Data input method and device based on named entity recognition, equipment and medium

Abstract

The invention relates to the field of text processing, and discloses a data entry method and device based on named entity recognition, equipment and a medium, and the method comprises the steps: receiving a to-be-entered data file, carrying out the data extraction of the data file, and obtaining a plurality of pieces of to-be-entered customer data; inputting the plurality of pieces of to-be-input customer data into a preset named entity recognition model, and recognizing to obtain named entity information in each piece of customer data; performing entity type identification on the named entity information in each piece of customer data according to a preset identification switching rule through a state machine mode to obtain an entity type in the named entity information; and according to the entity type in the named entity information, inputting the corresponding entity text into the corresponding field position in the standard table to complete client data input. By performing data extraction and named entity identification on the data file, the entities in the customer data are accurately identified and classified and are automatically input into the standard table, so that the manual input cost is reduced, and the data input accuracy and efficiency are improved.

Description

Technical Field

[0001] The present invention relates to the fields of text processing technology and financial technology, and in particular to a data entry method, device, equipment and medium based on named entity recognition. Background Art

[0002] At present, the business scenario of insurance companies entering group insurance information in batches mainly relies on manual input. The specific process is that the customer first communicates with the agent about the insurance-related information. After the customer confirms, the insured person's information is sorted into a table and sent to the agent. The agent manually enters the form content provided by the customer into the standard form and uploads it to the system.

[0003] However, the number of insured group customers is large, generally between 10 and 300 people, which means that the time required for manual data entry by agents is long, which is not only inefficient but also prone to errors. In addition, the forms provided by customers involve a variety of information and are in various formats, making it difficult to automatically identify and enter the form content through traditional Java technology. Therefore, the current batch entry efficiency and accuracy of insurance information are low, which in turn affects business processing efficiency. Summary of the invention

[0004] In view of the above-mentioned deficiencies in the prior art, the purpose of the present invention is to provide a data entry method, device, equipment and medium based on named entity recognition that can be applied to financial technology or other related fields. Its main purpose is to improve the accuracy and efficiency of batch data entry and meet the efficient entry requirements in the business scenario of batch entry of group insurance information.

[0005] The technical solution of the present invention is as follows:

[0006] A first aspect of the present invention provides a data entry method based on named entity recognition, comprising:

[0007] Receiving a data file to be entered, extracting data from the data file, and obtaining a plurality of customer data to be entered;

[0008] Inputting the plurality of customer data to be entered into a preset named entity recognition model to identify and obtain named entity information in each customer data;

[0009] By using the state machine mode, the entity type of the named entity information in each piece of customer data is identified according to the preset identification switching rule to obtain the entity type in the named entity information;

[0010] According to the entity type in the named entity information, the corresponding entity text is entered into the corresponding field position in the standard form to complete the entry of the customer data.

[0011] A second aspect of the present invention provides a data entry device based on named entity recognition, comprising:

[0012] A receiving and extracting module, used for receiving a data file to be entered, extracting data from the data file, and obtaining a plurality of customer data to be entered;

[0013] A named entity recognition module, used to input the plurality of customer data to be entered into a preset named entity recognition model, and identify the named entity information in each customer data;

[0014] An entity type identification module, used to identify the entity type of the named entity information in each customer data according to a preset identification switching rule through a state machine mode, and obtain the entity type in the named entity information;

[0015] The data entry module is used to enter the corresponding entity text into the corresponding field position in the standard form according to the entity type in the named entity information, so as to complete the entry of the customer data.

[0016] A third aspect of the present invention provides a computer device, comprising at least one processor; and

[0017] a memory communicatively connected to the at least one processor; wherein,

[0018] The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the above-mentioned data entry method based on named entity recognition.

[0019] A fourth aspect of the present invention provides a non-volatile computer-readable storage medium, which stores computer-executable instructions. When the computer-executable instructions are executed by one or more processors, the one or more processors can execute the above-mentioned data entry method based on named entity recognition.

[0020] Beneficial effects: The present invention discloses a data entry method, device, equipment and medium based on named entity recognition. Compared with the prior art, the embodiment of the present invention receives a data file to be entered, extracts data from the data file, and obtains multiple customer data to be entered; inputs the multiple customer data to be entered into a preset named entity recognition model to identify the named entity information in each customer data; through a state machine mode, the named entity information in each customer data is identified according to the preset recognition switching rules to obtain the entity type in the named entity information; according to the entity type in the named entity information, the corresponding entity text is entered into the corresponding field position in the standard form to complete the entry of the customer data. By performing automated data extraction and named entity recognition on irregular data files, the entity information in the customer data is accurately identified and classified and automatically entered into the standard form. Through accurate and standardized data entry, not only the manual entry cost is reduced, but also the accuracy and efficiency of data entry are improved, meeting the efficient entry requirements in the business scenario of batch entry of group insurance information. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] In order to more clearly illustrate the scheme in the present invention, a brief introduction is given below to the drawings required for use in the description of the embodiments of the present invention. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0022] Figure 1 A schematic diagram of an application environment of a data entry method based on named entity recognition provided by an embodiment of the present invention;

[0023] Figure 2 A flow chart of a data entry method based on named entity recognition provided by an embodiment of the present invention;

[0024] Figure 3 A flow chart of step S202 in the data entry method based on named entity recognition provided by an embodiment of the present invention;

[0025] Figure 4 A flow chart of step S204 in the data entry method based on named entity recognition provided by an embodiment of the present invention;

[0026] Figure 5 A schematic diagram of functional modules of a data entry device based on named entity recognition provided by an embodiment of the present invention;

[0027] Figure 6 A schematic diagram of the hardware structure of a computer device provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0028] In order to make the purpose, technical solution and effect of the present invention clearer and more specific, the present invention is further described in detail below. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention. The embodiments of the present invention are described below in conjunction with the accompanying drawings.

[0029] The data entry method based on named entity recognition provided by the embodiment of the present invention can be applied in the following aspects: Figure 1 In the application environment of FIG. 1 , it includes a first terminal device 101, a second terminal device 102, a third terminal device 103, a network 104 and a server 105. The network 104 is used to provide a medium for a communication link between the first terminal device 101, the second terminal device 102, the third terminal device 103 and the server 105. The network 104 may include various connection types, such as wired and / or wireless communication links, etc.

[0030] The user can use the first terminal device 101, the second terminal device 102, and the third terminal device 103 to interact with the server 105 through the network 104 to receive or send messages, etc. Various communication client applications can be installed on the first terminal device 101, the second terminal device 102, and the third terminal device 103, such as knowledge reading applications, web browser applications, search applications, instant messaging tools, email clients, and / or social platform software, etc. (only as examples).

[0031] The first terminal device 101, the second terminal device 102, and the third terminal device 103 may be various electronic devices having display screens and supporting web browsing, including but not limited to smart phones, tablet computers, laptop computers, desktop computers, and the like.

[0032] The server 105 may be a server that provides various services, such as a background server that provides support for the content browsed by the user using the first terminal device 101, the second terminal device 102, and the third terminal device 103 (for example only). The background server may analyze and process the received data such as the user request, and feed back the processing results (such as web pages, information, or data obtained or generated according to the user request, etc.) to the terminal device. The server 105 may be a cloud server, also known as a cloud computing server or a 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 ("Virtual Private Server", or "VPS" for short). The server 105 may also be a server for a distributed system, or a server combined with a blockchain.

[0033] It should be noted that the data entry method based on named entity recognition provided in the embodiment of the present application can generally be executed by the first terminal device 101, the second terminal device 102 or the third terminal device 103. Accordingly, the data entry device based on named entity recognition provided in the embodiment of the present invention can also be set in the first terminal device 101, the second terminal device 102 or the third terminal device 103. Alternatively, the data entry method based on named entity recognition provided in the embodiment of the present invention can also generally be executed by the server 105. Accordingly, the data entry device based on named entity recognition provided in the embodiment of the present invention can generally be set in the server 105.

[0034] It should be understood that the numbers of the above terminal devices, networks and servers are only illustrative. Any number of terminal devices, networks and servers may be provided according to implementation requirements.

[0035] like Figure 2 As shown, the data entry method based on named entity recognition provided by the embodiment of the present invention specifically includes the following steps:

[0036] S201, receiving a data file to be entered, extracting data from the data file, and obtaining a plurality of customer data to be entered.

[0037] In this embodiment, the data file to be entered can be a file in a non-uniform format such as an Excel spreadsheet, a CSV or a plain text file provided by the customer, which contains multiple pieces of customer data to be entered in a non-standard format. After receiving the data file to be entered, a structured data extraction process is performed on it, and multiple pieces of customer data to be entered are uniformly read from the non-standard format file. By automatically extracting data from the data file, it is convenient to reduce the workload of subsequent manual entry, and structuring the extracted data is conducive to improving the efficiency of subsequent data entry.

[0038] Specifically, in the business scenario of batch input of group insurance information, the historical process of agent processing can be investigated and counted in advance. About 99% of the scenarios use information carriers such as excel tables, csv or plain text, and each customer information is a separate line, including but not limited to the name of the insured, ID number, position, remarks, etc. Therefore, when extracting data from the data file, the data file can be cleaned first to remove blank lines and invalid data, and handle abnormal values ​​in the data such as incorrectly formatted ID numbers. Structured data extraction is performed on the cleaned data file, and each line of data is extracted as a customer data to be entered. Among them, for excel files, the read_excel function in the third-party library of python can be used to read the entire table and extract each line of data; for csv files, each line of data in the csv file can be read through the read_csv function in the third-party library of python; for plain text files, an empty list can be defined first, and then the file content can be read line by line according to the preset delimiter, and each line of text content read can be stored in the empty list.

[0039] S202: Input the plurality of customer data to be entered into a preset named entity recognition model to identify and obtain the named entity information in each customer data.

[0040] In this embodiment, the multiple extracted customer data to be entered are input into a preset named entity recognition model. Specifically, the named entity recognition model is used to identify entity information such as names of people, places, and organization names in text materials. The named entity model can be, for example, a rule-based model, a statistical learning model (such as a hidden Markov model, a conditional random field model), or a deep learning-based model (such as BiLSTM, BiLSTM-CRF, BERT, etc.). The named entity model is used to identify and mark the named entity information in each piece of customer data.

[0041] For example, the Excel file provided by the customer contains the group insurance information of all employees of a company. This table contains the following insurance data: employee ID, name, gender, date of birth, ID card, department, position, insurance type, etc. After extracting each employee insurance data, it is input into the trained named entity recognition model. According to the insurance entry requirements, the named entity information in each employee insurance data is identified, such as name, ID card number, insurance type, etc., so as to further filter out the entity information required for batch entry from the customer data, providing a reliable and accurate data foundation for efficient automatic entry.

[0042] S203 . Perform entity type identification on the named entity information in each piece of customer data according to a preset identification switching rule in a state machine mode to obtain the entity type in the named entity information.

[0043] In this embodiment, after identifying the named entity information in each customer data, based on the constructed state machine, the entity type of the named entity information in each customer data is identified through the state machine mode according to the preset identification switching rules. The specific identification switching rules include preset states (such as start, identify name, identify ID number, etc.) and transfer rules (such as first identify name, then identify ID number and the rules for each identification, etc.), so as to complete the identification of a single named entity information and obtain the entity type of each field content in each named entity information. The state machine mode ensures that each identified named entity information can be correctly classified, and the identification switching rules can also flexibly define and adjust the entity type identification logic, reduce errors in manual classification and entry, and the state machine mode is usually lighter than the deep learning model, which improves the processing speed of entity type identification.

[0044] S204: According to the entity type in the named entity information, the corresponding entity text is entered into the corresponding field position in the standard form to complete the entry of the customer data.

[0045] In this embodiment, after further entity type classification of the identified named entity information, the corresponding entity text is automatically filled into the corresponding field position in the preset standard form based on the entity text of each entity type. For example, when batch entering group insurance information, a standard insurance form is pre-set, which contains fields required for insurance such as the insured's name, ID number, insurance type, etc. According to the entity type, the entity text of the entity type "name" can be automatically filled into the "bidder name" field of the insurance form, and the entity text of the entity type "ID number" can be automatically filled into the insurance form. The entity text of the entity type "Insurance Type" is automatically filled into the "Insurance Type" field of the insurance form, thereby realizing batch entry of data after named entity recognition of customer data. Automated data entry reduces errors caused by human operations, ensures that all entered data follows the same format and standards, improves data consistency, and batch entry also improves the speed and efficiency of data entry. Therefore, in the business scenario of batch entry of group insurance information, efficient and accurate processing of group insurance information can be achieved, thereby improving the business processing capabilities and customer satisfaction of insurance companies.

[0046] In the above embodiment, the present invention discloses a data entry method based on named entity recognition, which receives a data file to be entered, extracts data from the data file, and obtains multiple customer data to be entered; inputs the multiple customer data to be entered into a preset named entity recognition model to identify the named entity information in each customer data; through a state machine mode, the named entity information in each customer data is identified according to the preset recognition switching rules to obtain the entity type in the named entity information; according to the entity type in the named entity information, the corresponding entity text is entered into the corresponding field position in the standard form to complete the entry of the customer data. By performing automated data extraction and named entity recognition on irregular data files, the entity information in the customer data is accurately identified and classified and automatically entered into the standard form. Through accurate and standardized data entry, not only the manual entry cost is reduced, but also the accuracy and efficiency of data entry are improved, meeting the efficient entry requirements in the business scenario of batch entry of group insurance information.

[0047] In one embodiment, before step S202, the method further includes:

[0048] Collecting corpus samples from a corpus in a specified business field, and annotating the corpus samples according to a preset annotation standard;

[0049] The corpus sample with entity annotations is input into a named entity recognition model constructed based on a pre-trained language model for training and fine-tuning to obtain a named entity recognition model adapted to the specified business field.

[0050] In this embodiment, in order to achieve accurate entity recognition of customer data in irregular document files, a named entity recognition model is trained and fine-tuned in advance according to the needs of the business field. The specific fine-tuning model is based on the pre-trained language model BERT based on Transformer technology, and is trained and fine-tuned with a small amount of data, so as to achieve the adaptation of the named entity recognition model to the specified business field.

[0051] First, collect corpus samples from the corpus of the specified business field, such as collecting text data from business databases, customer service records, insurance documents and other sources to ensure that the corpus covers all areas and scenarios of group insurance. Based on the defined entity categories such as name, ID number, insurance type, amount, date, etc., perform entity annotation on the corpus samples according to the BIOES annotation method, where B-entity word start, I-entity word inside, O-other words, E-entity end, S-a single word is an entity, so as to obtain training samples that are highly associated with the specified business field.

[0052] The corpus samples with entity annotations are input into the named entity recognition model built based on the pre-trained language model for training and fine-tuning. Specifically, the BERT-BiLSTM-CRF model is used as the named entity recognition model. By adjusting the parameters of the model, it can be better adapted to the specific terms and entity types in the field of group insurance. At the same time, the performance of the model can be tested on the validation set to ensure that it accurately recognizes various entities and obtains a named entity recognition model adapted to the specified business field. The fine-tuned named entity recognition model has improved recognition accuracy in specific fields, can better understand and identify specific entities in the field of group insurance, and thus accurately identify key information from the text, providing support for subsequent data processing and analysis.

[0053] In one embodiment, Figure 3 As shown, step S202 includes:

[0054] S301, inputting the plurality of customer data to be entered into a preset named entity recognition model, and identifying entity annotations in each customer data;

[0055] S302: Merge entity annotations in each piece of customer data according to a preset annotation standard to obtain corresponding named entity information.

[0056] In this embodiment, each line of customer data to be entered is extracted and input into the fine-tuned named entity recognition model. The named entity recognition model is trained using a corpus sample with a preset annotation standard during fine-tuning, and is specifically trained using the BIOES annotation method. Therefore, the named entity recognition model can identify entities in each piece of customer data, such as name, ID number, position, etc., and assign a BIOES label to each entity as an entity annotation. The BIOES annotation provides more fine-grained entity boundary information, which is helpful for subsequent processing. The entity annotations identified by the model are merged and merged according to the BIOES annotation standard. For example, if an entity is divided into multiple parts, a B-label is used to mark the beginning of the entity, an I-label is used to mark the inside of the entity, and an E-label is used to mark the end of the entity. The identified annotation results are merged according to the BIOES annotation standard, and the merged entity annotations are organized into a standard format, thereby obtaining more accurate named entity information for subsequent data entry.

[0057] In one embodiment, step S203 specifically includes:

[0058] Through the state machine mode, entity type identification is performed on the named entity information in each customer data according to the state switching sequence of identifying the name, identifying the ID number, and identifying other information, or according to the state switching sequence of identifying the ID number, identifying the name, and identifying other information, to obtain the entity type in the named entity information.

[0059] In this embodiment, when implementing entity type recognition, a state machine is pre-built, and multiple states such as "identifying name", "identifying ID number", "identifying other information" and their transition conditions are defined. A processing function is defined for each state to process the entity recognition task in the current state. In addition, the state transition logic is configured to implement the conversion logic between states, such as switching and identifying entity types through the state transition logic of start->identifying name->identifying ID number->identifying other information->end or start->identifying ID number->identifying name->identifying other information->end, thereby classifying the named entity information of each customer data and obtaining the entity type contained therein, so that it can be accurately filled into the correct field position when inputting, thereby improving the reliability of automatic data entry.

[0060] In one embodiment, Figure 4 As shown, step S204 includes:

[0061] S401, matching the entity type in the named entity information with the fields in the standard table;

[0062] S402: extracting entity text of the corresponding entity type according to the field matching result, and entering the entity text into the matching field position to complete the entry of the customer data.

[0063] In this embodiment, mapping rules between entity types and standard form fields are defined according to business needs. For example, the "name" entity is mapped to the "customer name" field, and the "document number" entity is mapped to the "identity card number" field. After named entity recognition and type recognition, the entity type is matched with the fields in the standard form according to the mapping rules based on the recognized entity type, and the corresponding field position to be filled is found for each entity type. The entity text of the corresponding entity type is extracted based on the field matching result, and the entity text is entered into the matching field position. For example, an entity type "name" in the named entity information is identified, which matches the "customer name" field in the standard form. At this time, the specific name text under the entity type "name" is extracted and filled into the position of the "customer name" field in the standard form, thereby completing the entry of the customer name data. The data entry of other entity types is similar, thereby realizing accurate and reliable automatic data entry on the basis of named entity recognition and classification of customer data, greatly improving the efficiency and accuracy of data entry.

[0064] In one embodiment, before step S203, the method further includes:

[0065] Obtain customer data to be corrected for named entity recognition failures;

[0066] Collecting statistics on entity type recognition results of all named entity information, extracting named entity information from the customer data to be corrected according to the statistical results, and obtaining named entity recognition results of the customer data to be corrected;

[0067] The standard form is corrected and supplemented according to the named entity recognition result of the customer data to be corrected.

[0068] In this embodiment, since the model may not be able to perform named entity recognition for some uncommon information, it is necessary to correct the named entity information in the case of recognition failure to ensure the integrity of the data entry. Specifically, after the named entity recognition model is processed, the data that has not been correctly recognized is collected. These data may fail to be recognized due to format problems, spelling errors, or vocabulary not covered by the model. The data are marked with corresponding to-be-corrected tags, so as to serve as customer data to be corrected for subsequent correction processing. During the correction process, it is implemented based on statistics. Statistics are performed on the entity types of other recognized named entity information. The patterns and regularities of various entities appearing in the data can be found through the statistical results. For example, if a certain entity type (such as a name) usually appears in a certain field position, then this position can be considered as a high-confidence index of the entity type, or the field position where a certain entity type (such as a name) has never appeared can also be determined by statistics. Then this field position in the customer data to be corrected is most likely not a name. Through statistical results, named entity information of the customer data to be revised is extracted to obtain the named entity recognition results of the customer data to be revised. Based on the corrected and extracted named entity recognition results, the standard form is supplemented and corrected to achieve complete and reliable data entry, thereby effectively processing customer data that failed to be recognized by the named entity recognition model and improving the overall data entry quality and reliability.

[0069] In one embodiment, the entity type recognition results of all named entity information are counted, and named entity information is extracted from the customer data to be corrected according to the statistical results to obtain the named entity information of the customer data to be corrected, including:

[0070] According to the entity type recognition results of all named entity information, count the field positions where each entity type appears in the customer data to obtain corresponding field index statistics results;

[0071] Establishing a confidence index for each entity type according to the field index statistics result;

[0072] Field information is extracted from a corresponding field position in the customer data to be corrected according to the confidence index of each entity type as a named entity recognition result of the customer data to be corrected.

[0073] In this embodiment, when performing entity recognition correction processing, the entity type recognition results of all identified named entity information are first counted, and the field position of each entity type in the customer data is specifically counted. For example, the field position of each entity type is recorded, such as "name" usually appears in field 1, "ID number" appears in field 2, etc., thereby generating corresponding field index statistics. For example, 80% of field 1 is "name" and 90% of field 2 is "ID number". By performing field index statistics on entity recognition results, the entity recognition model is optimized. Based on the field index statistics, an index is established for each entity type. For example, the confidence of each field position is calculated. For example, the confidence of "name" in field 1 is 0.8, and the confidence of "certificate number" is 0.1; the confidence of "certificate number" in field 2 is 0.9, and the confidence of "certificate name" is 0.1, etc. Based on the entity type with the highest confidence in each field position, a corresponding confidence index is established. These indexes represent the most likely position of the entity type, that is, the confidence index of the "name" entity type is field 1, and the confidence index of the "certificate number" entity type is field 2.

[0074] For customer data to be corrected where the entity cannot be identified, the corresponding field information is directly extracted from the field position corresponding to the customer data to be corrected through the confidence index of each entity type established based on the statistical results, as the named entity recognition result of the customer data to be corrected. For example, if the model fails to identify the name and ID number in a certain customer data, based on the confidence index obtained by the above statistics, the data of the first field is directly extracted from the customer data as the name, and the data of the second field is extracted as the ID number. The entity type is quickly located through the confidence index, thereby realizing automatic correction of unidentified data. This can effectively deal with the uncertainty in named entity recognition and improve the accuracy and completeness of the data.

[0075] It should be noted that there is not necessarily a certain order between the above-mentioned steps. A person skilled in the art can understand, based on the description of the embodiments of the present invention, that in different embodiments, the above-mentioned steps may have different execution orders, that is, they may be executed in parallel, may be executed interchangeably, and so on.

[0076] Further references Figure 5 , as a response to the above Figure 2 The present invention provides an embodiment of a data entry device based on named entity recognition, and the device embodiment is similar to Figure 2 Corresponding to the method embodiment shown, the device can be specifically applied to various electronic devices.

[0077] like Figure 5 As shown, the data entry device 50 based on named entity recognition described in this embodiment includes:

[0078] The receiving and extracting module 501 is used to receive a data file to be input, extract data from the data file, and obtain a plurality of customer data to be input;

[0079] A named entity recognition module 502, used to input the plurality of customer data to be entered into a preset named entity recognition model, and identify the named entity information in each customer data;

[0080] The entity type identification module 503 is used to identify the entity type of the named entity information in each customer data according to the preset identification switching rule through the state machine mode to obtain the entity type in the named entity information;

[0081] The data entry module 504 is used to enter the corresponding entity text into the corresponding field position in the standard form according to the entity type in the named entity information, so as to complete the entry of the customer data.

[0082] The module referred to in the present invention refers to a series of computer program instruction segments that can perform specific functions, which is more suitable for describing the data entry execution process based on named entity recognition than a program. For the specific implementation of each module, please refer to the corresponding method embodiment above, which will not be repeated here.

[0083] In one embodiment, the device 50 further includes:

[0084] A corpus collection module is used to collect corpus samples from a corpus in a specified business field and perform entity annotation on the corpus samples according to a preset annotation standard;

[0085] The model fine-tuning module is used to input the corpus samples with entity annotations into the named entity recognition model built based on the pre-trained language model for training and fine-tuning, so as to obtain a named entity recognition model adapted to the specified business field.

[0086] In one embodiment, the named entity recognition module 502 includes:

[0087] An identification unit, used to input the plurality of customer data to be entered into a preset named entity recognition model, and identify entity annotations in each customer data;

[0088] The merging unit is used to merge the entity annotations in each customer data according to a preset annotation standard to obtain corresponding named entity information.

[0089] In one embodiment, the entity type identification module 503 is specifically used to:

[0090] Through the state machine mode, entity type identification is performed on the named entity information in each customer data according to the state switching sequence of identifying the name, identifying the ID number, and identifying other information, or according to the state switching sequence of identifying the ID number, identifying the name, and identifying other information, to obtain the entity type in the named entity information.

[0091] In one embodiment, the data entry module 504 includes:

[0092] A field matching unit, used to match the entity type in the named entity information with the field in the standard table;

[0093] The data entry unit is used to extract the entity text of the corresponding entity type according to the field matching result, and enter the entity text into the matching field position to complete the entry of the customer data.

[0094] In one embodiment, the device 50 comprises:

[0095] An acquisition module, used to acquire customer data to be corrected for which named entity recognition failed;

[0096] A statistical correction module, used to collect statistics on entity type recognition results of all named entity information, extract named entity information from the customer data to be corrected according to the statistical results, and obtain named entity recognition results of the customer data to be corrected;

[0097] A supplementation module is used to complete and correct the information in the standard form according to the named entity recognition result of the customer data to be corrected.

[0098] In one embodiment, the statistics correction module includes:

[0099] A field index statistics unit, used to count the field positions of each entity type in the customer data according to the entity type recognition results of all named entity information, and obtain corresponding field index statistics results;

[0100] A confidence index building unit, used to build a confidence index for each entity type according to the field index statistical result;

[0101] The information extraction and correction unit is used to extract field information from the corresponding field position in the customer data to be corrected according to the confidence index of each entity type as the named entity recognition result of the customer data to be corrected.

[0102] In the above embodiment, the present invention discloses a data entry device based on named entity recognition, which receives a data file to be entered, extracts data from the data file, and obtains multiple customer data to be entered; inputs the multiple customer data to be entered into a preset named entity recognition model, and identifies the named entity information in each customer data; through a state machine mode, the named entity information in each customer data is identified according to the preset recognition switching rules to obtain the entity type in the named entity information; according to the entity type in the named entity information, the corresponding entity text is entered into the corresponding field position in the standard form to complete the entry of the customer data. By performing automated data extraction and named entity recognition on irregular data files, the entity information in the customer data is accurately identified and classified and automatically entered into the standard form. Through accurate and standardized data entry, not only the manual entry cost is reduced, but also the accuracy and efficiency of data entry are improved, meeting the efficient entry requirements in the business scenario of batch entry of group insurance information.

[0103] Another embodiment of the present invention provides a computer device, such as Figure 6 As shown, the computer device 60 includes:

[0104] One or more processors 601 and memory 602, Figure 6 A processor 601 is used as an example for description. The processor 601 and the memory 602 may be connected via a bus or other means. Figure 6 The example of connecting through bus is taken in the following.

[0105] Processor 601 is used to complete various control logics of computer device 60, and it can be a general-purpose processor, a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field programmable gate array (FPGA), a single-chip microcomputer, an ARM (Acorn RISC Machine) or other programmable logic device, a discrete gate or transistor logic, a discrete hardware component or any combination of these components. In addition, processor 601 can also be any traditional processor, microprocessor or state machine. Processor 601 can also be implemented as a combination of computing devices, for example, a combination of a DSP and a microprocessor, a plurality of microprocessors, one or more microprocessors combined with a DSP and / or any other such configuration.

[0106] The memory 602 is a non-volatile computer-readable storage medium that can be used to store non-volatile software programs, non-volatile computer executable programs and modules, such as program instructions corresponding to the data entry method based on named entity recognition in the embodiment of the present invention. The processor 601 executes various functional applications and data processing of the computer device 60 by running the non-volatile software programs, instructions and units stored in the memory 602, that is, implementing the data entry method based on named entity recognition in the above method embodiment.

[0107] The memory 602 may include a program storage area and a data storage area, wherein the program storage area may store an operating system, an application required by at least one function; the data storage area may store data created according to the use of the computer device 60, etc. In addition, the memory 602 may include a high-speed random access memory, and may also include a non-volatile memory, such as at least one disk storage device, a flash memory device, or other non-volatile solid-state storage device. In some embodiments, the memory 602 may optionally include a memory remotely arranged relative to the processor 601, and these remote memories may be connected to the computer device 60 via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and a combination thereof. One or more units are stored in the memory 602, and when executed by one or more processors 601, the steps of the data entry method based on named entity recognition in any of the above-mentioned method embodiments are executed.

[0108] In the above embodiment, the present invention discloses a computer device, which receives a data file to be entered, extracts data from the data file, and obtains multiple customer data to be entered; inputs the multiple customer data to be entered into a preset named entity recognition model to identify the named entity information in each customer data; through a state machine mode, the named entity information in each customer data is identified according to the preset recognition switching rules to obtain the entity type in the named entity information; according to the entity type in the named entity information, the corresponding entity text is entered into the corresponding field position in the standard form to complete the entry of the customer data. By performing automated data extraction and named entity recognition on irregular data files, the entity information in the customer data is accurately identified and classified and automatically entered into the standard form. Through accurate and standardized data entry, not only the manual entry cost is reduced, but also the accuracy and efficiency of data entry are improved, meeting the efficient entry requirements in the business scenario of batch entry of group insurance information.

[0109] An embodiment of the present invention provides a non-volatile computer-readable storage medium, which stores computer-executable instructions. When the computer-executable instructions are executed by one or more processors, the steps of the data entry method based on named entity recognition in any of the above method embodiments are executed.

[0110] In the above embodiment, the present invention discloses a non-volatile computer-readable storage medium, which receives a data file to be entered, extracts data from the data file, and obtains multiple customer data to be entered; inputs the multiple customer data to be entered into a preset named entity recognition model to identify the named entity information in each customer data; through a state machine mode, the named entity information in each customer data is identified according to the preset recognition switching rules to obtain the entity type in the named entity information; according to the entity type in the named entity information, the corresponding entity text is entered into the corresponding field position in the standard form to complete the entry of the customer data. By performing automated data extraction and named entity recognition on irregular data files, the entity information in the customer data is accurately identified and classified and automatically entered into the standard form. Through accurate and standardized data entry, not only the manual entry cost is reduced, but also the accuracy and efficiency of data entry are improved, meeting the efficient entry requirements in the business scenario of batch entry of group insurance information.

[0111] Through the description of the above implementation methods, those skilled in the art can clearly understand that the above-mentioned embodiment methods can be implemented by means of software plus a necessary general hardware platform, and of course by hardware, but in many cases the former is a better implementation method. Based on such an understanding, the technical solution of the present invention, or the part that contributes to the prior art, can be embodied in the form of a software product, which is stored in a storage medium (such as ROM / RAM, a magnetic disk, or an optical disk), and includes a number of instructions for enabling a terminal device (which can be a mobile phone, a computer, a server, an air conditioner, or a network device, etc.) to execute the methods described in each embodiment of the present invention.

[0112] The present invention can be used in many general or special computer system environments or configurations. For example: personal computers, server computers, handheld or portable devices, tablet devices, multiprocessor systems, microprocessor-based systems, set-top boxes, programmable consumer electronic devices, network PCs, minicomputers, mainframe computers, distributed computing environments including any of the above systems or devices, etc. The present invention can be described in the general context of computer-executable instructions executed by a computer, such as program modules. Generally, program modules include routines, programs, objects, components, data structures, etc. that perform specific tasks or implement specific abstract data types. The present invention can also be practiced in distributed computing environments, in which tasks are performed by remote processing devices connected through a communication network. In a distributed computing environment, program modules can be located in local and remote computer storage media including storage devices.

[0113] In summary, the data entry method, device, equipment and medium based on named entity recognition disclosed in the present invention include: receiving a data file to be entered, extracting data from the data file to obtain multiple customer data to be entered; inputting the multiple customer data to be entered into a preset named entity recognition model to identify the named entity information in each customer data; through a state machine mode, the named entity information in each customer data is identified according to the preset recognition switching rules to obtain the entity type in the named entity information; according to the entity type in the named entity information, the corresponding entity text is entered into the corresponding field position in the standard form to complete the entry of the customer data. By performing automated data extraction and named entity recognition on irregular data files, the entity information in the customer data is accurately identified and classified and automatically entered into the standard form. Through accurate and standardized data entry, not only the manual entry cost is reduced, but also the accuracy and efficiency of data entry are improved, meeting the efficient entry requirements in the business scenario of batch entry of group insurance information.

[0114] Of course, those skilled in the art can understand that all or part of the processes in the above-mentioned embodiments can be implemented by instructing related hardware (such as a processor, a controller, etc.) through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium, and the computer program can include the processes of the above-mentioned method embodiments when executed. The storage medium can be a memory, a disk, a floppy disk, a flash memory, an optical storage device, etc.

[0115] It should be noted that if any software tools or components other than those of the Company appear in the embodiments of the present application, they are merely used for illustration and do not represent actual use. It should be understood that the application of the present invention is not limited to the above examples, and that those skilled in the art can make improvements or changes based on the above description, and all such improvements and changes shall fall within the scope of protection of the claims attached to the present invention.

Claims

1. A data entry method based on named entity recognition, characterized in that: include: Receiving a data file to be entered, extracting data from the data file, and obtaining a plurality of customer data to be entered; Inputting the plurality of customer data to be entered into a preset named entity recognition model to identify and obtain named entity information in each customer data; By using the state machine mode, the entity type of the named entity information in each piece of customer data is identified according to the preset identification switching rule to obtain the entity type in the named entity information; According to the entity type in the named entity information, the corresponding entity text is entered into the corresponding field position in the standard form to complete the entry of the customer data.

2. The data entry method based on named entity recognition according to claim 1, characterized in that: Before inputting the plurality of customer data to be entered into a preset named entity recognition model to identify and obtain the named entity information in each customer data, the method further includes: Collecting corpus samples from a corpus in a specified business field, and annotating the corpus samples according to a preset annotation standard; The corpus sample with entity annotations is input into a named entity recognition model constructed based on a pre-trained language model for training and fine-tuning to obtain a named entity recognition model adapted to the specified business field.

3. The data entry method based on named entity recognition according to claim 1, characterized in that: The step of inputting the plurality of customer data to be entered into a preset named entity recognition model to identify and obtain the named entity information in each customer data includes: Inputting the plurality of customer data to be entered into a preset named entity recognition model to identify entity annotations in each customer data; The entity annotations in each customer data are merged according to the preset annotation standard to obtain the corresponding named entity information.

4. The data entry method based on named entity recognition according to claim 1, characterized in that: The state machine mode is used to identify the entity type of the named entity information in each piece of customer data according to the preset identification switching rule to obtain the entity type in the named entity information, specifically including: Through the state machine mode, entity type identification is performed on the named entity information in each customer data according to the state switching sequence of identifying the name, identifying the ID number, and identifying other information, or according to the state switching sequence of identifying the ID number, identifying the name, and identifying other information, to obtain the entity type in the named entity information.

5. The data entry method based on named entity recognition according to claim 1, characterized in that: The step of entering the corresponding entity text into the corresponding field position in the standard form according to the entity type in the named entity information to complete the entry of the customer data includes: Matching the entity type in the named entity information with the fields in the standard table; The entity text of the corresponding entity type is extracted according to the field matching result, and the entity text is entered into the matching field position to complete the entry of the customer data.

6. The data entry method based on named entity recognition according to claim 1, characterized in that: The method further comprises: Obtain customer data to be corrected for named entity recognition failures; Collecting statistics on entity type recognition results of all named entity information, extracting named entity information from the customer data to be corrected according to the statistical results, and obtaining named entity recognition results of the customer data to be corrected; The standard form is corrected and supplemented according to the named entity recognition result of the customer data to be corrected.

7. The data entry method based on named entity recognition according to claim 6, characterized in that: The step of collecting statistics on entity type recognition results of all named entity information, extracting named entity information from the customer data to be corrected according to the statistical results, and obtaining the named entity information of the customer data to be corrected includes: According to the entity type recognition results of all named entity information, count the field positions where each entity type appears in the customer data to obtain corresponding field index statistics results; Establishing a confidence index for each entity type according to the field index statistics result; Field information is extracted from a corresponding field position in the customer data to be corrected according to the confidence index of each entity type as a named entity recognition result of the customer data to be corrected.

8. A data entry device based on named entity recognition, characterized in that: include: A receiving and extracting module, used for receiving a data file to be entered, extracting data from the data file, and obtaining a plurality of customer data to be entered; A named entity recognition module, used to input the plurality of customer data to be entered into a preset named entity recognition model, and identify the named entity information in each customer data; An entity type identification module, used to identify the entity type of the named entity information in each customer data according to a preset identification switching rule through a state machine mode, and obtain the entity type in the named entity information; The data entry module is used to enter the corresponding entity text into the corresponding field position in the standard form according to the entity type in the named entity information, so as to complete the entry of the customer data.

9. A computer device, characterized in that: comprising at least one processor; and, a memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the data entry method based on named entity recognition as described in any one of claims 1-7.

10. A non-volatile computer-readable storage medium, characterized in that: The non-volatile computer-readable storage medium stores computer-executable instructions, which, when executed by one or more processors, enable the one or more processors to execute the data entry method based on named entity recognition as described in any one of claims 1-7.

Images