Data processing method and device, equipment and storage medium

Abstract

The embodiment of the invention provides a data processing method and device, equipment and a storage medium. The method comprises the following steps: acquiring customer complaint data to be processed, wherein the customer complaint data to be processed comprises at least one piece of customer complaint information from at least one service channel; for each piece of customer complaint information, determining entity data involved in the customer complaint information and the emergency degree of a user corresponding to the customer complaint information; according to the entity data and the emergency degree, alarm information corresponding to the customer complaint information is generated, and the alarm information reflects the alarm level of the customer complaint information and attribute information related to alarm; and the alarm information corresponding to the customer complaint information is pushed to the corresponding operation and maintenance department, so that the effect of quickly solving the customer complaint information is achieved.

Description

Technical Field

[0001] The present application relates to the field of operation and maintenance guarantee technology, and in particular to a data processing method, device, equipment and storage medium. Background Art

[0002] In the daily operation and maintenance work of the data center, we face the initial stage with low sensitivity to faults. At this time, the system data has not yet shown obvious fluctuations, has not reached the alarm judgment standard, and the system has not generated clear alarms. However, various abnormal situations such as interruptions and no response have occurred during customer use. Therefore, customers’ feedback on fault problems is more timely.

[0003] Currently, customer complaint information is basically handled manually, involving different channels and different types of issues, resulting in extremely high operating costs, and the timeliness and accuracy of problem handling are insufficient.

[0004] Therefore, how to quickly resolve and respond to customer complaints has become a technical problem that needs to be solved urgently. Summary of the invention

[0005] The embodiments of the present application provide a data processing method, apparatus, device and storage medium to achieve the effect of quickly resolving and responding to customer complaint information.

[0006] In a first aspect, an embodiment of the present application provides a data processing method, including:

[0007] Acquire customer complaint data to be processed, where the customer complaint data to be processed includes: at least one customer complaint information from at least one service channel;

[0008] For each customer complaint information, determine the urgency of the entity data involved in the customer complaint information and the user corresponding to the customer complaint information;

[0009] Generate alarm information corresponding to the customer complaint information according to the entity data and the urgency level, wherein the alarm information reflects the alarm level of the customer complaint information and attribute information involved in the alarm;

[0010] The alarm information corresponding to the customer complaint information is pushed to the corresponding operation and maintenance department.

[0011] In a possible implementation manner, determining the entity data involved in the customer complaint information includes:

[0012] Inputting the customer complaint information into a pre-trained word vector model to obtain optimized customer complaint information;

[0013] The optimized customer complaint information is input into a pre-trained named entity recognition model to obtain the entity data, which includes: target application name, target function point, time, and location.

[0014] In a possible implementation, determining the urgency level corresponding to the customer complaint information includes:

[0015] The customer complaint information is input into a pre-trained sentiment analysis model to obtain the urgency level, wherein the sentiment analysis model is trained based on at least one historical customer complaint information and the urgency level corresponding to the at least one historical customer complaint information.

[0016] In a possible implementation manner, generating the warning information corresponding to the customer complaint information according to the entity data and the urgency level includes:

[0017] Determining the alarm level in the alarm information according to the urgency and the application level corresponding to the target application name;

[0018] According to the target application name, the target function point, the time, and the location, attribute information related to the alarm in the alarm information is determined.

[0019] In a possible implementation manner, before pushing the alarm information corresponding to the customer complaint information to the corresponding operation and maintenance department, the method further includes:

[0020] Determine a target application system corresponding to the target application name in at least one application system;

[0021] A target application module corresponding to the target function point is determined from at least one application module in the target application system, and each application module corresponds to an operation and maintenance department.

[0022] In a possible implementation manner, before determining the entity data involved in the customer complaint information and the urgency of the user corresponding to the customer complaint information, the method further includes:

[0023] The customer complaint information is entity marked, and the marked customer complaint information is converted into a preset data format.

[0024] In a possible implementation, the method further includes:

[0025] Obtain feedback from operation and maintenance personnel regarding alarm information;

[0026] The effectiveness and accuracy of the warning information are determined based on the feedback information.

[0027] In a second aspect, an embodiment of the present application provides a data processing device, including:

[0028] An acquisition module, used to acquire customer complaint data to be processed, wherein the customer complaint data to be processed includes: at least one customer complaint information from at least one service channel;

[0029] A determination module, used to determine, for each customer complaint information, the urgency of the entity data involved in the customer complaint information and the user corresponding to the customer complaint information;

[0030] A generating module, configured to generate alarm information corresponding to the customer complaint information according to the entity data and the urgency level, wherein the alarm information reflects the alarm level of the customer complaint information and attribute information involved in the alarm;

[0031] The sending module is used to push the alarm information corresponding to the customer complaint information to the corresponding operation and maintenance department.

[0032] In a possible implementation manner, the determination module determines the entity data involved in the customer complaint information, specifically for:

[0033] Inputting the customer complaint information into a pre-trained word vector model to obtain optimized customer complaint information;

[0034] The optimized customer complaint information is input into a pre-trained named entity recognition model to obtain the entity data, which includes: target application name, target function point, time, and location.

[0035] In a possible implementation, the determination module determines the urgency level corresponding to the customer complaint information, specifically for:

[0036] The customer complaint information is input into a pre-trained sentiment analysis model to obtain the urgency level, wherein the sentiment analysis model is trained based on at least one historical customer complaint information and the urgency level corresponding to the at least one historical customer complaint information.

[0037] In a possible implementation manner, the generating module is specifically used to:

[0038] Determining the alarm level in the alarm information according to the urgency and the application level corresponding to the target application name;

[0039] According to the target application name, the target function point, the time, and the location, attribute information related to the alarm in the alarm information is determined.

[0040] In a possible implementation manner, before pushing the alarm information corresponding to the customer complaint information to the corresponding operation and maintenance department, the determining module is further used to:

[0041] Determine a target application system corresponding to the target application name in at least one application system;

[0042] A target application module corresponding to the target function point is determined from at least one application module in the target application system, and each application module corresponds to an operation and maintenance department.

[0043] In a possible implementation manner, before determining the entity data involved in the customer complaint information and the urgency of the user corresponding to the customer complaint information, the determination module is further configured to:

[0044] The customer complaint information is entity marked, and the marked customer complaint information is converted into a preset data format.

[0045] In a possible implementation manner, the determining module is further configured to:

[0046] Obtain feedback from operation and maintenance personnel regarding alarm information;

[0047] The effectiveness and accuracy of the warning information are determined based on the feedback information.

[0048] In a third aspect, an embodiment of the present application provides an electronic device, including: a memory, a processor;

[0049] The memory stores computer-executable instructions;

[0050] The processor executes the computer-executable instructions stored in the memory, so that the processor executes the above first aspect and / or various possible implementations of the first aspect.

[0051] In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium, in which computer-executable instructions are stored. When the computer-executable instructions are executed by a processor, they are used to implement the first aspect above and / or various possible implementations of the first aspect.

[0052] In a fifth aspect, an embodiment of the present application provides a computer program product, including a computer program, which, when executed by a processor, implements the above first aspect and / or various possible implementation methods of the first aspect.

[0053] The data processing method, apparatus, device and storage medium provided in the embodiments of the present application obtain customer complaint data to be processed, and the customer complaint data to be processed includes: at least one customer complaint information from at least one service channel; for each customer complaint information, determine the entity data involved in the customer complaint information and the urgency of the user corresponding to the customer complaint information; generate alarm information corresponding to the customer complaint information according to the entity data and the urgency, and the alarm information reflects the alarm level of the customer complaint information and the attribute information involved in the alarm; push the alarm information corresponding to the customer complaint information to the corresponding operation and maintenance department, so as to achieve the effect of quickly executing the solution to the customer complaint information. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present application and, together with the description, serve to explain the principles of the present application.

[0055] Figure 1 A schematic diagram of the system architecture involved in the fault location method provided in this application;

[0056] Figure 2 Schematic diagram of the data processing method provided for this application Figure 1 ;

[0057] Figure 3 Schematic diagram of the data processing method provided for this application Figure 2 ;

[0058] Figure 4 Schematic diagram of the data processing method provided for this application Figure 3 ;

[0059] Figure 5 A schematic diagram of the structure of the data processing device provided by this application;

[0060] Figure 6 A schematic diagram of the structure of the electronic device provided in this application.

[0061] The above drawings have shown clear embodiments of the present application, which will be described in more detail later. These drawings and text descriptions are not intended to limit the scope of the present application in any way, but to illustrate the concept of the present application to those skilled in the art by referring to specific embodiments. DETAILED DESCRIPTION

[0062] Exemplary embodiments will be described in detail herein, examples of which are shown in the accompanying drawings. When the following description refers to the drawings, the same numbers in different drawings represent the same or similar elements unless otherwise indicated. The implementations described in the following exemplary embodiments do not represent all implementations consistent with the present application. Instead, they are merely examples of devices and methods consistent with some aspects of the present application as detailed in the appended claims.

[0063] First, the terms used in this application are explained:

[0064] Named Entity Recognition (NER): A natural language processing task used to identify entities with specific meanings in text (such as names of people, places, places, companies, dates, etc.);

[0065] Preprocessing: Standardize the original customer complaint data, including denoising, word segmentation, part-of-speech tagging, etc.

[0066] Kafka: A high-throughput, low-latency distributed messaging system, often used to build real-time data pipelines. In this application, it is used to collect and transmit customer complaint data in real time.

[0067] Spark Structured Streaming: It provides a powerful streaming data processing interface and is used in the relevant systems of this application to build end-to-end data stream applications for real-time analysis and processing.

[0068] Jieba - Python library for Chinese word segmentation.

[0069] Natural Language Toolkit (NLTK): Python library for English word segmentation;

[0070] Spacy: a natural language processing library for preprocessing English customer complaint texts;

[0071] BIO / BIOES: a labeling system for sequence annotation, used for entity boundary marking in named entity recognition tasks;

[0072] Bidirectional Encoder Representations from Transformers (BERT): A pre-trained language model that captures contextual information through a bidirectional transformer; used for text feature extraction and generating context-sensitive representations;

[0073] Robustly optimized BERT approach (RoBERTa): An optimized version of BERT that better captures text features and is used to improve the quality of text representation to enhance named entity recognition.

[0074] DistilBERT: A lightweight version of the BERT model that provides performance close to BERT while reducing model parameters and computational overhead.

[0075] Bidirectional Long Short-Term Memory (BiLSTM): It can capture the forward and backward dependencies of sequence data and is used to extract sequence features from customer complaint texts.

[0076] Conditional Random Field (CRF): used for sequence labeling tasks to capture dependencies between labels; combined with BiLSTM, it is used for serialized labeling of named entities in text;

[0077] Horovod - framework for distributed deep learning, supporting parallel training of multiple deep learning frameworks.

[0078] Levenshtein Distance (LEV) algorithm: calculates the minimum number of edit operations required to convert between two strings; used for fuzzy matching of entities in customer complaint texts and backend system modules;

[0079] N-gram: A model that divides text into n consecutive words or characters, used to capture local features of text; in this application, it is used for fuzzy matching and text similarity calculation to determine the relevant system modules in customer complaints;

[0080] Secondly, the technical background technology involved in this application is described as follows:

[0081] In the daily operation and maintenance work of the data center, we face the initial stage with low sensitivity to faults. At this time, the system data has not yet shown obvious fluctuations, has not reached the alarm judgment standard, and the system has not generated clear alarms. However, various abnormal situations such as interruptions and no response have occurred during customer use. Therefore, customer feedback on fault problems may be more timely.

[0082] At the same time, customer complaints from remote businesses are currently handled basically manually, involving different channels and different types of issues, resulting in extremely high operating costs, and the timeliness and accuracy of problem handling are insufficient.

[0083] In order to solve the above technical problems, this application designs a data processing solution for converting customer complaint information into alarms, which can realize the immediate conversion of fault problems reported by customers into system alarms, and connect them to the data center - integrated operation and maintenance platform - monitoring center - fault diagnosis module for display.

[0084] Furthermore, this data processing solution can enable backend operation and maintenance personnel to obtain possible fault information as soon as possible, achieve efficient fault response and disposal, and improve business recovery efficiency; and it can break through the system barriers from remote business to data centers, optimize operation and maintenance processes, improve fault management levels, ensure timely response and problem solving, thereby enhancing system stability and improving service quality.

[0085] The technical solution of the present application and how the technical solution of the present application solves the above-mentioned technical problems are described in detail below with specific embodiments. The following specific embodiments can be combined with each other, and the same or similar concepts or processes may not be repeated in some embodiments. The embodiments of the present application will be described below in conjunction with the accompanying drawings.

[0086] Figure 1 A schematic diagram of the system architecture involved in the fault location method provided in this application, such as Figure 1 As shown, the system architecture includes: application layer, system logic layer, and data flow.

[0087] 1) For the application layer, it can include different remote services. Taking banking as an example, it can include: personal online banking, personal mobile banking, corporate online banking, telephone banking, SMS banking, branch counters, online payment, etc., as well as an integrated operation and maintenance platform (monitoring center-fault diagnosis);

[0088] Among them, for personal online banking, personal mobile banking, corporate online banking, etc., they can all include: online customer service, voice customer service, robot customer service, SMS customer service, email customer service, official media, etc.

[0089] 2) For the system logic layer, it can include: customer complaint data collection module (customer complaint data source, real-time data stream), customer complaint data preprocessing module (data cleaning, filtering, word segmentation, i.e. Jieba, NLTK, Spacy)-(entity tagging, i.e. BIO / BIOES, NLTK / Spacy)-(data standardization), NER model module (pre-trained model library, i.e. BERT, RoBERTa)-(main model, i.e. BiLSTM-ARF), sentiment analysis module (lightweight model, i.e. DistiBERT), alarm generation module (fuzzy matching, i.e. N-gram, Levenshtein Distance)-(alarm level calculation, i.e. application system level + sentiment analysis result)-(other field generation)-(monitoring platform docking), feedback and tracking module (data aggregation and analysis)-(analysis report), feedback and tracking module (data aggregation and analysis)-(analysis report).

[0090] Among them, the associated monitoring center: establishes a docking interface with the integrated operation and maintenance platform-monitoring center-fault diagnosis module, transmits alarm information to the bank's back-end system through the API interface, and triggers the corresponding processing flow. Use Spark structured stream processing to transmit alarm information in real time through the API interface, making the processing and transmission of streaming data more efficient.

[0091] 3) For data streams, it can include: Spark structured stream processing, Spark DataFrame, Spark SQL, distributed parallel computing, Spark Streaming real-time transmission, Spark data aggregation and analysis.

[0092] Based on the above system architecture, the technical solution of the present application and how the technical solution of the present application solves the above technical problems are described in detail below with specific embodiments. The following specific embodiments can be combined with each other, and the same or similar concepts or processes may not be repeated in some embodiments. The embodiments of the present application will be described below in conjunction with the accompanying drawings.

[0093] Figure 2 Schematic diagram of the data processing method provided for this application Figure 1 ,like Figure 2 As shown, the method includes:

[0094] Among them, the electronic device can be a physical device corresponding to the above system architecture.

[0095] Step 21: Obtain customer complaint data to be processed;

[0096] The customer complaint data to be processed includes: at least one customer complaint information from at least one service channel;

[0097] In this step, in order to improve the level of fault management, ensure timely response and problem solving, enhance system stability, and improve service quality, you can first obtain customer complaint data related to faults, service quality, etc., that is, customer complaint data to be processed.

[0098] Optionally, it can be implemented with the above-mentioned customer complaint data collection module, and real-time collection and processing of at least one customer complaint information can be performed based on Spark structured stream processing.

[0099] In one possible implementation, taking banking business as an example, the service channels can be: personal online banking, personal mobile banking, corporate online banking, online banking and other mobile and personal computer applications, online customer service, voice customer service, robot customer service, telephone customer service, email customer service, official media customer service, etc., as well as other channels such as online payment.

[0100] Step 22: for each customer complaint information, determine the urgency of the entity data involved in the customer complaint information and the user corresponding to the customer complaint information;

[0101] In this step, the acquired customer complaint information may be analyzed and processed to extract entity features involved in the customer complaint information and record them as entity data; and the urgency with which the user wishes to resolve the customer complaint in the customer complaint information may be extracted.

[0102] In a possible implementation, the entity data may be a target application name, a target function point, a time, and a location, etc.

[0103] Optionally, before executing step 22, the customer complaint information may be cleaned, filtered, and standardized, including denoising, filtering irrelevant information, word segmentation, part-of-speech tagging, etc.

[0104] That is, when the customer complaint information is in Chinese text, Jieba is used for word segmentation, part-of-speech tagging, and stop word removal. For English text, natural language processing tool libraries such as NLTK and Spacy are used.

[0105] Then, the customer complaint information is entity marked, and the marked customer complaint information is converted into a preset data format.

[0106] In a possible implementation, entity tagging is performed on the processed customer complaint information, and the NER tagging function provided by Spacy and NLTK is used, or BIO / BIOES is used to tag the data; and data standardization and data partitioning are performed, and the data format is unified and standardized based on Spark's DataFrame and SQL functions, and then step 22 is executed.

[0107] Optionally, in step 22, determining the entity data involved in the customer complaint information may include the following implementation:

[0108] Step 1: Input the customer complaint information into the pre-trained word vector model to obtain the optimized customer complaint information;

[0109] In this implementation, the pre-trained word vector model can be a model that extracts features from customer complaint information based on BERT or RoBERTa, etc., to obtain richer semantic information to improve the quality of text representation.

[0110] Step 2: Input the optimized customer complaint information into the pre-trained named entity recognition model to obtain entity data, which includes: target application name, target function point, time, and location.

[0111] In this implementation, based on the features obtained by the above BERT / RoBERTa (i.e., the optimized customer complaint information), entities can be extracted from the customer complaint information based on the BiLSTM-CRF named entity recognition model, including the target application name, target function point, time, and location.

[0112] Among them, BiLSTM is used to extract sentence-level features, and CRF is used to annotate serialized entities. At the same time, using Spark's distributed computing capabilities, BERT and BiLSTM-CRF models are distributed to accelerate the model training and prediction process. For the training of named entity recognition models, distributed deep learning frameworks such as Horovod can be used to integrate with Spark.

[0113] In a possible implementation, the above implementation process is as follows:

[0114] The context-sensitive representation of the customer complaint information obtained from the first layer of the BERT model is used as the output (optimized customer complaint information) (usually the hidden states of the last few layers), which is passed to the BiLSTM layer to learn the dependencies within the sentence and the bidirectional information. The final model output is passed through the CRF layer for label prediction and decoding to obtain entity data.

[0115] Combined with the above example, the process of the NER model module can be: input customer complaint information, load pre-trained word vector model, feature extraction, BiLSTM layer, CRF layer, entity recognition, entity labeling, and output entity data.

[0116] The structure of the NER model module can be: (customer complaint information) BERT (word vector), multiple Bi-LSTMs, Linear, and CRF.

[0117] Optionally, in step 22, determining the urgency level corresponding to the customer complaint information may include the following implementation: inputting the customer complaint information into a pre-trained sentiment analysis model to obtain the urgency level;

[0118] The sentiment analysis model is trained based on at least one historical customer complaint information and the urgency level corresponding to the at least one historical customer complaint information.

[0119] In this implementation, analyzing the sentiment tendency of customer complaints is to evaluate the urgency of the customer's event (customer complaint information), which serves as one of the judgment rules for the alarm level in the subsequent alarm information.

[0120] Furthermore, in order to ensure the timeliness of the alarm and reduce the model size and computational overhead while ensuring the high performance of the model, we can choose to use the lightweight version of the BERT model, DistilBERT, as the sentiment analysis model for sentiment analysis.

[0121] In a possible implementation, after obtaining the sentiment analysis result (ie, the urgency level), a more refined urgency level (level 5: low; level 4: medium-low; level 3: medium; level 2: medium-high; level 1: high) is generated according to the probability distribution of the sentiment analysis.

[0122] Similar to the NER module mentioned above, Spark's distributed processing capabilities can be used to parallelize calculations and improve processing speed.

[0123] Based on the above example, the process of the sentiment analysis module can be: input customer complaint data, load the DistilBERT model, extract features, classify sentiment, generate sentiment scores and levels, and output the degree of urgency.

[0124] Step 23: Generate warning information corresponding to the customer complaint information according to the entity data and the degree of urgency;

[0125] Among them, the alarm information reflects the alarm level of the customer complaint information and the attribute information involved in the alarm;

[0126] In this step, after the entity data and the degree of urgency are determined, an alarm level (ie, the level of the alarm) reflecting the customer complaint information and attribute information related to the alarm may be generated.

[0127] Attribute information may include: alarm type, alarm description, application system (module) number, application system (module) name, alarm occurrence time, alarm object, alarm rule, alarm professional name, monitoring access source, etc.

[0128] Step 24: Push the alarm information corresponding to the customer complaint information to the corresponding operation and maintenance department.

[0129] In this step, the corresponding alarm professional name, that is, the corresponding operation and maintenance department, is determined in the above alarm information. After that, the alarm information is pushed to the corresponding operation and maintenance department for quick processing.

[0130] Optionally, after step 24, the following implementation may also be performed:

[0131] Step 1: Obtain feedback from operation and maintenance personnel regarding alarm information;

[0132] In this implementation, at the operation and maintenance department, the corresponding operation and maintenance personnel can process the alarm information accordingly and provide feedback on the alarm information to obtain feedback information such as the accuracy of the alarm information, whether it has been resolved, and the resolution strategy.

[0133] Step 2: Determine the effectiveness and accuracy of the warning information based on the feedback information.

[0134] In this implementation, the effectiveness and accuracy of the system for alarm information are verified based on the actual feedback from the operation and maintenance team. The above feedback is then re-entered into the system for cyclic optimization to promote iterative improvement of the model.

[0135] In addition, data aggregation and analysis can be performed based on Spark to generate clear and intuitive statistical reports that display key issues and performance indicators.

[0136] The data processing method provided in the embodiment of the present application obtains customer complaint data to be processed, and the customer complaint data to be processed includes: at least one customer complaint information from at least one service channel; for each customer complaint information, the entity data involved in the customer complaint information and the urgency of the user corresponding to the customer complaint information are determined; according to the entity data and the urgency, alarm information corresponding to the customer complaint information is generated, and the alarm information reflects the alarm level of the customer complaint information and the attribute information involved in the alarm; the alarm information corresponding to the customer complaint information is pushed to the corresponding operation and maintenance department, so as to achieve the effect of quickly executing the solution to the customer complaint information.

[0137] Based on the above embodiments, Figure 3 Schematic diagram of the data processing method provided for this application Figure 2 ,like Figure 3 As shown, the above step 23 may include the following implementation:

[0138] Step 31: Determine the alarm level in the alarm information according to the urgency and the application level corresponding to the target application name;

[0139] In this step, based on a preset alarm level generation strategy, the alarm level of the alarm information may be generated by weighted calculation in combination with the application level corresponding to the target application name and the urgency level corresponding to the user emotion.

[0140] In a possible implementation, the urgency configuration coefficient is 60%, and the application level configuration coefficient is 40%, and the alarm level = urgency*60%+application level*40%.

[0141] Step 32: Determine the attribute information involved in the alarm in the alarm information according to the target application name, target function point, time, and location.

[0142] In this step, corresponding attribute information may be obtained based on information such as target application name, target function point, time, and location in the entity data.

[0143] In this implementation, the attributes in the entity data can be filled based on a preset information element template. For example, the information element template includes the following items: alarm type, alarm description, application system (module) number, application system (module) name, alarm occurrence time, alarm object, alarm rule, alarm professional name, and monitoring access source.

[0144] Optionally, taking a bank as an example, the attribute information may include any of the following:

[0145] 1) Alarm type: remote banking customer complaint alarm;

[0146] 2) Alarm description: alarm time, alarm system, alarm object, and brief description of the abnormality;

[0147] 3) Application system (module) number: the application system (module) number where the alarm occurs;

[0148] 4) Application system (module) name: the name of the application system (module) where the alarm occurs;

[0149] 5) Alarm occurrence time: the time when the alarm occurs;

[0150] 6) Alarm object: the resource object where the alarm occurs, obtain the resource object corresponding to the module from the configuration center;

[0151] 7) Alarm rules: Alarm triggering rules, which are displayed as customer complaint alarms by default;

[0152] 8) Alarm professional name: the profession to which the alarm occurs, located in the corresponding operation and maintenance professional office (host, open, security, equipment, etc.) according to the matching system and module;

[0153] 9) Monitor access source: The default display is remote bank.

[0154] It should be understood that entity data should also include: a brief description of the exception, etc.

[0155] The data processing method provided in the embodiment of the present application determines the alarm level in the alarm information according to the urgency and the application level corresponding to the target application name, and then determines the attribute information involved in the alarm in the alarm information according to the target application name, target function point, time, and location. This technical solution realizes more accurate and efficient determination of the alarm information corresponding to the generated customer complaint information.

[0156] Based on the above embodiments, Figure 4 Schematic diagram of the data processing method provided for this application Figure 3 ,like Figure 4 As shown, before the above step 24, the following implementation may also be included:

[0157] Step 41: Determine a target application system corresponding to the target application name in at least one application system;

[0158] In this step, the target application system corresponding to the target application name can be located by means of fuzzy matching.

[0159] Optionally, based on the customer complaint information, determine the corresponding target application name and the target application system.

[0160] In a possible implementation, a mapping of the backend application system may be established: a mapping table may be created, including various backend application systems and corresponding entities and keywords, as a basis for fuzzy matching.

[0161] Fuzzy matching: The fuzzy matching Levenshtein distance algorithm and N-gram model are used to evaluate the similarity between entities in the customer complaint text and entries in the mapping table of the backend application system.

[0162] After that, match the entities: for each entity extracted from the customer complaint information, determine the best match through the results of the selected fuzzy matching algorithm to obtain the target application system.

[0163] Step 42: Determine a target application module corresponding to the target function point from at least one application module in the target application system;

[0164] Among them, each application module has a corresponding operation and maintenance department.

[0165] In this step, the module involved in the specific customer complaint, ie, the target application module, is located by taking the target application system as the basis of the target function point.

[0166] Optionally, in combination with the above implementation, the execution process of the alarm generation module can be: collecting analysis results of customer complaint data to be processed (NER output, sentiment analysis output), locating the target application system and target application model (creating a mapping table, locating based on fuzzy matching), calculating the alarm level (application level, sentiment analysis level, weighting rules), generating alarm information (alarm description, application system (module) name / number, alarm object / level, alarm professional name, monitoring access source, alarm rules), and displaying alarm information (API interface transmission, monitoring platform-fault diagnosis module).

[0167] The data processing method provided in the embodiment of the present application determines the target application system corresponding to the target application name in at least one application system, and determines the target application module corresponding to the target function point in at least one application module in the target application system; wherein each application module corresponds to an operation and maintenance department. In this technical solution, the alarm information corresponding to the customer complaint information can be accurately pushed to the corresponding operation and maintenance department to achieve rapid processing.

[0168] Figure 5 A schematic diagram of the structure of the data processing device provided in this application, such as Figure 5 As shown, the data processing device provided in this embodiment includes:

[0169] The acquisition module 51 is used to acquire customer complaint data to be processed, where the customer complaint data to be processed includes: at least one customer complaint information from at least one service channel;

[0170] A determination module 52, for determining, for each customer complaint information, the entity data involved in the customer complaint information and the urgency of the user corresponding to the customer complaint information;

[0171] A generating module 53 is used to generate alarm information corresponding to the customer complaint information according to the entity data and the urgency level, wherein the alarm information reflects the alarm level of the customer complaint information and the attribute information involved in the alarm;

[0172] The sending module 54 is used to push the alarm information corresponding to the customer complaint information to the corresponding operation and maintenance department.

[0173] In a possible implementation, the determination module 52 determines the entity data involved in the customer complaint information, specifically for:

[0174] Input the customer complaint information into the pre-trained word vector model to obtain the optimized customer complaint information;

[0175] The optimized customer complaint information is input into a pre-trained named entity recognition model to obtain entity data, which includes: target application name, target function point, time, and location.

[0176] In a possible implementation, the determination module 52 determines the urgency level corresponding to the customer complaint information, specifically for:

[0177] The customer complaint information is input into a pre-trained sentiment analysis model to obtain the urgency level, where the sentiment analysis model is trained based on at least one historical customer complaint information and the urgency level corresponding to at least one historical customer complaint information.

[0178] In a possible implementation, the generating module 53 is specifically configured to:

[0179] Determine the alarm level in the alarm information according to the urgency and the application level corresponding to the target application name;

[0180] According to the target application name, target function point, time, and location, attribute information involved in the alarm in the alarm information is determined.

[0181] In a possible implementation manner, before pushing the alarm information corresponding to the customer complaint information to the corresponding operation and maintenance department, the determination module 52 is further configured to:

[0182] Determine a target application system corresponding to the target application name in at least one application system;

[0183] A target application module corresponding to the target function point is determined from at least one application module in the target application system, and each application module corresponds to an operation and maintenance department.

[0184] In a possible implementation, before determining the entity data involved in the customer complaint information and the urgency of the user corresponding to the customer complaint information, the determination module 52 is further configured to:

[0185] Entity mark the customer complaint information and convert the marked customer complaint information into a preset data format.

[0186] In a possible implementation manner, the determination module 52 is further configured to:

[0187] Obtain feedback from operation and maintenance personnel regarding alarm information;

[0188] Based on the feedback information, determine the effectiveness and accuracy of the warning information.

[0189] The data processing device provided in this embodiment can execute the method provided in the above method embodiment. Its implementation principle and technical effect are similar, and this embodiment will not be described in detail here.

[0190] Figure 6 This is a schematic diagram of the structure of the electronic device provided in this application. Figure 6 As shown, the electronic device 6 provided in this embodiment includes: at least one processor 61 and a memory 62.

[0191] Optionally, the device 6 further includes a communication component 63 , wherein the processor 61 , the memory 62 and the communication component 63 are connected via a bus 64 .

[0192] In a specific implementation process, at least one processor 61 executes the computer-executable instructions stored in the memory 62, so that at least one processor 61 executes the above method.

[0193] The specific implementation process of the processor 61 can be found in the above method embodiment, and its implementation principle and technical effect are similar, so this embodiment will not be repeated here.

[0194] In the above embodiments, it should be understood that the processor may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), etc. A general-purpose processor may be a microprocessor or any conventional processor. The steps of the method disclosed in the invention may be directly implemented as being executed by a hardware processor, or may be executed by a combination of hardware and software modules in the processor.

[0195] The memory may include a high-speed memory (Random Access Memory, RAM), and may also include a non-volatile memory (Non-volatile Memory, NVM), such as at least one disk memory.

[0196] The bus may be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, or an Extended Industry Standard Architecture (EISA) bus, etc. The bus may be divided into an address bus, a data bus, a control bus, etc. For ease of representation, the bus in the drawings of the present application is not limited to only one bus or one type of bus.

[0197] The present application also provides a computer program product, including a computer program, which implements the above method when executed by a processor.

[0198] The present application also provides a computer-readable storage medium, in which computer-executable instructions are stored. When a processor executes the computer-executable instructions, the above method is implemented.

[0199] The above-mentioned readable storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic disk or optical disk. The readable storage medium can be any available medium that can be accessed by a general or special-purpose computer.

[0200] An exemplary readable storage medium is coupled to a processor so that the processor can read information from the readable storage medium and write information to the readable storage medium. Of course, the readable storage medium can also be a component of the processor. The processor and the readable storage medium can be located in an application specific integrated circuit (Application Specific Integrated Circuits, referred to as: ASIC). Of course, the processor and the readable storage medium can also exist in the device as discrete components.

[0201] The division of units is only a logical function division, and there may be other divisions in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be an indirect coupling or communication connection through some interface, device or unit, which can be electrical, mechanical or other forms.

[0202] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0203] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0204] If the function is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions for a computer device (which can be a personal computer, server, or network device, etc.) to perform all or part of the steps of the methods of each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, etc. Various media that can store program codes.

[0205] Those skilled in the art can understand that all or part of the steps of implementing the above-mentioned method embodiments can be completed by hardware related to program instructions. The aforementioned program can be stored in a computer-readable storage medium. When the program is executed, the steps of the above-mentioned method embodiments are executed; and the aforementioned storage medium includes: ROM, RAM, disk or optical disk and other media that can store program codes.

[0206] Finally, it should be noted that those skilled in the art will readily conceive of other embodiments of the present invention after considering the specification and practicing the invention disclosed herein. The present invention is intended to cover any variations, uses or adaptations of the present invention, which follow the general principles of the present invention and include common knowledge or customary technical means in the art not disclosed by the present invention, are not limited to the precise structure described above and shown in the drawings, and may be modified and changed in various ways without departing from the scope thereof. The scope of the present invention is limited only by the appended claims.

Claims

1. A data processing method, characterized in that: include: Acquire customer complaint data to be processed, where the customer complaint data to be processed includes: at least one customer complaint information from at least one service channel; For each customer complaint information, determine the urgency of the entity data involved in the customer complaint information and the user corresponding to the customer complaint information; Generate alarm information corresponding to the customer complaint information according to the entity data and the urgency level, wherein the alarm information reflects the alarm level of the customer complaint information and attribute information involved in the alarm; The alarm information corresponding to the customer complaint information is pushed to the corresponding operation and maintenance department.

2. The method according to claim 1, characterized in that Determine the entity data involved in the customer complaint information, including: Inputting the customer complaint information into a pre-trained word vector model to obtain optimized customer complaint information; The optimized customer complaint information is input into a pre-trained named entity recognition model to obtain the entity data, which includes: target application name, target function point, time, and location.

3. The method according to claim 1, characterized in that Determine the urgency of the customer complaint information, including: The customer complaint information is input into a pre-trained sentiment analysis model to obtain the urgency level, wherein the sentiment analysis model is trained based on at least one historical customer complaint information and the urgency level corresponding to the at least one historical customer complaint information.

4. The method according to any one of claims 1 to 3, characterized in that: The generating, according to the entity data and the urgency, warning information corresponding to the customer complaint information includes: Determining the alarm level in the alarm information according to the urgency and the application level corresponding to the target application name; According to the target application name, the target function point, the time, and the location, attribute information related to the alarm in the alarm information is determined.

5. The method according to claim 2, characterized in that: Before pushing the alarm information corresponding to the customer complaint information to the corresponding operation and maintenance department, the method further includes: Determine a target application system corresponding to the target application name in at least one application system; A target application module corresponding to the target function point is determined from at least one application module in the target application system, and each application module corresponds to an operation and maintenance department.

6. The method according to any one of claims 1 to 3, characterized in that: Before determining the entity data involved in the customer complaint information and the urgency of the user corresponding to the customer complaint information, the method further includes: The customer complaint information is entity marked, and the marked customer complaint information is converted into a preset data format.

7. The method according to any one of claims 1 to 3, characterized in that: The method further comprises: Obtain feedback from operation and maintenance personnel regarding alarm information; The effectiveness and accuracy of the warning information are determined based on the feedback information.

8. A data processing device, characterized in that: include: An acquisition module, used to acquire customer complaint data to be processed, wherein the customer complaint data to be processed includes: at least one customer complaint information from at least one service channel; A determination module, used to determine, for each customer complaint information, the urgency of the entity data involved in the customer complaint information and the user corresponding to the customer complaint information; A generating module, configured to generate alarm information corresponding to the customer complaint information according to the entity data and the urgency level, wherein the alarm information reflects the alarm level of the customer complaint information and attribute information involved in the alarm; The sending module is used to push the alarm information corresponding to the customer complaint information to the corresponding operation and maintenance department.

9. An electronic device, characterized in that: include: Memory, processor; The memory stores computer-executable instructions; The processor executes the computer-executable instructions stored in the memory, so that the processor performs the method according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer-executable instructions, which are used to implement the method according to any one of claims 1 to 7 when executed by a processor.

Images