CN117234774A - Intelligent fault processing method, device, equipment and medium based on information system - Google Patents
Intelligent fault processing method, device, equipment and medium based on information system Download PDFInfo
- Publication number
- CN117234774A CN117234774A CN202311187892.6A CN202311187892A CN117234774A CN 117234774 A CN117234774 A CN 117234774A CN 202311187892 A CN202311187892 A CN 202311187892A CN 117234774 A CN117234774 A CN 117234774A
- Authority
- CN
- China
- Prior art keywords
- matching
- fault
- scene
- solution
- text
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000003672 processing method Methods 0.000 title claims abstract description 19
- 238000012423 maintenance Methods 0.000 claims abstract description 98
- 239000013598 vector Substances 0.000 claims abstract description 45
- 238000000034 method Methods 0.000 claims description 41
- 239000013589 supplement Substances 0.000 claims description 22
- 230000008569 process Effects 0.000 claims description 17
- 238000004590 computer program Methods 0.000 claims description 14
- 239000011159 matrix material Substances 0.000 claims description 13
- 238000006243 chemical reaction Methods 0.000 claims description 9
- 238000012163 sequencing technique Methods 0.000 claims description 6
- 230000008014 freezing Effects 0.000 claims description 4
- 238000007710 freezing Methods 0.000 claims description 4
- 238000012545 processing Methods 0.000 description 12
- 230000003993 interaction Effects 0.000 description 8
- 238000004891 communication Methods 0.000 description 7
- 238000003058 natural language processing Methods 0.000 description 7
- 238000010586 diagram Methods 0.000 description 6
- 230000009471 action Effects 0.000 description 4
- 238000004458 analytical method Methods 0.000 description 3
- 238000004422 calculation algorithm Methods 0.000 description 3
- 238000004364 calculation method Methods 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 3
- 238000007689 inspection Methods 0.000 description 3
- 238000012549 training Methods 0.000 description 3
- 230000006978 adaptation Effects 0.000 description 2
- 230000005540 biological transmission Effects 0.000 description 2
- 230000001419 dependent effect Effects 0.000 description 2
- 238000013461 design Methods 0.000 description 2
- 238000000605 extraction Methods 0.000 description 2
- 230000004044 response Effects 0.000 description 2
- 238000012216 screening Methods 0.000 description 2
- 206010063385 Intellectualisation Diseases 0.000 description 1
- 230000002159 abnormal effect Effects 0.000 description 1
- 238000009411 base construction Methods 0.000 description 1
- QVFWZNCVPCJQOP-UHFFFAOYSA-N chloralodol Chemical compound CC(O)(C)CC(C)OC(O)C(Cl)(Cl)Cl QVFWZNCVPCJQOP-UHFFFAOYSA-N 0.000 description 1
- 238000012790 confirmation Methods 0.000 description 1
- 238000013075 data extraction Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 230000000977 initiatory effect Effects 0.000 description 1
- 230000002452 interceptive effect Effects 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 230000008520 organization Effects 0.000 description 1
- 230000002085 persistent effect Effects 0.000 description 1
- 230000000750 progressive effect Effects 0.000 description 1
- 230000004043 responsiveness Effects 0.000 description 1
- 230000000153 supplemental effect Effects 0.000 description 1
Landscapes
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The application discloses an intelligent fault processing method, device, equipment and medium based on an information system, relating to the field of system operation and maintenance, comprising the following steps: receiving a fault scene text input by the front end, analyzing and calculating the fault scene text, matching a local operation and maintenance database based on the obtained scene subject words and scene sentence vectors, and generating a corresponding matching result; if the matching is successful, a temporary knowledge base is generated based on a plurality of target matching texts, and a first solution is generated based on the temporary knowledge base; if the matching fails, interacting with the front end through a preset language model to generate a second solution by using the supplemented fault scene text; the first solution or the second solution is saved to a local operation and maintenance database to update the solution. In this way, after determining that a fault has occurred, the fault can be automatically identified and a corresponding solution generated to help the operation and maintenance personnel respond to the fault quickly through the solution.
Description
Technical Field
The present invention relates to the field of system operation and maintenance, and in particular, to an intelligent fault processing method, apparatus, device and medium based on an information system.
Background
The safe and stable operation of the information system is an important foundation for guaranteeing the rapid and efficient development of banking business, the operation and maintenance of the information system in banking industry at present are gradually going to intensification, automation and intellectualization, and in the aspects of fault and emergency treatment, the information system has a complete flow mechanism, and the systemization and automation degree are continuously improved.
However, when the foreground service system fails, due to professional reasons, service personnel can hardly lock corresponding scenes in the project document quickly, obtain failure reasons and treatment suggestions, and also need to consult the background for the problem which can be solved by the simple operation of the foreground, so that the problem is highly dependent on operation and maintenance personnel, a great deal of time is consumed for solving the problem, and when the background personnel receives a fault alarm, because the emergency treatment tools are various in types, the emergency treatment tools are often required to be carried out across systems, platforms and equipment, the requirements on operation skills and operation experience are high, and the speed, efficiency and quality of the emergency treatment are difficult to ensure.
Disclosure of Invention
In view of the above, the present application aims to provide an intelligent fault processing method, apparatus, device and medium based on an information system, which can automatically identify a fault after determining that the fault occurs, and generate a corresponding solution, so as to help an operation and maintenance personnel to respond to the fault quickly through the solution. The specific scheme is as follows:
in a first aspect, the application discloses an intelligent fault processing method based on an information system, which is applied to a bank information system and comprises the following steps:
receiving a fault scene text input by a front end, analyzing and calculating the fault scene text, matching a local operation and maintenance database based on the obtained scene subject words and scene sentence vectors, and generating a corresponding matching result;
if the matching result represents successful matching, a temporary knowledge base is generated based on a plurality of target matching texts matched from the local operation and maintenance database, a first solution is generated based on the temporary knowledge base through a preset language model, and the first solution is returned to the client;
if the matching result represents that the matching fails, interacting with the front end based on preset priori knowledge through the preset language model to generate a second solution by utilizing the supplemented fault scene text input by the front end, and returning the second solution to the client;
And saving the first solution or the second solution to the local operation and maintenance database so as to update the solutions saved in the local operation and maintenance database.
Optionally, before the receiving the fault scene text input by the front end and analyzing and calculating the fault scene text to match the local operation and maintenance database based on the obtained scene subject term and the scene sentence vector and generate the corresponding matching result, the method further includes:
and collecting historical data and real-time data of the bank information system to integrate the historical data and the real-time data so as to obtain a local operation and maintenance database.
Optionally, the receiving the fault scene text input by the front end, and analyzing and calculating the fault scene text includes:
receiving a fault scene text input by a front end, and inputting the fault scene text into a preset document theme generation model so as to extract scene subject words in the fault scene text through the preset document theme generation model;
and inputting the fault scene text into a preset text conversion model so as to convert the fault scene text into scene sentence vectors through the preset text conversion model.
Optionally, the matching the local operation and maintenance database based on the obtained scene subject term and the scene sentence vector, and generating a corresponding matching result, includes:
if the local operation and maintenance database is matched based on the scene subject words and the scene sentence vectors and a plurality of matching texts are not obtained, generating a matching result representing failure of matching;
if the local operation and maintenance database is matched based on the scene subject words and the scene sentence vectors and a plurality of matching texts are obtained, a matching result representing successful matching is generated; respectively determining the similarity between the plurality of matching texts and the fault scene text, and sequencing the similarity values of the similarity to obtain a sequencing result; rejecting the matched text with the similarity value lower than a preset threshold value based on the sorting result to obtain a matched text after rejection; and extracting a preset number of matching texts with the similarity value not smaller than other matching texts from the matched texts after the removal so as to obtain a plurality of target matching texts.
Optionally, if the matching result characterizes that the matching is successful, generating a temporary knowledge base based on a plurality of target matching texts matched from the local operation and maintenance database, generating a first solution based on the temporary knowledge base through a preset language model, and before returning the first solution to the client, further including:
Adding a low-order bypass matrix for the ChatGLM model, and freezing the original weight of the ChatGLM model to perform fine tuning on the low-order bypass matrix and the low-order bypass matrix through a preset fine tuning data set so as to obtain the preset language model.
Optionally, if the matching result indicates that the matching fails, interacting with the front end based on preset priori knowledge through the preset language model to generate a second solution by using the post-supplement fault scene text input by the front end, and returning the second solution to the client, including:
if the matching result represents that the matching fails, generating a text supplement guide sentence based on preset priori knowledge through the preset language model, and sending the text supplement guide sentence to the front end, so that after the front end receives the text supplement guide sentence, a fault scene text after supplement is input;
analyzing and calculating the supplemented fault scene text to re-match the local operation and maintenance database so as to obtain a plurality of re-matched texts;
generating a second solution based on the plurality of re-matched texts through the preset language model, and returning the second solution to the client.
Optionally, the intelligent fault processing method based on the information system further includes:
a process manual is generated based on the solutions stored in the local operation and maintenance database and the historical operation and maintenance data to feed back the process manual to the front end upon receiving a fault alert.
In a second aspect, the present application discloses an intelligent fault handling device based on an information system, which is applied to a bank information system, and comprises:
the data matching module is used for receiving the fault scene text input by the front end, analyzing and calculating the fault scene text, matching the local operation and maintenance database based on the obtained scene subject words and scene sentence vectors, and generating corresponding matching results;
the first scheme generation module is used for generating a temporary knowledge base based on a plurality of target matching texts matched from the local operation and maintenance database if the matching result represents that the matching is successful, generating a first solution based on the temporary knowledge base through a preset language model, and returning the first solution to the client;
the second scheme generating module is used for interacting with the front end based on preset priori knowledge through the preset language model if the matching result represents that the matching is failed, so as to generate a second solution by utilizing the supplemented fault scene text input by the front end, and returning the second solution to the client;
And the scheme updating module is used for storing the first solution or the second solution into the local operation and maintenance database so as to update the solutions stored in the local operation and maintenance database.
In a third aspect, the present application discloses an electronic device, comprising:
a memory for storing a computer program;
and the processor is used for executing the computer program to realize the intelligent fault processing method based on the information system.
In a fourth aspect, the present application discloses a computer readable storage medium for storing a computer program, which when executed by a processor implements the foregoing intelligent fault handling method based on an information system.
In the application, firstly, a fault scene text input by a front end is received, analysis and calculation are carried out on the fault scene text, a local operation and maintenance database is matched based on an obtained scene subject word and a scene sentence vector, a corresponding matching result is generated, if the matching result represents that the matching is successful, a temporary knowledge base is generated based on a plurality of target matching texts matched in the local operation and maintenance database, a first solution is generated based on the temporary knowledge base through a preset language model, the first solution is returned to the client, if the matching result represents that the matching is failed, interaction is carried out with the front end through the preset language model based on preset priori knowledge, a second solution is generated by utilizing the supplemented fault scene text input by the front end, the second solution is returned to the client, and finally the first solution or the second solution is stored in the local operation and maintenance database, so that the solutions stored in the local operation and maintenance database are updated. Therefore, according to the intelligent fault processing method based on the information system, the input fault scene text can be analyzed and calculated to obtain scene subject words and scene vectors corresponding to the fault scene text, then the local operation and maintenance database is matched based on the obtained scene subject words and scene vectors, if the matching is successful, a plurality of target matching texts are obtained, a solution is generated through the plurality of target matching texts and returned to the front end, if the matching is failed, interaction is performed with the front end through a preset language model to obtain a fault scene text supplemented by the front end, and the solution is generated through the supplemented fault scene text and returned to the front end. In this way, after determining that a fault has occurred, the fault can be automatically identified and a corresponding solution generated to help the operation and maintenance personnel respond to the fault quickly through the solution.
Drawings
In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings that are required to be used in the embodiments or the description of the prior art will be briefly described below, and it is obvious that the drawings in the following description are only embodiments of the present application, and that other drawings can be obtained according to the provided drawings without inventive effort for a person skilled in the art.
FIG. 1 is a flow chart of an intelligent fault processing method based on an information system;
FIG. 2 is a timing diagram generated by one solution of the present disclosure;
FIG. 3 is a comparison of a prior art to and after application of the present disclosure;
FIG. 4 is a schematic diagram of an intelligent fault handling apparatus based on an information system according to the present disclosure;
fig. 5 is a block diagram of an electronic device according to the present disclosure.
Detailed Description
The following description of the embodiments of the present application will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present application, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the application without making any inventive effort, are intended to be within the scope of the application.
However, when the foreground service system fails, due to professional reasons, service personnel can hardly lock corresponding scenes in the project document quickly, obtain failure reasons and treatment suggestions, and also need to consult the background for the problem which can be solved by the simple operation of the foreground, so that the problem is highly dependent on operation and maintenance personnel, a great deal of time is consumed for solving the problem, and when the background personnel receives a fault alarm, because the emergency treatment tools are various in types, the emergency treatment tools are often required to be carried out across systems, platforms and equipment, the requirements on operation skills and operation experience are high, and the speed, efficiency and quality of the emergency treatment are difficult to ensure.
In order to solve the technical problems, the application provides an intelligent fault processing method, an intelligent fault processing device, intelligent fault processing equipment and an intelligent fault processing medium based on an information system, which can automatically identify faults after determining that the faults occur and generate corresponding solutions so as to help operation and maintenance personnel to respond to the faults rapidly through the solutions.
Referring to fig. 1, the embodiment of the application discloses an intelligent fault processing method based on an information system, which is applied to a bank information system and comprises the following steps:
And S11, receiving a fault scene text input by the front end, analyzing and calculating the fault scene text, matching the local operation and maintenance database based on the obtained scene subject words and scene sentence vectors, and generating a corresponding matching result.
In this embodiment, before the fault scene text input by the user at the front end is processed, a database for providing data support needs to be constructed, which specifically includes the following operations: and collecting historical data and real-time data of the bank information system to integrate the historical data and the real-time data so as to obtain a local operation and maintenance database. That is, in the bank information system, various information data of the bank information system, such as historical data and real-time data in terms of assets, hardware, systems, networks, logs, transactions, alarms and the like, need to be collected and integrated to generate a local operation and maintenance database for providing data support, and further, various rights of the system, such as system login rights and operation rights, can be integrated on the basis of the integrated data of the local database to form an operation and maintenance module based on big data, and the operation and maintenance module is used as a data support and execution tool of the bank information system.
In this embodiment, when a fault scene text input by a front-end user is received, the fault scene text needs to be analyzed, and the specific operation is as follows: receiving a fault scene text input by a front end, and inputting the fault scene text into a preset document theme generation model so as to extract scene subject words in the fault scene text through the preset document theme generation model; and inputting the fault scene text into a preset text conversion model so as to convert the fault scene text into scene sentence vectors through the preset text conversion model. That is, after receiving the fault scene text input by the user at the front end, the operation and maintenance knowledge base module needs to process the text, and based on the help document, the operation and maintenance experience document and various system component general technical documents existing in the system, the operation and maintenance knowledge base module builds a scene-operation and maintenance knowledge base by applying the design thought of scene, and can separate all text contents into two parts of scene and operation, so that all fault scene texts in the database can be associated to corresponding operations, and the method is specific: as shown in fig. 2, the received fault scene text may be input to a LDA (Latent Dirichlet Allocation) topic model technique for scene topic word extraction to obtain scene topic words corresponding to the input fault scene text, after the scene topic words of the fault scene text are obtained, each scene may be classified by the scene topic words, and as shown in fig. 2, the received fault scene text may be simultaneously input to a SimCSE (Simple Contrastive Learning of Sentence Embeddings) model to calculate sentence vectors of the fault scene text by means of a SimCSE model, and after the topic words or sentence vectors corresponding to the fault scene text are obtained, the obtained topic words and sentence vectors may be used as data supports of a scene-operation and maintenance knowledge base.
It should be noted that, after obtaining the scene subject words and sentence vectors, matching needs to be performed in the local operation and maintenance database based on the scene subject words and sentence vectors to determine whether the matching text can be matched with the corresponding matching text, which specifically includes the following steps: if the local operation and maintenance database is matched based on the scene subject words and the scene sentence vectors and a plurality of matching texts are not obtained, generating a matching result representing failure of matching; if the local operation and maintenance database is matched based on the scene subject words and the scene sentence vectors and a plurality of matching texts are obtained, a matching result representing successful matching is generated; respectively determining the similarity between the plurality of matching texts and the fault scene text, and sequencing the similarity values of the similarity to obtain a sequencing result; rejecting the matched text with the similarity value lower than a preset threshold value based on the sorting result to obtain a matched text after rejection; and extracting a preset number of matching texts with the similarity value not smaller than other matching texts from the matched texts after the removal so as to obtain a plurality of target matching texts. That is, the method can perform preliminary screening on a local operation and maintenance database according to the scene subject words, then perform matching according to sentence vectors input by a user and scenes of the preliminary screening results, finally obtain five matching results with highest correlation, then perform threshold judgment according to the matching results with highest correlation and the matching threshold input by the user, when the similarity is higher than the set matching threshold, it is indicated that the scene described by the user obtains matching in the knowledge base, at this time, a temporary knowledge base is formed based on the obtained five matching results, and it is required to be explained that when the input of the user does not obtain matching with higher responsiveness, the operation and maintenance knowledge base module does not perform any action, and finally, whether the matching result representing successful matching is generated is also required. Therefore, scene topics can be automatically extracted and scene sentence vector codes can be calculated through the LDA and SimCSE models, the knowledge base is automatically graded and layered on two levels of the topic words and the sentence meaning, the matching accuracy is improved, and the processing efficiency is effectively improved.
And step S12, if the matching result represents that the matching is successful, generating a temporary knowledge base based on a plurality of target matching texts matched from the local operation and maintenance database, generating a first solution based on the temporary knowledge base through a preset language model, and returning the first solution to the client.
In this embodiment, after the matching result is determined, a solution of response needs to be generated by processing the matching result, and a natural language processing module issues inspection and operation instructions in a natural language by combining a natural language processing (Natural Language Processing, NLP) technology with an operation and maintenance knowledge base, and converts the returned result into a natural language with a lower professional degree, so that on-duty operation and maintenance personnel with a lower degree of understanding on a corresponding system or business personnel with a limited professional level can directly reach a fault scene through a plurality of interactive consultations to enter a preset fault disposal process, thereby realizing quick inspection and disposal. The natural language processing module is realized based on a preset language model, and the specific generation process of the preset language model is as follows: adding a low-order bypass matrix for the ChatGLM model, and freezing the original weight of the ChatGLM model to perform fine tuning on the low-order bypass matrix and the low-order bypass matrix through a preset fine tuning data set so as to obtain the preset language model. In other words, in this embodiment, a ChatGLM (General Language Model, GLM) model is selected as an original model of a preset language model, in order to improve the processing performance of ChatGLM in the banking field, the ChatGLM may be directionally fine-tuned by a large language model Low-order adaptation method (Low-Rank Adaptation of Large Language Models, loRA) for the scenes of business banking, finance technology, IT operation and maintenance in finance industry, and the like. Specifically, the original structure of the ChatGLM is not changed, the low-order bypass matrix is only added in part of layers, all original weights of the ChatGLM are frozen when the model is fine-tuned, and only the low-order bypass matrix is trained, so that the calculated amount required by training is effectively reduced, and the phenomenon that preset knowledge is lost easily in a fine-tuned large language model is avoided. It should be noted that, the data in the preset fine tuning data set adopts solutions corresponding to common problems, alarms, regulations and the like in the banking industry, and special abbreviations in the banking industry, special synonyms and anti-ambiguities in the industry.
In this embodiment, if the matching is successful, a temporary knowledge base is required to be formed by five matching results of the matching, and as shown in fig. 2, the temporary knowledge base needs to be input into a preset language model, that is, through a fine-tuned ChatGLM model, so as to generate a first solution corresponding to the fault scene text input by the front-end user. It should be noted that the first solution is a solution that generates a second solution for the supplemental fault scenario text that is different from that entered by the front-end user.
It should be further noted that, for the generated solution, the solution may also be used as a data for training the ChatGLM, and the solution may be organized into an instruction-complex structure by a manual organization method, where instruction is a special word pair, and complex is a relationship between word pairs, and after the solution is converted into the data of the instruction-complex structure, the data may be input into the ChatGLM model for fine tuning.
And S13, if the matching result represents that the matching fails, interacting with the front end based on preset priori knowledge through the preset language model so as to generate a second solution by utilizing the supplemented fault scene text input by the front end, and returning the second solution to the client.
In this embodiment, if the matching fails, the description of the fault scene text input by the front-end user is not detailed enough, and interaction with the front-end user by using the preset language model is needed to guide the user to supplement the fault scene text, which has the following specific engineering: if the matching result represents that the matching fails, generating a text supplement guide sentence based on preset priori knowledge through the preset language model, and sending the text supplement guide sentence to the front end, so that after the front end receives the text supplement guide sentence, a fault scene text after supplement is input; analyzing and calculating the supplemented fault scene text to re-match the local operation and maintenance database so as to obtain a plurality of re-matched texts; generating a second solution based on the plurality of re-matched texts through the preset language model, and returning the second solution to the client. That is, when the matching fails, a text for guiding the front-end user to supplement the fault scene description, for example, "please see more details of the fault scene description," etc., needs to be generated based on preset priori knowledge, which is history data for training the preset language model, or history reply data, using the preset language model. After receiving the fault scene text supplemented by the front-end user, re-analyzing and calculating the supplemented fault scene text input by the front-end user to determine scene subject words and scene sentence vectors corresponding to the supplemented fault scene text, so as to match the scene subject words and the scene sentence vectors in the local operation and maintenance database again based on the scene subject words and the scene sentence vectors corresponding to the supplemented fault scene text, and re-generating a temporary knowledge base based on the matched matching result again to regenerate a second solution through the temporary knowledge base and feeding back the generated second solution to the front-end user.
And step S14, saving the first solution or the second solution to the local operation and maintenance database so as to update the solutions saved in the local operation and maintenance database.
In this embodiment, the generated first solution and the generated second solution may be saved in the local motion data base, so as to implement persistent saving of the solutions, so that after similar scene subject terms are obtained through analysis and calculation again, the saved corresponding solutions may be directly fed back to the front end.
Further, the intelligent fault processing method based on the information system further comprises the following steps: a process manual is generated based on the solutions stored in the local operation and maintenance database and the historical operation and maintenance data to feed back the process manual to the front end upon receiving a fault alert. That is, the existing solutions locally in the bank information system, the interaction situation with the user and the operation and maintenance big data information can be collected based on the fault handling manual module, and the handling manual for performing fault checking can be dynamically built, so that when an alarm or a fault occurs, parallel checking can be automatically initiated by using a correlation algorithm, fault reasons are intelligently analyzed, the handling manual is matched according to the possible fault reasons, so that a processor is guided to perform fault handling operation, and meanwhile, the influence range can be intelligently judged according to the fault situation, emergency initiation and handling suggestion are provided, and after the processor confirms, an emergency flow can be automatically initiated.
It can be seen that, in this embodiment, a fault scene text input by a front end is received first, analysis and calculation are performed on the fault scene text, so as to match a local operation and maintenance database based on an obtained scene subject word and a scene sentence vector, and generate a corresponding matching result, if the matching result characterizes that the matching is successful, a temporary knowledge base is generated based on a plurality of target matching texts matched from the local operation and maintenance database, a first solution is generated based on the temporary knowledge base through a preset language model, and the first solution is returned to the client, if the matching result characterizes that the matching is failed, interaction is performed with the front end through the preset language model based on preset priori knowledge, so as to generate a second solution by using the post-supplement fault scene text input by the front end, and return the second solution to the client, and finally, the first solution or the second solution is saved to the local operation and maintenance database, so as to update the solutions saved in the local operation and maintenance database. Therefore, according to the intelligent fault processing method based on the information system, the input fault scene text can be analyzed and calculated to obtain scene subject words and scene vectors corresponding to the fault scene text, then the local operation and maintenance database is matched based on the obtained scene subject words and scene vectors, if the matching is successful, a plurality of target matching texts are obtained, a solution is generated through the plurality of target matching texts and returned to the front end, if the matching is failed, interaction is performed with the front end through a preset language model to obtain a fault scene text supplemented by the front end, and the solution is generated through the supplemented fault scene text and returned to the front end. On the one hand, the scene subjects can be automatically extracted and the scene sentence vector codes can be calculated through the LDA and SimCSE models, and the knowledge base is automatically graded and layered on the two levels of the subject terms and the sentence meaning, so that the matching accuracy is improved, and the processing efficiency is effectively improved; on the other hand, through fine adjustment of the language model, the language model can better understand the input of a user so as to generate a reply with higher readability, and the use experience is effectively improved; in yet another aspect, upon determining a fault, the fault can be automatically identified and a corresponding solution generated to assist the service personnel in quickly responding to the fault via the solution.
Referring to fig. 3, the embodiment of the application discloses a traffic transmission method based on a stealth gateway, which comprises the following steps:
in the application, a fault disposal assistant can be constructed through a constructed operation and maintenance module, an operation and maintenance knowledge base module, a natural language processing module and a fault disposal manual module, and the constructed fault disposal assistant is applied to a bank information system, as shown in fig. 3, which is a comparison diagram before and after the fault disposal assistant is applied, wherein when a foreground finds that a service is unavailable, an abnormal situation can be described as a fault scene text, and the fault scene text is input into the fault disposal assistant, so that the fault disposal assistant processes the fault scene text through the operation and maintenance knowledge base module to extract scene subject words through an LDA subject model technology, calculates sentence vectors of the fault scene text through a SimCSE model, then matches the obtained scene subject words and sentence vectors in a local operation and maintenance database to obtain a plurality of matched texts, processes the obtained plurality of matched texts through the natural language processing module, generates a corresponding solution, namely a disposal proposal, synchronously informs the background, and ends a flow if the generated solution can solve the problem. If an alarm occurs in the background or the generated solution fails to solve the problem, then the fault scope needs to be automatically determined, and a corresponding process manual is provided by the fault handling manual module to provide a process recommendation by the process manual. For example, a service person may find out that the payment function is wrong, that is, may query various conventional inspection suggestions about the wrong payment through the fault handling assistant, which may include inspecting the network configuration of the terminal, inspecting the status of the user, and so on. If the problem can not be solved, the information can be timely fed back to the operation and maintenance personnel through an assistant. And the maintenance condition of the background can also help the business personnel to grasp the related condition through an assistant even if the maintenance condition is fed back to the front end, such as the interruption of the communication line of an operator, the interruption of the service of a third party and the like. As shown in fig. 3, the white rectangular box in the figure is a manual operation, and the gray rectangular box is an automatic operation that is performed automatically or only by manual confirmation. Therefore, the fault handling assistant constructed by the method can enable the processor to recognize the fault more quickly, avoid long-time waiting and unnecessary communication cost of the client, locate the fault quickly, improve emergency response capability and effectively improve processing efficiency.
Referring to fig. 4, the embodiment of the invention discloses an intelligent fault processing device based on an information system, which is applied to a bank information system and comprises:
the data matching module 11 is used for receiving the fault scene text input by the front end, analyzing and calculating the fault scene text, matching the local operation and maintenance database based on the obtained scene subject words and scene sentence vectors, and generating corresponding matching results;
a first scheme generating module 12, configured to generate a temporary knowledge base based on a plurality of target matching texts matched from the local operation and maintenance database if the matching result characterizes that the matching is successful, generate a first solution based on the temporary knowledge base through a preset language model, and return the first solution to the client;
a second solution generating module 13, configured to interact with the front end based on preset priori knowledge through the preset language model if the matching result indicates that the matching fails, so as to generate a second solution by using the post-supplement fault scene text input by the front end, and return the second solution to the client;
a solution updating module 14, configured to save the first solution or the second solution to the local operation and maintenance database, so as to update the solutions saved in the local operation and maintenance database.
Therefore, according to the intelligent fault processing method based on the information system, the input fault scene text can be analyzed and calculated to obtain scene subject words and scene vectors corresponding to the fault scene text, then the local operation and maintenance database is matched based on the obtained scene subject words and scene vectors, if the matching is successful, a plurality of target matching texts are obtained, a solution is generated through the plurality of target matching texts and returned to the front end, if the matching is failed, interaction is performed with the front end through a preset language model to obtain a fault scene text supplemented by the front end, and the solution is generated through the supplemented fault scene text and returned to the front end. On the one hand, the scene subjects can be automatically extracted and the scene sentence vector codes can be calculated through the LDA and SimCSE models, and the knowledge base is automatically graded and layered on the two levels of the subject terms and the sentence meaning, so that the matching accuracy is improved, and the processing efficiency is effectively improved; on the other hand, through fine adjustment of the language model, the language model can better understand the input of a user so as to generate a reply with higher readability, and the use experience is effectively improved; in yet another aspect, upon determining a fault, the fault can be automatically identified and a corresponding solution generated to assist the service personnel in quickly responding to the fault via the solution.
In some embodiments, the intelligent fault handling based on the information system may further include:
and the knowledge base construction unit is used for collecting historical data and real-time data of the bank information system so as to integrate the historical data and the real-time data to obtain a local operation and data database.
In some embodiments, the data matching module 11 may specifically include:
the text extraction unit is used for receiving fault scene texts input by the front end, inputting the fault scene texts into a preset document theme generation model and extracting scene subject words in the fault scene texts through the preset document theme generation model;
the data conversion unit is used for inputting the fault scene text into a preset text conversion model so as to convert the fault scene text into scene sentence vectors through the preset text conversion model.
In some embodiments, the data matching module 11 may specifically include:
the first matching result generating unit is used for generating a matching result representing failure of matching if the local operation and maintenance database is matched based on the scene subject words and the scene sentence vectors and a plurality of matching texts are not obtained;
The second matching result generating unit is used for generating a matching result representing successful matching if the local operation and maintenance database is matched based on the scene subject words and the scene sentence vectors and a plurality of matching texts are obtained;
the data ordering unit is used for respectively determining the similarity between the plurality of matching texts and the fault scene text and ordering the similarity values of the similarity to obtain an ordering result;
the data eliminating unit is used for eliminating the matched texts with the similarity value lower than a preset threshold value based on the sorting result so as to obtain eliminated matched texts;
and the data extraction unit is used for extracting a preset number of matching texts with the similarity value not smaller than other matching texts from the matched texts after the removal so as to obtain a plurality of target matching texts.
In some embodiments, the intelligent fault handling apparatus based on an information system may further include:
and the model fine tuning unit is used for adding a low-order bypass matrix for the ChatGLM model, freezing the original weight of the ChatGLM model, and carrying out fine tuning on the low-order bypass matrix through a preset fine tuning data set to obtain the preset language model.
In some embodiments, the second scheme generating module 13 may specifically include:
the interaction unit is used for generating a text supplement guide statement based on preset priori knowledge through the preset language model if the matching result represents that the matching is failed, and sending the text supplement guide statement to the front end, so that the front end receives the text supplement guide statement and then inputs a post-supplement fault scene text;
the data matching unit is used for analyzing and calculating the supplemented fault scene text so as to match the local operation and maintenance database again, so as to obtain a plurality of re-matched texts;
and the scheme generating unit is used for generating a second solution based on the plurality of re-matched texts through the preset language model and returning the second solution to the client.
In some embodiments, the intelligent fault handling apparatus based on an information system may further include:
and the data feedback unit is used for generating a processing manual based on the solutions stored in the local operation and maintenance database and the historical operation and maintenance data so as to feed back the processing manual to the front end after receiving the fault alarm.
Further, the embodiment of the present application further discloses an electronic device, and fig. 5 is a block diagram of an electronic device 20 according to an exemplary embodiment, where the content of the figure is not to be considered as any limitation on the scope of use of the present application.
Fig. 5 is a schematic structural diagram of an electronic device 20 according to an embodiment of the present application. The electronic device 20 may specifically include: at least one processor 21, at least one memory 22, a power supply 23, a communication interface 24, an input output interface 25, and a communication bus 26. Wherein the memory 22 is used for storing a computer program, which is loaded and executed by the processor 21 to implement the relevant steps in the information system-based intelligent fault handling method disclosed in any of the foregoing embodiments. In addition, the electronic device 20 in the present embodiment may be specifically an electronic computer.
In this embodiment, the power supply 23 is configured to provide an operating voltage for each hardware device on the electronic device 20; the communication interface 24 can create a data transmission channel between the electronic device 20 and an external device, and the communication protocol to be followed is any communication protocol applicable to the technical solution of the present application, which is not specifically limited herein; the input/output interface 25 is used for acquiring external input data or outputting external output data, and the specific interface type thereof may be selected according to the specific application requirement, which is not limited herein.
The memory 22 may be a carrier for storing resources, such as a read-only memory, a random access memory, a magnetic disk, or an optical disk, and the resources stored thereon may include an operating system 221, a computer program 222, and the like, and the storage may be temporary storage or permanent storage.
The operating system 221 is used for managing and controlling various hardware devices on the electronic device 20 and computer programs 222, which may be Windows Server, netware, unix, linux, etc. The computer program 222 may further include a computer program that can be used to perform other specific tasks in addition to the computer program that can be used to perform the information system-based intelligent fault handling method performed by the electronic device 20 as disclosed in any of the previous embodiments.
Further, the application also discloses a computer readable storage medium for storing a computer program; the computer program, when executed by the processor, implements the intelligent fault processing method based on the information system. For specific steps of the method, reference may be made to the corresponding contents disclosed in the foregoing embodiments, and no further description is given here.
In this specification, each embodiment is described in a progressive manner, and each embodiment is mainly described in a different point from other embodiments, so that the same or similar parts between the embodiments are referred to each other. For the device disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant points refer to the description of the method section.
Those of skill would further appreciate that the various illustrative elements and algorithm steps described in connection with the embodiments disclosed herein may be implemented as electronic hardware, computer software, or combinations of both, and that the various illustrative elements and steps are described above generally in terms of functionality in order to clearly illustrate the interchangeability of hardware and software. Whether such functionality is implemented as hardware or software depends upon the particular application and design constraints imposed on the solution. Skilled artisans may implement the described functionality in varying ways for each particular application, but such implementation decisions should not be interpreted as causing a departure from the scope of the present application.
The steps of a method or algorithm described in connection with the embodiments disclosed herein may be embodied directly in hardware, in a software module executed by a processor, or in a combination of the two. The software modules may be disposed in Random Access Memory (RAM), memory, read Only Memory (ROM), electrically programmable ROM, electrically erasable programmable ROM, registers, hard disk, a removable disk, a CD-ROM, or any other form of storage medium known in the art.
Finally, it is further noted that relational terms such as first and second, and the like are used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions. Moreover, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one … …" does not exclude the presence of other like elements in a process, method, article, or apparatus that comprises the element.
The foregoing has outlined rather broadly the more detailed description of the application in order that the detailed description of the application that follows may be better understood, and in order that the present principles and embodiments may be better understood; meanwhile, as those skilled in the art will have variations in the specific embodiments and application scope in accordance with the ideas of the present application, the present description should not be construed as limiting the present application in view of the above.
Claims (10)
1. An intelligent fault processing method based on an information system is characterized by being applied to a bank information system and comprising the following steps:
receiving a fault scene text input by a front end, analyzing and calculating the fault scene text, matching a local operation and maintenance database based on the obtained scene subject words and scene sentence vectors, and generating a corresponding matching result;
if the matching result represents successful matching, a temporary knowledge base is generated based on a plurality of target matching texts matched from the local operation and maintenance database, a first solution is generated based on the temporary knowledge base through a preset language model, and the first solution is returned to the client;
If the matching result represents that the matching fails, interacting with the front end based on preset priori knowledge through the preset language model to generate a second solution by utilizing the supplemented fault scene text input by the front end, and returning the second solution to the client;
and saving the first solution or the second solution to the local operation and maintenance database so as to update the solutions saved in the local operation and maintenance database.
2. The intelligent fault processing method based on the information system according to claim 1, wherein before the receiving the fault scene text input by the front end and analyzing and calculating the fault scene text to match the local operation and maintenance database based on the obtained scene subject term and the scene sentence vector and generate the corresponding matching result, further comprises:
and collecting historical data and real-time data of the bank information system to integrate the historical data and the real-time data so as to obtain a local operation and maintenance database.
3. The intelligent fault handling method based on the information system according to claim 1, wherein the receiving the fault scene text input by the front end and analyzing and calculating the fault scene text includes:
Receiving a fault scene text input by a front end, and inputting the fault scene text into a preset document theme generation model so as to extract scene subject words in the fault scene text through the preset document theme generation model;
and inputting the fault scene text into a preset text conversion model so as to convert the fault scene text into scene sentence vectors through the preset text conversion model.
4. The intelligent fault processing method based on the information system according to claim 1, wherein the matching the local operation and maintenance database based on the obtained scene subject term and the scene sentence vector, and generating a corresponding matching result, includes:
if the local operation and maintenance database is matched based on the scene subject words and the scene sentence vectors and a plurality of matching texts are not obtained, generating a matching result representing failure of matching;
if the local operation and maintenance database is matched based on the scene subject words and the scene sentence vectors and a plurality of matching texts are obtained, a matching result representing successful matching is generated; respectively determining the similarity between the plurality of matching texts and the fault scene text, and sequencing the similarity values of the similarity to obtain a sequencing result; rejecting the matched text with the similarity value lower than a preset threshold value based on the sorting result to obtain a matched text after rejection; and extracting a preset number of matching texts with the similarity value not smaller than other matching texts from the matched texts after the removal so as to obtain a plurality of target matching texts.
5. The intelligent fault handling method based on an information system according to claim 1, wherein if the matching result characterizes that the matching is successful, generating a temporary knowledge base based on a plurality of target matching texts matched from the local operation and maintenance database, generating a first solution based on the temporary knowledge base through a preset language model, and returning the first solution to the client, further comprising:
adding a low-order bypass matrix for the ChatGLM model, and freezing the original weight of the ChatGLM model to perform fine tuning on the low-order bypass matrix and the low-order bypass matrix through a preset fine tuning data set so as to obtain the preset language model.
6. The information system-based intelligent fault handling method according to claim 1, wherein if the matching result characterizes a matching failure, interacting with the front end based on preset priori knowledge through the preset language model to generate a second solution using the post-supplement fault scenario text input by the front end, and returning the second solution to the client, comprising:
if the matching result represents that the matching fails, generating a text supplement guide sentence based on preset priori knowledge through the preset language model, and sending the text supplement guide sentence to the front end, so that after the front end receives the text supplement guide sentence, a fault scene text after supplement is input;
Analyzing and calculating the supplemented fault scene text to re-match the local operation and maintenance database so as to obtain a plurality of re-matched texts;
generating a second solution based on the plurality of re-matched texts through the preset language model, and returning the second solution to the client.
7. The information system-based intelligent fault handling method of any of claims 1 to 6, further comprising:
a process manual is generated based on the solutions stored in the local operation and maintenance database and the historical operation and maintenance data to feed back the process manual to the front end upon receiving a fault alert.
8. An intelligent fault handling device based on an information system, which is characterized in that the intelligent fault handling device is applied to a bank information system and comprises:
the data matching module is used for receiving the fault scene text input by the front end, analyzing and calculating the fault scene text, matching the local operation and maintenance database based on the obtained scene subject words and scene sentence vectors, and generating corresponding matching results;
the first scheme generation module is used for generating a temporary knowledge base based on a plurality of target matching texts matched from the local operation and maintenance database if the matching result represents that the matching is successful, generating a first solution based on the temporary knowledge base through a preset language model, and returning the first solution to the client;
The second scheme generating module is used for interacting with the front end based on preset priori knowledge through the preset language model if the matching result represents that the matching is failed, so as to generate a second solution by utilizing the supplemented fault scene text input by the front end, and returning the second solution to the client;
and the scheme updating module is used for storing the first solution or the second solution into the local operation and maintenance database so as to update the solutions stored in the local operation and maintenance database.
9. An electronic device, comprising:
a memory for storing a computer program;
a processor for executing the computer program to implement the information system-based intelligent fault handling method as claimed in any one of claims 1 to 7.
10. A computer readable storage medium for storing a computer program which, when executed by a processor, implements the information system based intelligent fault handling method of any of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311187892.6A CN117234774A (en) | 2023-09-14 | 2023-09-14 | Intelligent fault processing method, device, equipment and medium based on information system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311187892.6A CN117234774A (en) | 2023-09-14 | 2023-09-14 | Intelligent fault processing method, device, equipment and medium based on information system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN117234774A true CN117234774A (en) | 2023-12-15 |
Family
ID=89088950
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311187892.6A Pending CN117234774A (en) | 2023-09-14 | 2023-09-14 | Intelligent fault processing method, device, equipment and medium based on information system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117234774A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117573852A (en) * | 2024-01-17 | 2024-02-20 | 深圳市伊登软件有限公司 | Task processing method, device, equipment and medium for intelligent office |
| CN117591659A (en) * | 2024-01-18 | 2024-02-23 | 卓望数码技术（深圳）有限公司 | Information processing method, device, equipment and medium based on ChatGLM operation and maintenance scene |
-
2023
- 2023-09-14 CN CN202311187892.6A patent/CN117234774A/en active Pending
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117573852A (en) * | 2024-01-17 | 2024-02-20 | 深圳市伊登软件有限公司 | Task processing method, device, equipment and medium for intelligent office |
| CN117573852B (en) * | 2024-01-17 | 2024-03-22 | 深圳市伊登软件有限公司 | Task processing method, device, equipment and medium for intelligent office |
| CN117591659A (en) * | 2024-01-18 | 2024-02-23 | 卓望数码技术（深圳）有限公司 | Information processing method, device, equipment and medium based on ChatGLM operation and maintenance scene |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN117234774A (en) | Intelligent fault processing method, device, equipment and medium based on information system | |
| CN109760041B (en) | Chat robot-based cloud management system and operation method thereof | |
| CN116737908A (en) | Knowledge question-answering method, device, equipment and storage medium | |
| EP4113357A1 (en) | Method and apparatus for recognizing entity, electronic device and storage medium | |
| CN111178537B (en) | Feature extraction model training method and device | |
| CN111125326A (en) | Method, device, medium and electronic equipment for realizing man-machine conversation | |
| CN113516340A (en) | Intelligent work order pushing method and device | |
| CN115168562A (en) | Method, device, equipment and medium for constructing intelligent question-answering system | |
| CN114048731A (en) | Text processing method and device, storage medium and computer equipment | |
| CN113591463A (en) | Intention recognition method and device, electronic equipment and storage medium | |
| CN112667791A (en) | Latent event prediction method, device, equipment and storage medium | |
| CN114372082B (en) | Data query method and device based on artificial intelligence, electronic equipment and medium | |
| CN116340831B (en) | Information classification method and device, electronic equipment and storage medium | |
| EP4254256A1 (en) | Spoken language processing method and apparatus, electronic device, and storage medium | |
| CN115906797A (en) | Text entity alignment method, device, equipment and medium | |
| CN113360672B (en) | Method, apparatus, device, medium and product for generating knowledge graph | |
| CN115221892A (en) | Work order data processing method and device, storage medium and electronic equipment | |
| CN114912541A (en) | Classification method, classification device, electronic equipment and storage medium | |
| AU2019229462B2 (en) | Monitoring an artificial intelligence (ai) based process | |
| CN114417881A (en) | Sensitive word detection method and device, electronic equipment and storage medium | |
| CN114141236A (en) | Language model updating method and device, electronic equipment and storage medium | |
| CN113743112A (en) | Keyword extraction method and device, electronic equipment and readable storage medium | |
| CN116011599B (en) | Intelligent reservation method and system for biological laboratory | |
| CN111091011A (en) | Domain prediction method, domain prediction device and electronic equipment | |
| CN114492456B (en) | Text generation method, model training method, device, electronic equipment and medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |