Big model-based rule recommendation and data governance method and related products

Abstract

The invention provides a rule recommendation and data governance method based on a large model and related products, and relates to the technical field of artificial intelligence, in particular to the technical fields of large models, cloud computing, big data and the like. The rule recommendation method based on the large model comprises the steps of obtaining target information of a target data table; the target information and preset prompt information are input into a target large model to obtain recommendation information output by the target large model, and the recommendation information comprises a target field in the target data table and an initial rule corresponding to the target field; the prompt information is used for prompting the target large model to obtain the initial rule according to a preset step; and obtaining a target rule corresponding to the target field based on the initial rule corresponding to the target field.

Description

Technical Field

[0001] The present disclosure relates to the field of artificial intelligence technology, specifically to the technical fields of big models, cloud computing, big data, etc., and in particular to a rule recommendation and data governance method based on big models and related products. Background Art

[0002] In the big data scenario, data quality issues are becoming increasingly prominent due to complex data sources and non-standard data collection processes, seriously affecting data analysis and application.

[0003] In related technologies, data quality rules are usually manually configured to ensure data quality, which is not only inefficient but also difficult to cover all possible data quality issues. Summary of the invention

[0004] The present disclosure provides a rule recommendation and data governance method based on a big model and related products.

[0005] According to one aspect of the present disclosure, a rule recommendation method based on a big model is provided, comprising: obtaining target information of a target data table; inputting the target information and preset prompt information into a target big model to obtain recommendation information output by the target big model, wherein the recommendation information comprises: a target field in the target data table, and an initial rule corresponding to the target field; the prompt information is used to prompt the target big model to obtain the initial rule according to preset steps; and based on the initial rule corresponding to the target field, obtaining the target rule corresponding to the target field.

[0006] According to another aspect of the present disclosure, a data governance method is provided, including: obtaining a data table to be governed; performing data governance on a target field in the data table to be governed based on a target rule corresponding to the target field; the target field and the target rule are determined using any of the methods described above.

[0007] According to another aspect of the present disclosure, a rule recommendation device based on a big model is provided, including: an acquisition module, used to acquire target information of a target data table; a recommendation module, used to input the target information and preset prompt information into a target big model to obtain recommendation information output by the target big model, wherein the recommendation information includes: a target field in the target data table, and an initial rule corresponding to the target field; the prompt information is used to prompt the target big model to acquire the initial rule according to preset steps; and a determination module, used to acquire the target rule corresponding to the target field based on the initial rule corresponding to the target field.

[0008] According to another aspect of the present disclosure, a data governance device is provided, including: an acquisition module for acquiring a data table to be governed; a governance module for performing data governance on a target field in the data table to be governed based on a target rule corresponding to the target field; the target field and the target rule are determined by using any of the methods described above.

[0009] According to another aspect of the present disclosure, an electronic device is provided, comprising: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable 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 any method as described in any of the above aspects.

[0010] According to another aspect of the present disclosure, a non-transitory computer-readable storage medium storing computer instructions is provided, wherein the computer instructions are used to enable the computer to execute any one of the methods according to any one of the above aspects.

[0011] According to another aspect of the present disclosure, a computer program product is provided, including a computer program, wherein when the computer program is executed by a processor, the computer program implements any one of the methods described in any one of the above aspects.

[0012] According to the embodiments of the present disclosure, rules can be obtained efficiently and with high quality.

[0013] It should be understood that the content described in this section is not intended to identify the key or important features of the embodiments of the present disclosure, nor is it intended to limit the scope of the present disclosure. Other features of the present disclosure will become easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] The accompanying drawings are used to better understand the present solution and do not constitute a limitation of the present disclosure.

[0015] Figure 1 is a schematic diagram according to a first embodiment of the present disclosure;

[0016] Figure 2 is a schematic diagram of an application scenario for implementing an embodiment of the present disclosure;

[0017] Figure 3 It is a schematic diagram of the target large model rule recommendation effect evaluation provided according to the embodiment of the present disclosure;

[0018] Figure 4 is a schematic diagram according to a second embodiment of the present disclosure;

[0019] Figure 5 is a schematic diagram according to a third embodiment of the present disclosure;

[0020] Figure 6 is a schematic diagram according to a fourth embodiment of the present disclosure;

[0021] Figure 7 is a schematic diagram according to a fifth embodiment of the present disclosure;

[0022] Figure 8 It is a schematic diagram of an electronic device used to implement the big model-based rule recommendation method or data governance method of the embodiment of the present disclosure. DETAILED DESCRIPTION

[0023] The following is a description of exemplary embodiments of the present disclosure in conjunction with the accompanying drawings, including various details of the embodiments of the present disclosure to facilitate understanding, which should be considered as merely exemplary. Therefore, it should be recognized by those of ordinary skill in the art that various changes and modifications may be made to the embodiments described herein without departing from the scope and spirit of the present disclosure. Similarly, for the sake of clarity and conciseness, descriptions of well-known functions and structures are omitted in the following description.

[0024] In the related art, data quality rules are usually manually configured, which is inefficient and lacks accuracy and comprehensiveness.

[0025] In order to obtain rules efficiently and with high quality, the present disclosure provides the following embodiments.

[0026] Figure 1 is a schematic diagram according to the first embodiment of the present disclosure. This embodiment provides a rule recommendation method based on a large model, such as Figure 1 As shown, the method includes:

[0027] 101. Obtain target information from the target data table.

[0028] 102. Input the target information and preset prompt information into the target big model to obtain the recommended information output by the target big model, wherein the recommended information includes: the target field in the target data table, and the initial rule corresponding to the target field; the prompt information is used to prompt the big model to obtain the initial rule according to preset steps.

[0029] 103. Based on the initial rule corresponding to the target field, obtain a target rule corresponding to the target field.

[0030] The target data table refers to the data table to be governed.

[0031] Specifically, the target data table may be input by the user, or alternatively, the user may be provided with selectable data tables through an interactive interface, and the user may select a desired target data table from these data tables. The target data table may be one or more.

[0032] Target information refers to the relevant information of the target data table, and the specific information type can be set. For example, target information can include the following information:

[0033] Metadata, such as table name, field name, etc.

[0034] Field (column) statistics, for example, including: total number of rows, field statistics, such as the number of unique values, maximum value, minimum value, etc.;

[0035] Business description information, for example, includes: overall description information, associations between tables, data quality rule descriptions, rule calculation logic, etc.

[0036] This embodiment can be executed by a rule recommendation platform, and the data governance platform performs data governance based on rules. The above target information can come from the data governance platform, for example, configuration information in the data governance platform, such as metadata and business description information, or is obtained by calculating the configuration information, such as column statistics.

[0037] In the data governance scenario, rules refer to data quality rules, which are used to describe data screening dimensions, which may include: completeness, consistency, validity, uniqueness, accuracy, timeliness, etc. That is, through data quality rules, data that meets the requirements of completeness and consistency can be retained, and the rest of the data that does not meet the requirements can be deleted.

[0038] The large model refers to the Large Language Model (LLM). LLM is a hot technology in the AI ​​field in recent years. LLM is a natural language processing model based on deep learning. It has a huge number of parameters and a complex structure, which enables it to process and understand a large amount of natural language data. Through the process of pre-training and fine-tuning, the LLM model can perform well in a variety of natural language processing tasks, including but not limited to text generation, language understanding, machine translation, etc.

[0039] The large model performs corresponding operations according to the prompt information.

[0040] The target big model refers to the big model used for rule recommendation.

[0041] The prompt information is used to prompt the target large model to obtain the initial rules according to the preset steps.

[0042] The preset steps can specifically include: field trimming -> rule screening -> rule matching -> rule generation, which can guide the target large model to recommend rules step by step and improve accuracy.

[0043] Taking the generation of Structured Query Language (SQL) rules as an example, the prompt information may be similar to “Please filter out the fields you think are important, and for the important fields, first perform rule filtering, then perform rule matching, and output the SQL rules corresponding to the important fields”.

[0044] After obtaining the above target information and prompt information, the target information and prompt information are input into the target macro model. The target macro model processes the target data table based on the prompt information and outputs the recommended information. The recommended information includes: the target field in the target data table and its corresponding rules. Furthermore, the recommended information may also include reasoning logic, that is, the reason for recommending the rule.

[0045] The target field refers to the field determined by the target big model, which can be all or part of the fields in the target data table, that is, the important fields in the target data table considered by the target big model.

[0046] The rules corresponding to the target field obtained by the target big model are called initial rules. The initial rules are part or all of the existing rules. For example, N rules can be pre-configured. For each target field, the target big model can select M rules from these N rules as the initial rules corresponding to the target field. Wherein, N and M are both positive integers, and M is less than or equal to N.

[0047] After obtaining the target field determined by the target large model and its corresponding initial rule, the target rule can be obtained based on the initial rule.

[0048] The target rule refers to the rule that is ultimately used in data governance. It can be the initial rule, or it can be an adjustment to the initial rule, with the adjusted rule being used as the target rule. For example, after obtaining the initial rule output by the target big model, the initial rule can be manually adjusted to obtain the target rule.

[0049] In this embodiment, the target field and its corresponding initial rule are determined based on the big model, and the excellent performance of the big model can be used to improve the comprehensiveness, accuracy and efficiency of rule recommendation. In addition, the accuracy of rule recommendation can be further improved by prompting the big model to obtain the initial rule according to the preset steps through prompt information.

[0050] In order to better understand the present disclosure, the application scenarios involved in the present disclosure are described as follows:

[0051] Figure 2 It is a schematic diagram of an application scenario for implementing the embodiment of the present disclosure.

[0052] like Figure 2 As shown, the system in this scenario includes:

[0053] The client 201 is used to display an interactive interface to the user, and the user can select a target data table in the interface.

[0054] The rule recommendation platform 202 is used to obtain target information of the target data table, input the target information and preset prompt information into the target macro model, and receive the recommendation information output by the target macro model, the recommendation information including: target fields in the target data table and their corresponding initial rules.

[0055] The target target large model 203 is used to output recommendation information according to the prompt information and target information input by the rule recommendation platform.

[0056] After the rule recommendation platform obtains the recommendation information output by the target large model, it can be displayed to the user through the client. Afterwards, the user can use the initial rule as the target rule of the target field as needed, or manually adjust the initial rule and use the adjusted rule as the target rule.

[0057] After obtaining the target rules, they can be stored in the data governance platform, which will perform data governance on the corresponding fields based on the rules.

[0058] The target information obtained by the rule recommendation platform may include metadata, column statistics, and business description information, which may come from the data governance platform. Furthermore, a private domain knowledge base may be established in advance, which records business documents in the target governance field (such as education, transportation, etc.). After determining the target data table, the private domain knowledge base is searched to obtain relevant documents, and the relevant documents are also input into the target big model for reference by the target big model.

[0059] The target big model can be obtained by fine-tuning the existing pre-trained target big model, and specifically, supervised fine-tuning (SFT) can be used. SFT uses a dataset with correct answers for fine-tuning, and the dataset with correct answers can be manually annotated or fed back. Fine-tuning can make the target big model better adapted to specific tasks or vertical fields, such as homework grading for education, travel service assistant for transportation, etc.

[0060] During SFT, all parameters can be fine-tuned; or, some parameters can be fine-tuned. Fine-tuning some parameters can be performed based on parameter-efficient methods (PEFT), specifically the Low-Rank Adaptation (LoRA) algorithm. The LoRA algorithm introduces low-rank matrix decomposition, which reduces computing resources and storage requirements while maintaining the initial performance of the pre-trained model, stabilizing the fine-tuning process, and reducing storage and deployment costs.

[0061] Furthermore, the data from manual adjustments to the initial rules can be used as annotation data for fine-tuning the target large model, so that the target large model can be continuously updated to improve the accuracy of the target large model's recommended rules.

[0062] After obtaining the initial rules recommended by the target large model, the recommendation effect of the target large model can be evaluated based on the initial rules and their corresponding true value rules.

[0063] Figure 3 It is a schematic diagram of the target large model rule recommendation effect evaluation provided according to the embodiment of the present disclosure.

[0064] like Figure 3 As shown in the figure, the rule recommendation effect of the target large model can be evaluated using two indicators: precision and recall.

[0065] Precision = TP / P;

[0066] Recall rate (Recall) = TP / T.

[0067] Where TP is the number of predicted correct rules;

[0068] TN is the number of unsuccessful predictions of the annotation rule;

[0069] PN is the number of predicted rules that exceed the annotation range;

[0070] T is the number of annotation rules, T = TP + TN;

[0071] P is the data of the prediction rule, P=TP+PN.

[0072] For example, for a certain field, its pre-annotated rules include the first rule and the second rule. If the recommended rule for the field output by the target large model is also the first rule and the second rule, it indicates that the recommended rule is a correct prediction rule; if the recommended rule is the first rule, it indicates that the recommended rule is a marked rule that has not been successfully predicted; if the recommended rule includes the first rule, the second rule and the third rule, it indicates that the recommended rule is a prediction rule that exceeds the annotation range.

[0073] In this way, the recommendation effect of the target large model can be evaluated through the evaluation indicators.

[0074] In combination with the above application scenarios, the present disclosure also provides the following embodiments.

[0075] Figure 4 is a schematic diagram according to a second embodiment of the present disclosure. This embodiment provides a rule recommendation method based on a large model, and the method includes:

[0076] 401. Fine-tune the existing pre-trained large model to obtain the target large model.

[0077] Among them, the pre-trained large model can adopt the existing natural language processing large model, such as the Wenxinyiyan large model.

[0078] The dataset used for fine-tuning can be a dataset containing multiple data to improve the versatility of the model. Specifically, the dataset can be the DuSQL dataset, which is a dataset for practical applications. It contains 200 databases and covers 164 fields. The problems cover common forms in practical applications such as matching, calculation, and reasoning. This dataset is closer to real application scenarios, has nothing to do with the model field or problem, and has the ability of calculation and reasoning.

[0079] During fine-tuning, SFT (supervised fine-tuning) can be used. At this time, the true value data (supervisory data) can be labeled manually or by other means.

[0080] When fine-tuning, you can adjust all parameters, or use algorithms such as LoRA to adjust some parameters.

[0081] In this embodiment, by fine-tuning the pre-trained large model to obtain the target large model, the performance of the target large model can be improved, thereby improving the rule recommendation effect.

[0082] 402. Obtain target information of the target data table.

[0083] The target data table is a data table for which rules need to be recommended, and may be selected by the user. For example, the rule recommendation platform may provide multiple candidate data tables, and the user may select one or more of the multiple candidate data tables as the target data table as needed.

[0084] After the rule recommendation platform obtains the target data table, it can obtain the preset type of target information of the target data table. Such target information includes, for example: metadata (such as table name, field name), column statistics (such as total number of rows), and business description information (such as application scenarios such as education).

[0085] 403. Input the target information and preset prompt information into the target macro model; the prompt information is used to prompt the target macro model to obtain initial rules according to preset steps.

[0086] The preset steps are, for example: field trimming -> rule screening -> rule matching -> rule generation. This can guide the target large model to recommend rules step by step to improve accuracy.

[0087] In addition, the prompt information may also include rule information, such as rule definitions, examples, rule recommendation requirements, etc., to guide the target large model to make appropriate rule recommendations.

[0088] In addition, the rule recommendation platform can also search from the preset private domain knowledge base to obtain relevant documents of the target data table, and input the relevant documents into the target big model for reference by the target big model when recommending rules.

[0089] 404. The target large model outputs recommendation information based on the target information and the prompt information, and the recommendation information includes: the target field in the target data table, and the initial rule corresponding to the target field.

[0090] For example, the target large model may determine that the first field corresponds to the first rule and the second rule, the second field corresponds to the third rule, and so on.

[0091] 405. Based on the initial rule corresponding to the target field, obtain a target rule corresponding to the target field.

[0092] The initial rule may be used as the target rule; or,

[0093] A manual adjustment result for the initial rule is obtained, and the manual adjustment result is used as the target rule.

[0094] For example, after the rule recommendation platform obtains the recommendation information output by the target large model, it can be displayed to the user through the client. The user can use the initial rule as the target rule according to the actual situation, or manually adjust the initial rule, such as adjusting the third rule corresponding to the second field to the fourth rule, and using the adjusted rule as the target rule.

[0095] In this embodiment, the target rules are obtained by manually adjusting the initial rules, which can combine large model recommendation and manual adjustment, and can fully utilize the knowledge contained in massive data while ensuring the accuracy and reliability of the recommendation results.

[0096] Furthermore, the manually adjusted annotation data can be used to fine-tune the target large model again, which can ensure that the target large model is continuously updated and improve the efficiency of obtaining annotation data during fine-tuning.

[0097] In some embodiments, after obtaining the initial rules of the target large model, the evaluation results of the target large model may also be obtained based on the initial rules and their corresponding true values.

[0098] For example, the evaluation results include precision and recall, combined with Figure 3 , the initial rule is the prediction rule, and the labeling rule is its corresponding true value. The precision and recall can be calculated based on the rule prediction situation.

[0099] In this embodiment, by obtaining the evaluation results, the performance of the target large model can be evaluated using the evaluation results, which can then be used as basic data for subsequent processing. For example, when one or more of the precision and recall rates are greater than a threshold, the target large model is formally deployed; otherwise, the target large model continues to be fine-tuned until the evaluation indicators meet the preset conditions.

[0100] After obtaining the target field and its corresponding target rule, you can use the target rule to perform data governance.

[0101] Figure 5 is a schematic diagram according to a third embodiment of the present disclosure. This embodiment provides a data governance method, which includes:

[0102] 501. Obtain the data table to be managed.

[0103] 502. For a target field in the to-be-governed data table, data governance is performed based on a target rule corresponding to the target field.

[0104] The target field and the target rule are determined by using the method described in any of the above embodiments.

[0105] In this embodiment, since comprehensive and accurate rules are adopted, the comprehensiveness and accuracy of data governance can be improved.

[0106] Figure 6 It is a schematic diagram according to the fourth embodiment of the present disclosure. This embodiment provides a rule recommendation device based on a large model. The device 600 includes: an acquisition module 601, a recommendation module 602 and a determination module 603.

[0107] The acquisition module 601 is used to acquire the target information of the target data table; the recommendation module 602 is used to input the target information and preset prompt information into the target big model to obtain the recommendation information output by the target big model, and the recommendation information includes: the target field in the target data table, and the initial rule corresponding to the target field; the prompt information is used to prompt the target big model to obtain the initial rule according to the preset steps; the determination module 603 is used to obtain the target rule corresponding to the target field based on the initial rule corresponding to the target field.

[0108] In this embodiment, the target field and its corresponding initial rule are determined based on the big model, and the excellent performance of the big model can be used to improve the comprehensiveness, accuracy and efficiency of rule recommendation. In addition, the accuracy of rule recommendation can be further improved by prompting the big model to obtain the initial rule according to the preset steps through prompt information.

[0109] In some embodiments, the determining module 603 is further configured to:

[0110] Taking the initial rule as the target rule; or,

[0111] A manual adjustment result for the initial rule is obtained, and the manual adjustment result is used as the target rule.

[0112] In this embodiment, the target rules are obtained by manually adjusting the initial rules, which can combine large model recommendation and manual adjustment, and can fully utilize the knowledge contained in massive data while ensuring the accuracy and reliability of the recommendation results.

[0113] In some embodiments, the target large model is generated after fine-tuning the pre-trained large model.

[0114] In this embodiment, by fine-tuning the pre-trained large model to obtain the target large model, the performance of the target large model can be improved, thereby improving the rule recommendation effect.

[0115] In some embodiments, the apparatus 600 further includes:

[0116] An updating module is used to update the target large model based on the manual adjustment result.

[0117] This ensures that the target large model is constantly updated and improves the efficiency of obtaining labeled data during fine-tuning.

[0118] In some embodiments, the apparatus 600 further includes:

[0119] An evaluation module is used to obtain an evaluation result of the target large model based on the initial rule and a preset true value corresponding to the initial rule.

[0120] In this embodiment, the performance of the target large model can be evaluated through the evaluation results, which can then be used as basic data for subsequent processing.

[0121] Figure 7 It is a schematic diagram according to the fifth embodiment of the present disclosure. This embodiment provides a data governance device, and the device 700 includes: an acquisition module 701 and a governance module 702.

[0122] The acquisition module 701 is used to obtain the data table to be managed; the management module 702 is used to perform data management on the target field in the data table to be managed based on the target rule corresponding to the target field; wherein the target field and the target rule are determined by the method described in any of the above embodiments.

[0123] In this embodiment, since comprehensive and accurate rules are adopted, the comprehensiveness and accuracy of data governance can be improved.

[0124] It can be understood that in the embodiments of the present disclosure, the same or similar contents in different embodiments can be referenced to each other.

[0125] It can be understood that the “first”, “second”, etc. in the embodiments of the present disclosure are only used for distinction and do not indicate the degree of importance, time sequence, etc.

[0126] It is understandable that unless there is any special limitation on the sequence of steps in the process, it means that the timing relationship between these steps is not limited.

[0127] In the technical solution of the present disclosure, the collection, storage, use, processing, transmission, provision and disclosure of user personal information involved are in compliance with the provisions of relevant laws and regulations and do not violate public order and good morals.

[0128] According to an embodiment of the present disclosure, the present disclosure also provides an electronic device, a readable storage medium and a computer program product.

[0129] Figure 8 A schematic block diagram of an example electronic device 800 that can be used to implement an embodiment of the present disclosure is shown. The electronic device 800 is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital assistants, cellular phones, smart phones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present disclosure described and / or required herein.

[0130] like Figure 8 As shown, the electronic device 800 includes a computing unit 801, which can perform various appropriate actions and processes according to a computer program stored in a read-only memory (ROM) 802 or a computer program loaded from a storage unit 808 into a random access memory (RAM) 803. In the RAM 803, various programs and data required for the operation of the electronic device 800 can also be stored. The computing unit 801, the ROM 802, and the RAM 803 are connected to each other via a bus 804. An input / output (I / O) interface 805 is also connected to the bus 804.

[0131] Multiple components in the electronic device 800 are connected to the I / O interface 805, including: an input unit 806, such as a keyboard, a mouse, etc.; an output unit 807, such as various types of displays, speakers, etc.; a storage unit 808, such as a disk, an optical disk, etc.; and a communication unit 809, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 809 allows the electronic device 800 to exchange information / data with other devices through a computer network such as the Internet and / or various telecommunication networks.

[0132] The computing unit 801 may be a variety of general and / or special processing components with processing and computing capabilities. Some examples of the computing unit 801 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various dedicated artificial intelligence (AI) computing chips, various computing units running machine learning model algorithms, digital signal processors (DSPs), and any appropriate processors, controllers, microcontrollers, etc. The computing unit 801 performs the various methods and processes described above, such as a rule recommendation method or a data governance method based on a large model. For example, in some embodiments, a rule recommendation method or a data governance method based on a large model may be implemented as a computer software program, which is tangibly included in a machine-readable medium, such as a storage unit 808. In some embodiments, part or all of the computer program may be loaded and / or installed on the electronic device 800 via the ROM 802 and / or the communication unit 809. When the computer program is loaded into the RAM 803 and executed by the computing unit 801, one or more steps of the rule recommendation method or the data governance method based on the large model described above may be performed. Alternatively, in other embodiments, the computing unit 801 may be configured to execute a big model-based rule recommendation method or a data governance method in any other appropriate manner (eg, by means of firmware).

[0133] Various implementations of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), complex programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various implementations can include: being implemented in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which can be a special purpose or general purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.

[0134] The program code for implementing the method of the present disclosure may be written in any combination of one or more programming languages. These program codes may be provided to a processor or controller of a general-purpose computer, a special-purpose computer, or other programmable task processing device, so that the program code, when executed by the processor or controller, implements the functions / operations specified in the flow chart and / or block diagram. The program code may be executed entirely on the machine, partially on the machine, partially on the machine and partially on a remote machine as a stand-alone software package, or entirely on a remote machine or server.

[0135] In the context of the present disclosure, a machine-readable medium may be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, device, or equipment. A machine-readable medium may be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium may include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, device, or equipment, or any suitable combination of the foregoing. A more specific example of a machine-readable storage medium may include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0136] To provide interaction with a user, the systems and techniques described herein can be implemented on a computer having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user can provide input to the computer. Other types of devices can also be used to provide interaction with the user; for example, the feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including acoustic input, voice input, or tactile input).

[0137] The systems and techniques described herein may be implemented in a computing system that includes back-end components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes front-end components (e.g., a user computer with a graphical user interface or a web browser through which a user can interact with implementations of the systems and techniques described herein), or a computing system that includes any combination of such back-end components, middleware components, or front-end components. The components of the system may be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: a local area network (LAN), a wide area network (WAN), and the Internet.

[0138] A computer system may include a client and a server. The client and the server are generally remote from each other and usually interact through a communication network. The relationship between the client and the server is generated by computer programs running on the corresponding computers and having a client-server relationship with each other. The server may be a cloud server, also known as a cloud computing server or cloud host, which is a host product in the cloud computing service system to solve the defects of difficult management and weak business scalability in traditional physical hosts and VPS services ("Virtual Private Server", or "VPS" for short). The server may also be a server of a distributed system, or a server combined with a blockchain.

[0139] It should be understood that the various forms of processes shown above can be used to reorder, add or delete steps. For example, the steps recorded in this disclosure can be executed in parallel, sequentially or in different orders, as long as the desired results of the technical solutions disclosed in this disclosure can be achieved, and this document does not limit this.

[0140] The above specific implementations do not constitute a limitation on the protection scope of the present disclosure. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modification, equivalent substitution and improvement made within the spirit and principle of the present disclosure shall be included in the protection scope of the present disclosure.

Claims

1. A rule recommendation method based on a large model, comprising: Get the target information of the target data table; Input the target information and preset prompt information into the target macro model to obtain the recommended information output by the target macro model, wherein the recommended information includes: the target field in the target data table, and the initial rule corresponding to the target field; the prompt information is used to prompt the target macro model to obtain the initial rule according to the preset steps; Based on the initial rule corresponding to the target field, a target rule corresponding to the target field is acquired.

2. The method according to claim 1, wherein: The acquiring, based on the initial rule corresponding to the target field, a target rule corresponding to the target field, comprises: Taking the initial rule as the target rule; or, A manual adjustment result for the initial rule is obtained, and the manual adjustment result is used as the target rule.

3. The method according to claim 1, wherein: The target large model is generated after fine-tuning the pre-trained large model.

4. The method according to claim 2, further comprising: The target large model is updated based on the manual adjustment result.

5. The method according to claim 1, further comprising: Based on the initial rule and the preset true value corresponding to the initial rule, an evaluation result of the target large model is obtained.

6. A data governance method, comprising: Get the data table to be managed; For the target field in the to-be-governed data table, data governance is performed based on the target rule corresponding to the target field; The target field and the target rule are determined by using the method according to any one of claims 1-5.

7. A rule recommendation device based on a large model, comprising: An acquisition module is used to acquire target information of a target data table; A recommendation module, used for inputting the target information and preset prompt information into the target macro model to obtain the recommendation information output by the target macro model, wherein the recommendation information includes: the target field in the target data table, and the initial rule corresponding to the target field; the prompt information is used to prompt the target macro model to obtain the initial rule according to the preset steps; A determination module is used to obtain a target rule corresponding to the target field based on the initial rule corresponding to the target field.

8. The device according to claim 7, wherein: The determination module is further used for: Taking the initial rule as the target rule; or, A manual adjustment result for the initial rule is obtained, and the manual adjustment result is used as the target rule.

9. The device according to claim 7, wherein: The target large model is generated after fine-tuning the pre-trained large model.

10. The apparatus according to claim 8, further comprising: An updating module is used to update the target large model based on the manual adjustment result.

11. The apparatus according to claim 7, further comprising: An evaluation module is used to obtain an evaluation result of the target large model based on the initial rule and a preset true value corresponding to the initial rule.

12. A data management device, comprising: The acquisition module is used to obtain the data table to be managed; A governance module, used for performing data governance on a target field in the data table to be governed based on a target rule corresponding to the target field; The target field and the target rule are determined by using the method according to any one of claims 1-5.

13. An electronic device comprising: at least one processor; as well as 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 to enable the at least one processor to perform the method according to any one of claims 1 to 6.

14. A non-transitory computer-readable storage medium storing computer instructions, wherein: The computer instructions are used to cause the computer to execute the method according to any one of claims 1-6.

15. A computer program product, comprising a computer program, which, when executed by a processor, implements the method according to any one of claims 1 to 6.

Images