A data matching method and apparatus

Abstract

The application discloses a data matching method and device, wherein the method comprises: obtaining target user's to-be-matched data, and performing data division on the to-be-matched data according to at least one entity attribute item to obtain to-be-matched sub-data corresponding to each entity attribute item; performing data matching on each entity attribute item and the corresponding to-be-matched sub-data based on a pre-constructed historical search record to obtain a data matching result; if the data matching result is not matched successfully, determining a target entity attribute item group from at least one entity attribute item group associated with the target user, and determining at least one candidate data corresponding to the target entity attribute item group from a pre-constructed target database; and determining target data corresponding to the to-be-matched data from the at least one candidate data according to to-be-determined similarity of each candidate data relative to the to-be-matched data. The result of quickly and accurately determining target data corresponding to the user in the target database is realized.

Description

Technical Field

[0001] This invention relates to the field of data processing technology, and in particular to a data matching method and apparatus. Background Technology

[0002] In the field of medical big data, accurately matching user data with patients is crucial for ensuring medical research, clinical trials, and personalized medical solutions.

[0003] Traditional user data matching is mostly based on manual matching methods, but this approach is often inefficient, time-consuming, and prone to errors. Furthermore, user data for the same patient is often linked to multiple data items, making it difficult to ensure the accuracy of data integration when merging patient data.

[0004] To solve the above problems, the data matching method needs to be improved. Summary of the Invention

[0005] This invention provides a data matching method and apparatus to solve the problems of low matching efficiency, long time consumption and easy error in the prior art when matching user data based on manual matching.

[0006] In a first aspect, embodiments of the present invention provide a data matching method, including:

[0007] Obtain the target user's data to be matched, and divide the data to be matched according to at least one entity attribute item to obtain the sub-data to be matched corresponding to each entity attribute item;

[0008] Based on pre-built historical search records, data matching is performed on each entity attribute item and the corresponding sub-data to be matched to obtain data matching results;

[0009] If the data matching result is unsuccessful, then a target entity attribute item group is determined from at least one entity attribute item group associated with the target user, and at least one candidate data corresponding to the target entity attribute item group is determined from the pre-constructed target database; wherein, the entity attribute item group includes at least one of the entity attribute items;

[0010] Based on the undetermined similarity between each candidate data and the data to be matched, target data corresponding to the data to be matched is determined from at least one candidate data.

[0011] Secondly, embodiments of the present invention also provide a data matching device, comprising:

[0012] The sub-data determination module is used to obtain the target user's data to be matched, and to divide the data to be matched according to at least one entity attribute item to obtain the sub-data to be matched corresponding to each entity attribute item.

[0013] The matching result determination module is used to perform data matching on each entity attribute item and the corresponding sub-data to be matched based on the pre-built historical search records, and to obtain the data matching result.

[0014] The candidate data determination module is configured to, if the data matching result is unsuccessful, determine a target entity attribute item group from at least one entity attribute item group associated with the target user, and determine at least one candidate data corresponding to the target entity attribute item group from a pre-constructed target database; wherein, the entity attribute item group includes at least one of the entity attribute items;

[0015] The target data determination module is used to determine the target data corresponding to the data to be matched from at least one of the candidate data based on the undetermined similarity between each candidate data and the data to be matched.

[0016] Thirdly, embodiments of the present invention also provide an electronic device, comprising:

[0017] At least one processor; and

[0018] A memory communicatively connected to the at least one processor; wherein,

[0019] The memory stores a computer program that can be executed by the at least one processor, the computer program being executed by the at least one processor to enable the at least one processor to perform the data matching method according to any embodiment of the present invention.

[0020] Fourthly, embodiments of the present invention also provide a computer-readable storage medium storing computer instructions that are used to cause a processor to execute the data matching method described in any embodiment of the present invention.

[0021] The technical solution of this invention involves acquiring target user's data to be matched, dividing the data according to at least one entity attribute item to obtain sub-data to be matched corresponding to each entity attribute item; performing data matching on each entity attribute item and the corresponding sub-data to be matched based on pre-built historical search records to obtain data matching results; if the data matching result is unsuccessful, determining a target entity attribute item group from at least one entity attribute item group associated with the target user, and determining at least one candidate data corresponding to the target entity attribute item group from a pre-built target database; and determining the target data corresponding to the data to be matched from the at least one candidate data based on the undetermined similarity between each candidate data and the data to be matched. This solves the problems of low matching efficiency, long time consumption, and easy error in the prior art when performing data matching on user data based on manual matching methods, and achieves the result of quickly and accurately determining the target data corresponding to the user in the target database.

[0022] It should be understood that the description in this section is not intended to identify key or essential features of the embodiments of the present invention, nor is it intended to limit the scope of the invention. Other features of the invention will become readily apparent from the following description. Attached Figure Description

[0023] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0024] Figure 1 This is a flowchart of a data matching method provided in Embodiment 1 of the present invention;

[0025] Figure 2 This is a flowchart of a data matching method provided according to Embodiment 2 of the present invention;

[0026] Figure 3 This is a flowchart of a data matching method provided according to Embodiment 2 of the present invention;

[0027] Figure 4 This is a schematic diagram of the structure of a data matching device provided in Embodiment 3 of the present invention. Detailed Implementation

[0028] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0029] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in sequences other than those illustrated or described herein.

[0030] Example 1

[0031] Figure 1 The flowchart of a data matching method provided in Embodiment 1 of the present invention is applicable to obtaining the target user's data to be matched. First, it is determined whether the historical search records contain the corresponding historical search data. If so, the target data corresponding to the target user is directly retrieved from the target database. If not, data matching is further performed on the data to be matched based on the target database. If there is user-related data in the target database that is consistent with the data to be matched, then the user-related data is used as the target user's data to be matched. This method can be executed by a data matching device, which can be implemented in hardware and / or software. The data matching device can be configured in a computing device capable of executing the data matching method.

[0032] like Figure 1 As shown, the method includes:

[0033] S110. Obtain the target user's data to be matched, and divide the data to be matched according to at least one entity attribute item to obtain the sub-data to be matched corresponding to each entity attribute item.

[0034] In practical applications, this technical solution can be used not only to integrate user data in medical institutions, but also to integrate user data in other fields. Therefore, target users can be understood as patients in medical institutions, business users in a business organization, or employees in an enterprise, etc.

[0035] In this technical solution, taking patients in medical institutions as the target user as an example, the data to be matched refers to the user data corresponding to the target user that needs to be integrated. For example, the data to be matched includes the target user's basic user information and patient visit-related data. Entity attribute items can be used to represent the data attributes corresponding to the sub-data in the data to be matched. It is understood that the acquired data to be matched corresponding to the target user may include multiple data pieces, and after data partitioning, at least one sub-data to be matched can be obtained.

[0036] For example, as shown in Table 1, a user information table is pre-set to populate user data. This table includes at least one entity attribute item, such as "user name," "contact information," and "medical history." The data to be matched corresponding to the target user includes "Zhang San + xxx + historical medical records." By partitioning the data to be matched based on at least one entity attribute item, the sub-data to be matched corresponding to each entity attribute item can be obtained. That is, the sub-data to be matched includes "Zhang San" corresponding to "user name," "xxx" corresponding to "contact information," and "historical medical records" corresponding to "medical history."

[0037] Table 1 User Information Table

[0038] Data source Username Contact information Medical history his1 Zhang San xxx Disease A, Disease B hisl Li Si xxx Disease C

[0039] S120. Based on the pre-built historical search records, perform data matching on each entity attribute item and the corresponding sub-data to be matched to obtain the data matching result.

[0040] Historical search records can be understood as historical information generated based on past search data. For example, taking the application scenario of matching user data of patients in medical institutions as an example, medical institutions receive a large number of patients every day, i.e., target users. When providing corresponding business services to each target user, it is necessary to obtain user-related data corresponding to the target user in advance. The more complete the user-related data, the more helpful it is to quickly provide corresponding personalized services to the target user.

[0041] In this technical solution, after obtaining the target user's matching data, in order to quickly perform data matching on the matching data to retrieve more complete user association data corresponding to the target user, a historical search record and a target database are pre-built. The historical search record can be used to record at least one historical search data. The target database is used to store user association data corresponding to at least one patient.

[0042] In practical applications, after obtaining the data to be matched, the data is divided according to the entity attribute items to obtain the sub-data to be matched corresponding to each entity attribute item. When performing data matching, medical staff can enter the corresponding sub-data to be matched according to the pre-set entity attribute items on the data search page to perform data matching based on the entity attribute items and the corresponding sub-data to be matched. For example, if the data search page includes "User Name", "Zhang San" can be entered in the edit box corresponding to "User Name". After the medical staff completes the input, the system searches the pre-built historical search records to see if there is any user-related data that matches each entity attribute item and the corresponding sub-data to be matched, and obtains the corresponding data matching results.

[0043] In a specific example, when performing data matching on the target user's data, firstly, at least one entity attribute item and the corresponding attribute value (i.e., the sub-data to be matched) are extracted from the data to be matched. This is then used as the query condition to perform a full attribute retrieval in the historical search records (i.e., a retrieval based on all the sub-data to be matched in the data to be matched). If a match is found (i.e., the historical search record), it means the data to be matched has been matched. Then, the target data corresponding to the target user is retrieved based on the standard index corresponding to the matched historical search data.

[0044] Optionally, data matching is performed on each entity attribute item and the corresponding sub-data to be matched based on pre-built historical search records to obtain data matching results, including: retrieving pre-built historical search records and determining at least one historical search data in the historical search records; and historical attribute association data corresponding to each historical entity attribute item; for each historical search data, if the current historical search data includes all entity attribute items and the sub-data to be matched corresponding to each entity attribute item, then the data matching result is determined to be a successful match; otherwise, the data matching result is determined to be a failed match.

[0045] The historical search data includes at least one historical search sub-data, each historical search sub-data corresponds to a historical entity attribute item, and the historical search sub-data corresponding to each historical entity attribute item is used as the corresponding historical attribute association data.

[0046] For example, the data to be matched includes five entity attribute items and corresponding sub-data items to be matched for each entity attribute item. When performing data matching on the data to be matched, the sub-data items to be matched corresponding to each entity attribute item are entered one by one on the data search page, and data matching is performed based on the sub-data items to be matched for each entity attribute item to obtain the data matching result.

[0047] Understandably, each additional entity attribute item that is entered as a matching sub-data item results in a higher degree of relevance between the displayed historical search data and the target user. For example, entering "Zhang San" as the matching sub-data item corresponding to "Username" might result in 10 historical search results for "Zhang San". Adding "Data Source 1" as the matching sub-data item corresponding to "Data Source" might result in 3 historical search results for "Zhang San + Data Source". Further adding matching sub-data items corresponding to entity attributes, such as "xxx" as the matching sub-data item for "Contact Information", will only display 1 historical search result. In other words, the more entity attribute items corresponding to the entered matching sub-data items, the fewer historical search results are displayed, but the higher the relevance of that data to the target user.

[0048] Furthermore, if there are matching sub-data items in the historical search records that correspond to all the input entity attribute items, then the data matching result can be determined to be a successful match; conversely, if as the number of matching sub-data items corresponding to entity attribute items increases, the corresponding data cannot be detected in the historical search records, it indicates that the data matching result is a failed match.

[0049] In practical applications, each time data matching is performed on the user-related data of a patient, corresponding historical search data is generated, and a historical search record can be generated based on at least one historical search data. When matching the data to be matched for a target user, a data query is first performed based on the historical search record. If the corresponding historical search data exists in the historical search record, the user-related data corresponding to the target user can be directly retrieved. If not, further data matching based on the target database is required to obtain the user-related data of the target user.

[0050] The advantage of this setup is that it first searches for at least one sub-data item to be matched based on historical search records, which can improve the speed of data matching and find user-related data corresponding to the target user more quickly.

[0051] Optionally, in practical applications, at least one equivalent attribute item can be determined from at least one entity attribute item based on pre-set equivalent configuration attributes, and the equivalent attribute data corresponding to each equivalent attribute item can be determined; for each historical search data, if the current historical search data includes at least one equivalent attribute item and the equivalent attribute data corresponding to each equivalent attribute item, then the data matching result is determined to be a successful match; otherwise, the data matching result is determined to be a failed match.

[0052] In this context, equivalent attribute items can be understood as entity attribute items that can be directly used to represent the identity information of the target user. Equivalent attribute data is the data corresponding to the equivalent attribute items. For example, if the equivalent attribute item can be the target user's ID card number, then the corresponding equivalent attribute data is the specific data of the ID card number.

[0053] In a specific example, it is also possible to obtain the pre-defined configuration attributes for equivalent retrieval (i.e., equivalent configuration attributes), and extract the corresponding raw material attribute values ​​based on these configuration attributes (i.e., the equivalent configuration attributes determine at least one equivalent attribute item from at least one entity match, and determine the equivalent attribute data corresponding to each equivalent attribute). Based on this, a search is performed in the historical search records. If the historical search records contain equivalent attribute items and their corresponding equivalent attribute data, the data matching result is determined to be a successful match; otherwise, the data matching result is determined to be a failed match.

[0054] The advantage of this setup is that it allows for retrieval of historical search records based on a small number of entity attribute items and the corresponding sub-data, in order to determine whether there is any corresponding historical search data, thereby further improving the data matching speed.

[0055] S130. If the data matching result is unsuccessful, then determine the target entity attribute item group from at least one entity attribute item group associated with the target user, and determine at least one candidate data corresponding to the target entity attribute item group from the pre-built target database.

[0056] The entity attribute item group includes at least one entity attribute item. The target entity attribute item can be understood as an entity attribute item group containing all entity attribute items in the data to be matched. It should be noted that the target database described in this technical solution is used to store the latest user association data corresponding to the patients. That is, after storing the user association data of the patients, if an update to the patient's data is detected, and the updated data is more complete, then the user association data in the target database is updated based on the latest user association data. Candidate data can be understood as user association data retrieved from the target database that matches all entity attribute items in the target entity attribute item group.

[0057] In practical applications, if the data matching result is unsuccessful, it indicates that no historical search data corresponding to the data to be matched was found in the historical search records. In this case, a further step is needed to perform data matching on the data to be matched based on the target database.

[0058] Optionally, determining the target entity attribute item group from at least one entity attribute item group associated with the target user includes: combining at least one entity attribute item according to a pre-set entity attribute influence degree to obtain at least one entity attribute item group; retrieving the target database and traversing at least one piece of data to be compared in the target database; for each entity attribute item group, if at least one piece of data to be compared includes all current entity attribute items in the current entity attribute item group and the matching sub-data corresponding to all current entity attribute items, then determining the current entity attribute item group as a candidate entity attribute item group; from at least one candidate entity attribute item group, determining the candidate entity attribute item group corresponding to the most candidate entity attribute items as the target entity attribute item group.

[0059] The entity attribute influence can be used to characterize the importance of an entity attribute item in identifying the identity information of the target user. The data to be compared can be understood as user-related data stored in the target database corresponding to at least one patient. The candidate entity attribute item group includes at least one candidate entity attribute item.

[0060] In a specific example, when performing data matching on the data to be matched, the number of entity attribute items is increased sequentially to perform data matching in the target database based on the sub-data to be matched corresponding to each entity attribute item. The target entity attribute item group refers to a combination of at least one entity attribute item. For example, if the data to be matched includes sub-data to be matched corresponding to three entity attribute items, namely attribute item 1, attribute item 2, and attribute item 3, then at least one entity attribute item group can be obtained based on at least one entity attribute item, including: {attribute item 1}{attribute item 2}{attribute item 3}{attribute item 1; attribute item 2}{attribute item 1; attribute item 3}{attribute item 2; attribute item 3} and {attribute item 1; attribute item 2; attribute item 3}.

[0061] Furthermore, candidate entity attribute items are determined from at least one group of entity attribute items. It is understandable that the target database contains a large amount of user-related data; comparing each user-related data item individually would be time-consuming and slow. Therefore, for each entity attribute group, if the data to be compared contains all the sub-data corresponding to the current entity attribute items in the current entity attribute group, then the current entity attribute group can be used as a candidate entity attribute group. Further, the candidate attribute group containing the largest number of candidate entity attribute items is selected as the target entity attribute group.

[0062] For example, if {attribute 1}{attribute 2}{attribute 3}{attribute 2; attribute 3} and {attribute 1; attribute 2; attribute 3} in the above example are determined to be candidate entity attribute item groups based on at least one data to be compared in the target database, then {attribute 1; attribute 2; attribute 3} will be determined as the target entity attribute item group.

[0063] Optionally, at least one candidate data corresponding to the target entity attribute item group is determined from the pre-built target database, including: determining the target entity attribute item group containing at least one target entity attribute item and the data to be compared corresponding to each target entity attribute item as the candidate data corresponding to the target entity attribute item group.

[0064] Building upon the example above, to determine whether target data corresponding to the target user exists in the target database, a data retrieval is performed in the target database based on at least one entity attribute item in the target entity attribute items and the corresponding sub-data to be matched. If no corresponding user-related data is retrieved, it indicates that target data corresponding to the target user does not exist in the target database. If a unique user-related data item is retrieved, this user-related data can be used as the target data corresponding to the target user. If multiple user-related data items matching the target entity attribute item group (i.e., data to be compared) are retrieved, the retrieved user-related data items are used as candidate data.

[0065] S140. Based on the undetermined similarity between each candidate data and the data to be matched, determine the target data corresponding to the data to be matched from at least one candidate data.

[0066] Here, the undetermined similarity can be understood as the similarity between candidate data and data to be matched, used to characterize the similarity between the data to be compared and the data to be matched. The target data can be understood as the candidate data corresponding to the highest undetermined similarity, that is, the user-related data in the target database that has the highest matching degree with the data to be matched.

[0067] The technical solution of this invention involves acquiring the target user's data to be matched and dividing the data according to at least one entity attribute item to obtain sub-data to be matched corresponding to each entity attribute item. By dividing the data to be matched using at least one entity attribute item, at least one sub-data to be matched is obtained. Further, data matching is performed on each entity attribute item and the corresponding sub-data to be matched based on pre-built historical search records to obtain a data matching result. That is, when matching the data to be matched, it is first determined whether corresponding historical search data exists in the historical search records. If it exists, the target data corresponding to the target user can be directly retrieved from the target database based on the index information of the historical search data. If it does not exist, further data matching based on the target database is required. Based on this, if the data matching result is unsuccessful, a target entity attribute item group is determined from at least one entity attribute item group associated with the target user, and at least one candidate data corresponding to the target entity attribute item group is determined from the pre-built target database. In practical applications, the entity attribute item group containing the most entity attribute items is taken as the target entity attribute item group, and all entity attribute items in the target entity attribute item group and their corresponding user-associated data in the target database are taken as candidate data. If there are multiple candidate data points, the target data corresponding to the data to be matched is determined from at least one candidate data point based on the undetermined similarity between each candidate data point and the data to be matched. This solves the problems of low matching efficiency, long time consumption, and easy error in the existing technology of matching user data based on manual matching, and realizes the result of quickly and accurately determining the target data corresponding to the user in the target database.

[0068] Example 2

[0069] Figure 2 The flowchart of a data matching method provided in Embodiment 2 of the present invention is as follows: Optionally, if the data matching result is unsuccessful, a target entity attribute item group is determined from at least one entity attribute item group associated with the target user, and at least one candidate data corresponding to the target entity attribute item group is determined from the pre-built target database for refinement.

[0070] like Figure 2 As shown, the method includes:

[0071] S210. Obtain the target user's data to be matched, and divide the data to be matched according to at least one entity attribute item to obtain the sub-data to be matched corresponding to each entity attribute item.

[0072] S220. Based on the pre-built historical search records, perform data matching on each entity attribute item and the corresponding sub-data to be matched to obtain the data matching result;

[0073] S230. If the data matching result is unsuccessful, then determine the target entity attribute item group from at least one entity attribute item group associated with the target user, and determine at least one candidate data corresponding to the target entity attribute item group from the pre-built target database.

[0074] S240. Based on at least one similarity algorithm, determine the undetermined similarity between at least one candidate data and the data to be matched.

[0075] The similarity algorithm includes at least one of the following: cosine similarity algorithm, text similarity algorithm, feature similarity algorithm, and Euclidean distance similarity algorithm.

[0076] In this embodiment, the feature similarity algorithm is used as an example to calculate the undetermined similarity between each candidate data and the data to be matched.

[0077] In a specific example, such as Figure 3 As shown, patient raw data (i.e., the target user's data to be matched) is obtained, and the raw data library is searched based on full attributes (i.e., the data to be matched corresponding to all entity attributes in the data to be matched is queried in the historical search records), or the raw data library is searched based on equivalence (i.e., the data to be matched corresponding to the target user's equivalent attribute items is queried in the historical search records). The standard library is searched one by one according to the patient attributes, and a patient standard data set is obtained (i.e., the data to be matched is searched in the target database based on the sub-data to be matched corresponding to each entity attribute item). If the set is empty, it means that there is no user-related data corresponding to the target user in the target database. In this case, the data to be matched is packaged into raw data, and the packaged data is written into the patient standard library (i.e., the target database). If the set is not empty, the standard data set is subjected to an intersection operation to filter the patient standard set with the most attribute intersections (i.e., the target entity attribute item group is determined). Based on this, at least one candidate data associated with the target entity attribute item group is determined.

[0078] Furthermore, a key-value pair A is constructed based on the data attributes and attribute values ​​to be matched, where K is the field + field value (data attribute item + attribute value to be matched), and V is a set of deduplicated data (this set is empty at this point because the standard set to be matched has not yet been traversed). Specifically, in this technical solution, a key-value pair A is constructed based on the raw material (i.e., the data to be matched), and the key-value pair data structure is a Map. <String,Set <string>In summary, it involves initializing a data structure called a Map using the data to be matched. <String,Set <string>> key-value pairs.

[0079] The structure after initialization is as follows:

[0080] Key (attribute + attribute value) Grouping set (patient standard unique index set) name = Zhao San [] sex = male [] Birthdate = 1955 / 08 / 22 [] address = Region A []

[0081] Furthermore, the attributes and attribute values ​​of standard patient records (i.e., entity attribute items and corresponding user-related data in the user-related data of the patient) are extracted from the standard dataset (i.e., the target database), and corresponding (KV) key-value pairs are constructed, where the key can be understood as the standard data attribute item + attribute value, and the value is the unique identifier index of the patient's data to be compared.

[0082] The key-value pairs corresponding to each patient are sequentially placed into A. If the keys are identical, their unique identifier index is added to the deduplication set, resulting in the following grouping results. Here, A is the data structure initialized above based on the attribute items and values ​​of the data to be matched: Map. <String,Set <string>Key-value pairs. As shown in Table 2:

[0083] Table 2 Grouping Results

[0084] raw material bond Grouping set (patient standard unique index set) name = Zhao San [1，2] sex = male [1，2，3，4，12，13] Birthdate = 1955 / 08 / 22 [1，2，3，4] address = Region A [3，4]

[0085] In Table 2, the same number in the group set represents the same entity attribute item, and different numbers represent different entity attribute items.

[0086] By performing operations on the above set, we can find the standard record set that appears most frequently (i.e., the entity attribute item with the largest number of occurrences). In this example, we filter the set [1,2,3,4,12,13] to obtain the set [1,2,3] (i.e., the target entity attribute item group).

[0087] Furthermore, data similarity is calculated based on the feature similarity algorithm (Fellegi-Sunter, FS). The FS algorithm can be understood as a probabilistic statistical model that uses pre-defined feature attributes of multiple patients for data identification and matching.

[0088] The specific calculation formula for the Fellegi-Sunter model is as follows (for illustrative purposes only):

[0089] a. Record Feature Weight Calculation: For each record feature i, calculate its weight w(i), which represents the importance of the feature in determining a record match. These features can be name, address, ID number, date of birth, etc.

[0090] b. Calculation of feature comparison results:

[0091] For each feature i, compare the feature value a(i) of the record to be matched A with the feature value b(i) of the target record B. If they are the same, m(i) = 1; otherwise, m(i) = 0.

[0092] c. Similarity calculation:

[0093] Calculate the similarity s(A,B) between the record to be matched A and the target record B. This similarity is calculated based on the feature comparison results and feature weights.

[0094] s(A,B)=Σ(w(i)*m(i)) / Σ(w(i))

[0095] Where Σ(w(i)*m(i)) represents the sum of the weights of all features multiplied by the comparison results, and Σ(w(i)) represents the sum of the weights of all features.

[0096] A threshold can be determined based on the similarity s(A,B). If s(A,B) exceeds the set threshold, the two records are considered to be a match; otherwise, the two records are considered to be a mismatch.

[0097] It should be noted that when using the FS algorithm model, parameters such as feature weights and thresholds may need to be adjusted and optimized to adapt to actual needs and data characteristics.

[0098] S250. The candidate data corresponding to the highest similarity to be determined is determined as the target data to be matched.

[0099] In practical applications, see again Figure 3 The higher the similarity to be determined, the higher the similarity between the candidate data and the data to be matched. Therefore, the candidate data corresponding to the highest similarity to be determined is taken as the target data corresponding to the data to be matched (i.e., the TOP1 is taken as the matching result).

[0100] Optionally, if the similarity to be determined is less than the preset similarity threshold, the data to be matched will be pushed to the manual review system for data review.

[0101] For details, see again. Figure 3 If the similarity to be determined is less than the preset similarity threshold, it indicates that the similarity between the candidate data and the data to be matched is low. In order to ensure the data matching result, it is necessary to further review the candidate data based on the manual review system to determine whether the candidate data is the target data of the data to be matched.

[0102] It should be noted that a minimum similarity threshold can also be set in this technical solution, and the minimum similarity threshold is less than the preset similarity threshold. When the similarity to be determined corresponding to the candidate data is less than the minimum similarity threshold, in order to reduce data interference and improve the speed of data matching, the candidate data that is lower than the minimum similarity threshold can be directly removed.

[0103] Optionally, if the data integrity of the data to be matched is higher than that of the target data, then the corresponding target data in the target database is updated based on the data to be matched.

[0104] In practical applications, it is necessary to determine whether there are differences in the attributes of the data to be matched and the target data, that is, to detect the data integrity of the data to be matched and the target data.

[0105] To ensure that the user association data in the target database is the latest user association data, when the data integrity of the target user's data to be matched is higher than that of the target data, attribute merging is performed according to the preset attribute merging strategy, that is, the target data is updated based on the data to be matched.

[0106] Optionally, if the data matching result is a successful match, the target historical search data corresponding to the data to be matched is determined from at least one historical search record; based on the target index information corresponding to the target historical search data, the target data corresponding to the target index information is retrieved from the target database.

[0107] It should be noted that, in order to improve the efficiency of data matching, this technical solution pre-establishes a correspondence between each historical search record and the target database, and uses index information to represent the relationship between the two. After determining the target historical search data corresponding to the data to be matched, the corresponding target data, that is, the target data corresponding to the data to be matched, can be retrieved from the target database according to the target index information corresponding to the target historical search data.

[0108] The technical solution of this invention involves acquiring the target user's data to be matched and dividing the data according to at least one entity attribute item to obtain sub-data to be matched corresponding to each entity attribute item. By dividing the data to be matched using at least one entity attribute item, at least one sub-data to be matched is obtained. Further, data matching is performed on each entity attribute item and the corresponding sub-data to be matched based on pre-built historical search records to obtain a data matching result. That is, when matching the data to be matched, it is first determined whether corresponding historical search data exists in the historical search records. If it exists, the target data corresponding to the target user can be directly retrieved from the target database based on the index information of the historical search data. If it does not exist, further data matching based on the target database is required. Based on this, if the data matching result is unsuccessful, a target entity attribute item group is determined from at least one entity attribute item group associated with the target user, and at least one candidate data corresponding to the target entity attribute item group is determined from the pre-built target database. In practical applications, the entity attribute item group containing the most entity attribute items is taken as the target entity attribute item group, and all entity attribute items in the target entity attribute item group and their corresponding user-associated data in the target database are taken as candidate data. If there are multiple candidate data points, the target data corresponding to the data to be matched is determined from at least one candidate data point based on the undetermined similarity between each candidate data point and the data to be matched. This solves the problems of low matching efficiency, long time consumption, and easy error in the existing technology of matching user data based on manual matching. It enables the rapid and accurate determination of the target data corresponding to the user in the target database after obtaining the user's user association data.

[0109] Example 3

[0110] Figure 4 This is a schematic diagram of a data matching device provided in Embodiment 3 of the present invention. Figure 4 As shown, the device includes: a sub-data determination module 310, a matching result determination module 320, a candidate data determination module 330, and a target data determination module 340.

[0111] The sub-data determination module 310 is used to obtain the target user's data to be matched, and to divide the data to be matched according to at least one entity attribute item to obtain the sub-data to be matched corresponding to each entity attribute item.

[0112] The matching result determination module 320 is used to perform data matching on each entity attribute item and the corresponding sub-data to be matched based on the pre-built historical search records to obtain the data matching result;

[0113] The candidate data determination module 330 is used to determine a target entity attribute item group from at least one entity attribute item group associated with the target user if the data matching result is unsuccessful, and to determine at least one candidate data corresponding to the target entity attribute item group from a pre-built target database; wherein, the entity attribute item group includes at least one entity attribute item;

[0114] The target data determination module 340 is used to determine the target data corresponding to the data to be matched from at least one candidate data based on the undetermined similarity between each candidate data and the data to be matched.

[0115] The technical solution of this invention involves acquiring target user's data to be matched, dividing the data according to at least one entity attribute item to obtain sub-data to be matched corresponding to each entity attribute item; performing data matching on each entity attribute item and the corresponding sub-data to be matched based on pre-built historical search records to obtain data matching results; if the data matching result is unsuccessful, determining a target entity attribute item group from at least one entity attribute item group associated with the target user, and determining at least one candidate data corresponding to the target entity attribute item group from a pre-built target database; and determining the target data corresponding to the data to be matched from the at least one candidate data based on the undetermined similarity between each candidate data and the data to be matched. This solves the problems of low matching efficiency, long time consumption, and easy error in the prior art when performing data matching on user data based on manual matching methods, and achieves the result of quickly and accurately determining the target data corresponding to the user in the target database.

[0116] Optionally, the sub-data determination module includes: a historical search data determination sub-module, used to retrieve pre-built historical search records and determine at least one piece of historical search data in the historical search records; wherein, the historical search data includes at least one historical entity attribute item, and historical attribute association data corresponding to each historical entity attribute item;

[0117] The first matching result determination submodule is used to determine the data matching result as a successful match if the current historical search data includes all entity attribute items and the corresponding sub-data to be matched for each entity attribute item.

[0118] The second matching result determination submodule is used to determine that the data matching result is unsuccessful if otherwise.

[0119] Optionally, the matching result determination module includes: an equivalent attribute item determination unit, used to determine at least one equivalent attribute item from at least one entity attribute item based on a pre-set equivalent configuration attribute, and to determine the equivalent attribute data corresponding to each equivalent attribute item;

[0120] The first matching result determination unit is used to determine the data matching result as a successful match if the current historical search data includes at least one equivalent attribute item and the equivalent attribute data corresponding to each equivalent attribute item.

[0121] The second matching result determination unit is used to determine the data matching result as unsuccessful if otherwise.

[0122] Optionally, the candidate data determination module includes: an entity attribute item group determination submodule, used to combine at least one entity attribute item according to a pre-set entity attribute influence degree to obtain at least one entity attribute item group;

[0123] The submodule for determining the data to be compared is used to retrieve the target database and iterate through at least one piece of data to be compared in the target database;

[0124] The candidate entity attribute item group determination submodule is used to determine the current entity attribute item group as a candidate entity attribute item group if at least one piece of data to be compared includes all the current entity attribute items in the current entity attribute item group and the matching sub-data corresponding to all the current entity attribute items; wherein, the candidate entity attribute item group includes at least one candidate entity attribute item.

[0125] The target entity attribute item group determination submodule is used to determine the candidate entity attribute item group corresponding to the most candidate entity attribute item from at least one candidate entity attribute item group as the target entity attribute item group.

[0126] Optionally, the candidate data determination module includes: a candidate data determination submodule, used to determine at least one target entity attribute item in the target entity attribute item group and the comparison data corresponding to each target entity attribute item as candidate data corresponding to the target entity attribute item group.

[0127] Optionally, the target data determination module includes a similarity determination submodule, used to determine the undetermined similarity between at least one candidate data and the data to be matched based on at least one similarity algorithm; wherein, the similarity algorithm includes at least one of cosine similarity algorithm, text similarity algorithm, feature similarity algorithm, and Euclidean distance similarity algorithm.

[0128] The target data determination submodule is used to determine the candidate data corresponding to the highest similarity to be determined as the target data to be matched.

[0129] Optionally, the data matching device may also include a data push module, which is used to push the data to be matched to the manual review system for data review if the similarity to be determined is less than a preset similarity threshold.

[0130] Optionally, the data matching device further includes a data update module, used to update the corresponding target data in the target database based on the data to be matched if the data integrity of the data to be matched is higher than the data integrity of the target data.

[0131] Optionally, the data matching device further includes: a target historical search data determination module, used to determine the target historical search data corresponding to the data to be matched from at least one historical search record if the data matching result is a successful match;

[0132] The data retrieval module is used to retrieve target data corresponding to the target index information from the target database based on the target index information corresponding to the target historical search data.

[0133] The data matching device provided in the embodiments of the present invention can execute the data matching method provided in any embodiment of the present invention, and has the corresponding functional modules and beneficial effects of the method execution.

[0134] It should be understood that the various forms of processes shown above can be used, with steps reordered, added, or deleted. For example, the steps described in this invention can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution of this invention can be achieved, and this is not limited herein.

[0135] The specific embodiments described above do not constitute a limitation on the scope of protection of this invention. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this invention should be included within the scope of protection of this invention.< / string> < / string> < / string>

Claims

1. A data matching method, characterized in that, include: Obtain the target user's data to be matched, and divide the data to be matched according to at least one entity attribute item to obtain the sub-data to be matched corresponding to each entity attribute item; Based on pre-built historical search records, data matching is performed on each entity attribute item and the corresponding sub-data to be matched to obtain data matching results; If the data matching result is unsuccessful, then a target entity attribute item group is determined from at least one entity attribute item group associated with the target user, and at least one candidate data corresponding to the target entity attribute item group is determined from the pre-constructed target database; wherein, the entity attribute item group includes at least one of the entity attribute items; Based on the undetermined similarity between each candidate data and the data to be matched, target data corresponding to the data to be matched is determined from at least one of the candidate data; Determining the target entity attribute item group from at least one entity attribute item group associated with the target user includes: Based on a pre-set entity attribute influence degree, at least one entity attribute item is combined to obtain at least one entity attribute item group, wherein the entity attribute influence degree is used to characterize the importance of the entity attribute item in identifying the identity information of the target user; Retrieve the target database and iterate through at least one record to be compared in the target database; For each entity attribute item group, if the at least one piece of data to be compared includes all current entity attribute items in the current entity attribute item group and the sub-data to be matched corresponding to all the current entity attribute items, then the current entity attribute item group is determined to be a candidate entity attribute item group; wherein, the candidate entity attribute item group includes at least one candidate entity attribute item; From at least one candidate entity attribute item group, the candidate entity attribute item group corresponding to the most candidate entity attribute item is determined as the target entity attribute item group.

2. The method according to claim 1, characterized in that, The process of matching data for each entity attribute item and the corresponding sub-data to be matched based on pre-built historical search records to obtain data matching results includes: Retrieve pre-built historical search records and determine at least one historical search data in the historical search records; wherein, the historical search data includes at least one historical entity attribute item, and historical attribute association data corresponding to each historical entity attribute item; For each historical search data, if the current historical search data includes all the entity attribute items and the sub-data to be matched corresponding to each entity attribute item, then the data matching result is determined to be a successful match; Otherwise, the data matching result is determined to be unsuccessful.

3. The method according to claim 1, characterized in that, The process of matching data for each entity attribute item and the corresponding sub-data to be matched based on pre-built historical search records to obtain data matching results includes: Based on the pre-set equivalent configuration attributes, at least one equivalent attribute item is determined from at least one entity attribute item, and equivalent attribute data corresponding to each equivalent attribute item is determined; For each historical search data, if the current historical search data includes at least one of the equivalent attribute items and the equivalent attribute data corresponding to each of the equivalent attribute items, then the data matching result is determined to be a successful match; Otherwise, the data matching result is determined to be unsuccessful.

4. The method according to claim 1, characterized in that, The step of determining at least one candidate data item corresponding to the target entity attribute item group from a pre-built target database includes: The target entity attribute item containing at least one target entity attribute item from the target entity attribute item group and the comparison data corresponding to each target entity attribute item are determined as candidate data corresponding to the target entity attribute item group.

5. The method according to claim 1, characterized in that, The step of determining the target data corresponding to the data to be matched from at least one of the candidate data based on the undetermined similarity of each candidate data to the data to be matched includes: Based on at least one similarity algorithm, determine the undetermined similarity between at least one candidate data and the data to be matched; wherein, the similarity algorithm includes at least one of cosine similarity algorithm, text similarity algorithm, feature similarity algorithm, and Euclidean distance similarity algorithm; The candidate data corresponding to the highest similarity to be determined is determined as the target data of the data to be matched.

6. The method according to claim 5, characterized in that, Also includes: If the similarity to be determined is less than the preset similarity threshold, the data to be matched will be pushed to the manual review system for data review.

7. The method according to claim 1, characterized in that, Also includes: If the data integrity of the data to be matched is higher than that of the target data, then the corresponding target data in the target database is updated based on the data to be matched.

8. The method according to claim 1, characterized in that, Also includes: If the data matching result is a successful match, then the target historical search data corresponding to the data to be matched is determined from at least one historical search record; Based on the target index information corresponding to the target historical search data, the target data corresponding to the target index information is retrieved from the target database.

9. A data matching device, characterized in that, include: The sub-data determination module is used to obtain the target user's data to be matched, and to divide the data to be matched according to at least one entity attribute item to obtain the sub-data to be matched corresponding to each entity attribute item. The matching result determination module is used to perform data matching on each entity attribute item and the corresponding sub-data to be matched based on the pre-built historical search records, and to obtain the data matching result. The candidate data determination module is configured to, if the data matching result is unsuccessful, determine a target entity attribute item group from at least one entity attribute item group associated with the target user, and determine at least one candidate data corresponding to the target entity attribute item group from a pre-constructed target database; wherein, the entity attribute item group includes at least one of the entity attribute items; The target data determination module is used to determine, based on the undetermined similarity between each candidate data and the data to be matched, the target data corresponding to the data to be matched from at least one candidate data; The candidate data determination module includes: an entity attribute item group determination submodule, which is used to combine at least one entity attribute item according to a pre-set entity attribute influence degree to obtain at least one entity attribute item group, wherein the entity attribute influence degree is used to characterize the importance of the entity attribute item to the identification of the target user's identity information; The submodule for determining the data to be compared is used to retrieve the target database and iterate through at least one piece of data to be compared in the target database; The candidate entity attribute item group determination submodule is used to determine the current entity attribute item group as a candidate entity attribute item group if at least one piece of data to be compared includes all the current entity attribute items in the current entity attribute item group and the matching sub-data corresponding to all the current entity attribute items; wherein, the candidate entity attribute item group includes at least one candidate entity attribute item. The target entity attribute item group determination submodule is used to determine the candidate entity attribute item group corresponding to the most candidate entity attribute item from at least one candidate entity attribute item group as the target entity attribute item group.

Images