Information processing method based on operation and maintenance system

Abstract

The invention discloses an information processing method based on an operation and maintenance system and relates to the field of information technology management. The problem that safety is not high when an operation and maintenance robot processes abnormal events in an operation and maintenance system is solved, and the method comprises the steps that S1, based on historical fault data and user type data, a text data object processing model is obtained through training; s2, obtaining target text information, and converting the target text information into formatted text data; s3, analyzing and identifying the formatted text data to obtain analyzed text data; s4, inputting the analyzed text data into the text data object processing model to obtain information processing results corresponding to different object types; s5, performing identity recognition on the target object, and outputting an information processing result corresponding to the target object; and S6, sending an information processing result to a related department and a manager so as to process a fault corresponding to the target text information. The method has a good prospect in the field of operation and maintenance management.

Description

Technical Field

[0001] The present invention relates to the field of information technology management, and in particular to an information processing method based on an operation and maintenance system. Background Art

[0002] With the continuous development of technologies such as artificial intelligence, the application of AI-based robots in the operation and maintenance systems of telecom operators is increasing. Using operation and maintenance robots to handle abnormal events can significantly improve the efficiency and quality of abnormal event handling.

[0003] However, in current related technologies, when handling abnormal events, maintenance robots usually use HTML5 page responses to feedback and display the processing results. This method has an important disadvantage: HTML5 page responses push the processing results uniformly to all users in the group, which threatens the security of maintenance robots when handling abnormal events. The fundamental reason is that HTML5 page responses cannot achieve personalized information transmission, resulting in an increased risk of information leakage, thus posing a potential threat to system security. Summary of the invention

[0004] In order to solve the problem that the operation and maintenance robot has low security when handling abnormal events in the operation and maintenance system, the present invention provides the following solution: an information processing method based on the operation and maintenance system, the method comprising:

[0005] S1. Based on historical fault data and user type data, a text data object processing model is trained;

[0006] S2, obtaining target text information, and converting the target text information into formatted text data;

[0007] S3, analyzing and identifying the formatted text data to obtain analyzed text data;

[0008] S4, inputting the analyzed text data into the text data object processing model to obtain information processing results corresponding to different object types;

[0009] S5, identify the target object and output the information processing result corresponding to the target object;

[0010] S6. Send the information processing result to the relevant departments and managers so as to process the fault corresponding to the target text information.

[0011] Further, in S1, the historical fault data includes: network connection failure data, bandwidth shortage data, equipment failure data, software upgrade error data and security problem data.

[0012] Furthermore, in S1, the user type data is data on user identity classification obtained after classification using a k-nearest neighbor algorithm.

[0013] Furthermore, S2 specifically includes:

[0014] S21, removing meaningless text marks from the target text information to obtain normalized text information;

[0015] S22, performing stem extraction on the normalized text information to obtain normalized stem data;

[0016] S23. One-hot encode the normalized stem data to obtain the formatted text data.

[0017] Furthermore, S3 specifically includes:

[0018] S31, performing text cleaning on the formatted text data to obtain cleaned text data;

[0019] S32, performing word segmentation on the cleaned text data to obtain split words or phrases;

[0020] S33, performing part-of-speech tagging on the split words or phrases to obtain the tagged words or phrases;

[0021] S34, performing named entity recognition on the annotated words or phrases to identify entities with specific meanings;

[0022] S35, performing grammatical analysis on the entity of the specific meaning, analyzing the structure and grammatical relationship of the entity;

[0023] S36, performing semantic analysis on the entities of the specific meaning, analyzing the meanings and semantic relationships of the entities;

[0024] S37. Perform sentiment analysis on the cleaned text data according to the structure and grammatical relationship of the entity and the meaning and semantic relationship of the entity to obtain analyzed text data.

[0025] Furthermore, S5 specifically includes:

[0026] S51, identifying the target object to obtain rights information;

[0027] S52, determining the object type to which the claim information belongs according to the claim information, thereby outputting the corresponding information processing result;

[0028] S53: If the object type to which the target object belongs cannot obtain an information response, directly reply to the user that he has no viewing permission.

[0029] The beneficial effects of the present invention are:

[0030] 1. When the operation and maintenance robot in the operation and maintenance system processes fault information, it uses the pre-collected historical fault data and user type data to train the text data processing model offline, which reduces manual intervention and improves system operation efficiency.

[0031] 2 In the operation and maintenance system, after the operation and maintenance robot processes the fault information, it converts the collected target text information into formatted text data, and after analysis and recognition, generates analysis results for expressing fault information query requests. This process improves the automation, efficiency and accuracy of fault information processing.

[0032] 3. In the operation and maintenance system, after the operation and maintenance robot processes the fault information, it inputs the analyzed text data into the pre-trained text data processing model to obtain the fault information processing results for different objects. Then, the target object is identified and the corresponding information processing results are output according to its identity information. This improves the personalization and security of fault information processing. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] Figure 1 It is a flow chart of an information processing method based on an operation and maintenance system;

[0034] Figure 2 This is a schematic diagram of the operation and maintenance system. DETAILED DESCRIPTION

[0035] In order to make the technical solutions and advantages of the embodiments of the present invention more clearly understood, the exemplary embodiments of the present invention are further described in detail below in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than an exhaustive list of all the embodiments. It should be noted that the embodiments of the present invention and the features in the embodiments can be combined with each other without conflict.

[0036] Example: Combination Figure 1 - Figure 2 This embodiment describes an information processing method based on an operation and maintenance system, the method comprising:

[0037] S1. Based on historical fault data and user type data, a text data object processing model is trained;

[0038] S2, obtaining target text information, and converting the target text information into formatted text data;

[0039] S3, analyzing and identifying the formatted text data to obtain analyzed text data;

[0040] S4, inputting the analyzed text data into the text data object processing model to obtain information processing results corresponding to different object types;

[0041] S5, identify the target object and output the information processing result corresponding to the target object;

[0042] S6. Send the information processing result to the relevant departments and managers so as to process the fault corresponding to the target text information.

[0043] Specifically, through Figure 2 It can be seen that the operation and maintenance system includes target objects, operation and maintenance robots and related departments.

[0044] Among them, the target object is used to send the target text information to the operation and maintenance robot.

[0045] After the operation and maintenance robot is used to execute the information processing method of steps 1-6 above, it forwards the information processing results to the target object and relevant departments.

[0046] To illustrate this solution, in a certain system abnormality event, the group includes three people, A, B, and C. According to the known identity classification, they are the person in charge of the business system, the person in charge of the infrastructure, and the leader. When A@the robot, a message is sent to describe the failure of the business system, for example: "The business system has failed during operation, please check whether the basic equipment has failed / how is the progress of infrastructure repair?" Or, "The business system has failed during operation, please check the specific situation." The robot will display the business interface and the progress of the failure, but will not display the detailed failure details to avoid A's difficulty in understanding. Instead, it will display summary information such as the progress of the failure maintenance so that A can understand the progress of the business system recovery.

[0047] If A does not request to check the infrastructure status, the robot does not need to display the fault progress. Then, the robot mentions B, prompts him to participate in the fault maintenance, and presents the infrastructure interface and A's requirements. Since B knows that A pays more attention to the progress, he can speed up the maintenance in a targeted manner.

[0048] In S1, the historical fault data includes: network connection failure data, bandwidth shortage data, equipment failure data, software upgrade error data and security problem data.

[0049] In S1, the user type data is data on user identity classification obtained after classification by a k-nearest neighbor algorithm.

[0050] Specifically, the k-nearest neighbor algorithm includes:

[0051] 1. Calculate the distance between the known user category and the current user, that is, calculate the similarity or distance between them based on user characteristics.

[0052] 2. Sort the calculated distances in ascending order and find the first k users whose distances are closest to the current user.

[0053] 3. Select k users with the smallest distance to the current user as neighbors.

[0054] 4. Count the frequency of the categories to which these k neighbors belong, that is, which category appears the most times.

[0055] 5. The category with the highest frequency is taken as the predicted category of the current user, that is, the current user and these neighbors are considered to belong to the same category.

[0056] Through this algorithm, users in a group can be effectively authenticated and classified to ensure that they can obtain HTML5 information that matches their identity characteristics.

[0057] S2 specifically includes:

[0058] S21, removing meaningless text marks from the target text information to obtain normalized text information;

[0059] S22, performing stem extraction on the normalized text information to obtain normalized stem data;

[0060] S23. One-hot encode the normalized stem data to obtain the formatted text data.

[0061] S3 specifically includes:

[0062] S31, performing text cleaning on the formatted text data to obtain cleaned text data;

[0063] S32, performing word segmentation on the cleaned text data to obtain split words or phrases;

[0064] S33, performing part-of-speech tagging on the split words or phrases to obtain the tagged words or phrases;

[0065] S34, performing named entity recognition on the annotated words or phrases to identify entities with specific meanings;

[0066] S35, performing grammatical analysis on the entity of the specific meaning, analyzing the structure and grammatical relationship of the entity;

[0067] S36, performing semantic analysis on the entities of the specific meaning, analyzing the meanings and semantic relationships of the entities;

[0068] S37. Perform sentiment analysis on the cleaned text data according to the structure and grammatical relationship of the entity and the meaning and semantic relationship of the entity to obtain analyzed text data.

[0069] Specifically, the text cleaning includes: removing special characters, punctuation marks, numbers and HTML tags, and converting to lowercase letter format.

[0070] Word segmentation refers to the process of breaking text into words, phrases, or smaller units, called tags or tokens. Word segmentation can be performed using rule-based methods, such as splitting based on spaces or punctuation marks, or using machine learning methods, such as word segmentation based on statistical models or neural networks.

[0071] Part-of-speech tagging refers to tagging each word with its part of speech in a sentence, such as noun, verb, adjective, etc. Part-of-speech tagging can be performed using rule-based methods, such as tagging based on dictionaries and rules, or using machine learning methods, such as tagging based on hidden Markov models or conditional random fields.

[0072] Named entity recognition refers to identifying entities with specific meanings in text, such as names of people, places, names of organizations, etc. Named entity recognition can use rule-based methods, such as matching based on dictionaries and rules, or machine learning methods, such as recognition based on conditional random fields or recursive neural networks.

[0073] The grammatical analysis refers to analyzing the structure and grammatical relations of a sentence, such as subject-verb-object structure, modification relations, etc. The grammatical analysis can be performed using rule-based methods, such as analysis based on context-free grammar, or machine learning methods, such as dependency syntactic analysis based on transition or graph algorithms.

[0074] The semantic analysis refers to understanding the meaning and semantic relationship of a sentence, such as word sense disambiguation, coreference resolution, relationship extraction, etc. Semantic analysis can use rule-based methods, such as analysis based on semantic role labeling, or machine learning methods, such as semantic role labeling or relationship classification based on neural networks.

[0075] Sentiment analysis refers to analyzing the sentiment tendency of a text, such as positive, negative or neutral. Sentiment analysis can use rule-based methods, such as sentiment calculation based on a dictionary, or machine learning methods, such as sentiment classification based on a support vector machine or a recurrent neural network.

[0076] S5 specifically includes:

[0077] S51, identifying the target object to obtain rights information;

[0078] S52, determining the object type to which the claim information belongs according to the claim information, thereby outputting the corresponding information processing result;

[0079] S53: If the object type to which the target object belongs cannot obtain an information response, directly reply to the user that he has no viewing permission.

[0080] Although the present invention has been described according to a limited number of embodiments, it will be apparent to those skilled in the art, with the benefit of the above description, that other embodiments may be envisioned within the scope of the invention thus described. In addition, it should be noted that the language used in this specification is selected primarily for readability and teaching purposes, rather than for explaining or defining the subject matter of the present invention. Therefore, many modifications and variations will be apparent to those skilled in the art without departing from the scope and spirit of the appended claims. The disclosure of the present invention is illustrative, not restrictive, with respect to the scope of the present invention, which is defined by the appended claims.

Claims

1. An information processing method based on an operation and maintenance system, characterized in that: The method comprises: S1. Based on historical fault data and user type data, a text data object processing model is trained; S2, obtaining target text information, and converting the target text information into formatted text data; S3, analyzing and identifying the formatted text data to obtain analyzed text data; S4, inputting the analyzed text data into the text data object processing model to obtain information processing results corresponding to different object types; S5, identify the target object and output the information processing result corresponding to the target object; S6. Send the information processing result to the relevant departments and managers so as to process the fault corresponding to the target text information.

2. The information processing method based on the operation and maintenance system according to claim 1, characterized in that: In S1, the historical fault data includes: network connection failure data, bandwidth shortage data, equipment failure data, software upgrade error data and security problem data.

3. The information processing method based on the operation and maintenance system according to claim 1, characterized in that: In S1, the user type data is data on user identity classification obtained after classification by a k-nearest neighbor algorithm.

4. The information processing method based on the operation and maintenance system according to claim 1, characterized in that: S2 specifically includes: S21, removing meaningless text marks from the target text information to obtain normalized text information; S22, performing stem extraction on the normalized text information to obtain normalized stem data; S23. One-hot encode the normalized stem data to obtain the formatted text data.

5. The information processing method based on the operation and maintenance system according to claim 1, characterized in that: S3 specifically includes: S31, performing text cleaning on the formatted text data to obtain cleaned text data; S32, performing word segmentation on the cleaned text data to obtain split words or phrases; S33, performing part-of-speech tagging on the split words or phrases to obtain the tagged words or phrases; S34, performing named entity recognition on the annotated words or phrases to identify entities with specific meanings; S35, performing grammatical analysis on the entity of the specific meaning, analyzing the structure and grammatical relationship of the entity; S36, performing semantic analysis on the entities of the specific meaning, analyzing the meanings and semantic relationships of the entities; S37. Perform sentiment analysis on the cleaned text data according to the structure and grammatical relationship of the entity and the meaning and semantic relationship of the entity to obtain analyzed text data.

6. The information processing method based on the operation and maintenance system according to claim 1, characterized in that: S5 specifically includes: S51, identifying the target object to obtain rights information; S52, determining the object type to which the claim information belongs according to the claim information, thereby outputting the corresponding information processing result; S53: If the object type to which the target object belongs cannot obtain an information response, directly reply to the user that he has no viewing permission.

Images