Information analysis method and device based on artificial intelligence, computer equipment and medium

Abstract

The invention belongs to the technical field of artificial intelligence, and relates to an artificial intelligence-based information analysis method, which comprises the following steps of: obtaining an input quotation information text; preprocessing the quotation information text to obtain a specified quotation information text; calling a quotation analysis large model, and inputting the specified quotation information text into the quotation analysis large model; analyzing the specified quotation information text according to the target prompt text through a quotation analysis large model to obtain quotation information; performing data verification on the quotation information to obtain a data verification result; and performing output processing on the target quotation information passing the data verification in all the quotation information based on the data verification result. The invention further provides an information analysis device based on artificial intelligence, computer equipment and a storage medium. In addition, the target offer information may be stored in a blockchain. According to the invention, based on the use of the quotation analysis large model, the analysis effect of the quotation information text is improved, and the accuracy of the quotation information generated by analysis is improved.

Description

Technical Field

[0001] The present application relates to the field of artificial intelligence development technology and the field of financial technology, and in particular to information analysis methods, devices, computer equipment and storage media based on artificial intelligence. Background Art

[0002] In today's era of rapid information development, text parsing technology plays a vital role as a key link in information processing and data analysis. Text parsing technology can accurately extract valuable information from massive amounts of text information and output it in a structured manner, thereby greatly improving the efficiency and accuracy of information processing. Especially in financial fields such as debt ecology business, the application of text parsing technology is indispensable.

[0003] In the bond ecosystem business, the analysis of bond quotation text is one of the core links. Traditional bond quotation text analysis mainly relies on conventional solutions based on rule templates. Although this solution has achieved automatic text analysis to a certain extent, it has exposed many drawbacks in actual application.

[0004] First, the bond quotation text parsing solution based on rule templates has a strong dependence on the standard quotation template. When encountering quotation texts that are significantly different from the standard template, the parsing effect is often greatly reduced, resulting in poor parsing results. In addition, the solution lacks reasoning ability. When processing complex quotation texts, relying solely on fixed rule templates often fails to accurately understand the deep meaning and logical relationship in the text, which leads to low accuracy in quotation information parsing. Summary of the invention

[0005] The purpose of the embodiments of the present application is to propose an information parsing method, device, computer equipment and storage medium based on artificial intelligence to solve the technical problems that the existing bond quotation text parsing scheme based on rule templates has poor parsing effect and low accuracy of quotation information parsing.

[0006] In order to solve the above technical problems, the embodiment of the present application provides an information analysis method based on artificial intelligence, which adopts the following technical solutions:

[0007] Get the input quotation information text;

[0008] Preprocessing the quotation information text to obtain a corresponding specified quotation information text;

[0009] Calling a pre-trained quotation parsing model, and inputting the specified quotation information text into the quotation parsing model;

[0010] By using the quotation parsing model, the specified quotation information text is parsed according to the built-in target prompt text to obtain corresponding quotation information; wherein the number of the quotation information includes multiple;

[0011] Performing data verification on the quotation information to obtain corresponding data verification results;

[0012] Based on the data verification result, the target quotation information that has passed the data verification in all the quotation information is output.

[0013] Furthermore, the step of performing data verification on the quotation information to obtain a corresponding data verification result specifically includes:

[0014] Acquire a specified keyword from specified quotation information; wherein the specified quotation information is any one of all the quotation information;

[0015] Based on a preset search verification strategy, searching and verifying the specified keyword in the quotation information text;

[0016] If the specified keyword passes the search verification, generating a first data verification result indicating that the specified quotation information passes the data verification;

[0017] If the designated keyword fails the search verification, a second data verification result indicating that the designated quotation information fails the data verification is generated.

[0018] Furthermore, before the step of calling the pre-trained quotation parsing model and inputting the specified quotation information text into the quotation parsing model, it also includes:

[0019] Get pre-built quote sample data;

[0020] Calling a preset initial large model and determining target model parameters corresponding to the initial large model;

[0021] Obtaining a preset designated prompt text corresponding to the initial large model;

[0022] The initial large model is trained based on the quotation sample data and the specified prompt text to obtain a corresponding initial large model;

[0023] Based on the first version of the large model, the quotation sample data is subjected to case screening and processing to obtain corresponding target quotation cases;

[0024] Constructing target quotation sample data based on the target quotation case and the quotation sample data, and performing secondary training on the initial version of the large model based on the target quotation sample data and the specified prompt text to obtain a final version of the large model that meets the preset construction requirements;

[0025] The final version big model is used as the quotation analysis big model.

[0026] Furthermore, the step of obtaining pre-built quotation sample data specifically includes:

[0027] Export the original chat records containing quotation information from the preset business system;

[0028] Performing data cleaning on the original chat record to obtain corresponding first quotation data;

[0029] Marking the first quotation data based on a preset marking strategy to obtain marked second quotation data;

[0030] The quotation sample data is generated based on the second quotation data.

[0031] Furthermore, the step of determining the target model parameters corresponding to the initial large model specifically includes:

[0032] Determining initial model parameters corresponding to the initial large model;

[0033] Get pre-built parameter optimization algorithms;

[0034] Based on the parameter optimization algorithm, the optimal solution parameters of the initial large model are iteratively calculated to obtain the optimal parameter combination that makes the preset analysis result have the highest accuracy rate;

[0035] The optimal parameter combination is used as the target model parameters.

[0036] Furthermore, the step of performing case screening processing on the quotation sample data based on the first version large model to obtain corresponding target quotation cases specifically includes:

[0037] Analyze and process the quotation sample data based on the first version large model to obtain corresponding analysis results;

[0038] Analyze the parsing results and select third quotation data that meets the parsing failure condition;

[0039] Arrange and process the third quotation data to obtain corresponding arranged data;

[0040] The sorted data is used as the target quotation case.

[0041] Further, after the step of outputting the target quotation information that has passed the data verification in all the quotation information based on the data verification result, the method further includes:

[0042] Get the preset specified conversion format;

[0043] Performing data conversion processing on the target quotation information based on a specified conversion format to obtain corresponding processed data;

[0044] Get the specified storage method;

[0045] The processed data is stored based on the specified storage method.

[0046] In order to solve the above technical problems, the embodiment of the present application further provides an information analysis device based on artificial intelligence, which adopts the following technical solution:

[0047] The first acquisition module is used to acquire the input quotation information text;

[0048] A preprocessing module, used for preprocessing the quotation information text to obtain a corresponding specified quotation information text;

[0049] A first calling module is used to call a pre-trained quotation parsing model and input the specified quotation information text into the quotation parsing model;

[0050] A parsing module, used to parse the specified quotation information text according to the built-in target prompt text through the quotation parsing large model to obtain corresponding quotation information; wherein the number of the quotation information includes multiple;

[0051] A verification module, used to perform data verification on the quotation information to obtain corresponding data verification results;

[0052] The output module is used to output the target quotation information that has passed the data verification among all the quotation information based on the data verification result.

[0053] In order to solve the above technical problems, the embodiment of the present application further provides a computer device, which adopts the following technical solution:

[0054] Get the input quotation information text;

[0055] Preprocessing the quotation information text to obtain a corresponding specified quotation information text;

[0056] Calling a pre-trained quotation parsing model, and inputting the specified quotation information text into the quotation parsing model;

[0057] By using the quotation parsing model, the specified quotation information text is parsed according to the built-in target prompt text to obtain corresponding quotation information; wherein the number of the quotation information includes multiple;

[0058] Performing data verification on the quotation information to obtain corresponding data verification results;

[0059] Based on the data verification result, the target quotation information that has passed the data verification in all the quotation information is output.

[0060] In order to solve the above technical problems, the embodiment of the present application further provides a computer-readable storage medium, which adopts the following technical solution:

[0061] Get the input quotation information text;

[0062] Preprocessing the quotation information text to obtain a corresponding specified quotation information text;

[0063] Calling a pre-trained quotation parsing model, and inputting the specified quotation information text into the quotation parsing model;

[0064] By using the quotation parsing model, the specified quotation information text is parsed according to the built-in target prompt text to obtain corresponding quotation information; wherein the number of the quotation information includes multiple;

[0065] Performing data verification on the quotation information to obtain corresponding data verification results;

[0066] Based on the data verification result, the target quotation information that has passed the data verification in all the quotation information is output.

[0067] Compared with the prior art, the embodiments of the present application have the following beneficial effects:

[0068] The present application first obtains the input quotation information text; then pre-processes the quotation information text to obtain the corresponding designated quotation information text; then calls the pre-trained quotation parsing model, and inputs the designated quotation information text into the quotation parsing model; subsequently, the designated quotation information text is parsed according to the built-in target prompt text through the quotation parsing model to obtain the corresponding quotation information; wherein the number of the quotation information includes multiple; further data verification is performed on the quotation information to obtain the corresponding data verification result; finally, based on the data verification result, the target quotation information that passes the data verification in all the quotation information is output and processed. The present application obtains the input quotation information text, pre-processes the quotation information text to obtain the designated quotation information text, and then uses the pre-trained quotation parsing model with stronger text understanding and reasoning capabilities, so as to realize fast and accurate parsing of the designated quotation information text, effectively improve the parsing effect and parsing accuracy of the quotation information text, and improve the accuracy of the parsed quotation information. BRIEF DESCRIPTION OF THE DRAWINGS

[0069] In order to more clearly illustrate the scheme in the present application, a brief introduction is given below to the drawings required for use in the description of the embodiments of the present application. Obviously, the drawings described below are some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0070] Figure 1 is an exemplary system architecture diagram to which the present application may be applied;

[0071] Figure 2 A flowchart of an embodiment of an information parsing method based on artificial intelligence according to the present application;

[0072] Figure 3 is a schematic structural diagram of an embodiment of an information analysis device based on artificial intelligence according to the present application;

[0073] Figure 4 It is a structural diagram of an embodiment of a computer device according to the present application. DETAILED DESCRIPTION

[0074] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those commonly understood by technicians in the technical field of the present application; the terms used in the specification of the application herein are only for the purpose of describing specific embodiments and are not intended to limit the present application; the terms "including" and "having" and any variations thereof in the specification and claims of the present application and the above-mentioned drawings are intended to cover non-exclusive inclusions. The terms "first", "second", etc. in the specification and claims of the present application or the above-mentioned drawings are used to distinguish different objects, not to describe a specific order.

[0075] Reference to "embodiments" herein means that a particular feature, structure, or characteristic described in conjunction with the embodiments may be included in at least one embodiment of the present application. The appearance of the phrase in various locations in the specification does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment that is mutually exclusive with other embodiments. It is explicitly and implicitly understood by those skilled in the art that the embodiments described herein may be combined with other embodiments.

[0076] In order to enable those skilled in the art to better understand the solution of the present application, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the accompanying drawings.

[0077] like Figure 1As shown, the system architecture 100 may include a terminal device 101, a network 102 and a server 103. The terminal device 101 may be a laptop 1011, a tablet computer 1012 or a mobile phone 1013. The network 102 is used to provide a medium for a communication link between the terminal device 101 and the server 103. The network 102 may include various connection types, such as wired, wireless communication links or optical fiber cables, etc.

[0078] The user can use the terminal device 101 to interact with the server 103 through the network 102 to receive or send messages, etc. Various communication client applications can be installed on the terminal device 101, such as web browser applications, shopping applications, search applications, instant messaging tools, email clients, social platform software, etc.

[0079] The terminal device 101 can be any electronic device with a display screen and supporting web browsing. In addition to the laptop computer 1011, tablet computer 1012 or mobile phone 1013, the terminal device 101 can also be an e-book reader, an MP3 player (Moving Picture Experts Group Audio Layer III), an MP4 (Moving Picture Experts Group Audio Layer IV), a laptop computer, a desktop computer, etc.

[0080] The server 103 may be a server that provides various services, such as a background server that provides support for a web page displayed on the terminal device 101 .

[0081] It should be noted that the artificial intelligence-based information analysis method provided in the embodiment of the present application is generally executed by a server / terminal device, and accordingly, the artificial intelligence-based information analysis device is generally set in the server / terminal device.

[0082] It should be understood that Figure 1 The number of terminal devices, networks and servers in the embodiment is only for illustration. Any number of terminal devices, networks and servers may be provided according to implementation requirements.

[0083] Continue to refer Figure 2, showing a flow chart of an embodiment of an information analysis method based on artificial intelligence according to the present application. According to different requirements, the order of the steps in the flow chart can be changed, and some steps can be omitted. The information analysis method based on artificial intelligence provided in the embodiment of the present application can be applied to any scenario where information analysis is required, and the information analysis method based on artificial intelligence can be applied to products in these scenarios, for example, information analysis in the field of finance and insurance. The information analysis method based on artificial intelligence includes the following steps:

[0084] Step S201, obtaining the input quotation information text.

[0085] In this embodiment, the information parsing method based on artificial intelligence is executed on the electronic device (for example Figure 1 The server / terminal device shown in the figure) can obtain the quotation information text through a wired connection or a wireless connection. It should be noted that the above-mentioned wireless connection method may include but is not limited to 3G / 4G / 5G connection, WiFi connection, Bluetooth connection, WiMAX connection, Zigbee connection, UWB (ultra wideband) connection, and other wireless connection methods currently known or developed in the future. The executor of the present application may be an information parsing system, or simply referred to as a system. The present application can be applied to business scenarios of parsing and processing bond quotation information in the financial field. The above-mentioned quotation information text can provide users or business systems with texts containing bond quotation information, which may be obtained from chat records, emails, market announcements and other channels.

[0086] Step S202: pre-process the quotation information text to obtain a corresponding designated quotation information text.

[0087] In this embodiment, the quotation information text can be processed by removing irrelevant characters, unifying the digital format, and cleaning the text to ensure that the specified quotation information text meets the input request of the quotation parsing model. Among them, a function can be written to remove irrelevant characters in the text, such as punctuation marks, special symbols, etc. These characters may affect the parsing accuracy of the quotation parsing model. In addition, the above-mentioned unified digital format includes: unifying the digital format in the text into a standard format, such as converting percentages to decimals, removing commas from digital formats with commas, etc. The above-mentioned text cleaning includes: using string processing functions to remove blank characters, line breaks, etc. in the text to ensure the neatness of the text.

[0088] Step S203, calling a pre-trained quotation parsing model, and inputting the specified quotation information text into the quotation parsing model.

[0089] In this embodiment, a pre-trained quotation parsing model can be loaded by using a suitable framework or library, and the pre-processed specified quotation information text can be passed as input to the quotation parsing model. The model building process of the above-mentioned quotation parsing model will be further described in detail in the subsequent specific embodiments of this application, and will not be elaborated on here.

[0090] Step S204, parsing the specified quotation information text according to the built-in target prompt text through the quotation parsing large model to obtain corresponding quotation information; wherein the number of the quotation information includes multiple.

[0091] In this embodiment, the quotation parsing model will parse the input specified quotation information text according to the built-in target prompt text (Prompt formula) and the parsing ability obtained through training, and extract the corresponding bond quotation information, i.e. the above quotation information. The quotation information may include keywords such as coupon rate + BP, absolute rate of return, expected transaction size (billion), etc.

[0092] Step S205, performing data verification on the quotation information to obtain corresponding data verification results.

[0093] In this embodiment, the specific implementation process of performing data verification on the quotation information to obtain the corresponding data verification result will be further described in detail in subsequent specific embodiments of the present application, and will not be elaborated on here.

[0094] Step S206: Based on the data verification result, output processing is performed on the target quotation information that has passed the data verification in all the quotation information.

[0095] In this embodiment, the data verification result includes passing the data verification or failing the data verification. The data verification result is analyzed to filter out the target quotation information that passes the data verification from all quotation information, and then the target quotation information is output to the provider corresponding to the quotation information text, such as a related user or business system.

[0096] The present application first obtains the input quotation information text; then pre-processes the quotation information text to obtain the corresponding designated quotation information text; then calls the pre-trained quotation parsing model, and inputs the designated quotation information text into the quotation parsing model; subsequently, the designated quotation information text is parsed according to the built-in target prompt text through the quotation parsing model to obtain the corresponding quotation information; wherein the number of the quotation information includes multiple; further data verification is performed on the quotation information to obtain the corresponding data verification result; finally, based on the data verification result, the target quotation information that passes the data verification in all the quotation information is output and processed. The present application obtains the input quotation information text, pre-processes the quotation information text to obtain the designated quotation information text, and then uses the pre-trained quotation parsing model with stronger text understanding and reasoning capabilities, so as to realize fast and accurate parsing of the designated quotation information text, effectively improve the parsing effect and parsing accuracy of the quotation information text, and improve the accuracy of the parsed quotation information.

[0097] In some optional implementations, step S205 includes the following steps:

[0098] Acquire a specified keyword from specified quotation information; wherein the specified quotation information is any one of all the quotation information.

[0099] In this embodiment, the above-mentioned specified keywords may at least include keywords such as face rate+BP, absolute rate of return, and expected transaction scale (billion).

[0100] Based on a preset search verification strategy, the designated keyword is searched and verified in the quotation information text.

[0101] In this embodiment, the policy content of the above-mentioned search verification policy includes: defining a corresponding regular expression according to the specified keyword, locating the position of the specified keyword in the quotation information text, and matching and extracting the located keyword to ensure that the specified keyword does exist and is correct in the original quotation information text.

[0102] If the designated keyword passes the search verification, a first data verification result indicating that the designated quotation information passes the data verification is generated.

[0103] In this embodiment, if data information corresponding to the specified keyword is searched in the quotation information text, it is determined that the specified keyword passes the search verification, and a first data verification result indicating that the specified quotation information passes the data verification is generated.

[0104] If the designated keyword fails the search verification, a second data verification result indicating that the designated quotation information fails the data verification is generated.

[0105] In this embodiment, if the data information corresponding to the specified keyword is not searched in the quotation information text, it is determined that the specified keyword has not passed the search verification, and then a second data verification result is generated that the specified quotation information has not passed the data verification. In addition, the result content of the specified keyword can be assigned to empty and marked as erroneous data. The detailed information of the erroneous data can also be recorded at the same time for subsequent analysis and improvement. In addition, an error handling strategy can be formulated according to business needs. For example, the large quotation parsing model can be retrained to improve the parsing accuracy, or the erroneous data can be manually corrected.

[0106] The present application obtains specified keywords from specified quotation information; wherein the specified quotation information is any one of all the quotation information; subsequently, based on a preset search verification strategy, the specified keywords are searched and verified in the quotation information text; if the specified keywords pass the search verification, a first data verification result is generated indicating that the specified quotation information passes the data verification; and if the specified keywords fail the search verification, a second data verification result is generated indicating that the specified quotation information fails the data verification. The present application obtains specified keywords from specified quotation information, and then searches and verifies the specified keywords in the quotation information text based on the use of a search verification strategy, and then analyzes the obtained search verification results, thereby achieving rapid and accurate completion of data verification of the quotation information and generating corresponding data verification results, thereby improving the processing efficiency of data verification of the quotation information and ensuring the data accuracy of the obtained data verification results.

[0107] In some optional implementations of this embodiment, before step S203, the electronic device may further perform the following steps:

[0108] Get pre-built quote sample data.

[0109] In this embodiment, the specific implementation process of obtaining the pre-built quotation sample data will be further described in detail in subsequent specific embodiments of the present application, and will not be elaborated in detail here.

[0110] A preset initial large model is called, and target model parameters corresponding to the initial large model are determined.

[0111] In this embodiment, there is no limitation on the selection of the initial large model, which can be determined according to actual business needs. The specific implementation process of determining the target model parameters corresponding to the initial large model will be further described in detail in the subsequent specific embodiments of this application, and will not be elaborated here.

[0112] Obtain a preset designated prompt text corresponding to the initial large model.

[0113] In this embodiment, the above-mentioned specified prompt text (Prompt formula) is a specific format of the prompt. The construction process of the specified prompt text includes: 1) Determine the role: According to business needs, determine the role that the big model should play when parsing quotation information, such as a bond business manager. 2) Clarify the task (clearly and concisely state the content that the prompt requires the model to generate): clearly define the task of the big model, that is, to extract bond quotation information. Specifically, it may include keywords such as bond abbreviation, name of the investor, face rate + BP, absolute rate of return, expected transaction size, etc. 3) Formulate instructions (instructions that the big model should follow when generating text): formulate detailed instructions to guide the big model on how to identify and process quotation information. Specifically, it may include keyword recognition rules (such as regular expressions), number format conversion rules (such as converting percentages to decimals), text segmentation rules, etc.

[0114] Preferably, the content of the specified prompt text may specifically include the following content: Role: You are a bond business manager, help me extract bond quotation information from the text information. Task: Bond quotation keywords include: bond abbreviation, name of the investor, coupon rate + BP, absolute yield, expected transaction size (billion); incomplete keywords also need to be identified, and the display will not be reflected if the keywords are not identified. Instruction: The absolute yield must be a number without - before and after, for example, 3.8%. If it is 5.1-5.6%, it does not belong to the absolute yield and the output is not reflected. The expected transaction size (billion) is rounded to 4 significant decimal places, and values ​​exceeding 4 significant decimal places are discarded: 26,000 = 2.6*0.0001 = 0.0002 billion, 3.1w = 3.1*0.0001 = 0.0003 billion, 106,000 = 10.6*0.0001 = 0.0010 billion, 30.4w = 30.4*0.0001 = 0.0040 billion, 509,000 = 50.9*0.000 1=0.00500 million; when converting 10,000Ww to 100 million, the number should be multiplied by 0.0001, for example, 8 million=800*0.0001=0.0800 million, 1000W=1000*0.0001=0.100 million, 2000w=2000*0.0001=0.200 million, 3kw=3*1000*0.0001=0.300 million, 5KW=5*1000*0.0001=0.500 million. Do not explain your results, and do not make up information that does not appear in the input information.

[0115] The initial large model is trained based on the quotation sample data and the specified prompt text to obtain a corresponding initial large model.

[0116] In this embodiment, according to the Prompt specification of the large model, with the accuracy of the parsing result as the task goal, through multiple training iterations, the bond quotation prompt with the best result accuracy is confirmed, and the initial training of the quotation parsing large model is completed to obtain the initial large model. Specifically, the process of training the initial large model includes: first, the marked quotation sample data is divided into a training set and a verification set, and it is ensured that the training set and the verification set are representative and can reflect the diversity of quotation information in actual business. Then, the initial large model is trained using the designed specified prompt text and training set. A suitable training algorithm can be selected, and the initial large model is parameterized according to the above-mentioned target model parameters. After that, the performance of the initial large model is evaluated using the verification set, such as the accuracy of the parsing result. And according to the evaluation results, the specified prompt text and model parameters are adjusted to improve the accuracy of the parsing result. The above processing steps are repeated later, and the specified prompt text and model parameters are optimized through multiple iterative training until a satisfactory parsing result accuracy is achieved, thereby completing the training of the initial large model and obtaining a trained initial large model.

[0117] Based on the first version of the large model, the quotation sample data is subjected to case screening and processing to obtain corresponding target quotation cases.

[0118] In this embodiment, the above-mentioned specific implementation process of case screening and processing of the quotation sample data based on the initial version of the large model to obtain the corresponding target quotation case will be further described in detail in the subsequent specific embodiments of this application, and will not be elaborated in detail here.

[0119] Constructing target quotation sample data based on the target quotation case and the quotation sample data, and performing secondary training on the initial version of the large model based on the target quotation sample data and the specified prompt text to obtain a final version of the large model that meets the preset construction requirements;

[0120] In this embodiment, the target quotation case and the quotation sample data can be merged to form new target quotation sample data, so as to ensure that the obtained target quotation sample data has sufficient diversity and representativeness and can reflect the quotation information changes in the actual business. Among them, the specific implementation process of the secondary training of the initial large model based on the target quotation sample data and the specified prompt text can refer to the aforementioned processing process of training the initial large model based on the quotation sample data and the specified prompt text, and no further details are given here.

[0121] The final version big model is used as the quotation analysis big model.

[0122] The present application obtains pre-constructed quotation sample data; then calls a preset initial large model, and determines the target model parameters corresponding to the initial large model; and obtains the preset designated prompt text corresponding to the initial large model; then trains the initial large model based on the quotation sample data and the designated prompt text to obtain the corresponding first edition large model; subsequently, performs case screening on the quotation sample data based on the first edition large model to obtain the corresponding target quotation case; further constructs target quotation sample data based on the target quotation case and the quotation sample data, and performs secondary training on the first edition large model based on the target quotation sample data and the designated prompt text to obtain a final edition large model that meets the preset construction requirements; finally, the final edition large model is used as the quotation parsing large model. The present application obtains a preliminary version of a large model by training an initial large model based on quotation sample data and specified prompt text, and then performs case screening and processing on the quotation sample data based on the preliminary version of the large model to obtain a target quotation case, and then intelligently constructs the target quotation sample data based on the target quotation case and the quotation sample data, and performs secondary training on the preliminary version of the large model based on the target quotation sample data and the specified prompt text, thereby intelligently and accurately obtaining a quotation parsing large model that meets the construction requirements, effectively improving the construction intelligence and construction accuracy of the quotation parsing large model, and improving the parsing effect of the generated quotation parsing large model.

[0123] In some optional implementations, the obtaining of pre-built quotation sample data includes the following steps:

[0124] Export the original chat record containing quotation information from the preset business system.

[0125] In this embodiment, the business system is a system that provides text containing bond quotation information, such as a banking system, a loan system, and the like.

[0126] The original chat record is cleaned to obtain corresponding first quotation data.

[0127] In this embodiment, the original chat records are cleaned to remove duplicate, invalid or incorrectly formatted records to ensure that the text fields in the data are complete and readable, thereby obtaining the corresponding first quotation data.

[0128] The first quotation data is marked based on a preset marking strategy to obtain marked second quotation data.

[0129] In this embodiment, according to the above-mentioned labeling strategy, a group of experienced bond business experts or data labelers are assigned to manually review the first quotation data. The labelers need to identify the text containing the quotation information and mark it as quotation sample data. In addition, quality control measures such as double labeling and cross-validation can be further implemented to ensure the accuracy and consistency of the labeling. The labeling results are regularly reviewed and evaluated to identify and correct any errors or inconsistencies.

[0130] The quotation sample data is generated based on the second quotation data.

[0131] In this embodiment, the generated second quotation data is used as the above-mentioned quotation sample data.

[0132] The present application exports the original text of the chat record containing quotation information from a preset business system; then performs data cleaning on the original text of the chat record to obtain the corresponding first quotation data; then performs labeling processing on the first quotation data based on a preset labeling strategy to obtain the labeled second quotation data; and subsequently generates the quotation sample data based on the second quotation data. The present application exports the original text of the chat record containing quotation information from a business system, performs data cleaning on the original text of the chat record to obtain the first quotation data, and then labels the first quotation data based on the labeling strategy, so that the required quotation sample data can be constructed quickly and accurately, thereby improving the construction efficiency of the quotation sample data and ensuring the data accuracy of the obtained quotation sample data.

[0133] In some optional implementations, determining the target model parameters corresponding to the initial large model comprises the following steps:

[0134] Initial model parameters corresponding to the initial large model are determined.

[0135] In this embodiment, according to the project requirements (the large model focuses on extracting keyword results, and the large model's ability to generate text diversity is not required. The accuracy of the parsing results is the task goal), the parameter settings of the initial large model are determined in advance, which may include temperature, top_k, top_p, repetition_penalty, num_generate_sequences, etc. And according to the model characteristics of the large model, a set of reasonable parameter values ​​are preliminarily set by referring to relevant literature, experience values ​​or conducting preliminary experiments.

[0136] Get pre-built parameter optimization algorithms.

[0137] In this embodiment, by taking the accuracy of the analysis result as the target and the accuracy of the analysis result as the task target, a parameter optimization algorithm corresponding to the large model is constructed. Specifically, the parameter optimization algorithm may include methods based on grid search, random search, Bayesian optimization, etc.

[0138] Based on the parameter optimization algorithm, the optimal solution parameters of the initial large model are iteratively calculated to obtain the optimal parameter combination that makes the preset analysis result most accurate.

[0139] In this embodiment, the process of the above-mentioned iterative calculation processing may include: first, setting up an experimental environment, including necessary hardware and software resources, and validation set data for evaluating the model performance of the large model. Then, use the constructed parameter optimization algorithm to iteratively calculate the model performance under different parameter combinations. In each iteration, the accuracy of the model's parsing results is recorded, and the optimal parameter combination is saved. Afterwards, according to the set stop conditions (such as the number of iterations, the performance improvement, etc.), it is determined whether the parameter iteration has converged. If converged, the optimal parameter combination is output; otherwise, the iterative calculation continues. Subsequently, the validation set data is used to verify the model under the optimal parameter combination to ensure its performance in practical applications. Preferably, the parameters temperature = 0.1, top_k = 1, top_p = 1, repetition_penalty = 2, num_generate_sequences = 1 are finally confirmed, so that the large model uses greedy search as the main working mode for quotation parsing to ensure the accuracy of the parsing results.

[0140] The optimal parameter combination is used as the target model parameters.

[0141] The present application determines the initial model parameters corresponding to the initial large model; then obtains a pre-constructed parameter optimization algorithm; then iteratively calculates the optimal solution parameters of the initial large model based on the parameter optimization algorithm, thereby obtaining the optimal parameter combination that makes the preset analysis result most correct; and subsequently uses the optimal parameter combination as the target model parameters. The present application determines the initial model parameters corresponding to the initial large model; then iteratively calculates the optimal solution parameters of the initial large model based on the use of the parameter optimization algorithm, thereby obtaining the optimal parameter combination that makes the preset analysis result most correct, and uses it as the final target model parameters, so that the initial large model is subsequently trained based on the target model parameters and the final quotation analysis large model is generated, which can effectively improve the performance and effect of the generated quotation analysis large model.

[0142] In some optional implementations of this embodiment, the case screening process is performed on the quotation sample data based on the initial large model to obtain the corresponding target quotation case, including the following steps:

[0143] The quotation sample data is parsed based on the initial large model to obtain corresponding parsing results.

[0144] In this embodiment, the quotation sample data is parsed by using the generated first edition large model, and the corresponding parsing results are recorded, wherein the parsing results include correctly parsed texts and incorrectly parsed texts.

[0145] The analysis result is analyzed to filter out third quotation data that meets the analysis failure condition.

[0146] In this embodiment, the incorrectly parsed text contained in the above parsing result can be analyzed, and the quotation data that cannot be parsed by the specified prompt text can be screened out, so as to obtain the third quotation data that meets the parsing failure condition.

[0147] The third quotation data is sorted and processed to obtain corresponding sorted data.

[0148] In this embodiment, all the third quotation data obtained can be sorted and analyzed to form corresponding typical case data, that is, the sorted data mentioned above. The sorted data should cover various possible quotation text structures and complex quotation information to ensure that the large model can be fully learned.

[0149] The sorted data is used as the target quotation case.

[0150] This application parses and processes the quotation sample data based on the first edition large model to obtain the corresponding parsing results; then analyzes the parsing results to screen out the third quotation data that meets the parsing failure conditions; then organizes and processes the third quotation data to obtain the corresponding organized data; and subsequently uses the organized data as the target quotation case. This application parses and processes the quotation sample data based on the first edition large model to obtain the parsing results, then analyzes the parsing results to screen out the third quotation data that meets the parsing failure conditions, and then organizes and processes the third quotation data, so that the required target quotation cases can be quickly and accurately screened out from the quotation sample data, thereby improving the screening efficiency of the target quotation cases. It is beneficial to accurately conduct secondary training of the first edition large model based on the obtained target quotation cases in the future, thereby helping to improve the parsing effect of the generated quotation parsing large model.

[0151] In some optional implementations of this embodiment, after step S206, the electronic device may further perform the following steps:

[0152] Gets the preset specified conversion format.

[0153] In this embodiment, the above-mentioned specified conversion format may be a structured data format designed according to actual business needs, which may specifically include field names, data types, etc., to ensure that the parsing results output by the quotation parsing model can be easily used by the business system.

[0154] The target quotation information is subjected to data conversion processing based on a specified conversion format to obtain corresponding processed data.

[0155] In this embodiment, the target quotation information may be converted into the above-mentioned specified conversion format, which may specifically include operations such as data cleaning and format conversion, so as to obtain processed data.

[0156] Get the specified storage method.

[0157] In this embodiment, there is no specific limitation on the selection of the above-mentioned designated storage method. For example, database storage, file system storage, blockchain storage, etc. can be used.

[0158] The processed data is stored based on the specified storage method.

[0159] In this embodiment, the obtained processed data can be stored in a designated storage location corresponding to the designated storage method according to the above-mentioned designated storage method. In the storage process, the integrity and accuracy of the transmitted processed data must be ensured so that the business system can read and use it correctly.

[0160] The present application obtains a preset designated conversion format; then performs data conversion processing on the target quotation information based on the designated conversion format to obtain corresponding processed data; then obtains a designated storage method; and subsequently stores and processes the processed data based on the designated storage method. After obtaining the target quotation information that has been verified, the present application will also intelligently perform data conversion processing on the target quotation information based on the preset designated conversion format to obtain corresponding processed data, thereby facilitating the business system to correctly read and use it. In addition, the processed data will also be intelligently stored and processed based on the designated storage method, thereby ensuring the security of the generated processed data.

[0161] In some optional implementations, the user information obtained is subject to the user's consent and complies with relevant laws and policies.

[0162] In addition, the non-Company software tools or components that appear in the embodiments of the present application are merely examples and do not represent actual use.

[0163] In addition, after the development of the quotation parsing large model was completed, a parallel running comparison test was carried out with the quotation text parsing model of the conventional solution. The results showed that the bond quotation parsing large model provided by this application has stronger text understanding and reasoning capabilities, and can identify quotation information with relatively complex text structures such as step quotations and names of 2 or more bonds; the parsing ability of the quotation parsing large model is greatly improved (20%) compared with the conventional solution; the bond quotation parsing large model has good compatibility with interference information, and can accurately identify the quotation content of quotations containing some special characters in the original text of the chat record quotation; the bond quotation parsing large model training iteration is faster, and it only takes one week to update and iterate a model version.

[0164] It should be understood that the order of execution of the steps in the above embodiment does not necessarily mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiment of the present invention.

[0165] It should be emphasized that in order to further ensure the privacy and security of the above target quotation information, the above target quotation information can also be stored in a node of a blockchain.

[0166] The blockchain referred to in this application is a new application model of computer technologies such as distributed data storage, peer-to-peer transmission, consensus mechanism, encryption algorithm, etc. Blockchain is essentially a decentralized database, a string of data blocks generated by cryptographic methods. Each data block contains a batch of network transaction information, which is used to verify the validity of its information (anti-counterfeiting) and generate the next block. Blockchain can include the underlying blockchain platform, platform product service layer, and application service layer.

[0167] The embodiments of the present application can acquire and process relevant data based on artificial intelligence technology. Among them, artificial intelligence (AI) is the theory, method, technology and application system that uses digital computers or machines controlled by digital computers to simulate, extend and expand human intelligence, perceive the environment, acquire knowledge and use knowledge to obtain the best results.

[0168] AI basic technologies generally include sensors, dedicated AI chips, cloud computing, distributed storage, big data processing technology, operation / interaction systems, mechatronics, etc. AI software technologies mainly include computer vision technology, robotics technology, biometrics technology, speech processing technology, natural language processing technology, and machine learning / deep learning.

[0169] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiments can be implemented by instructing the relevant hardware through computer-readable instructions, and the computer-readable instructions can be stored in a computer-readable storage medium. When the program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, the aforementioned storage medium can be a non-volatile storage medium such as a disk, an optical disk, a read-only memory (ROM), or a random access memory (RAM).

[0170] It should be understood that, although the steps in the flowchart of the accompanying drawings are displayed in sequence as indicated by the arrows, these steps are not necessarily executed in sequence in the order indicated by the arrows. Unless otherwise specified herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least a part of the steps in the flowchart of the accompanying drawings may include multiple sub-steps or multiple stages, and these sub-steps or stages are not necessarily executed at the same time, but can be executed at different times, and their execution order is not necessarily sequential, but can be executed in turn or alternately with other steps or at least a part of the sub-steps or stages of other steps.

[0171] Further references Figure 3 , as a response to the above Figure 2 The present application provides an embodiment of an information analysis device based on artificial intelligence, 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.

[0172] like Figure 3 As shown, the information analysis device 300 based on artificial intelligence described in this embodiment includes: a first acquisition module 301, a preprocessing module 302, a first calling module 303, a parsing module 304, a verification module 305 and an output module 306. Among them:

[0173] The first acquisition module 301 is used to acquire the input quotation information text;

[0174] A preprocessing module 302 is used to preprocess the quotation information text to obtain a corresponding specified quotation information text;

[0175] The first calling module 303 is used to call a pre-trained quotation parsing model and input the specified quotation information text into the quotation parsing model;

[0176] The parsing module 304 is used to parse the specified quotation information text according to the built-in target prompt text through the quotation parsing model to obtain corresponding quotation information; wherein the number of the quotation information includes multiple;

[0177] Verification module 305, used to perform data verification on the quotation information to obtain corresponding data verification results;

[0178] The output module 306 is used to output the target quotation information that has passed the data verification among all the quotation information based on the data verification result.

[0179] In this embodiment, the operations performed by the above modules or units correspond one-to-one to the steps of the artificial intelligence-based information analysis method in the aforementioned implementation mode, and will not be repeated here.

[0180] In some optional implementations of this embodiment, the verification module 305 includes:

[0181] A first acquisition submodule is used to acquire a specified keyword from specified quotation information; wherein the specified quotation information is any one of all the quotation information;

[0182] A verification submodule, configured to search and verify the specified keywords in the quotation information text based on a preset search verification strategy;

[0183] A first generating submodule, configured to generate a first data verification result indicating that the specified quotation information passes the data verification if the specified keyword passes the search verification;

[0184] The second generating submodule is used to generate a second data verification result indicating that the designated quotation information has not passed the data verification if the designated keyword has not passed the search verification.

[0185] In this embodiment, the operations performed by the above modules or units correspond one-to-one to the steps of the artificial intelligence-based information analysis method in the aforementioned implementation mode, and will not be repeated here.

[0186] In some optional implementations of this embodiment, the information analysis device based on artificial intelligence also includes:

[0187] The second acquisition module is used to acquire pre-built quotation sample data;

[0188] A second calling module is used to call a preset initial large model and determine target model parameters corresponding to the initial large model;

[0189] A third acquisition module is used to acquire a preset designated prompt text corresponding to the initial large model;

[0190] A training module, used for training the initial large model based on the quotation sample data and the specified prompt text to obtain a corresponding initial large model;

[0191] A screening module, used to perform case screening processing on the quotation sample data based on the first version large model to obtain corresponding target quotation cases;

[0192] A construction module, used to construct target quotation sample data based on the target quotation case and the quotation sample data, and to perform secondary training on the initial version of the large model based on the target quotation sample data and the specified prompt text to obtain a final version of the large model that meets the preset construction requirements;

[0193] A determination module is used to use the final version large model as the quotation parsing large model.

[0194] In this embodiment, the operations performed by the above modules or units correspond one-to-one to the steps of the artificial intelligence-based information analysis method in the aforementioned implementation mode, and will not be repeated here.

[0195] In some optional implementations of this embodiment, the second acquisition module includes:

[0196] The export submodule is used to export the original chat records containing quotation information from the preset business system;

[0197] A cleaning submodule, used for cleaning the original text of the chat record to obtain corresponding first quotation data;

[0198] A marking submodule, used for marking the first quotation data based on a preset marking strategy to obtain marked second quotation data;

[0199] The third generating submodule is used to generate the quotation sample data based on the second quotation data.

[0200] In this embodiment, the operations performed by the above modules or units correspond one-to-one to the steps of the artificial intelligence-based information analysis method in the aforementioned implementation mode, and will not be repeated here.

[0201] In some optional implementations of this embodiment, the second calling module includes:

[0202] A first determination submodule, used to determine initial model parameters corresponding to the initial large model;

[0203] The second acquisition submodule is used to acquire a pre-built parameter optimization algorithm;

[0204] A calculation submodule, used for iteratively calculating the optimal solution parameters of the initial large model based on the parameter optimization algorithm, so as to obtain the optimal parameter combination that makes the preset analysis result have the highest accuracy rate;

[0205] The second determination submodule is used to use the optimal parameter combination as the target model parameter.

[0206] In this embodiment, the operations performed by the above modules or units correspond one-to-one to the steps of the artificial intelligence-based information analysis method in the aforementioned implementation mode, and will not be repeated here.

[0207] In some optional implementations of this embodiment, the screening module includes:

[0208] A parsing submodule, used for parsing the quotation sample data based on the first version of the large model to obtain corresponding parsing results;

[0209] A screening submodule, used for analyzing the parsing result and screening out third quotation data that meets the parsing failure condition;

[0210] A sorting submodule, used for sorting the third quotation data to obtain corresponding sorted data;

[0211] The third determination submodule is used to use the sorted data as the target quotation case.

[0212] In this embodiment, the operations performed by the above modules or units correspond one-to-one to the steps of the artificial intelligence-based information analysis method in the aforementioned implementation mode, and will not be repeated here.

[0213] In some optional implementations of this embodiment, the information parsing device based on artificial intelligence further includes:

[0214] A fourth acquisition module, used to acquire a preset specified conversion format;

[0215] A conversion module, used to perform data conversion processing on the target quotation information based on a specified conversion format to obtain corresponding processed data;

[0216] A fifth acquisition module, used to acquire a specified storage mode;

[0217] The storage module is used to store and process the processed data based on the specified storage method.

[0218] In this embodiment, the operations performed by the above modules or units correspond one-to-one to the steps of the artificial intelligence-based information analysis method in the aforementioned implementation mode, and will not be repeated here.

[0219] To solve the above technical problems, the present application also provides a computer device. Figure 4 , Figure 4 This is a basic structural block diagram of the computer device in this embodiment.

[0220] The computer device 4 includes a memory 41, a processor 42, and a network interface 43 that are interconnected through a system bus. It should be noted that the figure only shows a computer device 4 with components 41-43, but it should be understood that it is not required to implement all the components shown, and more or fewer components can be implemented instead. Among them, those skilled in the art can understand that the computer device here is a device that can automatically perform numerical calculations and / or information processing according to pre-set or stored instructions, and its hardware includes but is not limited to microprocessors, application specific integrated circuits (Application Specific Integrated Circuit, ASIC), programmable gate arrays (Field-Programmable Gate Array, FPGA), digital processors (Digital Signal Processor, DSP), embedded devices, etc.

[0221] The computer device may be a desktop computer, a notebook, a PDA, a cloud server, etc. The computer device may interact with the user through a keyboard, a mouse, a remote controller, a touch pad, or a voice control device.

[0222] The memory 41 includes at least one type of readable storage medium, and the readable storage medium includes flash memory, hard disk, multimedia card, card-type memory (for example, SD or DX memory, etc.), random access memory (RAM), static random access memory (SRAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), programmable read-only memory (PROM), magnetic memory, disk, optical disk, etc. In some embodiments, the memory 41 can be an internal storage unit of the computer device 4, such as a hard disk or memory of the computer device 4. In other embodiments, the memory 41 can also be an external storage device of the computer device 4, such as a plug-in hard disk equipped on the computer device 4, a smart memory card (Smart Media Card, SMC), a secure digital (Secure Digital, SD) card, a flash card (FlashCard), etc. Of course, the memory 41 can also include both the internal storage unit of the computer device 4 and its external storage device. In this embodiment, the memory 41 is generally used to store the operating system and various application software installed on the computer device 4, such as computer-readable instructions based on artificial intelligence information parsing methods, etc. In addition, the memory 41 can also be used to temporarily store various types of data that have been output or are to be output.

[0223] The processor 42 may be a central processing unit (CPU), a controller, a microcontroller, a microprocessor, or other data processing chips in some embodiments. The processor 42 is generally used to control the overall operation of the computer device 4. In this embodiment, the processor 42 is used to run computer-readable instructions stored in the memory 41 or process data, such as computer-readable instructions for running the information parsing method based on artificial intelligence.

[0224] The network interface 43 may include a wireless network interface or a wired network interface. The network interface 43 is generally used to establish a communication connection between the computer device 4 and other electronic devices.

[0225] The present application also provides another embodiment, namely, providing a computer-readable storage medium, wherein the computer-readable storage medium stores computer-readable instructions, and the computer-readable instructions can be executed by at least one processor to enable the at least one processor to perform the steps of the above-mentioned artificial intelligence-based information parsing method.

[0226] 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 application, 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, magnetic disk, optical disk), and includes a number of instructions for a terminal device (which can be a mobile phone, computer, server, air conditioner, or network device, etc.) to execute the methods described in each embodiment of the present application.

[0227] Obviously, the embodiments described above are only some embodiments of the present application, rather than all embodiments. The preferred embodiments of the present application are given in the accompanying drawings, but they do not limit the patent scope of the present application. The present application can be implemented in many different forms. On the contrary, the purpose of providing these embodiments is to make the understanding of the disclosure of the present application more thorough and comprehensive. Although the present application is described in detail with reference to the aforementioned embodiments, for those skilled in the art, it is still possible to modify the technical solutions recorded in the aforementioned specific implementation methods, or to perform equivalent replacement of some of the technical features therein. Any equivalent structure made using the contents of the specification and drawings of this application, directly or indirectly used in other related technical fields, is similarly within the scope of patent protection of this application.

Claims

1. An information analysis method based on artificial intelligence, characterized in that: The steps include: Get the input quotation information text; Preprocessing the quotation information text to obtain a corresponding specified quotation information text; Calling a pre-trained quotation parsing model, and inputting the specified quotation information text into the quotation parsing model; By using the quotation parsing model, the specified quotation information text is parsed according to the built-in target prompt text to obtain corresponding quotation information; wherein the number of the quotation information includes multiple; Performing data verification on the quotation information to obtain corresponding data verification results; Based on the data verification result, the target quotation information that has passed the data verification in all the quotation information is output.

2. The information analysis method based on artificial intelligence according to claim 1, characterized in that: The step of performing data verification on the quotation information to obtain a corresponding data verification result specifically includes: Acquire a specified keyword from specified quotation information; wherein the specified quotation information is any one of all the quotation information; Based on a preset search verification strategy, searching and verifying the specified keyword in the quotation information text; If the specified keyword passes the search verification, generating a first data verification result indicating that the specified quotation information passes the data verification; If the designated keyword fails the search verification, a second data verification result indicating that the designated quotation information fails the data verification is generated.

3. The information analysis method based on artificial intelligence according to claim 1, characterized in that: Before the step of calling the pre-trained quotation parsing model and inputting the specified quotation information text into the quotation parsing model, the step further includes: Get pre-built quote sample data; Calling a preset initial large model and determining target model parameters corresponding to the initial large model; Obtaining a preset designated prompt text corresponding to the initial large model; The initial large model is trained based on the quotation sample data and the specified prompt text to obtain a corresponding initial large model; Based on the first version of the large model, the quotation sample data is subjected to case screening and processing to obtain corresponding target quotation cases; Constructing target quotation sample data based on the target quotation case and the quotation sample data, and performing secondary training on the initial version of the large model based on the target quotation sample data and the specified prompt text to obtain a final version of the large model that meets the preset construction requirements; The final version big model is used as the quotation analysis big model.

4. The information analysis method based on artificial intelligence according to claim 3 is characterized in that: The step of obtaining pre-built quotation sample data specifically includes: Export the original chat records containing quotation information from the preset business system; Performing data cleaning on the original chat record to obtain corresponding first quotation data; Marking the first quotation data based on a preset marking strategy to obtain marked second quotation data; The quotation sample data is generated based on the second quotation data.

5. The information analysis method based on artificial intelligence according to claim 3 is characterized in that: The step of determining the target model parameters corresponding to the initial large model specifically includes: Determining initial model parameters corresponding to the initial large model; Get pre-built parameter optimization algorithms; Based on the parameter optimization algorithm, the optimal solution parameters of the initial large model are iteratively calculated to obtain the optimal parameter combination that makes the preset analysis result have the highest accuracy rate; The optimal parameter combination is used as the target model parameters.

6. The information analysis method based on artificial intelligence according to claim 3 is characterized in that: The step of performing case screening processing on the quotation sample data based on the first version large model to obtain corresponding target quotation cases specifically includes: Analyze and process the quotation sample data based on the first version large model to obtain corresponding analysis results; Analyze the parsing results and filter out third quotation data that meets the parsing failure condition; Arrange and process the third quotation data to obtain corresponding arranged data; The sorted data is used as the target quotation case.

7. The information analysis method based on artificial intelligence according to claim 1 is characterized in that: After the step of outputting the target quotation information that has passed the data verification among all the quotation information based on the data verification result, the method further includes: Get the preset specified conversion format; Performing data conversion processing on the target quotation information based on a specified conversion format to obtain corresponding processed data; Get the specified storage method; The processed data is stored based on the specified storage method.

8. An information analysis device based on artificial intelligence, characterized in that: include: The first acquisition module is used to acquire the input quotation information text; A preprocessing module, used for preprocessing the quotation information text to obtain a corresponding specified quotation information text; A first calling module is used to call a pre-trained quotation parsing model and input the specified quotation information text into the quotation parsing model; A parsing module, used to parse the specified quotation information text according to the built-in target prompt text through the quotation parsing large model to obtain corresponding quotation information; wherein the number of the quotation information includes multiple; A verification module, used to perform data verification on the quotation information to obtain corresponding data verification results; The output module is used to output the target quotation information that has passed the data verification among all the quotation information based on the data verification result.

9. A computer device, characterized in that: It comprises a memory and a processor, wherein the memory stores computer-readable instructions, and when the processor executes the computer-readable instructions, the steps of the information parsing method based on artificial intelligence as described in any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer-readable instructions, which, when executed by a processor, implement the steps of the artificial intelligence-based information parsing method according to any one of claims 1 to 7.

Images