CN116050359A

CN116050359A - Policy escrow entry method, system, terminal equipment and storage medium

Info

Publication number: CN116050359A
Application number: CN202310074890.XA
Authority: CN
Inventors: 李鉴庭
Original assignee: Shenzhen Paikrypton Technology Co ltd
Current assignee: Shenzhen Paikrypton Technology Co ltd
Priority date: 2023-01-12
Filing date: 2023-01-12
Publication date: 2023-05-02

Abstract

The invention is applicable to the technical field of computers, and provides a policy escrow entry method, a policy escrow entry system, terminal equipment and a storage medium, wherein the policy escrow entry method comprises the following steps: configuring structural information and corresponding first text; generating a first semantic feature vector according to the first text and establishing a feature vector index; performing word recognition on the policy picture to generate a second text; obtaining a plurality of text relation pairs according to the second text, and obtaining a second semantic feature vector and a plurality of corresponding similar text feature vectors according to the text relation pairs; obtaining a plurality of similarity values of the text relation pairs; and setting the Value text part in the text relation pair with the similarity Value larger than the preset similarity threshold Value in the plurality of similarity values as the text Value corresponding to the structured information. The invention not only supports the direct entry of the insurance policy in the formats of the pictures and PDF documents, but also can enter the insurance policy after the second comparison and confirmation by the insurance broker, thereby improving the entry efficiency of the insurance policy and reducing the error rate of manual entry.

Description

Policy escrow entry method, system, terminal equipment and storage medium

Technical Field

The invention belongs to the technical field of computers, and particularly relates to a policy escrow entry method, a policy escrow entry system, terminal equipment and a storage medium.

Background

The development of insurance business makes insurance customers more and more, and insurance customers generally generate various insurance policies when insurance is applied by different insurance companies, such as: electronic insurance policies, paper insurance policies, handwritten insurance policies and the like, and because data among insurance companies are not shared, insurance clients cannot see all insurance policy data in the application of the same insurance company, and therefore, when insurance brokers uniformly manage insurance policies of the insurance clients, insurance brokers need to integrate and host data of various insurance policies provided by the insurance clients. Insurance brokers typically host insurance policies for insurance customers by way of manual entry, but such manual entry methods are inefficient and prone to error.

Disclosure of Invention

In view of the above, the embodiment of the invention provides a policy escrow entry method, a policy escrow entry system, terminal equipment and a storage medium, which are used for solving the problems of low efficiency and easy error when an insurance broker manually enters a policy in the prior art.

A first aspect of an embodiment of the present invention provides a policy escrow entry method, applied to a policy escrow entry system, including:

configuring structural information and corresponding first text;

generating a first semantic feature vector according to the first text and establishing a feature vector index;

performing word recognition on the policy picture to generate a second text;

obtaining a plurality of text relation pairs according to the second text, and obtaining a second semantic feature vector and a plurality of corresponding similar text feature vectors according to the text relation pairs;

obtaining a plurality of similarity values of the text relation pairs;

and setting the Value text part in the text relation pair with the similarity Value larger than the preset similarity threshold Value as the text Value corresponding to the structured information.

A second aspect of an embodiment of the present invention provides a policy escrow entry system, including:

the information configuration module is used for configuring the structured information and the corresponding first text;

the index generation module is used for generating a first semantic feature vector according to the first text and establishing a feature vector index;

the picture identification module is used for carrying out character identification on the policy picture and generating a second text;

the text extraction module is used for obtaining a plurality of text relation pairs according to the second text, and obtaining a second semantic feature vector and a plurality of corresponding similar text feature vectors according to the text relation pairs;

the text similarity calculation module is used for obtaining a plurality of similarity values of the text relation pairs;

and the text Value setting module is used for setting the Value text part in the text relation pair with the similarity Value larger than the preset similarity threshold Value in the plurality of similarity values as the text Value corresponding to the structural information.

A third aspect of the embodiment of the present invention provides a terminal device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, where the processor implements the policy escrow entry method according to the first aspect of the embodiment of the present invention when the processor executes the computer program.

A fourth aspect of the embodiments of the present invention provides a computer readable storage medium storing a computer program which, when executed by a processor, implements a policy escrow entry method according to the first aspect of the embodiments of the present invention.

The policy escrow input method provided by the first aspect of the embodiment of the invention configures the structured information and the corresponding first text; generating a first semantic feature vector according to the first text and establishing a feature vector index; performing word recognition on the policy picture to generate a second text; obtaining a plurality of text relation pairs according to the second text, and obtaining a second semantic feature vector and a plurality of corresponding similar text feature vectors according to the text relation pairs; obtaining a plurality of similarity values of the text relation pairs; and setting the Value text part in the text relation pair with the similarity Value larger than the preset similarity threshold Value as the text Value corresponding to the structured information. The method not only supports the direct entry of the insurance policy in the formats of the pictures and PDF documents, but also can enter the insurance policy after the second comparison and confirmation by the insurance broker, thereby not only improving the entry efficiency of the insurance policy, but also reducing the error rate of manual entry.

It will be appreciated that the advantages of the second to fourth aspects may be found in the relevant description of the first aspect and are not repeated here.

Drawings

In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings that are needed in the embodiments or the description of the prior art will be briefly introduced below, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and that other drawings can be obtained according to these drawings without inventive effort for a person skilled in the art.

Fig. 1 is a schematic diagram of a first flow of a policy escrow entry method according to an embodiment of the present invention;

fig. 2 is a schematic diagram of a second flow of a policy hosting and entering method according to an embodiment of the present invention;

fig. 3 is a schematic diagram of a third flow of a policy escrow entry method according to an embodiment of the present invention;

fig. 4 is a schematic diagram of a policy escrow entry system provided by an embodiment of the present invention;

fig. 5 is a schematic diagram of a terminal device according to an embodiment of the present invention.

Detailed Description

In the following description, for purposes of explanation and not limitation, specific details are set forth such as the particular system architecture, techniques, etc., in order to provide a thorough understanding of the embodiments of the present invention. It will be apparent, however, to one skilled in the art that the present invention may be practiced in other embodiments that depart from these specific details. In other instances, detailed descriptions of well-known systems, devices, circuits, and methods are omitted so as not to obscure the description of the present invention with unnecessary detail.

It should be understood that the terms "comprises" and/or "comprising," when used in this specification and the appended claims, specify the presence of stated features, integers, steps, operations, elements, and/or components, but do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and/or groups thereof.

Furthermore, the terms "first," "second," "third," and the like in the description of the present specification and in the appended claims, are used for distinguishing between descriptions and not necessarily for indicating or implying a relative importance.

Reference in the specification to "one embodiment" or "some embodiments" or the like means that a particular feature, structure, or characteristic described in connection with the embodiment is included in one or more embodiments of the invention. Thus, appearances of the phrases "in one embodiment," "in some embodiments," "in other embodiments," and the like in the specification are not necessarily all referring to the same embodiment, but mean "one or more but not all embodiments" unless expressly specified otherwise. The terms "comprising," "including," "having," and variations thereof mean "including but not limited to," unless expressly specified otherwise.

The insurance plays a great role in guaranteeing social stability, promoting economic development and external trade, not only can guarantee normal operation of social reproduction and actively promote circulation and consumption of commodities, but also is beneficial to balance of financial and credit income, increase foreign exchange income and enhance national payment capacity. Insurance helps individuals or institutions reduce economic hazards by compensating for economic losses of insured persons, enhances their risk management awareness, and facilitates the balance of individual and household consumption. Insurance includes property insurance, life insurance, liability insurance, credit guarantee insurance, etc., and insurance clients generally generate various different insurance policies when performing insurance application by different insurance companies, such as: electronic insurance policy, paper insurance policy, handwriting insurance policy and the like, and because data among insurance companies are not shared, insurance clients cannot see all insurance policy data in the application of the same insurance company, so that the insurance clients have a wish to give various insurance policies of themselves to an insurance broker for hosting, services such as insurance policy payment, expiration reminding and the like are obtained, and the insurance broker can better analyze the insurance requirements of the clients by knowing the insurance policy components of the insurance clients, and further recommend the insurance clients to purchase proper insurance. When the insurance broker manages the insurance policy of the insurance client in a unified way, the insurance broker first integrates data of various insurance policies provided by the insurance client so as to better serve the insurance client.

The policy escrow input method provided by the embodiment of the invention configures the structured information and the corresponding first text; generating a first semantic feature vector according to the first text and establishing a feature vector index; performing word recognition on the policy picture to generate a second text; obtaining a plurality of text relation pairs according to the second text, and obtaining a second semantic feature vector and corresponding similar text feature vectors according to the text relation pairs, and obtaining a plurality of similarity values of the text relation pairs; and setting the Value text part in the text relation pair with the similarity Value larger than the preset similarity threshold Value as the text Value corresponding to the structured information. The policy escrow input method provided by the invention not only supports the direct input of the policy in the formats of pictures and PDF documents, but also can input the policy after the secondary comparison and confirmation by the insurance broker, thereby not only improving the input efficiency of the policy, but also reducing the error rate of manual input.

As shown in fig. 1, the policy escrow entry method provided by the embodiment of the invention includes steps S100 to S105 as follows:

step S100, configuring the structured information and the corresponding first text, and proceeding to step S101.

In the application, the configuration structural information and the corresponding first text are as follows: { "guarantor number": [ "policy number", "contract number" ], the "insurance name": [ "insurance name", "risk name" ], an "insurance amount": [ "insurance amount", "premium" ], the "applicant name": [ "applicant name", "applicant" ], "applicant certificate type": [ "applicant certificate type" ], "applicant certificate number": [ "insuring person number" ], "insured name": [ "insured name" ], "insured certificate number": [ "insured life certificate number" ], beneficiary ": [ "beneficiary" ], an "insurance onset": [ "insurance onset" ], an "insurance expiration": [ "insurance expiration" ], a "payment method": [ "Payment method" ], a "Payment period": [ "Payment period" ] }.

In the application, the structured information is "insurance number", "insurance name", "insurance amount", "insurance applicant name", "insurance applicant certificate type", "insurance applicant certificate number", "insured person name", "insured person certificate number", "beneficiary", "insurance onset", "insurance expiration date", "payment mode", "payment period", etc. The first text corresponding to the structured information is [ "insurance number", "contract number" ], [ "insurance name", "risk name" ], [ "insurance amount", "insurance fee" ], [ "insurance applicant name", "insurance applicant" ], [ "insurance applicant certificate type" ], [ "insurance applicant license number" ], [ "protected person name" ], [ "insured certificate number" ], [ "beneficiary" ], [ "insurance onset" ], [ "insurance expiration" ], [ "payment mode" ], [ "payment period" ], and the like.

Step S101, according to the first text, generating a first semantic feature vector, establishing a feature vector index, and entering step S102.

In the application, a first text [ "insurance policy number", "contract number" ], [ (insurance name "," risk name "], [" insurance amount "," insurance fee "], [" insurance applicant name "," insurance applicant "], [" insurance applicant certificate type "], [" insurance applicant certificate number "], [" insured person name "], [" insured person certificate number "], [" beneficiary "], [" insurance onset "], [" insurance expiration date "], [" payment manner "], [" payment period "], and the like are converted into a first semantic feature vector through a data tower in the semantic matching double-tower model.

As shown in fig. 2, in one embodiment, step S101 includes the following steps S1011 to S1012:

step S1011, converting the first text into a first semantic feature vector through a semantic matching double-tower model, and entering step S1012.

In the application, after the structural information and the corresponding first text are configured, the first text is converted into a first semantic feature vector through semantically matching data towers in the double-tower model.

Step S1012, establishing a feature vector index through HNSW algorithm.

In the application, after the first text is converted into the first semantic feature vector, a feature vector index is established through an HNSW algorithm. The HNSW algorithm is a further optimized version on top of the NSW algorithm. The core is to introduce a skip list to realize layering on the basis of NSW algorithm, so as to further optimize the retrieval speed of the target vector. The NSW algorithm is an approximate optimization based on Delaunay triangulation, which references the form of triangulation, and the search results returned by the NSW algorithm are in fact only approximate results.

Step S102, performing word recognition on the policy pictures to generate a second text, and entering step S103.

In application, various forms to be input into the form hosting system are converted into pictures, after all forms are converted into pictures, optical character (Optical Character Recognition, OCR) recognition is carried out on the pictures, and finally, the pictures can be converted into second texts.

In application, the OCR technology is adopted, the data elements of the picture policy can be automatically extracted, the workload of the insurance broker for manually inputting the policy data can be reduced, the repeated labor is reduced, the error rate is reduced, and especially after the insurance broker is combined with the policy hosting and inputting system, the original work of manually inputting the policy data by the insurance broker can be replaced, the data on the policy is automatically identified by a computer, and the inputting efficiency of the policy data is greatly improved.

In application, rejection rate, false recognition rate, recognition speed, user interface friendliness, product stability, usability, feasibility and the like are generally used as evaluation bases of OCR technology.

In one embodiment, step S102 includes:

converting the non-picture format policy document into a picture;

and carrying out character recognition on the picture to generate the second text.

In the application, the insurance policy in the picture format provided by the insurance client can be subjected to OCR text recognition directly to generate a second text; for non-picture format policy, such as PDF document format policy, provided by an insurance client, the non-picture format policy document is first converted into a picture format, and then a second text is generated by OCR recognition technology.

Step S103, obtaining a plurality of text relation pairs according to the second text, obtaining a second semantic feature vector and a plurality of corresponding similar text feature vectors according to the text relation pairs, and entering step S104.

In the application, after a plurality of Key-Value text relation pairs are obtained according to the second text, converting a Key text part in the Key-Value relation pair into a second semantic feature vector through a query tower in a semantic matching double-tower model; a plurality of similar text feature vectors that are similar to the second semantic feature vector are then obtained by the feature vector index query.

As shown in fig. 3, in one embodiment, step S103 includes the following steps S1031 to S1034:

step S1031 is to extract a relation pair from the second text data, and the process advances to step S1032.

In the application, after the second text is obtained through OCR technology, the second text data is extracted by Key-Value relation pairs.

Step S1032, generating a Key-Value relation pair text list, and entering step S1033.

In the application, after the second text data is extracted by the Key-Value relation pair, a text list of the Key-Value relation pair is generated according to the extracted Key-Value relation pair.

Step S1033, converting the Key text portion in the Key-Value relation pair text list into a second semantic feature vector, and entering step S1034.

In the application, the Key text part in the Key-Value relation pair text list is converted into a second semantic feature vector through a query tower in the semantic matching double-tower model.

Step S1034, querying and obtaining a plurality of similar text feature vectors corresponding to the second semantic feature vector through the feature vector index.

In an application, after the second semantic feature vector is obtained, a plurality of similar text feature vectors corresponding to the second semantic feature vector may be obtained through feature vector indexing.

In the application, extracting a Key-Value relation pair from the second text to generate a Key-Value relation pair text list: { "contract number": "SH00001", "applicant": "Zhang San", "insurance certificate type": "identity card", "insurance certificate number": "110000000000000000", "insurance beneficiary": "insured life", "insurance start time": "month 1 of 2022", "insurance end time": "day 1 of 12, 2023". Key text parts ("contract number", "applicant", "insurance certificate type", "insurance number", "insurance beneficiary", "insurance start time", "insurance end time", etc.) in the Key-Value relation pair are converted into second semantic feature vectors through query towers in the semantic matching double-tower model, and then TopN (N=5) similar text feature vectors similar to the second semantic feature vectors are obtained through feature vector index query.

Step S104, obtaining a plurality of similarity values of the text relation pairs, and proceeding to step S105.

In the application, the plurality of similarity values of the second semantic feature vector and the TopN (n=5) similar text feature vector can be obtained through semantic similarity model calculation, and specifically, the plurality of similarity values of the second semantic feature vector and the TopN (n=5) similar text feature vector can be obtained through a plurality of semantic similarity calculation methods such as an inner product method, a cosine method, a Dice coefficient method, a Jaccard coefficient method and the like.

In one embodiment, step S104 includes:

respectively obtaining a plurality of similarity values of each text relation pair;

the plurality of similarity values of the plurality of text relationship pairs are ordered in ascending or descending order.

In the application, after the similarity between the second semantic feature vector and the TopN (n=5) similar text feature vector is obtained through the semantic similarity model, the similarity of a plurality of similar texts can be arranged in a descending order, and the similarity of the first similar text is the highest; the similarity of the plurality of similar texts can be arranged in an ascending order, and the similarity of the last similar text is highest.

Step S105, setting a Value text portion in the text relation pair with the similarity Value larger than a preset similarity threshold Value as a text Value corresponding to the structured information.

In the application, after the similarity of the similar texts is arranged in a descending order, if the similarity of the first similar texts is greater than a preset similarity threshold, setting a Value text part in a text relation pair as a text Value corresponding to the structural information; after the similarity of the similar texts is arranged in an ascending order, if the similarity of the similar texts arranged at the last position is larger than a preset similarity threshold Value, setting the Value text part in the text relation pair as a text Value corresponding to the structural information.

In one embodiment, the preset similarity threshold is 0.6, and step S105 includes:

and setting a Value text part in the text relation pair with the target similarity Value larger than a preset similarity threshold as a text Value corresponding to the structured information, wherein the target similarity Value is the maximum similarity Value in the similarity values of the text relation pairs.

In the application, the preset similarity threshold Value is 0.6, if the target similarity Value in the similarity values is greater than the preset similarity threshold Value 0.6, the Value text part in the text relation pair of the target similarity is set as the text Value corresponding to the structured information, and the structured data is finally obtained as follows: { "guarantor number": "SH00001", "applicant name": "Zhang san", "applicant certificate type": "identity card", "insurance person license number": "110000000000000000", "beneficiary": "insured life", "insured life": "12 th year, 1 day 2022", "insurance expiration date": "day 1 of 12, 2023".

In one embodiment, step S105 is followed by:

and outputting a text value corresponding to the structured information.

In the application, when the finally obtained structured data is displayed, the second text is also displayed. The insurance broker performs a secondary validation of the second text and the presented structured data, and if incorrect structured data is found, reassigns the structured data by way of manual input.

As shown in fig. 4, the policy escrow entry system 200 provided by the embodiment of the invention includes an information configuration module 201, an index generation module 202, a picture identification module 203, a text extraction module 204, a text similarity calculation module 205, and a text value setting module 206.

An information configuration module 201, configured to configure the structured information and the corresponding first text;

an index generation module 202, configured to generate a first semantic feature vector according to the first text and establish a feature vector index;

the picture recognition module 203 is configured to perform text recognition on the policy picture to generate a second text;

the text extraction module 204 is configured to obtain a plurality of text relation pairs according to the second text, and obtain a second semantic feature vector and a plurality of corresponding similar text feature vectors according to the text relation pairs;

a text similarity calculation module 205, configured to obtain a plurality of similarity values of the plurality of text relationship pairs;

the text Value setting module 206 is configured to set a Value text portion in the text relationship pair that is greater than a preset similarity threshold Value in the plurality of similarity values as a text Value corresponding to the structured information.

In an application, the index generation module 202 is specifically configured to: converting the first text and generating a first semantic feature vector through a semantic matching double-tower model; and establishing a feature vector index through an HNSW algorithm.

In application, the semantic matching double-tower model is used, so that the similar text can be more accurately matched; the HNSW algorithm is used for establishing the feature vector index, so that the time complexity can be reduced. In computer science, the temporal complexity of an algorithm is a function that quantitatively describes the runtime of the algorithm.

In application, the picture recognition module 203 is specifically configured to: converting the non-picture format policy document into a picture; and carrying out character recognition on the picture to generate a second text.

In application, the text extraction module 204 is specifically configured to:

extracting a relation pair from the second text data;

generating a Key-Value relation pair text list;

converting the Key text part in the Key-Value relation pair text list into a second semantic feature vector;

and querying and obtaining a plurality of similar text feature vectors corresponding to the second semantic feature vector through the feature vector index.

In an application, the text similarity calculation module 205 is specifically configured to:

the plurality of similarity values of the plurality of text relationship pairs are sorted in ascending or descending order.

In an application, the text value setting module 206 is specifically configured to: and setting the Value text part in the text relation pair with the target similarity Value larger than the preset similarity threshold as the text Value corresponding to the structured information, wherein the target similarity Value is the maximum similarity Value in the similarity values of the text relation pairs.

As shown in fig. 5, the embodiment of the present application further provides a terminal device 3, including: at least one processor 31 (only one processor is shown in fig. 5), a memory 32 and a computer program 33 stored in the memory 32 and executable on the at least one processor 31, the steps of the various method embodiments described above being implemented when the computer program 33 is executed by the processor 31.

In an application, the terminal device may include, but is not limited to, a processor, a memory. It will be appreciated by those skilled in the art that fig. 5 is merely an example of a terminal device and is not intended to be limiting, and that a terminal device may include more or less components than those illustrated, or may combine some components, or may include different components, for example, input/output devices, network access devices, etc., which may include displays, keyboards, mice, etc., and network access devices, which may include wired or wireless communication modules.

In application, the display may be a display screen, which may be a thin film transistor liquid crystal display (Thin Film Transistor Liquid Crystal Display, TFT-LCD), a liquid crystal display (Liquid Crystal Display, LCD), an Organic Light-Emitting Diode (OLED), a quantum dot Light Emitting Diode (Quantum Dot Light Emitting Diodes, QLED) display screen, a seven-segment or eight-segment nixie tube, or the like.

In an application, the communication module may comprise a wireless communication module or a wired communication module. The wireless communication module may include at least one of a wireless fidelity (WiFi) unit, a Bluetooth (Bluetooth) unit, a Zigbee (Zigbee) unit, a mobile communication network unit, a global navigation satellite system (Global Navigation Satellite System, GNSS) unit, a frequency modulation (Frequency Modulation, FM) unit, a near field wireless communication technology (Near Field Communication, NFC) unit, etc., and the wired communication module may include at least one of an Ethernet (Ethernet) unit, an asymmetric digital subscriber line (Asymmetric Digital Subscriber Line, ADSL) unit, a network fiber to the home (Fiber To The Home, FTTH) unit, etc.

In application, the processor may be a central processing unit (Central Processing Unit, CPU), which may also be other general purpose processors, digital signal processors (Digital Signal Processor, DSP), application specific integrated circuits (Application Specific Integrated Circuit, ASIC), field-programmable gate arrays (Field-Programmable Gate Array, FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, or the like. A general purpose processor may be a microprocessor or the processor may be any conventional processor or the like.

In applications, the memory may in some embodiments be an internal storage unit of the terminal device, such as a hard disk or a memory of the terminal device. The memory may in other embodiments also be an external storage device of the terminal device, for example a plug-in hard disk, a Smart Media Card (SMC), a Secure Digital (SD) Card, a Flash Card (Flash Card) or the like, which are provided on the terminal device. Further, the memory may also include both an internal storage unit of the terminal device and an external storage device. The memory is used to store an operating system, application programs, boot Loader (Boot Loader), data, and other programs, etc., such as program code for a computer program, etc. The memory may also be used to temporarily store data that has been output or is to be output.

It should be noted that, because the content of information interaction and execution process between the modules/units is based on the same concept as the embodiment of the present invention, specific functions and technical effects thereof may be referred to in the embodiment of the present invention, and will not be described herein.

It will be apparent to those skilled in the art that, for convenience and brevity of description, only the above-described division of the functional units and modules is illustrated, and in practical application, the above-described functional distribution may be performed by different functional units and modules according to needs, i.e. the internal structure of the apparatus is divided into different functional units or modules to perform all or part of the above-described functions. The functional units and modules in the embodiment may be integrated in one processing unit, or each unit may exist alone physically, or two or more units may be integrated in one unit, where the integrated units may be implemented in a form of hardware or a form of a software functional unit. In addition, specific names of the functional units and modules are only for convenience of distinguishing from each other, and are not used for limiting the protection scope of the present application. The specific working process of the units and modules in the above system may refer to the corresponding process in the foregoing method embodiment, which is not described herein again.

The embodiments of the present application also provide a computer readable storage medium storing a computer program, where the computer program can implement the steps in the above-mentioned method embodiments when executed by a processor.

The embodiments of the present application provide a computer program product which, when run on a terminal device, causes the terminal device to perform the steps of the method embodiments described above.

The integrated units, if implemented in the form of software functional units and sold or used as stand-alone products, may be stored in a computer readable storage medium. Based on such understanding, the present application implements all or part of the flow of the method of the above embodiments, and may be implemented by a computer program to instruct related hardware, where the computer program may be stored in a computer readable storage medium, where the computer program may implement the steps of each of the method embodiments described above when executed by a processor. Wherein the computer program comprises computer program code, which may be in the form of source code, object code, executable files or in some intermediate form, etc. The computer readable medium may include at least: any entity or device capable of carrying computer program code to an apparatus/terminal device, recording medium, computer Memory, read-Only Memory (ROM), random access Memory (RAM, random Access Memory), electrical carrier signals, telecommunications signals, and software distribution media. Such as a U-disk, removable hard disk, magnetic or optical disk, etc. In some jurisdictions, computer readable media may not be electrical carrier signals and telecommunications signals in accordance with legislation and patent practice.

In the foregoing embodiments, the descriptions of the embodiments are emphasized, and in part, not described or illustrated in any particular embodiment, reference is made to the related descriptions of other embodiments.

Those of ordinary skill in the art will appreciate that the various illustrative elements and algorithm steps described in connection with the embodiments disclosed herein may be implemented as electronic hardware, or combinations of computer software and electronic hardware. Whether such functionality is implemented as hardware or software depends upon the particular application and design constraints imposed on the solution. Skilled artisans may implement the described functionality in varying ways for each particular application, but such implementation decisions should not be interpreted as causing a departure from the scope of the present application.

In the embodiments provided in the present application, it should be understood that the disclosed apparatus and method may be implemented in other manners. For example, the apparatus embodiments described above are merely illustrative, e.g., the division of modules or units is merely a logical functional division, and there may be additional divisions when actually implemented, e.g., multiple units or components may be combined or integrated into another system, or some features may be omitted or not performed. Alternatively, the coupling or direct coupling or communication connection shown or discussed may be an indirect coupling or communication connection via interfaces, devices or units, which may be in electrical, mechanical or other forms.

The units described as separate units may or may not be physically separate, and units shown as units may or may not be physical units, may be located in one place, or may be distributed over a plurality of network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

The above embodiments are only for illustrating the technical solution of the present application, and are not limiting thereof; although the present application has been described in detail with reference to the foregoing embodiments, it should be understood by those of ordinary skill in the art that: the technical scheme described in the foregoing embodiments can be modified or some technical features thereof can be replaced by equivalents; such modifications and substitutions do not depart from the spirit and scope of the technical solutions of the embodiments of the present application, and are intended to be included in the scope of the present application.

Claims

1. A policy escrow entry method, characterized in that it is applied to a policy escrow entry system, the method comprising:

configuring structural information and corresponding first text;

performing word recognition on the policy picture to generate a second text;

obtaining a plurality of similarity values of the text relation pairs;

2. The policy escrow entry method of claim 1, wherein generating a first semantic feature vector and establishing a feature vector index from the first text comprises:

converting the first text into a first semantic feature vector through a semantic matching double-tower model;

and establishing a feature vector index through an HNSW algorithm.

3. The policy escrow entry method of claim 1, wherein the text recognition of the policy picture to generate the second text comprises:

converting the non-picture format policy document into a picture;

and carrying out character recognition on the picture to generate a second text.

4. The policy escrow entry method of claim 1, wherein obtaining a plurality of text-relation pairs from the second text and obtaining a second semantic feature vector and a corresponding plurality of similar text feature vectors from the text-relation pairs comprises:

extracting a relation pair from the second text data;

generating a Key-Value relation pair text list;

and inquiring and obtaining a plurality of similar text feature vectors corresponding to the second semantic feature vector through the feature vector index.

5. The policy escrow entry method of claim 1, wherein obtaining a plurality of similarity values for the plurality of text-relation pairs comprises:

6. The policy escrow entry method of claim 1, wherein the preset similarity threshold is 0.6;

the setting the Value text part in the text relation pair with the similarity Value larger than the preset similarity threshold Value as the text Value corresponding to the structured information includes:

7. The policy hosting and entering method as claimed in claim 1 or 6, wherein after setting a Value text portion in a text relationship pair greater than a preset similarity threshold Value in a plurality of similarity values as a text Value corresponding to the structured information, the method includes:

and outputting a text value corresponding to the structured information.

8. A policy escrow entry system, comprising:

9. A terminal device comprising a memory, a processor and a computer program stored in the memory and executable on the processor, wherein the processor implements the policy escrow entry method of any of claims 1 to 7 when the computer program is executed.

10. A computer readable storage medium storing a computer program, wherein the computer program when executed by a processor implements the policy escrow entry method of any of claims 1 to 7.