Data matching method and device, electronic equipment and storage medium

Abstract

The invention discloses a data matching method and device, electronic equipment and a storage medium, and relates to the technical field of data processing. The data matching method comprises the following steps: respectively inputting a to-be-tested form and field basic information in an obtained business domain into a pre-created event trigger word recognition model and a business object element recognition model to respectively obtain a corresponding to-be-tested event trigger word set, a to-be-tested business object set and a to-be-tested object element set; obtaining a corresponding to-be-tested event type, a to-be-tested object ontology and a to-be-tested object element ontology by using a pre-created mapping rule; the actual matching confidence coefficient is determined by using the preset matching confidence coefficient rule, and the to-be-tested form and the collaborative form field corresponding to the to-be-tested field are determined according to the actual matching confidence coefficient and the preset matching confidence coefficient threshold, so that powerful support is provided for business analysis and decision, and the business structure knowledge reuse efficiency is improved. The problems of poor traceability and low agility in a business system are solved, and cost reduction, efficiency improvement and business management of enterprises are realized.

Description

Technical Field

[0001] The present invention relates to the field of data processing technology, and in particular to a data matching method, device, electronic equipment and storage medium. Background Art

[0002] The business system is the business links required for an enterprise to achieve its positioning, the roles played by various partners, and the ways and contents of cooperation and transactions among stakeholders. The business system is complex and diverse. Different business systems involve different information technology terminology such as master data, metadata, data standards, sharing and communication, etc., and it is even more difficult to establish connections between different businesses of different enterprises.

[0003] At present, the existing data matching method obtains business requirements and creates a business domain based on the business requirements, creates business objects and business relationships in the business domain, obtains entity objects and entity relationships based on the business objects and business relationships, and then builds a graph database model based on the entity objects and entity relationships to perform matching operations on the database.

[0004] However, existing technologies cannot meet the growing demand for existing enterprise data storage, cannot integrate different data sources in a standardized form, have low efficiency in reusing business structure knowledge, have poor traceability and low agility, and have high enterprise governance costs. Summary of the invention

[0005] The present invention provides a data matching method, device, electronic device and storage medium to provide strong support for business analysis and decision-making, improve the efficiency of business structure knowledge reuse, solve the problems of poor traceability and low agility in business systems, and achieve cost reduction, efficiency improvement and business governance of enterprises.

[0006] In a first aspect, an embodiment of the present invention provides a data matching method, comprising:

[0007] Obtain the basic information of the form to be tested and / or the basic information of the field to be tested in the business domain;

[0008] Input the basic information of the form to be tested into the pre-created event trigger word recognition model to obtain the corresponding event trigger word set to be tested, and input the basic information of the field to be tested into the pre-created business object element recognition model to obtain the corresponding business object set to be tested and the object element set to be tested;

[0009] Using the pre-created mapping rules, obtain the event type to be tested corresponding to the set of trigger words of the event to be tested, and using the pre-created mapping rules, obtain the object ontology to be tested and the object element ontology to be tested corresponding to the set of business objects to be tested and the set of object elements to be tested, respectively;

[0010] Determine the actual matching confidence between the event type to be tested and the tested event type corresponding to each form basic information in the pre-created business domain database by using a preset matching confidence rule as a first actual matching confidence, and determine the actual matching confidence between the business object ontology to be tested and the element ontology of the object to be tested and the tested business object ontology and the element ontology of the tested object corresponding to each field basic information in the pre-created business domain database by using a preset matching confidence rule as a second actual matching confidence;

[0011] The collaborative form corresponding to the form to be tested is determined according to the first actual matching confidence and the preset matching confidence threshold, and the collaborative field corresponding to the field to be tested is determined according to the second actual matching confidence and the preset matching confidence threshold.

[0012] In a second aspect, an embodiment of the present invention further provides a data matching device, including:

[0013] An information acquisition module is used to acquire basic information of the form to be tested and / or basic information of the field to be tested in the business domain;

[0014] A set acquisition module is used to input the basic information of the form to be tested into a pre-created event trigger word recognition model to obtain a corresponding set of event trigger words to be tested, and input the basic information of the field to be tested into a pre-created business object element recognition model to obtain a corresponding set of business objects to be tested and a set of object elements to be tested;

[0015] An ontology acquisition module is used to obtain the event type to be tested corresponding to the set of trigger words of the event to be tested by using the pre-created mapping rules, and to obtain the object ontology to be tested and the object element ontology to be tested corresponding to the set of business objects to be tested and the set of object elements to be tested respectively by using the pre-created mapping rules;

[0016] An actual matching confidence acquisition module is used to determine the actual matching confidence between the event type to be tested and the tested event type corresponding to each form basic information in the pre-created business domain database using a preset matching confidence rule as a first actual matching confidence, and to determine the actual matching confidence between the business object ontology to be tested and the element ontology of the object to be tested and the tested business object ontology and the element ontology of the object to be tested corresponding to each field basic information in the pre-created business domain database using a preset matching confidence rule as a second actual matching confidence;

[0017] The collaborative form field acquisition module is used to determine the collaborative form corresponding to the form to be tested according to the first actual matching confidence and the preset matching confidence threshold, and to determine the collaborative field corresponding to the field to be tested according to the second actual matching confidence and the preset matching confidence threshold.

[0018] In a third aspect, an embodiment of the present invention further provides an electronic device, including:

[0019] at least one processor; and

[0020] a memory communicatively connected to at least one processor; wherein,

[0021] The memory stores a computer program that can be executed by at least one processor. The computer program is executed by at least one processor so that the at least one processor can execute any one of the data matching methods in the embodiments of the present invention.

[0022] In a fourth aspect, an embodiment of the present invention further provides a computer-readable storage medium, wherein the computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a processor to implement any one of the data matching methods in the embodiments of the present invention when executed.

[0023] According to the technical solution of the embodiment of the present invention, by inputting the basic information of the forms and fields to be tested in the acquired business domain into the pre-created event trigger word recognition model and business object element recognition model respectively, the corresponding event trigger word set to be tested, the business object set to be tested and the object element set to be tested are obtained respectively; the corresponding event type to be tested, the object ontology to be tested and the object element ontology to be tested are obtained by using the pre-created mapping rules; the actual matching confidence is determined by using the preset matching confidence rules, and the collaborative forms and fields corresponding to the forms to be tested and the fields to be tested are determined according to the actual matching confidence and the preset matching confidence threshold. By identifying business objects and semi-automatically implementing the design ontology system in a bottom-up manner, a mapping relationship between the business domain data model and the real world is established, and different data sources are integrated in a standardized form, which provides strong support for business analysis and decision-making, improves the efficiency of business structure knowledge reuse, solves the problems of poor traceability and low agility in the business system, and realizes cost reduction, efficiency improvement and business governance of the enterprise.

[0024] It should be understood that the contents described in this section are not intended to identify the key or important features of the embodiments of the present invention, nor are they intended to limit the scope of the present invention. Other features of the present invention will become easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0026] Figure 1 A flowchart of a data matching method provided according to an embodiment of the present invention;

[0027] Figure 2 A flowchart of another data matching method provided according to an embodiment of the present invention;

[0028] Figure 3 A flowchart of a method for mining collaborative relationships of business domain form fields based on ontology according to an embodiment of the present invention;

[0029] Figure 4 A flowchart of establishing and predicting an event trigger word and business object element recognition model according to an embodiment of the present invention;

[0030] Figure 5 A flowchart of another data matching method provided according to an embodiment of the present invention;

[0031] Figure 6 A flowchart of a bottom-up combing ontology system provided according to an embodiment of the present invention;

[0032] Figure 7 A flowchart of mining collaborative form fields provided according to an embodiment of the present invention;

[0033] Figure 8 A schematic diagram of the structure of a data matching device provided according to an embodiment of the present invention;

[0034] Fig. 9 The present invention is a schematic diagram of the structure of an electronic device provided according to an embodiment of the present invention. DETAILED DESCRIPTION

[0035] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings 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 should fall within the scope of protection of the present invention.

[0036] It should be noted that the terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0037] In one embodiment, Figure 1 A flow chart of a data matching method provided in an embodiment of the present invention. This embodiment can be applied to data matching in a business system. The method can be executed by a data matching device. The data matching device can be implemented in the form of hardware and / or software. The data matching device can be configured in an electronic device.

[0038] like Figure 1 As shown, a data matching method provided in this embodiment may include:

[0039] S110: Obtain basic information of the form to be tested and / or basic information of the field to be tested in the business domain.

[0040] Specifically, basic information of business forms and / or basic information of fields that require data matching can be obtained in the business domain composed of business system data. The basic information of the form can include the form name, form description information, etc., and the basic information of the field can include the field name, field description information, etc.

[0041] Exemplarily, the name of the form to be tested in the business domain is obtained as "Summary of Allowances for the Elderly", and the form description information is "Basic Information on the Summary of Allowances for the Elderly"; the name of the field to be tested in the business domain is obtained as "Applicant's ID Number", and the field description information is empty.

[0042] S120, input the basic information of the form to be tested into a pre-created event trigger word recognition model to obtain a corresponding set of event trigger words to be tested, and input the basic information of the field to be tested into a pre-created business object element recognition model to obtain a corresponding set of business objects to be tested and a set of object elements to be tested.

[0043] In this embodiment, the event trigger words to be tested may include words or phrases that can represent a type of event in the form information, for example, payment, review, subsidy, allowance, etc.; the event trigger word recognition model may include a deep learning model that analyzes and identifies the input basic information of the form to be tested to obtain words or phrases that can represent a type of event in the form information; the business object to be tested may include a business entity class with business significance in the form and its field information, for example, payer, defaulting vehicle, etc.; the object elements to be tested may include corresponding attributes of the business object, for example, age, vehicle model, color, etc.; the business object element recognition model may include a deep learning model that analyzes and identifies the input basic information of the field to be tested to obtain a business entity class with business significance and corresponding attributes.

[0044] Specifically, the basic information of the business form that needs to be matched with data can be input into the event trigger word recognition model that has been trained in advance. After model recognition, a set of words or phrases in the form to be tested that can represent a type of event is obtained, that is, the event trigger word set to be tested, which may include words or short sentences such as payment, review, subsidy, and allowance. If there are multiple identical trigger words in the obtained event trigger word set to be tested, the duplicate trigger words can be removed, leaving only one of the multiple identical trigger words. The basic information of the form may include the form name, form description information, etc.; the basic information of the business field that needs to be matched with data can be input into the business object element recognition model that has been trained in advance. After model recognition, a set of business entity classes with business significance in the field to be tested is obtained, which may include payers, defaulting vehicles, etc., as well as a set of corresponding attributes of business objects, which may include age, vehicle model, color, etc. The basic information of the field may include the field name, field description information, etc.

[0045] Exemplarily, the form name "Summary of allowances for the elderly" and its description information "Basic information on the summary of allowances for the elderly" are input into a pre-created event trigger word recognition model to obtain the corresponding event trigger word set to be tested ["allowances"]; the field name "Applicant's ID number" and its corresponding field description information are empty and input into the business object element recognition model to obtain its business object set ["Applicant"] and the business object element set ["ID number"].

[0046] S130, using pre-created mapping rules to obtain the event type corresponding to the event trigger word set, and using pre-created mapping rules to obtain the object ontology and object element ontology corresponding to the business object set and object element set respectively.

[0047] In this embodiment, the object ontology can include four major object ontologies: people, places, objects, organizations and their event systems. The object ontology system can cover ontology concepts, ontology attributes and ontology hierarchical and sub-concepts. The object element ontology can include the attribute information corresponding to the object ontology itself.

[0048] Specifically, after obtaining the set of trigger words for events to be tested, the set of business objects to be tested, and the set of object elements to be tested, the type of event to be tested corresponding to the set of trigger words for events to be tested can be obtained according to the mapping rules set in advance according to the actual situation. The set of trigger words for events to be tested can include a set of words such as payer, defaulting vehicle, etc., and the type of event to be tested can include subsidies, penalties, etc.; the object ontology to be tested and the object element ontology to be tested corresponding to the set of business objects to be tested and the set of object elements to be tested can also be obtained according to the mapping rules set in advance according to the actual situation. The set of business objects to be tested can include a set of business entity classes such as payer, defaulting vehicle, etc., and the object elements to be tested can include a set of corresponding attributes of business objects such as age, car model, color, etc.

[0050] S140. Determine the actual matching confidence between the event type to be tested and the tested event type corresponding to each form basic information in a pre-created business domain database by using a preset matching confidence rule as a first actual matching confidence, and determine the actual matching confidence between the business object ontology to be tested and the object element ontology to be tested and the tested business object ontology and the tested object element ontology corresponding to each field basic information in a pre-created business domain database by using a preset matching confidence rule as a second actual matching confidence.

[0051] In this embodiment, the preset matching confidence rules may include rules set in advance according to actual conditions to obtain actual matching confidence values ​​by comparing the event type to be tested, the business object ontology to be tested, the element ontology of the object to be tested with the event type to be tested, the business object ontology to be tested, and the element ontology of the object to be tested with the measured event type, the measured business object ontology, and the measured object element ontology.

[0052] Specifically, the matching confidence rules set in advance according to actual conditions can be used to determine the actual matching confidence between the event type to be tested and the tested event type corresponding to the basic information of each form in the database established in advance containing all the business data in the business system, and this value is used as the first actual matching confidence. The event type may include subsidies, penalties, etc., and the basic information of the form may include the form name, form description information, etc.; the matching confidence rules set in advance according to actual conditions can also be used to determine the actual matching confidence between the business object ontology to be tested and the object element ontology to be tested and the tested business object ontology and the tested object element ontology corresponding to the basic information of each field in the database established in advance containing all the business data in the business system, and this value is used as the second actual matching confidence. The business object ontology may include four major object ontologies: people, land, objects, organizations and their event systems. The object element ontology may include the attributes corresponding to the object ontology, and the field basic information may include the field name, field description information, etc.

[0053] In one embodiment, the preset matching confidence rules include:

[0054] The matching confidence when the list intersection of the event type to be tested and the event type to be tested is not empty is greater than the matching confidence when the list intersection is empty;

[0055] The matching confidence that the business object ontology to be tested and the element ontology of the object to be tested are the same as the corresponding business object ontology and the element ontology of the object to be tested is greater than the matching confidence that the business object ontology to be tested is the same as the corresponding business object ontology to be tested, and the element ontology of the object to be tested is different from the element ontology of the object to be tested;

[0056] The matching confidence that the business object ontology to be tested is the same as the corresponding business object ontology already tested, and the element ontology of the object to be tested is different from the element ontology of the already tested object, is greater than the matching confidence that both the business object ontology to be tested and the element ontology of the object to be tested are different from the corresponding business object ontology and the element ontology of the already tested object.

[0057] Specifically, when the list intersection of the event type to be tested and the event type to be tested is not empty, that is, the same event type exists, the matching confidence at this time is greater than the matching confidence when the list intersection is empty, that is, the same event type does not exist. For example, the event type to be tested is "subsidy", the event type to be tested ("subsidy", "penalty"), the list intersection is not empty, the event type to be tested is "subsidy", the event type to be tested is ("penalty"), the list intersection is empty, and the matching confidence of the former is greater than the latter; when the business object ontology to be tested and the element ontology of the object to be tested are the same as the corresponding business object ontology to be tested and the element ontology of the object to be tested are the same, the matching confidence is greater than the matching confidence when the business object ontology to be tested is the same as the corresponding business object ontology to be tested, but the element ontology of the object to be tested is different from the element ontology of the object to be tested. For example, the business object ontology to be tested is "person", and the element ontology of the object to be tested is "ID card". , the measured business object ontology "person", the measured object element ontology "identity card", the matching confidence at this time is greater than the matching confidence of the measured business object ontology "person" and the measured object element ontology "blood type"; when the business object ontology to be tested is the same as the corresponding measured business object ontology, but the element ontology of the object to be tested is different from the element ontology of the measured object, the matching confidence is greater than the matching confidence when the business object ontology to be tested and the element ontology of the object to be tested are different from the corresponding measured business object ontology and the element ontology of the measured object. For example, the business object ontology to be tested is "person", the measured business object ontology is also "person", and the element ontology of the object to be tested is "identity card", and the element ontology of the measured object is "blood type". At this time, the matching confidence is greater than the matching confidence of the business object ontology to be tested is "child", the measured business object ontology is "teacher", the element ontology of the object to be tested is "identity card", and the element ontology of the measured object is "blood type".

[0058] S150: Determine a collaborative form corresponding to the form to be tested according to the first actual matching confidence and the preset matching confidence threshold, and determine a collaborative field corresponding to the field to be tested according to the second actual matching confidence and the preset matching confidence threshold.

[0059] In this embodiment, the preset matching confidence threshold may include a tested form or field set in advance according to actual conditions, which can be used as a critical value of the actual matching confidence of the form to be tested, the collaborative form or field corresponding to the field to be tested.

[0060] Specifically, whether the tested form is the collaborative form corresponding to the form to be tested can be determined based on the actual matching confidence between the event type of the form to be tested determined by using the preset matching confidence rules and the event type of the tested form, and the matching confidence threshold set in advance according to the actual situation. Alternatively, whether the tested field is the collaborative field corresponding to the field to be tested can be determined based on the actual matching confidence between the business object ontology of the field to be tested and the object element ontology of the field to be tested determined by using the preset matching confidence rules and the business object ontology and the object element ontology of the tested field, and the comparison between the actual matching confidence and the matching confidence threshold set in advance according to the actual situation.

[0061] In one embodiment, determining a collaborative form corresponding to the form to be tested according to a first actual matching confidence and a preset matching confidence threshold, and determining a collaborative field corresponding to the field to be tested according to a second actual matching confidence and a preset matching confidence threshold, includes:

[0062] The candidate form whose first actual matching confidence is less than the preset matching confidence threshold is used as the collaborative form corresponding to the form to be tested;

[0063] The candidate field whose second actual matching confidence is less than the preset matching confidence threshold is used as the collaborative field corresponding to the field to be tested.

[0064] Specifically, when the actual matching confidence between the event type of the form to be tested and the event type of the tested form determined by the preset matching confidence rule, that is, the first actual matching confidence is less than the matching confidence threshold set in advance according to the actual situation, the tested form in the pre-created business domain database, that is, the candidate form, can be used as the collaborative form corresponding to the form to be tested; when the actual matching confidence between the business object ontology of the field to be tested and the object element ontology of the field to be tested and the business object ontology and the tested object element ontology of the field to be tested determined by the preset matching confidence rule, that is, the second actual matching confidence is less than the matching confidence threshold set in advance according to the actual situation, the tested field in the pre-created business domain database, that is, the candidate field, can be used as the collaborative field corresponding to the field to be tested.

[0065] The technical solution provided by the embodiment of the present invention obtains the basic information of the form fields to be tested in the business domain by inputting them into the pre-created event trigger word recognition model and the business object element recognition model, respectively, to obtain the corresponding event trigger word set to be tested, the business object set to be tested and the object element set to be tested; uses the pre-created mapping rules to obtain the corresponding event type to be tested, the object ontology to be tested and the object element ontology to be tested; uses the preset matching confidence rule to determine the actual matching confidence, and determines the collaborative form fields corresponding to the form to be tested and the field to be tested according to the actual matching confidence and the preset matching confidence threshold, and semi-automatically realizes the design of the ontology system in a bottom-up manner by identifying business objects, establishes the mapping relationship between the business domain data model and the real world, and integrates different data sources in a standardized form, which provides strong support for business analysis and decision-making, improves the efficiency of business structure knowledge reuse, solves the problems of poor traceability and low agility in the business system, and realizes cost reduction, efficiency improvement and business governance of the enterprise.

[0066] Based on the above embodiments, event trigger words and event types present a many-to-one mapping relationship; business objects and their object elements present a one-to-one mapping relationship with business object ontology and its object element ontology.

[0067] Specifically, business event trigger words and business event types can present a many-to-one mapping relationship, that is, multiple event trigger words can correspond to the same event type. For example, the event trigger words "grant" and "subsidy" both correspond to the event type "subsidy"; business objects and their object elements and business object ontology and its object element ontology present a one-to-one mapping relationship, that is, a business object and its object element can only correspond to one business object ontology and its object element ontology.

[0068] In one embodiment, Figure 2 A flowchart of another data matching method provided for an embodiment of the present invention. Based on the above embodiments, this embodiment further optimizes and expands the training process of the event trigger word recognition model and the business object element recognition model.

[0069] like Figure 2 As shown, a method for establishing an event trigger word and business object element identification model provided by this embodiment may include:

[0070] S210, obtaining basic information of training forms and basic information of training fields in different business domain databases.

[0071] Specifically, the basic information of the form and the basic information of the field used for model training can be obtained from the database containing data from different business systems. The basic information of the form can include the form name, form description information, etc., and the basic information of the field can include the field name, field description information, etc.

[0072] S220: Annotate the basic information of the training form and the basic information of the training field respectively to obtain corresponding annotation event trigger words, annotation business objects, and annotation object elements.

[0073] Specifically, after obtaining the basic information of the training form and the basic information of the training fields, the basic information of the training form and the basic information of the training fields can be annotated respectively to obtain the annotated event trigger words, business objects and object elements. The event trigger words can include words or phrases that can represent a type of event in the form information, such as payment, review, subsidy, allowance, etc. Business objects can include business entity classes with business significance in the form and its field information, such as payer, defaulting vehicle, etc. Object elements can include corresponding attributes of the business object, such as age, vehicle model, color, etc.

[0074] S230, based on the marked event trigger words, the basic information of the training form is input as a training sample into the information extraction model for iterative training to obtain the corresponding event trigger word recognition model, and based on the marked business objects and marked object elements, the basic information of the training fields is input as a training sample into the information extraction model for iterative training to obtain the corresponding business object element recognition model.

[0075] Specifically, based on the marked event trigger words and the basic information of the form used for model training, iterative training can be performed by inputting the basic information of the form used for model training into the information extraction model for iterative training, and the corresponding event trigger word recognition model can be obtained; based on the marked business objects and object elements, and the basic information of the fields used for model training can be input into the information extraction model for iterative training, the corresponding business object element recognition model can be obtained. The event trigger word recognition model and the business object element recognition model can use a deep learning algorithm model to extract entities.

[0076] S240, input the basic information of the form to be tested into a pre-created event trigger word recognition model to obtain a corresponding set of event trigger words to be tested, and input the basic information of the field to be tested into a pre-created business object element recognition model to obtain a corresponding set of business objects to be tested and a set of object elements to be tested.

[0077] S250, using pre-created mapping rules to obtain the event type corresponding to the event trigger word set, and using pre-created mapping rules to obtain the object ontology and object element ontology corresponding to the business object set and object element set respectively.

[0078] S260. Determine the actual matching confidence between the event type to be tested and the tested event type corresponding to each form basic information in a pre-created business domain database by using a preset matching confidence rule as a first actual matching confidence, and determine the actual matching confidence between the business object ontology to be tested and the object element ontology to be tested and the tested business object ontology and the tested object element ontology corresponding to each field basic information in a pre-created business domain database by using a preset matching confidence rule as a second actual matching confidence.

[0079] S270: Determine the collaborative form corresponding to the form to be tested according to the first actual matching confidence and the preset matching confidence threshold, and determine the collaborative field corresponding to the field to be tested according to the second actual matching confidence and the preset matching confidence threshold.

[0080] The technical solution provided by the embodiment of the present invention obtains the basic information of the training form and the basic information of the training field in different business domain databases, and respectively annotates the basic information of the training form and the basic information of the training field to obtain the corresponding annotated event trigger words and the annotated business objects and the annotated object elements based on the annotated event trigger words, based on the annotated event trigger words, the annotated business objects and the annotated object elements, and inputs the training form and the basic information of the field as training samples into the information extraction model for iterative training, so as to obtain an accurate event trigger word recognition model and a business object element recognition model, and obtain the corresponding event type to be tested, the object ontology to be tested and the object element ontology to be tested by using the pre-created mapping rules; the actual matching confidence is determined by using the preset matching confidence rule, and the collaborative form and field corresponding to the form to be tested and the field to be tested are determined according to the actual matching confidence and the preset matching confidence threshold, so as to more accurately identify the business event trigger words, business objects and object elements in the business system, improve the accuracy and efficiency of data matching, provide strong support for business analysis and decision-making, improve the efficiency of business structure knowledge reuse, solve the problems of poor traceability and low agility in the business system, and realize the cost reduction and efficiency improvement and business governance of the enterprise.

[0081] Based on the above embodiment, the basic information of the training form includes: the training form name and form description information; based on the labeled event trigger words, the basic information of the training form is input as a training sample into the information extraction model for iterative training to obtain the corresponding event trigger word recognition model, including:

[0082] Concatenate the training form name and form description information to obtain the corresponding training form input fragment;

[0083] Input the training form input fragment into the pre-trained language model for encoding and vectorization to obtain the corresponding input feature vector;

[0084] Input the input feature vector as input feature to the information extraction model to obtain the corresponding output feature vector;

[0085] The output feature vector is decoded using a semi-pointer and semi-annotation method to obtain the corresponding actual event trigger word;

[0086] When the actual event trigger word does not match the labeled event trigger word, return to the step of inputting the training form input fragment into the pre-trained language model for encoding and vectorization, so as to perform iterative training until the actual event trigger word matches the labeled event trigger word, and obtain the corresponding event trigger word recognition model.

[0087] Figure 3 A flowchart of a method for mining collaborative relationships of business domain form fields based on ontology provided by an embodiment of the present invention is shown in FIG. Figure 3 As shown on the left, when determining the collaborative relationship between business domain form fields through data matching, it is necessary to identify the leave type and business object elements of the form and field to be tested, and obtain the leave type, business object elements and ontology of the form and field to be tested through mapping rules; Figure 3 As shown on the right, it also includes an offline processing module, namely, the establishment of an event trigger word recognition model and a business object element recognition model. It is necessary to perform structured processing on all forms and fields in the business domain form database, and use labeled form and field sample sets for model training.

[0088] Specifically, Figure 4 A flowchart of establishing and predicting an event trigger word and business object element recognition model provided by an embodiment of the present invention, such as Figure 4As shown, based on the basic information of the training form including the training form name and form description information, and based on the labeled event trigger words, the basic information of the training form can be input as a training sample into the information extraction model for iterative training to obtain a corresponding event trigger word recognition model. The specific steps may include: splicing the training form name and the form description information to obtain the corresponding training form input segment in the event trigger word recognition model. For example, the form name is "Summary of allowances for the elderly", and the description information is "Basic situation of the summary of allowances for the elderly". The spliced ​​training form input segment is "[CLS] Summary of allowances for the elderly is the basic situation of the summary of allowances for the elderly"; the training form input segment is Input into the pre-trained language model for encoding and vectorization to obtain a corresponding input feature vector. The pre-trained language model may include BERT, RoBERTa, etc. For example, "[CLS] The summary table of subsidies for the elderly is the basic information of the summary of subsidies for the elderly" is input into the BERT pre-trained language model for encoding and vectorization to obtain a 128-dimensional feature vector [7.0709, -0.8330, ..., -1.1320, -1.3722]; the input feature vector is input as an input feature into the information extraction model to obtain a corresponding output feature vector. The information extraction model may include a deep learning model. For example, the input feature is processed through a dropout layer, a layer normalization layer, The fully connected layer is converted into the final 64-dimensional output feature; the output feature vector is decoded using a half-pointer and half-annotation method to obtain the corresponding actual event trigger word. If there are multiple identical trigger words in the actual event trigger word set, the duplicate trigger words can be removed, leaving only one of the multiple identical trigger words. Exemplarily, the half-pointer and half-annotation method is used for decoding to obtain the entity type "trigger word" The corresponding sequence length is 32 bits, the starting position pointer sequence start_ids = [0,0,0,0,0,1,0,0,0,0,0,0,0,0,0,1,...,0,0] and the ending position pointer sequence end_ids = [0,0,0 ,0,0,0,0,1,0,0,0,0,0,0,0,0,0,0,0,1,...,0,0], so the entity set of entity type "trigger word" obtained by decoding is ["Jinzhu", "Jinzhu"], and after removing duplications, the trigger word set obtained by the trigger word recognition model for the form and its description information is ["Jinzhu"]; when the actual event trigger word obtained does not match the labeled event trigger word, we can return to the step of inputting the training form input fragment into the pre-trained language model for encoding and vectorization, and continue to iteratively train the event trigger word recognition model until the actual event trigger word matches the labeled event trigger word, so as to obtain a high-precision event trigger word recognition model.

[0089] Based on the above embodiment, the basic information of the training field includes: the name of the training field and the field description information; based on the annotated business object and the annotated object element, the basic information of the training field is input as a training sample into the information extraction model for iterative training to obtain the corresponding business object element recognition model, including:

[0090] Concatenate the training field name and field description information to obtain the corresponding training field input segment;

[0091] Input the training field input segment into the pre-trained language model for encoding and vectorization to obtain a corresponding first input feature vector;

[0092] The part-of-speech sequence features of the training field input segment are one-hot encoded and vectorized through a bidirectional LSTM model to obtain the corresponding second input feature vector;

[0093] Concatenate the first input feature vector and the second input feature vector to obtain a corresponding final input feature vector;

[0094] The final input feature vector is input as the input feature into the information extraction model to obtain the corresponding output feature vector;

[0095] The output feature vector is decoded using a semi-pointer and semi-annotation method to obtain the corresponding actual business object and actual object elements;

[0096] When the actual business object does not match the labeled business object, and the actual object requirement does not match the labeled object element, return to the step of inputting the training field input fragment into the pre-trained language model for encoding and vectorization, so as to perform iterative training until the actual business object matches the labeled business object, and the actual object requirement matches the labeled object element, and the corresponding business object element recognition model is obtained.

[0097] Specifically, Figure 4As shown, based on the basic information of the training field including the training field name and the field description information, and based on the annotated business object and the annotated object element, the basic information of the training field can be input as a training sample into the information extraction model for iterative training to obtain the corresponding business object element recognition model. The specific steps may include: splicing the training field name and the field description information to obtain the corresponding training field input segment in the business object element recognition model. For example, the field name is "applicant's ID number", and its corresponding field description information is empty, and the spliced ​​training field input segment is "[CLS] applicant's ID number"; the training field input can be The segment is input into a pre-trained language model for encoding and vectorization to obtain a corresponding first input feature vector. The pre-trained language model may include BERT, RoBERTa, etc. For example, "[CLS] Applicant ID number" is input into the BERT pre-trained language model for encoding and vectorization to obtain a 128-dimensional feature vector [5.8211, -0.1016, ..., -0.3537, 1.9704]; the part-of-speech sequence feature of the training field input segment is one-hot encoded, i.e., one-hot encoded, and vectorized through a bidirectional LSTM model, i.e., a BiLSTM model, to obtain a corresponding second input feature vector. The first input feature vector and the second input feature vector are concatenated to obtain the corresponding final input feature vector; the final input feature vector is input as an input feature into the information extraction model to obtain the corresponding output feature vector, and the information extraction model may include a deep learning model. Exemplarily, the input feature is converted into a final 64-dimensional output feature through a dropout layer, a layer normalization layer, and a fully connected layer; the output feature vector is decoded using a semi-pointer and semi-annotation method to obtain the corresponding actual business object and actual object element. Exemplarily, the semi-pointer and semi-annotation method is used to decode and obtain the corresponding sequence of the entity type "business object". The column lengths are both 32 bits, the starting position pointer sequence start_ids = [0,1,0,...,0,0] and the ending position pointer sequence end_ids = [0,0,0,1,0,...,0,0], so the entity set of the entity type "business object" obtained by decoding is ["applicant"], and the corresponding sequence lengths of the entity type "business object element" are both 32 bits, the starting position pointer sequence start_ids = [0,0,0,0,1,0,...,0,0] and the ending position pointer sequence end_ids = [0,0,0,0,0,0,0,0,1,0,...,0,0], so the entity set of the entity type "business object element" obtained by decoding is ["ID number"], that is, the business object set obtained by the business object element recognition model for this field and its description information is ["applicant"], and the business object element set is ["ID number"]; when the actual business object obtained does not match the labeled business object, and the actual object requirement does not match the labeled object element, it is possible to return to the step of inputting the training field input fragment into the pre-trained language model for encoding vectorization, and continue to iterate the training of the business object element recognition model until the actual business object matches the labeled business object, and the actual object requirement matches the labeled object element, so as to obtain a high-precision business object element recognition model. .

[0098] In one embodiment, Figure 5 A flowchart of another data matching method provided for an embodiment of the present invention. Based on the above embodiments, this embodiment is taken as a preferred embodiment, and takes the business domain database OpenSearch as an example to illustrate the process of determining data collaborative relationship forms and fields.

[0099] like Figure 5 As shown, another data matching method provided by this embodiment may include:

[0100] S310, querying the basic information of the form and the basic information of its fields from different business domain databases to obtain the form name and its description information, the field name and its description information data.

[0101] Specifically, the basic information of the form and its field basic information related data can be queried from different business domain databases to obtain the form name and its description information, the field name and its description information data, which may include the forms and their field information covered by the business systems of many enterprises or government units, including but not limited to the form name and field name information.

[0102] S320, annotate the form and its field samples, the annotated content includes event trigger words and business objects and corresponding elements, input the annotated samples into the information extraction model for training, and obtain the event trigger word recognition model and the business object element recognition model.

[0103] Specifically, the data can be structured according to the data distribution of the form and its field information, and the event trigger words, business objects and corresponding elements of the business data form and fields can be labeled respectively. The labeled samples can be input into the information extraction model for training to obtain the event trigger word recognition model and the business object element recognition model.

[0104] S330, using the event trigger word recognition model to identify the event trigger words of all business domain forms and their fields; using the business object element recognition model to identify the business objects and object elements of all business domain forms and their fields.

[0105] Specifically, the event trigger word recognition model and the business object element recognition model can be used to recognize all forms and their field information in the business domain form database to obtain corresponding event trigger word sets, business object sets, business object element sets and other information.

[0106] S340. Use a bottom-up approach to proofread, summarize, and sort out the ontology system of people, places, objects, organizations, and events involved in the business domain forms and their field databases; on this basis, encode and map rules, establish a one-to-one mapping relationship between the business domain data model and the ontology system, and store it in the OpenSearch database.

[0107] Specifically, Figure 6 A flowchart of a bottom-up combing ontology system provided by an embodiment of the present invention, such as Figure 6 As shown, a bottom-up semi-automatic approach can be used to sort out the ontology systems of people, places, objects, organizations, and events involved in the business domain data model, including the four major ontology objects of people, places, objects, organizations and their event systems, establish a one-to-one mapping relationship between the business domain data model and the ontology system, and store the corresponding mapping event types, ontologies, attributes, etc. obtained by the mapping rules in the OpenSearch database.

[0108] S350. When inputting any business domain form name and its description information, field name and its description information, the event trigger word recognition model and the business object element recognition model are used to obtain the event type, business object and its elements; the mapping rules are used to obtain the ontology and its attributes corresponding to the business domain data model data; the preset matching confidence rules are used to give the matching confidence of the business domain data model and the ontology element set of the form and its fields in the business domain database; the form and its fields that are less than the preset confidence threshold are used as the collaborative form and its fields corresponding to the business domain form and its fields.

[0109] In this embodiment, any business form and field information is the form and field to be retrieved for collaboration. The trigger word set is obtained by the event trigger word recognition model, the business object set and the business object element set are obtained by the business object element recognition model, and the corresponding event type, ontology and its attribute set are obtained by the mapping rules. The form and field name and description information are input to retrieve a list of candidate forms and field information with higher scores in OpenSearch, wherein the form and field information includes the corresponding trigger word set and event type set, business object and element set, ontology and attribute set; the preset matching confidence rules may include: the event type set between form fields is taken to have an intersection, and the confidence of the intersection being not empty is higher than the confidence of the intersection being empty; the confidence is high if the ontology of the business domain data model data mapping between form fields is the same, and the confidence is high if the attributes are the same; and the combination matching rules of the preset form field information of the collaborative field to be retrieved and the information of each candidate form field shown in Table 1.

[0110] Table 1 Preset combination matching rules of the form and field information of the collaborative field to be retrieved and the information of each candidate form and field

[0111]

[0112] Specifically, Figure 7 A flowchart of mining collaborative form fields provided by an embodiment of the present invention, such as Figure 7 As shown, any business domain form and basic field information can be input into the event trigger word recognition model and the business object element recognition model to obtain the event type, business object and its elements, and the ontology and its attributes corresponding to the business domain data model data can be obtained using the mapping rules; the preset matching confidence rules can be used to match any business domain form field information with the business domain data model and ontology element set of the form and its fields in the business domain database and give a matching confidence, and the form and its fields with a confidence threshold less than the preset confidence threshold will be used as the collaborative form and its fields corresponding to the business domain form and its fields, and the candidate form and field information with a confidence threshold greater than the preset confidence threshold will be discarded and not used as its collaborative form field.

[0113] The technical solution provided by the embodiment of the present invention semi-automatically designs an ontology system by identifying business objects and in a bottom-up manner. Compared with the traditional pure manual summary and combing method, it not only improves the accuracy of obtaining the ontology system in the business field, but also improves the combing efficiency. At the same time, the embodiment of the present invention establishes a mapping relationship between the business domain data model and the real world, and integrates different data sources in a standardized form, connects the data of the same business object in different business systems, and further establishes the connection between different business systems. It can not only provide strong support for business analysis and decision-making, but also improve the efficiency of business structure knowledge reuse, thereby solving the problems of poor traceability and low agility in the business system, and realizing cost reduction, efficiency improvement and business governance of the enterprise.

[0114] In one embodiment, Figure 8 1 is a schematic diagram of a data matching device provided by an embodiment of the present invention. This embodiment can execute the above implementation mode. This embodiment can be applied to data matching in a business system. The device can be implemented in hardware / software and can be configured in an electronic device.

[0115] like Figure 8 As shown, the information acquisition module 401, the set acquisition module 402, the ontology acquisition module 403, the actual matching confidence determination module 404, and the collaborative form field determination module 405 provided in this embodiment include:

[0116] The information acquisition module 401 is used to acquire basic information of the form to be tested and / or basic information of the field to be tested in the business domain.

[0117] The collection acquisition module 402 is used to input the basic information of the form to be tested into a pre-created event trigger word recognition model to obtain the corresponding event trigger word set to be tested, and input the basic information of the field to be tested into a pre-created business object element recognition model to obtain the corresponding business object set to be tested and the object element set to be tested.

[0118] The ontology acquisition module 403 is used to obtain the event type corresponding to the event trigger word set using the pre-created mapping rules, and to obtain the object ontology and object element ontology corresponding to the business object set and object element set respectively using the pre-created mapping rules.

[0119] The actual match confidence determination module 404 is used to use the preset match confidence rules to determine the actual match confidence between the event type to be tested and the measured event type corresponding to each form basic information in the pre-created business domain database, as the first actual match confidence, and to use the preset match confidence rules to determine the actual match confidence between the business object ontology to be tested and the object element ontology to be tested and the measured business object ontology and the measured object element ontology corresponding to each field basic information in the pre-created business domain database, as the second actual match confidence.

[0120] The collaborative form field determination module 405 is used to determine the collaborative form corresponding to the form to be tested according to the first actual matching confidence and the preset matching confidence threshold, and to determine the collaborative field corresponding to the field to be tested according to the second actual matching confidence and the preset matching confidence threshold.

[0121] The technical solution provided by the embodiment of the present invention obtains the basic information of the forms and fields to be tested in the business domain by inputting them into the pre-created event trigger word recognition model and the business object element recognition model, respectively, to obtain the corresponding event trigger word set to be tested, the business object set to be tested and the object element set to be tested; the corresponding event type to be tested, the object ontology to be tested and the object element ontology to be tested are obtained by using the pre-created mapping rules; the actual matching confidence is determined by using the preset matching confidence rules, and the collaborative forms and fields corresponding to the forms to be tested and the fields to be tested are determined according to the actual matching confidence and the preset matching confidence threshold, and the ontology system is semi-automatically designed in a bottom-up manner by identifying business objects, establishing a mapping relationship between the business domain data model and the real world, and integrating different data sources in a standardized form, which provides strong support for business analysis and decision-making, improves the efficiency of business structure knowledge reuse, solves the problems of poor traceability and low agility in the business system, and realizes cost reduction, efficiency improvement and business governance of the enterprise.

[0122] Based on the above embodiment, the data matching device further includes:

[0123] The basic information acquisition module is used to obtain the basic information of training forms and training fields in different business domain databases.

[0124] The standard module is used to mark the basic information of the training form and the basic information of the training field respectively, and obtain the corresponding marking event trigger words, marking business objects and marking object elements.

[0125] The iterative training module is used to iteratively train based on the labeled event trigger words and input the basic information of the training form as a training sample into the information extraction model to obtain the corresponding event trigger word recognition model, and to iteratively train based on the labeled business objects and labeled object elements and input the basic information of the training fields as a training sample into the information extraction model to obtain the corresponding business object element recognition model.

[0126] Based on the above embodiment, the basic information of the training form includes: the name of the training form and the form description information; based on the labeled event trigger words, the basic information of the training form is input as a training sample into the information extraction model for iterative training to obtain the corresponding event trigger word recognition model, which is specifically used for:

[0127] Concatenate the training form name and form description information to obtain the corresponding training form input fragment.

[0128] The training form input fragment is input into the pre-trained language model for encoding and vectorization to obtain the corresponding input feature vector.

[0129] The input feature vector is input as the input feature to the information extraction model to obtain the corresponding output feature vector.

[0130] The output feature vector is decoded using a semi-pointer and semi-annotation method to obtain the corresponding actual event trigger word.

[0131] When the actual event trigger word does not match the labeled event trigger word, return to the step of inputting the training form input fragment into the pre-trained language model for encoding and vectorization, so as to perform iterative training until the actual event trigger word matches the labeled event trigger word, and obtain the corresponding event trigger word recognition model.

[0132] Based on the above embodiment, the basic information of the training field includes: a training field name and field description information; based on the annotated business object and the annotated object element, the basic information of the training field is input as a training sample into the information extraction model for iterative training to obtain a corresponding business object element recognition model, which is specifically used for:

[0133] The training field name and field description information are concatenated to obtain the corresponding training field input segment.

[0134] The training field input segment is input into the pre-trained language model for encoding and vectorization to obtain the corresponding first input feature vector.

[0135] The part-of-speech sequence features of the training field input segment are one-hot encoded and vectorized through a bidirectional LSTM model to obtain the corresponding second input feature vector.

[0136] The first input feature vector and the second input feature vector are concatenated to obtain a corresponding final input feature vector.

[0137] The final input feature vector is input as the input feature to the information extraction model to obtain the corresponding output feature vector.

[0138] The output feature vector is decoded using a semi-pointer and semi-annotation method to obtain the corresponding actual business object and actual object elements.

[0139] When the actual business object does not match the labeled business object, and the actual object requirement does not match the labeled object element, return to the step of inputting the training field input fragment into the pre-trained language model for encoding and vectorization, so as to perform iterative training until the actual business object matches the labeled business object, and the actual object requirement matches the labeled object element, and the corresponding business object element recognition model is obtained.

[0140] Based on the above embodiments, event trigger words and event types present a many-to-one mapping relationship; business objects and their object elements present a one-to-one mapping relationship with business object ontology and its object element ontology.

[0141] Based on the above embodiment, the preset matching confidence rules include:

[0142] The matching confidence when the intersection of the lists of the event type to be tested and the event type to be tested is not empty is greater than the matching confidence when the intersection of the lists is empty.

[0143] The matching confidence that the business object ontology to be tested and the element ontology of the object to be tested are the same as the corresponding business object ontology and the element ontology of the object to be tested is greater than the matching confidence that the business object ontology to be tested is the same as the corresponding business object ontology to be tested, and the element ontology of the object to be tested is different from the element ontology of the object to be tested.

[0144] The matching confidence that the business object ontology to be tested is the same as the corresponding business object ontology already tested, and the element ontology of the object to be tested is different from the element ontology of the already tested object, is greater than the matching confidence that both the business object ontology to be tested and the element ontology of the object to be tested are different from the corresponding business object ontology and the element ontology of the already tested object.

[0145] Based on the above embodiment, the collaborative form field determination module 405 includes:

[0146] The collaborative form unit is used to take the candidate form whose first actual matching confidence is less than a preset matching confidence threshold as the collaborative form corresponding to the form to be tested.

[0147] The collaborative field unit is used to use the candidate field whose second actual matching confidence is less than a preset matching confidence threshold as the collaborative field corresponding to the field to be tested.

[0148] A data matching device provided in an embodiment of the present invention can execute any data matching method provided in an embodiment of the present invention, and has corresponding functional modules and beneficial effects of the execution method. For the contents not described in detail in this embodiment, reference can be made to the description in any method embodiment of the present invention.

[0149] In one embodiment, Fig. 9 A schematic diagram of the structure of an electronic device provided for an embodiment of the present invention. An electronic device 50 that can be used to implement an embodiment of the present invention is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device may also represent various forms of mobile devices, such as personal digital processing, cellular phones, smart phones, wearable devices (such as helmets, glasses, watches, etc.) and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present invention described and / or required herein.

[0150] like Fig. 9 As shown, the electronic device 50 includes at least one processor 51, and a memory connected to the at least one processor 51 in communication, such as a read-only memory (ROM) 52, a random access memory (RAM) 53, etc., wherein the memory stores a computer program that can be executed by at least one processor, and the processor 51 can perform various appropriate actions and processes according to the computer program stored in the read-only memory (ROM) 52 or the computer program loaded from the storage unit 58 to the random access memory (RAM) 53. In the RAM 53, various programs and data required for the operation of the electronic device 50 can also be stored. The processor 51, the RAM 52, and the RAM 53 are connected to each other via a bus 54. An input / output (I / O) interface 55 is also connected to the bus 54.

[0151] A number of components in the electronic device 50 are connected to the I / O interface 55, including: an input unit 55, such as a keyboard, a mouse, etc.; an output unit 57, such as various types of displays, speakers, etc.; a storage unit 58, such as a disk, an optical disk, etc.; and a communication unit 59, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 59 allows the electronic device 50 to exchange information / data with other devices through a computer network such as the Internet and / or various telecommunication networks.

[0152] The processor 51 may be a variety of general and / or special processing components with processing and computing capabilities. Some examples of the processor 51 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various dedicated artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, a digital signal processor (DSP), and any appropriate processor, controller, microcontroller, etc. The processor 51 performs the various methods and processes described above, such as a data matching method.

[0153] In some embodiments, the data matching method may be implemented as a computer program, which is tangibly contained in a computer-readable storage medium, such as a storage unit 58. In some embodiments, part or all of the computer program may be loaded and / or installed on the electronic device 50 via the ROM 52 and / or the communication unit 59. When the computer program is loaded into the RAM 53 and executed by the processor 51, one or more steps of the data matching method described above may be performed. Alternatively, in other embodiments, the processor 51 may be configured to perform the data matching method in any other appropriate manner (e.g., by means of firmware).

[0154] Various implementations of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), load programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various implementations can include: being implemented in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which can be a special purpose or general purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.

[0155] Computer programs for implementing the methods of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, so that when the computer program is executed by the processor, the functions / operations specified in the flow chart and / or block diagram are implemented. The computer program may be executed entirely on the machine, partially on the machine, partially on the machine and partially on a remote machine as a stand-alone software package, or entirely on a remote machine or server.

[0156] In the context of the present invention, a computer-readable storage medium may be a tangible medium that may contain or store a computer program for use by or in combination with an instruction execution system, device or equipment. A computer-readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices or equipment, or any suitable combination of the foregoing. Alternatively, a computer-readable storage medium may be a machine-readable signal medium. A more specific example of a machine-readable storage medium may include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0157] To provide interaction with a user, the systems and techniques described herein may be implemented on an electronic device having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and a pointing device (e.g., a mouse or trackball) through which the user can provide input to the electronic device. Other types of devices may also be used to provide interaction with the user; for example, the feedback provided to the user may be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user may be received in any form (including acoustic input, voice input, or tactile input).

[0158] The systems and techniques described herein may be implemented in a computing system that includes backend components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes frontend components (e.g., a user computer with a graphical user interface or a web browser through which a user can interact with implementations of the systems and techniques described herein), or a computing system that includes any combination of such backend components, middleware components, or frontend components. The components of the system may be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: a local area network (LAN), a wide area network (WAN), a blockchain network, and the Internet.

[0159] A computing system may include a client and a server. The client and the server are generally remote from each other and usually interact through a communication network. The client and server relationship is generated by computer programs running on the corresponding computers and having a client-server relationship with each other. The server may be a cloud server, also known as a cloud computing server or cloud host, which is a host product in the cloud computing service system to solve the defects of difficult management and weak business scalability in traditional physical hosts and VPS services.

[0160] It should be understood that the various forms of processes shown above can be used to reorder, add or delete steps. For example, the steps described in the present invention can be executed in parallel, sequentially or in different orders, as long as the desired results of the technical solution of the present invention can be achieved, and this document does not limit this.

[0161] The above specific implementations do not constitute a limitation on the protection scope of the present invention. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modification, equivalent substitution and improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A data matching method, characterized in that: include: Obtain the basic information of the form to be tested and / or the basic information of the field to be tested in the business domain; Input the basic information of the form to be tested into a pre-created event trigger word recognition model to obtain a corresponding set of event trigger words to be tested, and input the basic information of the field to be tested into a pre-created business object element recognition model to obtain a corresponding set of business objects to be tested and a set of object elements to be tested; Using a pre-created mapping rule to obtain a to-be-tested event type corresponding to the to-be-tested event trigger word set, and using a pre-created mapping rule to obtain an to-be-tested object ontology and an to-be-tested object element ontology corresponding to the to-be-tested business object set and the to-be-tested object element set, respectively; Determine, by using a preset matching confidence rule, an actual matching confidence between the event type to be tested and the tested event type corresponding to each form basic information in a pre-created business domain database, as a first actual matching confidence, and determine, by using a preset matching confidence rule, an actual matching confidence between the business object ontology to be tested and the element ontology of the object to be tested and the tested business object ontology and the element ontology of the tested object corresponding to each field basic information in a pre-created business domain database, as a second actual matching confidence; The collaborative form corresponding to the form to be tested is determined according to the first actual matching confidence and the preset matching confidence threshold, and the collaborative field corresponding to the field to be tested is determined according to the second actual matching confidence and the preset matching confidence threshold.

2. The method according to claim 1, characterized in that The method further comprises: Obtain basic information of training forms and training fields in different business domain databases; The training form basic information and the training field basic information are respectively annotated to obtain corresponding annotation event trigger words, annotation business objects, and annotation object elements; Based on the marked event trigger words, the basic information of the training form is input as a training sample into the information extraction model for iterative training to obtain a corresponding event trigger word recognition model; and based on the marked business objects and the marked object elements, the basic information of the training fields is input as a training sample into the information extraction model for iterative training to obtain a corresponding business object element recognition model.

3. The method according to claim 2, characterized in that The training form basic information includes: training form name and form description information; based on the labeled event trigger words, the training form basic information is input as a training sample into the information extraction model for iterative training to obtain the corresponding event trigger word recognition model, including: Concatenate the training form name and the form description information to obtain a corresponding training form input segment; Inputting the training form input segment into the pre-trained language model for encoding and vectorization to obtain a corresponding input feature vector; Inputting the input feature vector as an input feature into an information extraction model to obtain a corresponding output feature vector; Decoding the output feature vector using a semi-pointer and semi-annotation method to obtain a corresponding actual event trigger word; In the case that the actual event trigger word does not match the annotated event trigger word, return to the step of inputting the training form input fragment into the pre-trained language model for encoding and vectorization, so as to perform iterative training until the actual event trigger word matches the annotated event trigger word, thereby obtaining a corresponding event trigger word recognition model.

4. The method according to claim 2, characterized in that: The training field basic information includes: training field name and field description information; based on the annotated business object and the annotated object element, the training field basic information is input as a training sample into the information extraction model for iterative training to obtain the corresponding business object element recognition model, including: Concatenate the training field name and the field description information to obtain a corresponding training field input segment; Inputting the training field input segment into a pre-trained language model for encoding and vectorization to obtain a corresponding first input feature vector; One-hot encodes the part-of-speech sequence features of the training field input segment, and vectorizes them through a bidirectional LSTM model to obtain a corresponding second input feature vector; Concatenate the first input feature vector and the second input feature vector to obtain a corresponding final input feature vector; Inputting the final input feature vector as an input feature into an information extraction model to obtain a corresponding output feature vector; Decoding the output feature vector using a semi-pointer and semi-annotation method to obtain a corresponding actual business object and an actual object element; When the actual business object does not match the annotated business object, and the actual object requirement does not match the annotated object element, return to the step of inputting the training field input segment into the pre-trained language model for encoding and vectorization, so as to perform iterative training until the actual business object matches the annotated business object, and the actual object requirement matches the annotated object element, thereby obtaining a corresponding business object element recognition model.

5. The method according to claim 1, characterized in that The event trigger word and the event type present a many-to-one mapping relationship; the business object and its object elements present a one-to-one mapping relationship with the business object ontology and its object element ontology.

6. The method according to claim 1, characterized in that The preset matching confidence rule includes: The matching confidence level when the intersection of the lists of the event type to be tested and the event type to be tested is not empty is greater than the matching confidence level when the intersection of the lists is empty; The matching confidence that the business object ontology to be tested and the element ontology of the object to be tested are identical to the corresponding business object ontology and element ontology of the object to be tested is greater than the matching confidence that the business object ontology to be tested is identical to the corresponding business object ontology to be tested, and the element ontology of the object to be tested is different from the element ontology of the object to be tested; The matching confidence that the business object ontology to be tested is identical to the corresponding business object ontology already tested, and that the element ontology of the object to be tested is different from the element ontology of the object already tested, is greater than the matching confidence that both the business object ontology to be tested and the element ontology of the object to be tested are different from the corresponding business object ontology already tested and the element ontology of the object already tested.

7. The method according to claim 1, characterized in that The step of determining the collaborative form corresponding to the form to be tested according to the first actual matching confidence and the preset matching confidence threshold, and determining the collaborative field corresponding to the field to be tested according to the second actual matching confidence and the preset matching confidence threshold, includes: The candidate form whose first actual matching confidence is less than a preset matching confidence threshold is used as the collaborative form corresponding to the form to be tested; The candidate field whose second actual matching confidence is less than a preset matching confidence threshold is used as a collaborative field corresponding to the field to be tested.

8. A data matching device, characterized in that: The device comprises: An information acquisition module is used to acquire basic information of the form to be tested and / or basic information of the field to be tested in the business domain; A set acquisition module is used to input the basic information of the form to be tested into a pre-created event trigger word recognition model to obtain a corresponding set of event trigger words to be tested, and input the basic information of the field to be tested into a pre-created business object element recognition model to obtain a corresponding set of business objects to be tested and a set of object elements to be tested; An ontology acquisition module, used to obtain the event type to be tested corresponding to the set of trigger words of the event to be tested by using a pre-created mapping rule, and to obtain the object ontology to be tested and the object element ontology to be tested corresponding to the set of business objects to be tested and the set of object elements to be tested respectively by using a pre-created mapping rule; An actual matching confidence determination module is used to determine the actual matching confidence between the event type to be tested and the tested event type corresponding to each form basic information in the pre-created business domain database using a preset matching confidence rule as a first actual matching confidence, and to determine the actual matching confidence between the business object ontology to be tested and the element ontology of the object to be tested and the tested business object ontology and the element ontology of the tested object corresponding to each field basic information in the pre-created business domain database using a preset matching confidence rule as a second actual matching confidence; The collaborative form field determination module is used to determine the collaborative form corresponding to the form to be tested according to the first actual matching confidence and the preset matching confidence threshold, and to determine the collaborative field corresponding to the field to be tested according to the second actual matching confidence and the preset matching confidence threshold.

9. An electronic device, characterized in that: The electronic device comprises: at least one processor; and a memory communicatively connected to the at least one processor; wherein, The memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can perform the data matching method according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a processor to implement the data matching method according to any one of claims 1 to 7 when executed.

Images