A database update method, system, and device based on Internet AI outbound calling

Abstract

The present invention discloses a database updating method, system and device based on Internet AI outbound calling, the method comprising: obtaining first update data required in the database of an Internet AI outbound calling platform, and obtaining second update data after processing the first update data in accordance with the database data format requirements; adding the second update data to the database of the Internet AI outbound calling platform, and verifying whether the second update data meets the update verification requirements during the operation of the Internet AI outbound calling platform, and if the update verification requirements are met, determining to add the second update data to the database of the Internet AI outbound calling platform; compared with the prior art, the present invention optimizes the data structure of the database without affecting the normal use of the Internet AI outbound calling platform, improves the accuracy and practicability of database updating, and at the same time, reduces the operation load of the database and improves the response speed of database retrieval.

Description

Technical Field

[0001] The present invention belongs to the field of artificial intelligence and database updating, and in particular to a database updating method, system and device based on Internet AI outbound calling. Background Art

[0002] With the rapid development of modern science and technology, especially Internet technology and artificial intelligence, great technological changes have taken place in all walks of life. Work that previously required a lot of manpower has gradually been replaced by intelligent machines. Against the background of new breakthroughs in intelligent algorithms, voice recognition, image recognition, simulation and other technologies, in the technical field of customer service outbound calls, AI outbound call platforms based on artificial intelligence have also emerged to help companies screen potential customers and replace manual customer service. Moreover, by taking advantage of the Internet's big data, the Internet-based AI outbound call platform has gradually been widely used in the service industry, e-commerce industry, financial industry, real estate industry, automobile industry, education and training and other fields.

[0003] Although the Internet-based AI outbound call platform is widely used, there is still a problem that when the database is updated, the updated data added to the database contains a lot of invalid, duplicate, erroneous and other abnormal data, which makes the database structure redundant, the workload is heavy, and the response speed is slow. In addition, the updated data is directly added to the database for use without being verified in actual use, and the accuracy and practicality of the updated data cannot be guaranteed, which further increases the operating burden of the database and reduces the working efficiency of the database. Summary of the invention

[0004] In view of the above problems, the present application provides a database update method, system and device based on Internet AI outbound calls to solve the above technical problems.

[0005] The present invention provides the following technical solutions:

[0006] In a first aspect, the present invention provides a database updating method based on Internet AI outbound calling, the method comprising:

[0007] Obtain the first update data required in the database of the Internet AI outbound call platform, and obtain the second update data after processing the first update data in accordance with the database data format requirements;

[0008] The second updated data is added to the database of the Internet AI outbound call platform, and during the operation of the Internet AI outbound call platform, it is verified whether the second updated data meets the update verification requirements. If the update verification requirements are met, it is determined to add the second updated data to the database of the Internet AI outbound call platform;

[0009] The first updated data is data that is not in the existing database of the Internet AI outbound call platform, or is data that is already in the existing database but needs to be updated;

[0010] Specifically, the first updated data is only used as data to be verified before being added to the database. After being processed into the second updated data that meets the requirements, it still needs to undergo update verification during the subsequent use of the Internet AI outbound call platform. After meeting the requirements of the update verification, it is added to the database as confirmed update data;

[0011] The verification of whether the second updated data meets the update verification requirements is to determine whether the second updated data meets the minimum requirements for entering the database through multiple searches during the operation of the Internet AI outbound call platform before adding the second updated data to the database of the Internet AI outbound call platform;

[0012] The update verification requirement is the minimum requirement for the second update data to pass the update verification and determine that it can enter the database;

[0013] The database updating method based on Internet AI outbound calling of the present invention can immediately add the updated data to the database for use after obtaining the updated data, but does not directly determine to join the existing database. Instead, it confirms to join the database only after verification during the subsequent use of the Internet AI outbound calling platform. Before the database determines the updated data, a further verification link is added, and the verification method is determined based on the subsequent actual use effect as the standard. In this way, the updated data can not only be used directly on the Internet AI outbound calling platform immediately, but also its accuracy can be further verified during the use process. This not only takes into account the efficiency of database updating, but also further improves the accuracy of database update data.

[0014] Further, the obtaining of the first update data required in the database of the Internet AI outbound call platform and processing the first update data in accordance with the database data format requirements to obtain the second update data includes:

[0015] Obtain the first updated data required in the database of the Internet AI outbound calling platform;

[0016] The structure of the data stored in the database is a graph data structure, that is, the database is a graph database, or other databases similar to the graph data structure.

[0017] After performing data cleaning on the first update data, third update data is obtained;

[0018] Data cleaning is the process of discovering and correcting identifiable errors in data files, including checking data consistency, handling invalid values ​​and missing values, and re-examining and verifying data, with the aim of deleting duplicate information, correcting existing errors, and providing data consistency;

[0019] After extracting key information from the third update data, fourth update data is obtained;

[0020] The key information extraction is to extract the key information contained in the third update data and convert it into data conforming to the graph data structure as the fourth update data;

[0021] After feature-marking the fourth updated data, second updated data is obtained;

[0022] The feature annotation is to check the update type of the fourth update data, and perform different processing according to the specific update type as follows:

[0023] (a) If the update type of the fourth update data is a newly added node type, then according to the existing data level classification information, a corresponding level value is added to the fourth update data, and the data content updated by the fourth update data is marked as pending verification, so as to obtain the second update data;

[0024] (b) if the update type of the fourth update data is a modified data type, marking the data content updated by the fourth update data as pending verification to obtain the second update data;

[0025] The update type of the fourth update data is a new node type, which means that when performing the operation of updating data, a new node needs to be added to the existing database, and at least one edge connecting the new node and other nodes needs to be added;

[0026] The existing data level classification information is the existing classification information in a specific field;

[0027] The level value indicates the level at which the fourth updated data is located in the data structure of this field. The larger the level value, the lower the level in the data structure, that is, the smaller the category and the finer the classification. The level value of the fourth updated data is the level value of the obtained second updated data.

[0028] The update type of the fourth update data is a data modification type, which means that when performing the operation of updating data, there is no need to add a new node in the existing database, and only data needs to be added or modified in the existing database;

[0029] The present invention will obtain the first update data required in the database of the Internet AI outbound call platform, and after data cleaning, remove the erroneous or abnormal data therein, obtain the third update data, and then perform key information extraction, screen out the core information, and obtain the fourth update data. Then, the specific update type of the fourth update data is determined, and different processing methods are adopted according to the different update types. If it is a modified data type, it is sufficient to make a to-be-verified mark. If it is a newly added node type, in addition to the to-be-verified mark, the corresponding level value needs to be added. In this way, not only can the acquired update data be screened layer by layer until it meets the requirements of the database, but also before the subsequent verification, the identification information to be verified can be marked in advance to prepare for the subsequent verification process, making the subsequent verification process more efficient and convenient.

[0030] Further, the adding of the second updated data to the database of the Internet AI outbound calling platform, and verifying whether the second updated data meets the update verification requirements during the operation of the Internet AI outbound calling platform, and if the update verification requirements are met, determining to add the second updated data to the database of the Internet AI outbound calling platform includes:

[0031] The second updated data marked as pending verification is added to the database of the Internet AI outbound calling platform for subsequent use by the Internet AI outbound calling platform;

[0032] During the operation of the Internet AI outbound calling platform, obtaining search data related to the second updated data in the database;

[0033] Verifying whether the second update data meets the update verification requirement for entering the database by obtaining the search data related to the second update data in the database;

[0034] If the requirement for updating verification of entering the database is met, the to-be-verified mark of the second updated data is modified to a verified mark;

[0035] If the requirement for the update verification of entering the database is not met, the pending verification mark of the second update data is maintained;

[0036] The modification of the pending verification mark of the second updated data to the verified mark indicates that the updated data is determined to be added to the database of the Internet AI outbound call platform;

[0037] Although the present invention adds the second update data to the database for direct use after acquiring it, it does not determine that the second update data will be finally added to the database. Instead, before determining that the second update data will be added to the database, the second update data is marked as pending verification, so that the state of the second update data before meeting the requirements for update verification for entering the database can be clearly identified. At the same time, in the subsequent operation of the Internet AI outbound call platform, the marked second update data is continuously verified, and after meeting the requirements for update verification for entering the database, the pending verification mark of the second update data is modified to a verified mark; in this way, during the update process of the database, not only can the actual impact of the update data operation on the use of the Internet AI outbound call platform be effectively reduced, but also the update data can be reasonably verified through actual use, thereby improving the accuracy and practicality of the update data. At the same time, the timely marking and modification of the update data status also makes the database update process traceable, providing convenience for other subsequent database operations.

[0038] Furthermore, during the operation of the Internet AI outbound calling platform, obtaining the search data related to the second updated data in the database includes:

[0039] During the operation of the Internet AI outbound call platform, the total number of times the Internet AI outbound call platform performs a search operation on the data in the field to which the second updated data belongs in the database after the second updated data is added is obtained, and at the same time, the number of times the second updated data is retrieved during the search operation, and the number of times the second updated data is retrieved and verified;

[0040] Further, the obtaining of the search data related to the second update data in the database to verify whether the second update data meets the update verification requirement for entering the database includes:

[0041] The method for determining the update verification requirement for entering the database is that the update verification rate of the second update data meets a preset threshold, and the specific model of the update verification rate of the second update data is as follows:

[0042]

[0043] Wherein, μ is the update verification rate of the second update data;

[0044] s0 is the total number of retrieval operations performed by the Internet AI outbound call platform on the data in the field to which the second updated data belongs in the database after the second updated data is added during the verification of the second updated data;

[0045] s1 is the number of times the second updated data is retrieved by the Internet AI outbound call platform when performing a search operation in the database after the second updated data is added during the process of verifying the second updated data;

[0046] s2 is the number of times that the second updated data is retrieved and verified by the Internet AI outbound call platform during the search operation in the database after the second updated data is added during the verification of the second updated data;

[0047] β1 is the first verification weight coefficient;

[0048] β2 is the second verification weight coefficient, and satisfies the relationship β1+β2=1;

[0049] k0 is the level value of the second update data;

[0050] m is the total number of levels in the field to which the second updated data belongs, that is, the total number of levels obtained according to the existing level classification information of the field to which it belongs;

[0051] After the second updated data is acquired, the present invention adds it to the database of the Internet AI outbound call platform, and verifies its accuracy and practicability in the subsequent actual use of the platform, collects relevant information of the retrieval operation on the data in the field to which the second updated data belongs, and then uses the model of the update verification rate of the second updated data to quantify the verification index of the second updated data, and reasonably judges whether the second updated data is determined to enter the database of the Internet AI outbound call platform; this not only has no impact on the use of the updated data, but also verifies its accuracy and practicability through the actual use of the updated data, which is more scientific and reasonable, further reduces various abnormal data such as invalid data, duplicate data, erroneous data after the database is updated, reduces the operating load of the database, optimizes the data structure of the database, and improves the response speed of database retrieval, as well as the accuracy of retrieval and the practicability of data.

[0052] Furthermore, after determining to add the second updated data to the database of the Internet AI outbound calling platform, the method further includes:

[0053] For the second updated data of the database confirmed to join the Internet AI outbound call platform, data anomaly monitoring is carried out, and if an abnormal situation is found in the monitoring, an early warning is issued;

[0054] Furthermore, the second updated data of the database determined to be added to the Internet AI outbound call platform is monitored for data anomalies, and if an abnormality is found in the monitoring, an early warning is issued, including:

[0055] Acquire information related to the second updated data in the database, including the level value of the second updated data, the total number of nodes directly connected to the first node, and the level value of each node directly connected to the first node;

[0056] Determining the data abnormality of the second data in the database according to the information related to the second updated data in the database obtained;

[0057] If the data abnormality is greater than or equal to the preset abnormality warning value, a warning is issued;

[0058] The first node is the node used as the value of k0 in the second update data;

[0059] Further, determining the data abnormality of the second data in the database according to the acquired information related to the second updated data in the database includes:

[0060] The method for determining the data abnormality of the second data in the database is to calculate using a model of the data abnormality of the second data in the database. The specific model is as follows:

[0061]

[0062] Wherein, θ is the data anomaly degree of the second updated data in the database;

[0063] k0 is the level value of the second update data;

[0064] i is the i-th node directly connected to the first node;

[0065] The first node is the node used as the value of k0 in the second update data;

[0066] n is the total number of nodes directly connected to the first node;

[0067] k i is the level value of the i-th node directly connected to the first node;

[0068] δ is the level deviation from the baseline value;

[0069] k max is the highest value among the level values ​​of all nodes directly connected to the first node;

[0070] k min is the lowest value among the level values ​​of all nodes directly connected to the first node;

[0071] After verifying the second updated data, the present invention also monitors the data anomaly from the relationship between the second updated data and other data directly connected to it in the database, and issues an early warning in a timely manner. This not only ensures the accuracy and practicality of the updated data, but also makes a technical analysis and judgment on whether the structure of the updated data in the database is reasonable. On the basis of the updated data, the stability of the database structure is further guaranteed.

[0072] In a second aspect, the present invention provides a database update system based on Internet AI outbound calling, comprising:

[0073] A data acquisition module, used to obtain the first update data required in the database of the Internet AI outbound call platform, and obtain the second update data after processing the first update data in accordance with the database data format requirements;

[0074] A data verification module, used for adding the second updated data to the database of the Internet AI outbound calling platform, and verifying whether the second updated data meets the update verification requirements during the operation of the Internet AI outbound calling platform, and if the update verification requirements are met, determining to add the second updated data to the database of the Internet AI outbound calling platform;

[0075] The data monitoring module is used to monitor the second updated data of the database determined to join the Internet AI outbound call platform for data anomalies, and issue an early warning if an abnormality is found in the monitoring;

[0076] The first updated data is data that is not in the existing database of the Internet AI outbound call platform, or is data that is already in the existing database but needs to be updated;

[0077] Specifically, the first updated data is only used as data to be verified before being added to the database. After being processed into the second updated data that meets the requirements, it still needs to undergo update verification during the subsequent use of the Internet AI outbound call platform. After meeting the requirements of the update verification, it is added to the database as confirmed update data;

[0078] The verification of whether the second updated data meets the update verification requirements is to determine whether the second updated data meets the minimum requirements for entering the database through multiple searches during the operation of the Internet AI outbound call platform before adding the second updated data to the database of the Internet AI outbound call platform;

[0079] The update verification requirement is the minimum requirement for the second update data to pass the update verification and to determine its entry into the database.

[0080] In a third aspect, the present invention provides a computer device comprising a memory and a processor; the memory is used to store a computer program; the processor is used to implement the method described in the first aspect when executing the computer program.

[0081] In a fourth aspect, the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein the program, when executed by a processor, implements the method described in the first aspect.

[0082] Compared with the prior art, the present invention optimizes the data structure of the database and improves the accuracy and practicality of database updates without affecting the normal use of the Internet AI outbound calling platform. At the same time, it reduces the operating load of the database and improves the response speed of database retrieval. BRIEF DESCRIPTION OF THE DRAWINGS

[0083] For ease of explanation, the present invention is described in detail with reference to the following specific implementations and the accompanying drawings.

[0084] Figure 1 One of the schematic flow charts of the method of the present invention;

[0085] Figure 2 This is the second schematic diagram of the method flow of the present invention;

[0086] Figure 3 The third schematic diagram of the method flow of the present invention;

[0087] Figure 4 This is the fourth schematic diagram of the method flow of the present invention;

[0088] Figure 5 It is a schematic diagram of the system structure of the present invention. DETAILED DESCRIPTION

[0089] The following will be combined with the figures in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0090] The overall idea of ​​the technical solution in the embodiment of the present invention is as follows:

[0091] During the operation of the Internet AI outbound call platform, when it needs to perform a data update operation, the required update data in the database of the Internet AI outbound call platform is obtained, and after processing the update data, it is added to the database of the Internet AI outbound call platform for direct use. At the same time, in the subsequent use of the Internet AI outbound call platform, it is verified whether the update data meets the requirements. If the requirements are met, it is determined to add the second update data to the database of the Internet AI outbound call platform; therefore, without affecting the normal use of the Internet AI outbound call platform, the process of updating database data is optimized, the accuracy and practicality of database updates are improved, and at the same time, the operating load of the database is reduced, the response speed of database retrieval is improved, and the problems of redundant database structure, heavy workload, slow response speed, and inability to guarantee the accuracy and practicality of updated data are solved.

[0092] In order to better understand the above technical solution, the above technical solution will be described in detail below in conjunction with the accompanying drawings and specific implementation methods.

[0093] Example 1

[0094] like Figure 1 As shown, the present invention provides a database updating method based on Internet AI outbound calling, the method comprising:

[0095] Obtain the first update data required in the database of the Internet AI outbound call platform, and obtain the second update data after processing the first update data in accordance with the database data format requirements;

[0096] The second updated data is added to the database of the Internet AI outbound call platform, and during the operation of the Internet AI outbound call platform, it is verified whether the second updated data meets the update verification requirements. If the update verification requirements are met, it is determined to add the second updated data to the database of the Internet AI outbound call platform;

[0097] The first updated data is data that is not in the existing database of the Internet AI outbound call platform, or is data that is already in the existing database but needs to be updated;

[0098] Specifically, the first updated data is only used as data to be verified before being added to the database. After being processed into the second updated data that meets the requirements, it still needs to undergo update verification during the subsequent use of the Internet AI outbound call platform. After meeting the requirements of the update verification, it is added to the database as confirmed update data;

[0099] The verification of whether the second updated data meets the update verification requirements is to determine whether the second updated data meets the minimum requirements for entering the database through multiple searches during the operation of the Internet AI outbound call platform before adding the second updated data to the database of the Internet AI outbound call platform;

[0100] The update verification requirement is the minimum requirement for the second update data to pass the update verification and determine that it can enter the database;

[0101] The database updating method based on Internet AI outbound calling of the present invention can immediately add the updated data to the database for use after obtaining the updated data, but does not directly determine to join the existing database. Instead, it confirms to join the database only after verification during the subsequent use of the Internet AI outbound calling platform. Before the database determines the updated data, a further verification link is added, and the verification method is determined based on the subsequent actual use effect as the standard. In this way, the updated data can not only be used directly on the Internet AI outbound calling platform immediately, but also its accuracy can be further verified during the use process. This not only takes into account the efficiency of database updating, but also further improves the accuracy of database update data.

[0102] Further, such as Figure 2 As shown, the first update data required in the database of the Internet AI outbound call platform is obtained, and the first update data is processed in accordance with the database data format requirements to obtain the second update data, including:

[0103] Obtain the first update data required in the database of the Internet AI outbound calling platform;

[0104] Specifically, the first update data is obtained by, including but not limited to, the following:

[0105] (1) During an outbound call conversation, if no matching data is found after the first data search in the platform's existing database, the Internet big data is used for a second search, and the matching data of this search is used as the first updated data;

[0106] (2) matching data retrieved from historical outbound call conversation data on the Internet AI outbound call platform is used as the first update data;

[0107] (3) manually added update data, as the first update data;

[0108] The structure of the data stored in the database is a graph data structure, that is, the database is a graph database, or other databases similar to the graph data structure.

[0109] Compared with traditional relational databases (such as SQL), the main difference between graph databases (such as Neo4j) and traditional relational databases (such as SQL) is that the storage structure is significantly different from that of traditional relational databases. Graph databases do not set specific data in various tables and then associate them through foreign keys, but directly connect nodes and their associated nodes through edges (i.e. relationships). In this way, when you need to query or retrieve the associated nodes of a node, you can directly follow the links from the node to retrieve step by step, without having to retrieve all the nodes as in the past as with traditional relational databases. In terms of retrieval time and response speed, they are significantly better than traditional relational databases. For example, for specific data in a certain field in the database of an Internet AI outbound call platform, the graph database will treat the entities in the data as nodes, and the nodes are connected through different edges (i.e. various relationships);

[0110] Therefore, the advantage of graph databases in the application of Internet AI outbound calls is that when using graph databases for retrieval operations, the workload of the retrieval only needs to include the retrieved node and its related nodes (i.e., a local graph), without having to retrieve all the nodes. In this way, when the number of nodes in the entire database increases or the graph structure expands, the retrieval workload for a certain retrieval operation does not change much. In addition, in a graph database, even if the retrieval level becomes deeper, it only adds one level of local graph and one level of nodes. Compared with traditional relational databases, it not only does not increase the complexity of the database and the workload of data retrieval, but the retrieval logic is relatively clear. This is also the reason why graph databases are more suitable than traditional relational databases for the application of Internet AI outbound calls.

[0111] Specifically, the database is a database used to retrieve and call data during the operation of the Internet AI outbound calling platform. It can be a local database in the Internet AI outbound calling platform server, or a cloud database, or a combination of the two databases. The database storage method can be centralized or distributed, such as a centralized local database or a distributed cloud database.

[0112] After performing data cleaning on the first update data, third update data is obtained;

[0113] Data cleaning is the process of discovering and correcting identifiable errors in data files, including checking data consistency, handling invalid values ​​and missing values, and re-examining and verifying data, with the aim of deleting duplicate information, correcting existing errors, and providing data consistency;

[0114] After extracting key information from the third update data, fourth update data is obtained;

[0115] The key information extraction is to extract the key information contained in the third update data and convert it into data conforming to the graph data structure as the fourth update data;

[0116] After feature-marking the fourth updated data, second updated data is obtained;

[0117] The feature annotation is to check the update type of the fourth update data, and perform different processing according to the specific update type as follows:

[0118] (a) If the update type of the fourth update data is a newly added node type, then according to the existing data level classification information, a corresponding level value is added to the fourth update data, and the data content updated by the fourth update data is marked as pending verification, so as to obtain the second update data;

[0119] (b) if the update type of the fourth update data is a modified data type, marking the data content updated by the fourth update data as pending verification to obtain the second update data;

[0120] The update type of the fourth update data is a new node type, which means that when performing the operation of updating data, a new node needs to be added to the existing database, and at least one edge connecting the new node and other nodes needs to be added;

[0121] Specifically, the data content updated by the fourth update data of the new node type is specifically the data contained in the new node (i.e., entity) (such as node category or node label, node attribute, etc.), and the data contained in the edge (i.e., relationship) connecting the new node and other nodes in the database (such as the direction of the relationship, the attribute of the relationship, etc.);

[0122] The existing data level classification information is the existing classification information in a specific field;

[0123] Specifically, first determine which field the information (i.e., node category or node label) contained in the fourth updated data belongs to, and then add a specific level value after comparing it with the fourth updated data based on the existing hierarchical information of the field;

[0124] The level value indicates the level at which the fourth updated data is located in the data structure of this field. The larger the level value, the lower the level in the data structure, that is, the smaller the category and the finer the classification. The level value of the fourth updated data is the level value of the obtained second updated data.

[0125] Specifically, the node category or node label is completed in the process of obtaining the fourth update data after extracting the key information from the third update data, that is, in this process, the node category or node label is extracted while extracting the key information;

[0126] The update type of the fourth update data is a data modification type, which means that when performing the operation of updating data, there is no need to add a new node in the existing database, and only data needs to be added or modified in the existing database;

[0127] Specifically, the update data content included in the fourth update data of the update type being the modified data type includes but is not limited to the following situations:

[0128] (1) Modify the relevant data of the node in the existing database; for example, modify the node category or node label, modify the node attributes, etc.;

[0129] (2) Modify the edge-related data in the existing database; for example, modify the edge relationship direction, modify the edge relationship attributes, etc.;

[0130] (3) Add data to nodes in the existing database; for example, add new categories or new labels to nodes, add new attributes to nodes, etc.

[0131] (4) Add data to the edges in the existing database; for example, add new relationships to the edges, add new relationship attributes to the edges, etc.

[0132] The pending verification mark only indicates that the data with the pending verification mark is in a state of waiting for further verification and does not participate in actual data interaction;

[0133] Specifically, each fourth update data can only be any one type of adding a node type or modifying data type, and only contains one specific update data. For example, an update operation of modifying data type needs to modify the category of an existing node and also needs to modify the attributes of the existing node. Then, the update operation includes two fourth update data, namely, the fourth update data A for modifying the category of the existing node, and the fourth update data B for modifying the attributes of the existing node, and so on. Similarly, an update operation of adding a node type needs to add multiple nodes, so the update operation requires multiple corresponding fourth update data.

[0134] The present invention will obtain the first update data required in the database of the Internet AI outbound call platform, and after data cleaning, remove the erroneous or abnormal data therein, obtain the third update data, and then perform key information extraction, screen out the core information, and obtain the fourth update data. Then, the specific update type of the fourth update data is determined, and different processing methods are adopted according to the different update types. If it is a modified data type, it is sufficient to make a to-be-verified mark. If it is a newly added node type, in addition to the to-be-verified mark, the corresponding level value needs to be added. In this way, not only can the acquired update data be screened layer by layer until it meets the requirements of the database, but also before the subsequent verification, the identification information to be verified can be marked in advance to prepare for the subsequent verification process, making the subsequent verification process more efficient and convenient.

[0135] Further, such as Figure 3 As shown, the second updated data is added to the database of the Internet AI outbound call platform, and during the operation of the Internet AI outbound call platform, the second updated data is verified whether it meets the update verification requirements. If the update verification requirements are met, it is determined that the second updated data is added to the database of the Internet AI outbound call platform, including:

[0136] The second updated data marked as pending verification is added to the database of the Internet AI outbound calling platform for subsequent use by the Internet AI outbound calling platform;

[0137] During the operation of the Internet AI outbound calling platform, obtaining search data related to the second updated data in the database;

[0138] Verifying whether the second update data meets the update verification requirement for entering the database by obtaining the search data related to the second update data in the database;

[0139] If the requirement for updating verification of entering the database is met, the to-be-verified mark of the second updated data is modified to a verified mark;

[0140] If the requirement for the update verification of entering the database is not met, the pending verification mark of the second update data is maintained;

[0141] The modification of the pending verification mark of the second updated data to the verified mark indicates that the updated data is determined to be added to the database of the Internet AI outbound call platform;

[0142] Although the present invention adds the second update data to the database for direct use after acquiring it, it does not determine that the second update data will be finally added to the database. Instead, before determining that the second update data will be added to the database, the second update data is marked as pending verification, so that the state of the second update data before meeting the requirements for update verification for entering the database can be clearly identified. At the same time, in the subsequent operation of the Internet AI outbound call platform, the marked second update data is continuously verified, and after meeting the requirements for update verification for entering the database, the pending verification mark of the second update data is modified to a verified mark; in this way, during the update process of the database, not only can the actual impact of the update data operation on the use of the Internet AI outbound call platform be effectively reduced, but also the update data can be reasonably verified through actual use, thereby improving the accuracy and practicality of the update data. At the same time, the timely marking and modification of the update data status also makes the database update process traceable, providing convenience for other subsequent database operations.

[0143] Furthermore, during the operation of the Internet AI outbound calling platform, obtaining the search data related to the second updated data in the database includes:

[0144] During the operation of the Internet AI outbound call platform, the total number of times the Internet AI outbound call platform performs a search operation on the data in the field to which the second updated data belongs in the database after the second updated data is added is obtained, and at the same time, the number of times the second updated data is retrieved during the search operation, and the number of times the second updated data is retrieved and verified;

[0145] Specifically, the second updated data is retrieved by the Internet AI outbound call platform directly or indirectly retrieving the second updated data in the process of verifying the second updated data; the direct retrieval means that the searched keyword and the second updated data meet the search matching requirements and are directly retrieved; the indirect retrieval means that although the searched keyword and the second updated data do not meet the search matching requirements, other data directly related to the second updated data are retrieved;

[0146] The second updated data is retrieved and verified, which means that on the basis of retrieving the second updated data, verification is performed on the Internet AI outbound call platform during an outbound call dialogue process. When the verification result is that the retrieved data is correct, the verification is passed;

[0147] Further, the obtaining of the search data related to the second update data in the database to verify whether the second update data meets the update verification requirement for entering the database includes:

[0148] The method for determining the update verification requirement for entering the database is that the update verification rate of the second update data meets a preset threshold, and the specific model of the update verification rate of the second update data is as follows:

[0149]

[0150] Wherein, μ is the update verification rate of the second update data;

[0151] Specifically, the update verification rate of the second update data is a quantitative indicator used to characterize whether the second update data meets the minimum requirement for entering the database through multiple searches during the operation of the Internet AI outbound call platform before entering the database; that is, when μ is greater than or equal to the preset threshold of the update verification rate, the update verification requirement for entering the database is met;

[0152] s0 is the total number of retrieval operations performed by the Internet AI outbound call platform on the data in the field to which the second updated data belongs in the database after the second updated data is added during the verification of the second updated data;

[0153] s1 is the number of times the second updated data is retrieved by the Internet AI outbound call platform when performing a search operation in the database after the second updated data is added during the process of verifying the second updated data;

[0154] s2 is the number of times that the second updated data is retrieved and verified by the Internet AI outbound call platform during the search operation in the database after the second updated data is added during the verification of the second updated data;

[0155] Specifically, the second updated data is retrieved by the Internet AI outbound call platform directly or indirectly retrieving the second updated data in the process of verifying the second updated data; the direct retrieval means that the searched keyword and the second updated data meet the search matching requirements and are directly retrieved; the indirect retrieval means that although the searched keyword and the second updated data do not meet the search matching requirements, other data directly related to the second updated data are retrieved;

[0156] The second updated data is retrieved and verified, which means that on the basis of retrieving the second updated data, verification is performed on the Internet AI outbound call platform during an outbound call dialogue process. When the verification result is that the retrieved data is correct, the verification is passed;

[0157] β1 is the first verification weight coefficient;

[0158] β2 is the second verification weight coefficient, and satisfies the relationship β1+β2=1;

[0159] Specifically, the first verification weight coefficient β1 represents the influence weight of the situation of retrieving the second update data on the update verification rate μ of the second update data, that is, the larger β1 is, the greater the influence of this situation on the update verification rate of the second update data is, and μ is also larger;

[0160] The second verification weight coefficient β2 represents the influence weight of the situation that the second update data is retrieved and verified on the update verification rate μ of the second update data, that is, the larger β2 is, the greater the influence of this situation on the update verification rate of the second update data, and μ is also larger;

[0161] k0 is the level value of the second update data;

[0162] Specifically, there are two cases for the value of k0, as follows:

[0163] (1) If the update type of the second update data is a newly added node type, k0 is the level value added by the second update data;

[0164] (2) If the update type of the second update data is a modified data type, k0 needs to be set according to the specific content of the update;

[0165] ① When the specific content of the update only involves a single node (i.e., entity), such as a node category or label, or a node attribute, k0 is the level value of the node;

[0166] ② When the specific content of the update only involves a single edge (i.e., relationship), such as the direction or attribute of the edge, k0 is the smaller level value of the two nodes directly connected by the edge;

[0167] m is the total number of levels in the field to which the second updated data belongs, that is, the total number of levels obtained according to the existing level classification information of the field to which it belongs;

[0168] After the second updated data is acquired, the present invention adds it to the database of the Internet AI outbound call platform, and verifies its accuracy and practicability in the subsequent actual use of the platform, collects relevant information of the retrieval operation on the data in the field to which the second updated data belongs, and then uses the model of the update verification rate of the second updated data to quantify the verification index of the second updated data, and reasonably judges whether the second updated data is determined to enter the database of the Internet AI outbound call platform; this not only has no impact on the use of the updated data, but also verifies its accuracy and practicability through the actual use of the updated data, which is more scientific and reasonable, further reduces various abnormal data such as invalid data, duplicate data, erroneous data after the database is updated, reduces the operating load of the database, optimizes the data structure of the database, and improves the response speed of database retrieval, as well as the accuracy of retrieval and the practicability of data.

[0169] Further, such as Figure 4 As shown, after the determination to add the second update data to the database of the Internet AI outbound call platform, it also includes:

[0170] For the second updated data of the database confirmed to join the Internet AI outbound call platform, data anomaly monitoring is carried out, and if an abnormal situation is found in the monitoring, an early warning is issued;

[0171] Furthermore, the second updated data of the database determined to be added to the Internet AI outbound call platform is monitored for data anomalies, and if an abnormality is found in the monitoring, an early warning is issued, including:

[0172] Acquire information related to the second updated data in the database, including the level value of the second updated data, the total number of nodes directly connected to the first node, and the level value of each node directly connected to the first node;

[0173] Determining the data abnormality of the second data in the database according to the information related to the second updated data in the database obtained;

[0174] If the data abnormality is greater than or equal to the preset abnormality warning value, a warning is issued;

[0175] The first node is the node used as the value of k0 in the second update data;

[0176] Specifically, the data anomaly monitoring of the second data is carried out on the basis of the existing database, and the purpose is to judge whether the hierarchical position allocated to the determined updated data in the database is within an appropriate range; the issuance of the warning is to issue specific warning information within the Internet AI outbound call platform. At the same time, in addition to the content of the abnormal updated data, the warning information may also include the location of the abnormal updated data, and the information and location of other directly connected nodes, so as to facilitate the viewing and modification of data during background maintenance.

[0177] Further, determining the data abnormality of the second data in the database according to the acquired information related to the second updated data in the database includes:

[0178] The method for determining the data abnormality of the second data in the database is to calculate using a model of the data abnormality of the second data in the database. The specific model is as follows:

[0179]

[0180] Wherein, θ is the data anomaly degree of the second updated data in the database;

[0181] k0 is the level value of the second update data;

[0182] i is the i-th node directly connected to the first node;

[0183] The first node is the node used as the value of k0 in the second update data;

[0184] n is the total number of nodes directly connected to the first node;

[0185] k i is the level value of the i-th node directly connected to the first node;

[0186] δ is the level deviation from the baseline value;

[0187] k max is the highest value among the level values ​​of all nodes directly connected to the first node;

[0188] k min is the lowest value among the level values ​​of all nodes directly connected to the first node;

[0189] After verifying the second updated data, the present invention also monitors the data anomaly from the relationship between the second updated data and other data directly connected to it in the database, and issues an early warning in a timely manner. This not only ensures the accuracy and practicality of the updated data, but also makes a technical analysis and judgment on whether the structure of the updated data in the database is reasonable. On the basis of the updated data, the stability of the database structure is further guaranteed.

[0190] Example 2

[0191] like Figure 5 As shown, the present invention provides a database update system based on Internet AI outbound calling, the system comprising:

[0192] A data acquisition module, used to obtain the first update data required in the database of the Internet AI outbound call platform, and obtain the second update data after processing the first update data in accordance with the database data format requirements;

[0193] A data verification module, used for adding the second updated data to the database of the Internet AI outbound calling platform, and verifying whether the second updated data meets the update verification requirements during the operation of the Internet AI outbound calling platform, and if the update verification requirements are met, determining to add the second updated data to the database of the Internet AI outbound calling platform;

[0194] The data monitoring module is used to monitor the second updated data of the database determined to join the Internet AI outbound call platform for data anomalies, and issue an early warning if an abnormality is found in the monitoring;

[0195] The first updated data is data that is not in the existing database of the Internet AI outbound call platform, or is data that is already in the existing database but needs to be updated;

[0196] Specifically, the first updated data is only used as data to be verified before being added to the database. After being processed into the second updated data that meets the requirements, it still needs to undergo update verification during the subsequent use of the Internet AI outbound call platform. After meeting the requirements of the update verification, it is added to the database as confirmed update data;

[0197] The verification of whether the second updated data meets the update verification requirements is to determine whether the second updated data meets the minimum requirements for entering the database through multiple searches during the operation of the Internet AI outbound call platform before adding the second updated data to the database of the Internet AI outbound call platform;

[0198] The update verification requirement is the minimum requirement for the second update data to pass the update verification and determine that it can enter the database;

[0199] Further, such as Figure 2 As shown, the first update data required in the database of the Internet AI outbound call platform is obtained, and the first update data is processed in accordance with the database data format requirements to obtain the second update data, including:

[0200] Obtain the first update data required in the database of the Internet AI outbound calling platform;

[0201] Specifically, the first update data is obtained by, including but not limited to, the following:

[0202] (1) During an outbound call conversation, if no matching data is found after the first data search in the platform's existing database, the Internet big data is used for a second search, and the matching data of this search is used as the first updated data;

[0203] (2) matching data retrieved from historical outbound call conversation data on the Internet AI outbound call platform is used as the first update data;

[0204] (3) manually added update data, as the first update data;

[0205] The structure of the data stored in the database is a graph data structure, that is, the database is a graph database, or other databases similar to the graph data structure.

[0206] Specifically, the database is a database used to retrieve and call data during the operation of the Internet AI outbound calling platform. It can be a local database in the Internet AI outbound calling platform server, or a cloud database, or a combination of the two databases. The database storage method can be centralized or distributed, such as a centralized local database or a distributed cloud database.

[0207] After performing data cleaning on the first update data, third update data is obtained;

[0208] Data cleaning is the process of discovering and correcting identifiable errors in data files, including checking data consistency, handling invalid values ​​and missing values, and re-examining and verifying data, with the aim of deleting duplicate information, correcting existing errors, and providing data consistency;

[0209] After extracting key information from the third update data, fourth update data is obtained;

[0210] The key information extraction is to extract the key information contained in the third update data and convert it into data conforming to the graph data structure as the fourth update data;

[0211] After feature-marking the fourth updated data, second updated data is obtained;

[0212] The feature annotation is to check the update type of the fourth update data, and perform different processing according to the specific update type as follows:

[0213] (a) If the update type of the fourth update data is a newly added node type, then according to the existing data level classification information, a corresponding level value is added to the fourth update data, and the data content updated by the fourth update data is marked as pending verification, so as to obtain the second update data;

[0214] (b) if the update type of the fourth update data is a modified data type, marking the data content updated by the fourth update data as pending verification to obtain the second update data;

[0215] The update type of the fourth update data is a new node type, which means that when performing the operation of updating data, a new node needs to be added to the existing database, and at least one edge connecting the new node and other nodes needs to be added;

[0216] Specifically, the data content updated by the fourth update data of the new node type is specifically the data contained in the new node (i.e., entity) (such as node category or node label, node attribute, etc.), and the data contained in the edge (i.e., relationship) connecting the new node and other nodes in the database (such as the direction of the relationship, the attribute of the relationship, etc.);

[0217] The existing data level classification information is the existing classification information in a specific field;

[0218] Specifically, first determine which field the information (i.e., node category or node label) contained in the fourth updated data belongs to, and then add a specific level value after comparing it with the fourth updated data based on the existing hierarchical information of the field;

[0219] The level value indicates the level at which the fourth updated data is located in the data structure of this field. The larger the level value, the lower the level in the data structure, that is, the smaller the category and the finer the classification. The level value of the fourth updated data is the level value of the obtained second updated data.

[0220] Specifically, the node category or node label is completed in the process of obtaining the fourth update data after extracting the key information from the third update data, that is, in this process, the node category or node label is extracted while extracting the key information;

[0221] The update type of the fourth update data is a data modification type, which means that when performing the operation of updating data, there is no need to add a new node in the existing database, and only data needs to be added or modified in the existing database;

[0222] Specifically, the update data content included in the fourth update data of the update type being the modified data type includes but is not limited to the following situations:

[0223] (1) Modify the relevant data of the node in the existing database; for example, modify the node category or node label, modify the node attributes, etc.;

[0224] (2) Modify the edge-related data in the existing database; for example, modify the edge relationship direction, modify the edge relationship attributes, etc.;

[0225] (3) Add data to nodes in the existing database; for example, add new categories or new labels to nodes, add new attributes to nodes, etc.

[0226] (4) Add data to the edges in the existing database; for example, add new relationships to the edges, add new relationship attributes to the edges, etc.

[0227] The pending verification mark only indicates that the data with the pending verification mark is in a state of waiting for further verification and does not participate in actual data interaction;

[0228] Specifically, each fourth update data can only be any one type of adding a node type or modifying data type, and only contains one specific update data. For example, an update operation of modifying data type needs to modify the category of an existing node and also needs to modify the attributes of the existing node. Then, the update operation includes two fourth update data, namely, the fourth update data A for modifying the category of the existing node, and the fourth update data B for modifying the attributes of the existing node, and so on. Similarly, an update operation of adding a node type needs to add multiple nodes, so the update operation requires multiple corresponding fourth update data.

[0229] Further, such as Figure 3 As shown, the second updated data is added to the database of the Internet AI outbound call platform, and during the operation of the Internet AI outbound call platform, the second updated data is verified whether it meets the update verification requirements. If the update verification requirements are met, it is determined that the second updated data is added to the database of the Internet AI outbound call platform, including:

[0230] The second updated data marked as pending verification is added to the database of the Internet AI outbound calling platform for subsequent use by the Internet AI outbound calling platform;

[0231] During the operation of the Internet AI outbound calling platform, obtaining search data related to the second updated data in the database;

[0232] Verifying whether the second update data meets the update verification requirement for entering the database by obtaining the search data related to the second update data in the database;

[0233] If the requirement for updating verification of entering the database is met, the to-be-verified mark of the second updated data is modified to a verified mark;

[0234] If the requirement for the update verification of entering the database is not met, the pending verification mark of the second update data is maintained;

[0235] The modification of the pending verification mark of the second updated data to the verified mark indicates that the updated data is determined to be added to the database of the Internet AI outbound call platform;

[0236] Furthermore, during the operation of the Internet AI outbound calling platform, obtaining the search data related to the second updated data in the database includes:

[0237] During the operation of the Internet AI outbound call platform, the total number of times the Internet AI outbound call platform performs a search operation on the data in the field to which the second updated data belongs in the database after the second updated data is added is obtained, and at the same time, the number of times the second updated data is retrieved during the search operation, and the number of times the second updated data is retrieved and verified;

[0238] Specifically, the second updated data is retrieved by the Internet AI outbound call platform directly or indirectly retrieving the second updated data in the process of verifying the second updated data; the direct retrieval means that the searched keyword and the second updated data meet the search matching requirements and are directly retrieved; the indirect retrieval means that although the searched keyword and the second updated data do not meet the search matching requirements, other data directly related to the second updated data are retrieved;

[0239] The second updated data is retrieved and verified, which means that on the basis of retrieving the second updated data, verification is performed on the Internet AI outbound call platform during an outbound call dialogue process. When the verification result is that the retrieved data is correct, the verification is passed;

[0240] Further, the obtaining of the search data related to the second update data in the database to verify whether the second update data meets the update verification requirement for entering the database includes:

[0241] The method for determining the update verification requirement for entering the database is that the update verification rate of the second update data meets a preset threshold, and the specific model of the update verification rate of the second update data is as follows:

[0242]

[0243] Wherein, μ is the update verification rate of the second update data;

[0244] Specifically, the update verification rate of the second update data is a quantitative indicator used to characterize whether the second update data meets the minimum requirement for entering the database through multiple searches during the operation of the Internet AI outbound call platform before entering the database; that is, when μ is greater than or equal to the preset threshold of the update verification rate, the update verification requirement for entering the database is met;

[0245] s0 is the total number of retrieval operations performed by the Internet AI outbound call platform on the data in the field to which the second updated data belongs in the database after the second updated data is added during the verification of the second updated data;

[0246] s1 is the number of times the second updated data is retrieved by the Internet AI outbound call platform when performing a search operation in the database after the second updated data is added during the process of verifying the second updated data;

[0247] s2 is the number of times that the second updated data is retrieved and verified by the Internet AI outbound call platform during the search operation in the database after the second updated data is added during the verification of the second updated data;

[0248] Specifically, the second updated data is retrieved by the Internet AI outbound call platform directly or indirectly retrieving the second updated data in the process of verifying the second updated data; the direct retrieval means that the searched keyword and the second updated data meet the search matching requirements and are directly retrieved; the indirect retrieval means that although the searched keyword and the second updated data do not meet the search matching requirements, other data directly related to the second updated data are retrieved;

[0249] The second updated data is retrieved and verified, which means that on the basis of retrieving the second updated data, verification is performed on the Internet AI outbound call platform during an outbound call dialogue process. When the verification result is that the retrieved data is correct, the verification is passed;

[0250] β1 is the first verification weight coefficient;

[0251] β2 is the second verification weight coefficient, and satisfies the relationship β1+β2=1;

[0252] Specifically, the first verification weight coefficient β1 represents the influence weight of the situation of retrieving the second update data on the update verification rate μ of the second update data, that is, the larger β1 is, the greater the influence of this situation on the update verification rate of the second update data is, and μ is also larger;

[0253] The second verification weight coefficient β2 represents the influence weight of the situation that the second update data is retrieved and verified on the update verification rate μ of the second update data, that is, the larger β2 is, the greater the influence of this situation on the update verification rate of the second update data, and μ is also larger;

[0254] k0 is the level value of the second update data;

[0255] Specifically, there are two cases for the value of k0, as follows:

[0256] (1) If the update type of the second update data is a newly added node type, k0 is the level value added by the second update data;

[0257] (2) If the update type of the second update data is a modified data type, k0 needs to be set according to the specific content of the update;

[0258] ① When the specific content of the update only involves a single node (i.e., entity), such as a node category or label, or a node attribute, k0 is the level value of the node;

[0259] ② When the specific content of the update only involves a single edge (i.e., relationship), such as the direction or attribute of the edge, k0 is the smaller level value of the two nodes directly connected by the edge;

[0260] m is the total number of levels in the field to which the second updated data belongs, that is, the total number of levels obtained according to the existing level classification information of the field to which it belongs;

[0261] Further, such as Figure 4 As shown, after the determination to add the second update data to the database of the Internet AI outbound call platform, it also includes:

[0262] For the second updated data of the database confirmed to join the Internet AI outbound call platform, data anomaly monitoring is carried out, and if an abnormal situation is found in the monitoring, an early warning is issued;

[0263] Furthermore, the second updated data of the database determined to be added to the Internet AI outbound call platform is monitored for data anomalies, and if an abnormality is found in the monitoring, an early warning is issued, including:

[0264] Acquire information related to the second updated data in the database, including the level value of the second updated data, the total number of nodes directly connected to the first node, and the level value of each node directly connected to the first node;

[0265] Determining the data abnormality of the second data in the database according to the information related to the second updated data in the database obtained;

[0266] If the data abnormality is greater than or equal to the preset abnormality warning value, a warning is issued;

[0267] The first node is the node used as the value of k0 in the second update data;

[0268] Specifically, the data anomaly monitoring of the second data is carried out on the basis of the existing database, and the purpose is to judge whether the hierarchical position allocated to the determined updated data in the database is within an appropriate range; the issuance of the warning is to issue specific warning information within the Internet AI outbound call platform. At the same time, in addition to the content of the abnormal updated data, the warning information may also include the location of the abnormal updated data, and the information and location of other directly connected nodes, so as to facilitate the viewing and modification of data during background maintenance.

[0269] Further, determining the data abnormality of the second data in the database according to the acquired information related to the second updated data in the database includes:

[0270] The method for determining the data abnormality of the second data in the database is to calculate using a model of the data abnormality of the second data in the database. The specific model is as follows:

[0271]

[0272] Wherein, θ is the data anomaly degree of the second updated data in the database;

[0273] k0 is the level value of the second update data;

[0274] i is the i-th node directly connected to the first node;

[0275] The first node is the node used as the value of k0 in the second update data;

[0276] n is the total number of nodes directly connected to the first node;

[0277] k i is the level value of the i-th node directly connected to the first node;

[0278] δ is the level deviation from the baseline value;

[0279] k max is the highest value among the level values ​​of all nodes directly connected to the first node;

[0280] k min is the lowest value among the level values ​​of all nodes directly connected to the first node;

[0281] Example 3

[0282] The present invention provides a computer-readable storage medium having a computer program stored thereon, wherein the program, when executed by a processor, implements the method described in the above-mentioned embodiment 1.

[0283] Example 4

[0284] The present invention provides a computer device, characterized in that it includes a memory and a processor; the memory is used to store a computer program; the processor is used to implement the method described in the above embodiment 1 when executing the computer program.

[0285] Compared with the prior art, the present invention optimizes the data structure of the database and improves the accuracy and practicality of database updates without affecting the normal use of the Internet AI outbound calling platform. At the same time, it reduces the operating load of the database and improves the response speed of database retrieval.

[0286] Those skilled in the art can clearly understand that, for the convenience and simplicity of description, the specific working processes of the systems, media, devices, modules and units described above can refer to the corresponding processes in the aforementioned method embodiments, and will not be repeated here. In addition, each functional module or unit in each embodiment of the present application can be integrated into a processing module or unit, or each module or unit can exist physically separately, or two or more modules or units can be integrated into one module or unit. The above-mentioned integrated modules or units can be implemented in the form of hardware or in the form of software functional units.

[0287] The integrated system, module, unit, etc., if implemented in the form of a software functional unit and sold or used as an independent product, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application is essentially or the part that contributes to the prior art or all or part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions to enable a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes: various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a disk or an optical disk.

[0288] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit it. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A database updating method based on Internet AI outbound calling, characterized in that: The method comprises: Obtain the first update data required in the database of the Internet AI outbound call platform, and obtain the second update data after processing the first update data in accordance with the database data format requirements; The second updated data is added to the database of the Internet AI outbound call platform, and during the operation of the Internet AI outbound call platform, it is verified whether the second updated data meets the update verification requirements. If the update verification requirements are met, it is determined to add the second updated data to the database of the Internet AI outbound call platform; The first updated data is data that is not in the existing database of the Internet AI outbound call platform, or is data that is already in the existing database but needs to be updated; The verification of whether the second updated data meets the update verification requirements is to determine whether the second updated data meets the minimum requirements for entering the database through multiple searches during the operation of the Internet AI outbound call platform before adding the second updated data to the database of the Internet AI outbound call platform; The update verification requirement is the minimum requirement for the second update data to pass the update verification and to determine its entry into the database.

2. The database updating method based on Internet AI outbound calling according to claim 1 is characterized in that: The method of obtaining the first update data required in the database of the Internet AI outbound call platform and processing the first update data in accordance with the database data format requirements to obtain the second update data includes: Obtain the first update data required in the database of the Internet AI outbound calling platform; The structure of data stored in the database is a graph data structure; After performing data cleaning on the first update data, third update data is obtained; The data cleaning is the process of discovering and correcting identifiable errors in data files, including checking data consistency, processing invalid values ​​and missing values, and re-examining and verifying data; After extracting key information from the third update data, fourth update data is obtained; The key information extraction is to extract the key information contained in the third update data and convert it into data conforming to the graph data structure as the fourth update data; After feature-marking the fourth updated data, second updated data is obtained; The feature annotation is to check the update type of the fourth update data, and perform the following different processing according to the specific update type: (a) If the update type of the fourth update data is a newly added node type, then according to the existing data level classification information, a corresponding level value is added to the fourth update data, and the data content updated by the fourth update data is marked as pending verification, so as to obtain the second update data; (b) if the update type of the fourth update data is a modified data type, marking the data content updated by the fourth update data as pending verification to obtain the second update data; The update type of the fourth update data is a new node type, which means that when performing the operation of updating data, a new node needs to be added to the existing database, and at least one edge connecting the new node and other nodes needs to be added; The level value indicates the level of the fourth updated data in the data structure. The larger the level value, the lower the level in the data structure. The level value of the fourth updated data is used as the level value of the second updated data obtained. The update type of the fourth update data is a data modification type, which means that when performing the operation of updating data, there is no need to add a new node in the existing database, and only data needs to be added or modified in the existing database.

3. The database updating method based on Internet AI outbound calling according to claim 2 is characterized in that: The adding of the second updated data to the database of the Internet AI outbound calling platform and verifying whether the second updated data meets the update verification requirements during the operation of the Internet AI outbound calling platform, and if the update verification requirements are met, determining to add the second updated data to the database of the Internet AI outbound calling platform includes: The second updated data marked as pending verification is added to the database of the Internet AI outbound calling platform for subsequent use by the Internet AI outbound calling platform; During the operation of the Internet AI outbound calling platform, obtaining search data related to the second updated data in the database; Verifying whether the second update data meets the update verification requirement for entering the database by obtaining the search data related to the second update data in the database; If the requirement for updating verification of entering the database is met, the to-be-verified mark of the second updated data is modified to a verified mark; If the requirement for the update verification of entering the database is not met, the pending verification mark of the second update data is maintained; The modification of the pending verification mark of the second updated data to the verified mark indicates that the updated data is determined to be added to the database of the Internet AI outbound call platform.

4. The database updating method based on Internet AI outbound calling according to claim 3 is characterized in that: During the operation of the Internet AI outbound calling platform, obtaining the search data related to the second updated data in the database includes: During the operation of the Internet AI outbound call platform, the total number of retrieval operations performed by the Internet AI outbound call platform on the data in the field to which the second updated data belongs in the database after the second updated data is added is obtained, and at the same time, the number of times the second updated data is retrieved during the retrieval operation, and the number of times the second updated data is retrieved and verified is obtained.

5. The database updating method based on Internet AI outbound calling according to claim 4 is characterized in that: The step of verifying whether the second update data meets the update verification requirement for entering the database by obtaining the search data related to the second update data in the database includes: The method for determining the update verification requirement for entering the database is that the update verification rate of the second update data meets a preset threshold, and the specific model of the update verification rate of the second update data is as follows: Wherein, μ is the update verification rate of the second update data; s0 is the total number of retrieval operations performed by the Internet AI outbound call platform on the data in the field to which the second updated data belongs in the database after the second updated data is added during the verification of the second updated data; s1 is the number of times the second updated data is retrieved by the Internet AI outbound call platform when performing a search operation in the database after the second updated data is added during the process of verifying the second updated data; s2 is the number of times that the second updated data is retrieved and verified by the Internet AI outbound call platform during the search operation in the database after the second updated data is added during the verification of the second updated data; β1 is the first verification weight coefficient; β2 is the second verification weight coefficient, and satisfies β1+β2=1; k0 is the level value of the second update data; m is the total number of levels in the field to which the second update data belongs.

6. The database updating method based on Internet AI outbound calling according to claim 5 is characterized in that: After determining to add the second updated data to the database of the Internet AI outbound calling platform, the method further includes: The second updated data of the database confirmed to join the Internet AI outbound call platform is monitored for data anomalies. If any abnormality is found during the monitoring, an early warning is issued.

7. The database updating method based on Internet AI outbound calling according to claim 6 is characterized in that: The second updated data of the database determined to be added to the Internet AI outbound call platform is monitored for data anomalies, and if an abnormality is found during the monitoring, an early warning is issued, including: Acquire information related to the second updated data in the database, including the level value of the second updated data, the total number of nodes directly connected to the first node, and the level value of each node directly connected to the first node; Determining the data abnormality of the second data in the database according to the information related to the second updated data in the database obtained; If the data abnormality is greater than or equal to the preset abnormality warning value, a warning is issued; The first node is the node used as the value of k0 in the second update data.

8. The database updating method based on Internet AI outbound calling according to claim 7 is characterized in that: The step of determining the data abnormality of the second data in the database according to the acquired information related to the second updated data in the database includes: The method for determining the data abnormality of the second data in the database is to calculate using a model of the data abnormality of the second data in the database. The specific model is as follows: Wherein, θ is the data anomaly degree of the second updated data in the database; k0 is the level value of the second update data; i is the i-th node directly connected to the first node; The first node is the node used as the value of k0 in the second update data; n is the total number of nodes directly connected to the first node; k i is the level value of the i-th node directly connected to the first node; δ is the level deviation from the baseline value; k max is the highest value among the level values ​​of all nodes directly connected to the first node; k min It is the lowest value among the level values ​​of all nodes directly connected to the first node.

9. A database update system based on Internet AI outbound calling, characterized in that: The system comprises: A data acquisition module, used to obtain the first update data required in the database of the Internet AI outbound call platform, and obtain the second update data after processing the first update data in accordance with the database data format requirements; A data verification module, used for adding the second updated data to the database of the Internet AI outbound calling platform, and verifying whether the second updated data meets the update verification requirements during the operation of the Internet AI outbound calling platform, and if the update verification requirements are met, determining to add the second updated data to the database of the Internet AI outbound calling platform; The data monitoring module is used to monitor the second updated data of the database determined to join the Internet AI outbound call platform for data anomalies, and issue an early warning if an abnormality is found in the monitoring; The first updated data is data that is not in the existing database of the Internet AI outbound call platform, or data that is already in the existing database but needs to be updated; The verification of whether the second updated data meets the update verification requirements is to determine whether the second updated data meets the minimum requirements for entering the database through multiple searches during the operation of the Internet AI outbound call platform before adding the second updated data to the database of the Internet AI outbound call platform; The update verification requirement is the minimum requirement for the second update data to pass the update verification and to determine its entry into the database.

10. A computer device comprising a memory and a processor; the memory is used to store a computer program; the processor is used to implement the database update method based on Internet AI outbound calling as described in any one of claims 1 to 8 when executing the computer program.

Images