-ray disc, etc.), a magnetic disk storage medium or other magnetic storage device, or any other medium that can be used to carry or store the desired program code in the form of instructions or data structures and can be accessed by a computer, but is not limited thereto. The memory may exist independently and be connected to the processing unit via a bus. The memory may also be integrated with the processing unit and be a volatile storage medium in the GPU.

[0167] The memory 603 is used to store instructions for executing the solution of the present disclosure, and the execution is controlled by the processor 601. The processor 601 is used to execute the instructions stored in the memory 603, so as to realize the functions in the method of the present disclosure.

[0168] In a specific implementation, as an embodiment, the processor 601 may include one or more GPUs, such as Fig.14 GPU0 and GPU1 in .

[0169] In a specific implementation, as an embodiment, the knowledge graph construction device 60 may include multiple processors, such as Fig.14 601 and processor 607 in FIG. Each of these processors may be a single-core (single-CPU) processor or a multi-core (multi-GPU) processor. A processor herein may refer to one or more devices, circuits, and / or processing cores for processing data (e.g., computer program instructions).

[0170] In a specific implementation, as an embodiment, the knowledge graph construction device 60 may also include an output device 605 and an input device 606. The output device 605 communicates with the processor 601 and can display information in a variety of ways. For example, the output device 605 may be a liquid crystal display (LCD), a light emitting diode (LED) display device, a cathode ray tube (CRT) display device, or a projector. The input device 606 communicates with the processor 601 and can accept user input in a variety of ways. For example, the input device 606 may be a mouse, a keyboard, a touch screen device, or a sensor device.

[0171] Those skilled in the art will understand that Fig.14 The structure shown in does not constitute a limitation on the knowledge graph construction device 60, and may include more or fewer components than shown in the figure, or combine certain components, or adopt a different component arrangement.

[0172] The present disclosure also provides a computer-readable storage medium having instructions stored thereon. When the instructions in the storage medium are executed by a processor of an electronic device, the electronic device can execute the group communication method provided by the above-mentioned embodiment of the present disclosure.

[0173] The embodiments of the present disclosure also provide a computer program product containing instructions, which, when executed on an electronic device, enables the electronic device to execute the knowledge graph construction method provided by the embodiments of the present disclosure.

[0174] The present disclosure also provides a communication system. Figure 2 As shown, the system includes a server 11 and a client 12. The server 11 and the client 12 are respectively used to execute the corresponding steps in the above-mentioned embodiments of the present disclosure, so that the communication system solves the technical problems solved by the embodiments of the present disclosure and achieves the technical effects achieved by the embodiments of the present disclosure, which will not be repeated here.

[0175] Those skilled in the art will readily appreciate other embodiments of the present disclosure after considering the specification and practicing the invention disclosed herein. The present disclosure is intended to cover any variations, uses or adaptations of the present disclosure that follow the general principles of the present disclosure and include common knowledge or customary techniques in the art that are not disclosed in the present disclosure. The description and examples are to be considered exemplary only, and the true scope and spirit of the present disclosure are indicated by the following claims.

[0176] It should be understood that the present disclosure is not limited to the exact structures that have been described above and shown in the drawings, and that various modifications and changes may be made without departing from the scope thereof. The scope of the present disclosure is limited only by the appended claims.

Claims

1. A knowledge graph construction method, characterized in that: The method comprises: Acquire incremental data and a first historical database; the incremental data is data updated after a target timestamp, and the first historical database includes historical data before the target timestamp; Acquire a second historical database; wherein the data in the first historical database is data obtained by performing a first data processing on the data in the second historical database; Performing a second data processing on the incremental data to obtain target update data, and performing data fusion processing on the target update data and the second historical database to obtain a target database, wherein the second data processing is different from the first data processing in a processing manner, and the data processing includes at least one of the following: data processing, pattern mapping processing, and data classification processing; Acquire a first data group from the first historical database, and acquire a second data group from the target database; the first data group is historical data associated with the incremental data in the first historical database, and the second data group is data associated with the incremental data in the target database; Determine a target fusion data group according to a comparison result of data in the first data group and the second data group; The target fusion data group is fused with the data in the first historical database except the first data group to obtain a target fusion database, and a target knowledge graph is generated according to the target fusion database.

2. The method according to claim 1, characterized in that The acquiring the first data group from the first historical database and acquiring the second data group from the target database comprises: According to the attribute information of the incremental data, a first data group corresponding to the incremental data is obtained from the first historical database, and according to the attribute information, the second data group is obtained from the target database; the attribute information includes at least one of the following: a name identifier of the incremental data and an alias identifier of the incremental data.

3. The method according to claim 2, characterized in that The first historical database is a historical fusion database, and the historical fusion database is used to store historical data after the first data processing, and the first data group includes a third data group; The acquiring, from the first historical database according to the attribute information of the incremental data, a first data group corresponding to the incremental data comprises: According to the attribute information of the incremental data, a third data group corresponding to the incremental data is acquired from the historical fusion database.

4. The method according to claim 1, characterized in that: The first historical database is a historical knowledge graph, the historical knowledge graph is a knowledge graph generated according to a historical fusion database, the historical fusion database is used to store historical data after the first data processing, and the first data group includes a fourth data group; The acquiring the first data group from the first historical database comprises: According to the attribute information of the incremental data, a fourth data group corresponding to the incremental data is obtained from the historical knowledge graph, the data contained in the fourth data group is the data obtained after the data in the third data group is fused, and the third data group includes the historical data obtained after the first data processing in the historical fusion database.

5. The method according to claim 1, characterized in that The performing second data processing on the incremental data to obtain target update data includes: Determine the data to be processed according to the update time corresponding to each data in the incremental data, where the update time corresponding to the data to be processed is the time after the target timestamp; The target data to be processed is subjected to target processing to obtain the target update data, wherein the target processing includes at least one of the following: data processing, pattern mapping processing, and data classification processing.

6. The method according to claim 1, characterized in that The step of determining the target fusion data group according to the comparison result of the data in the first data group and the second data group comprises: Scoring the data included in the first data group and the second data group to determine the degree of association between the data included in the first data group and the second data group; The data in the first data group and the second data group whose scoring results are greater than or equal to a preset score are determined as the target fusion data group.

7. A knowledge graph construction device, characterized in that: include: An acquisition unit is configured to acquire incremental data and a first historical database; the incremental data is data updated after a target timestamp, and the first historical database includes historical data before the target timestamp; The acquisition unit is further configured to execute acquisition of a second historical database; wherein the data in the first historical database is data obtained by performing a first data processing on the data in the second historical database; a processing unit configured to perform a second data processing on the incremental data to obtain target update data, and perform a data fusion processing on the target update data and the second historical database to obtain a target database, wherein the second data processing is different from the first data processing in a processing manner, and the data processing includes at least one of the following: data processing, pattern mapping processing, and data classification processing; The acquisition unit is further configured to acquire a first data group from the first historical database and acquire a second data group from the target database; the first data group is historical data associated with the incremental data in the first historical database, and the second data group is data associated with the incremental data in the target database; The processing unit is further configured to determine a target fusion data group according to a comparison result of data in the first data group and the second data group; The processing unit is further configured to perform fusion processing on the target fusion data group and the data in the first historical database other than the first data group to obtain a target fusion database; A generating unit is configured to generate a target knowledge graph according to the target fusion database.

8. The knowledge graph construction device according to claim 7, characterized in that: The acquisition unit is specifically configured to execute, based on the attribute information of the incremental data, acquiring the first data group corresponding to the incremental data from the first historical database, and acquiring the second data group from the target database based on the attribute information; the attribute information includes at least one of the following: a name identifier of the incremental data and an alias identifier of the incremental data.

9. The knowledge graph construction device according to claim 8, characterized in that: The first historical database is a historical fusion database, and the historical fusion database is used to store historical data after the first data processing, and the first data group includes a third data group; The acquisition unit is specifically configured to acquire, from the historical fusion database according to the attribute information of the incremental data, a third data group corresponding to the incremental data.

10. The knowledge graph construction device according to claim 7, characterized in that: The first historical database is a historical knowledge graph, the historical knowledge graph is a knowledge graph generated according to a historical fusion database, the historical fusion database is used to store historical data after the first data processing, and the first data group includes a fourth data group; The acquisition unit is specifically configured to execute, based on the attribute information of the incremental data, to acquire a fourth data group corresponding to the incremental data from the historical knowledge graph, wherein the data contained in the fourth data group is data obtained after fusion of the data in the third data group, and the third data group includes historical data obtained after first data processing in the historical fusion database.

11. The knowledge graph construction device according to claim 7, characterized in that: The processing unit is specifically configured to determine the data to be processed according to the update time corresponding to each data in the incremental data, wherein the update time corresponding to the data to be processed is the time after the target timestamp; and perform target processing on the data to be processed to obtain the target updated data, wherein the target processing includes at least one of the following: data processing, pattern mapping processing, and data classification processing.

12. The knowledge graph construction device according to claim 7, characterized in that: The processing unit is specifically configured to perform scoring on the data included in the first data group and the second data group, determine the correlation between the data included in the first data group and the second data group; and determine the data in the first data group and the second data group whose scoring results are greater than or equal to a preset score as the target fusion data group.

13. An electronic device, characterized in that: The electronic device comprises: processor; a memory for storing instructions executable by the processor; Wherein, the processor is configured to execute the instructions to implement the knowledge graph construction method as described in any one of claims 1-6.

14. A computer-readable storage medium, characterized in that: Instructions are stored on the computer-readable storage medium. When the instructions in the computer-readable storage medium are executed by a processor of an electronic device, the electronic device is enabled to execute the knowledge graph construction method as described in any one of claims 1 to 6.

15. A computer program product, characterized in that The computer program product includes instructions, which, when executed by a processor, implement the knowledge graph construction method described in any one of claims 1 to 6.

Images