CN114765575A - Network fault cause prediction method and device and electronic equipment - Google Patents

Abstract

The invention provides a method and a device for predicting a network fault reason and electronic equipment, and solves the problem of low accuracy rate of the conventional network fault reason prediction. The method of the invention comprises the following steps: obtaining classification feature vectors in a fault work order, wherein the classification feature vectors comprise a first class feature vector and a second class feature vector; obtaining the target fault reason category to which the fault work order belongs according to the first class feature vector and the first classification prediction model; and obtaining a target fault cause sub-category of the fault work order in the target fault cause category according to the second class feature vector and a second class prediction model corresponding to the target fault cause category. According to the invention, through a two-step prediction method, namely, the large class of the fault reasons is predicted, and then the subdivided classes of the fault reasons in the large class of the fault reasons are predicted, so that the number of the predicted classes in each step can be effectively reduced, and the accuracy of the prediction result is improved.

Description

A kind of network fault cause prediction method, device and electronic equipment

technical field

The invention relates to the technical field of artificial intelligence, and in particular, to a method, device and electronic equipment for predicting the cause of a network fault.

Background technique

In a network system, there are many kinds of network elements, and the network structure is complex. In the process of network operation, various faults will inevitably occur. After a fault occurs, network operation and maintenance personnel need to troubleshoot the fault, find out the cause of the fault, and then take corresponding measures to help the existing network resume operation.

Specifically, during the operation of the existing network, after a fault occurs, the network device will generate an alarm and report it to the network management system. The network management system dispatches orders to the operation and maintenance personnel based on the received alarms and certain order dispatching rules. The operation and maintenance personnel investigate the cause of the failure based on the alarm and other information, and then take corresponding measures according to the cause of the failure. After that, map the cause of the failure and the handling measures to the corresponding work order.

In the existing technical solutions for predicting failure causes, there are many types of failure causes, some of which are relatively similar, and the accuracy of direct prediction is low.

SUMMARY OF THE INVENTION

The purpose of the present invention is to provide a method, device and electronic device for predicting the cause of a network failure, which are used to solve the problem of low accuracy of predicting the cause of a network failure in the prior art.

In order to achieve the above object, the present invention provides a method for predicting the cause of a network fault, including:

obtaining a classification feature vector in the trouble ticket, where the classification feature vector includes a first-class feature vector and a second-class feature vector;

According to the first type of feature vector and the first classification prediction model, obtain the target fault cause category to which the fault work order belongs;

According to the feature vector of the second type and the second classification prediction model corresponding to the target fault category, the target fault cause sub-category of the fault work order in the target fault cause category is obtained.

Wherein, the obtaining the classification feature vector in the trouble ticket includes:

Obtaining a fault ticket to be processed, the fields of the fault ticket include an alarm title, a network element name, a network element type, and a fault occurrence time;

Based on the correspondence between the fields of the trouble ticket and the feature vector and/or the feature extraction model, the classification feature vector in the trouble ticket is extracted.

Wherein, the classification feature vector includes:

a first feature vector for characterizing the alert title;

a second feature vector used to characterize the fault cause category corresponding to the alarm title;

a third feature vector for characterizing the network element type;

a fourth feature vector used to characterize the failure cause category corresponding to the network element type;

a fifth feature vector used to represent the alarm information associated with the trouble ticket;

a sixth eigenvector for characterizing the fault cause subcategory corresponding to the alarm title; and a seventh eigenvector for characterizing the fault cause subcategory corresponding to the network element type;

Wherein, the first type of eigenvectors include: the first eigenvector, the second eigenvector, the third eigenvector, the fourth eigenvector, and the fifth eigenvector;

The second type of eigenvectors include: the first eigenvector, the second eigenvector, the third eigenvector, the fourth eigenvector, the fifth eigenvector, and the sixth eigenvector and the seventh eigenvector.

Wherein, according to the first type of feature vector and the first classification prediction model, the target fault cause category to which the fault work order belongs is obtained, including:

Classify the first-class feature vector by using the first classification prediction model to obtain the probability value of each fault cause category;

The failure cause category corresponding to the maximum probability value among the probability values of each failure cause category is determined as the target failure cause category to which the failure work order belongs.

Wherein, according to the second type of feature vector and the second classification prediction model corresponding to the target fault cause category, the target fault cause subcategory of the fault work order in the target fault cause category is obtained, include:

Classify the feature vector of the second type by using the second classification prediction model to obtain the probability value of each fault cause sub-category of the fault work order in the target fault cause category;

The fault cause sub-category corresponding to the maximum probability value among the probability values of each fault cause sub-category is determined as the target fault cause sub-category.

Wherein, the method also includes:

Obtain multiple historical fault tickets and multiple historical alarm information. The fields of each historical fault ticket include the alarm title, NE name, NE type, fault occurrence time, fault cause category, and the fault corresponding to the fault cause category. Cause sub-category, each field of the historical alarm information includes the alarm title, network element name and alarm start time;

According to the field of the historical trouble ticket and the field of the historical alarm information, a classification feature vector is obtained, and the classification feature vector includes: a first feature vector used to represent the alarm title, used to represent the alarm title The second eigenvector of the corresponding fault cause category, used to characterize the third eigenvector of the network element type, used to characterize the fourth eigenvector of the fault cause category corresponding to the NE type, used to characterize the fault The fifth eigenvector of the alarm information associated with the work order, the sixth eigenvector used to represent the fault cause subcategory corresponding to the alarm title, and the seventh eigenvector used to represent the fault cause subcategory corresponding to the network element type;

According to the first eigenvector, the second eigenvector, the third eigenvector, the fourth eigenvector, the fifth eigenvector, and the category label of the fault cause category, perform model training to obtain the first Classification prediction model.

Wherein, after obtaining the classification feature vector according to the field of the historical trouble ticket and the field of the historical alarm information, the method further includes:

Grouping a plurality of the historical fault work orders according to the fault cause category to obtain multiple sets of historical fault work order data;

According to the category label and classification feature vector of the fault cause sub-category corresponding to each group of historical fault work order data, model training is performed on each group of historical fault work order data to obtain a plurality of second classification prediction models.

The present invention also provides a network fault cause prediction device, comprising:

a first obtaining module, configured to obtain a classification feature vector in the trouble ticket, where the classification feature vector includes a first-type feature vector and a second-class feature vector;

a first fault cause prediction module, configured to obtain the target fault cause category to which the fault work order belongs according to the first type feature vector and the first classification prediction model;

The second fault cause prediction module is configured to obtain the target of the fault work order in the target fault cause category according to the second type feature vector and the second classification prediction model corresponding to the target fault cause category Failure cause subcategory.

The present invention also provides an electronic device comprising a processor and a transceiver, the transceiver receiving and transmitting data under the control of the processor, and the processor is configured to perform the following operations:

obtaining a classification feature vector in the trouble ticket, where the classification feature vector includes a first-class feature vector and a second-class feature vector;

According to the first type of feature vector and the first classification prediction model, obtain the target fault cause category to which the fault work order belongs;

According to the second type of feature vector and the second classification prediction model, the target fault cause sub-category of the fault work order in the target fault cause category is obtained.

Wherein, the processor is also used for:

Obtaining a fault ticket to be processed, the fields of the fault ticket include an alarm title, a network element name, a network element type, and a fault occurrence time;

Based on the correspondence between the fields of the trouble ticket and the feature vector and/or the feature extraction model, the classification feature vector in the trouble ticket is extracted.

Wherein, the classification feature vector includes:

a first feature vector for characterizing the alert title;

a second feature vector used to characterize the fault cause category corresponding to the alarm title;

a third feature vector for characterizing the network element type;

a fourth feature vector used to characterize the failure cause category corresponding to the network element type;

a fifth feature vector used to represent the alarm information associated with the trouble ticket;

a sixth feature vector for characterizing the fault cause sub-category corresponding to the alarm title; and

a seventh feature vector used to characterize the fault cause sub-category corresponding to the network element type;

Wherein, the first type of eigenvectors include: the first eigenvector, the second eigenvector, the third eigenvector, the fourth eigenvector, and the fifth eigenvector;

The second type of eigenvectors include: the first eigenvector, the second eigenvector, the third eigenvector, the fourth eigenvector, the fifth eigenvector, and the sixth eigenvector and the seventh eigenvector.

Wherein, the processor is also used for:

Classify the first-class feature vector by using the first classification prediction model to obtain the probability value of each fault cause category;

The failure cause category corresponding to the maximum probability value among the probability values of each failure cause category is determined as the target failure cause category to which the failure work order belongs.

Wherein, the processor is also used for:

Classify the feature vector of the second type by using the second classification prediction model to obtain the probability value of each fault cause sub-category of the fault work order in the target fault cause category;

The fault cause sub-category corresponding to the maximum probability value among the probability values of each fault cause sub-category is determined as the target fault cause sub-category.

Wherein, the processor is also used for:

Obtain multiple historical fault tickets and multiple historical alarm information. The fields of each historical fault ticket include the alarm title, NE name, NE type, fault occurrence time, fault cause category, and the fault corresponding to the fault cause category. Cause sub-category, each field of the historical alarm information includes the alarm title, network element name and alarm start time;

According to the field of the historical trouble ticket and the field of the historical alarm information, a classification feature vector is obtained, and the classification feature vector includes: a first feature vector used to represent the alarm title, used to represent the alarm title The second eigenvector of the corresponding fault cause category, used to characterize the third eigenvector of the network element type, used to characterize the fourth eigenvector of the fault cause category corresponding to the NE type, used to characterize the fault The fifth eigenvector of the alarm information associated with the work order, the sixth eigenvector used to represent the fault cause subcategory corresponding to the alarm title, and the seventh eigenvector used to represent the fault cause subcategory corresponding to the network element type;

According to the first eigenvector, the second eigenvector, the third eigenvector, the fourth eigenvector, the fifth eigenvector, and the category label of the fault cause category, perform model training to obtain the first Classification prediction model.

Wherein, the processor is also used for:

Grouping a plurality of the historical fault work orders according to the fault cause category to obtain multiple sets of historical fault work order data;

According to the category label and classification feature vector of the fault cause sub-category corresponding to each group of historical fault work order data, model training is performed on each group of historical fault work order data to obtain a plurality of second classification prediction models.

The present invention also provides an electronic device, comprising a memory, a processor, and a program stored on the memory and running on the processor; the processor implements the above-mentioned network failure when executing the program Cause prediction method.

The present invention also provides a computer-readable storage medium on which a computer program is stored, and when the program is executed by a processor, implements the steps in the above-mentioned method for predicting a cause of a network failure.

The above-mentioned technical scheme of the present invention has at least the following beneficial effects:

In the embodiment of the present invention, by acquiring the classification feature vector in the fault work order, the classification feature vector includes a first-class feature vector and a second-class feature vector; according to the first-class feature vector and the first classification prediction model, Obtain the target fault cause category to which the fault work order belongs; obtain the target fault cause of the fault work order according to the second type feature vector and the second classification prediction model corresponding to the target fault cause category The target fault cause sub-category in the category, so, through the two-step prediction method, that is, first predict the fault cause category, and then predict the fault cause sub-category in the fault cause category, which can effectively reduce the number of faults. The number of categories predicted in one step to improve the accuracy of the prediction results.

Description of drawings

FIG. 1 shows one of the schematic flowcharts of a method for predicting a cause of a network fault according to an embodiment of the present invention;

2 shows a schematic diagram of a model training process of a first classification prediction model and a second classification prediction model according to an embodiment of the present invention;

FIG. 3 shows the second schematic flowchart of the method for predicting the cause of a network fault according to an embodiment of the present invention;

4 is a schematic diagram of a module of an apparatus for predicting a cause of a network failure according to an embodiment of the present invention;

FIG. 5 is a schematic structural diagram of an electronic device according to an embodiment of the present invention.

Detailed ways

In order to make the technical problems, technical solutions and advantages to be solved by the present invention more clear, the following will be described in detail with reference to the accompanying drawings and specific embodiments.

The present invention provides a method, device and electronic equipment for predicting the causes of network failures, aiming at the problem of low accuracy in predicting the causes of existing network failures.

As shown in FIG. 1 , it is a schematic flowchart of a method for predicting a cause of a network fault provided by an embodiment of the present invention. Specifically, the method includes:

Step 101, obtaining a classification feature vector in the trouble ticket, where the classification feature vector includes a first-type feature vector and a second-type feature vector;

In this step, the fault work order is generated based on the received alarm information and according to the preset order dispatch rules, and is a new work order, that is, the cause of the fault has not been recorded on the fault work order.

Step 102, according to the first type of feature vector and the first classification prediction model, obtain the target fault cause category to which the fault work order belongs;

In this step, the first classification prediction model is a pre-trained model, the first type feature vector of the fault work order is used as input, the first classification prediction model is input, and the target failure cause category to which the fault work order belongs is output, That is, the category of fault causes to which the fault ticket belongs.

The first classification prediction model is obtained by training based on historical fault tickets and alarm data.

Step 103 , according to the second type of feature vector and the second classification prediction model corresponding to the target fault cause category, obtain the target fault cause subcategory of the fault work order in the target fault cause category.

In this step, the second classification prediction model is a pre-trained model.

Here, the second type feature vector of the fault work order is used as input, the second classification prediction model is input, and the target fault cause sub-category of the fault work order in the target fault cause category is output, that is, the fault in the fault cause category Category of reason breakdown.

It should be noted that since the historical fault tickets include the failure causes and handling measures recorded by the operation and maintenance personnel after the fault is resolved, it contains a lot of operation and maintenance experience, so the information contained in the historical fault tickets is mined and the machine is trained. Learn the model to obtain the first classification prediction model and the second classification prediction model, and then automatically predict the cause of the failure in the work order information and alarm information in the new dispatch order, thereby saving the time of the operation and maintenance personnel to troubleshoot the cause of the failure.

In the method for predicting the cause of a network fault according to the embodiment of the present invention, the classification feature vector in the fault work order is obtained, and the classification feature vector includes a first-class feature vector and a second-class feature vector; according to the first-class feature vector and the second-class feature vector a classification prediction model, to obtain the target failure cause category to which the fault work order belongs; The target fault cause sub-category in the target fault cause category. In this way, through a two-step prediction method, that is, first predict the fault cause category, and then predict the sub-categories of the fault cause in the fault cause category. It can effectively reduce the number of categories predicted at each step and improve the accuracy of the prediction results.

As an optional implementation manner, step 101 in this embodiment of the present invention may specifically include:

Obtaining a fault ticket to be processed, the fields of the fault ticket include an alarm title, a network element name, a network element type, and a fault occurrence time;

In this step, after a network failure occurs, an alarm message will be generated. Based on the received alarm information, the electronic device generates a trouble ticket to be processed according to the preset order dispatch rule.

Generally, the fault ticket to be processed includes but is not limited to fields such as alarm title, network element name, network element type, and fault occurrence time. It should be noted that the fault cause that caused the network fault and the corresponding processing measures are not recorded in the fault ticket to be processed at this time.

Based on the correspondence between the fields of the trouble ticket and the feature vector and/or the feature extraction model, the classification feature vector in the trouble ticket is extracted.

Optionally, the classification feature vector includes:

a first feature vector for characterizing the alert title;

a second feature vector used to characterize the fault cause category corresponding to the alarm title;

a third feature vector for characterizing the network element type;

a fourth feature vector used to characterize the failure cause category corresponding to the network element type;

a fifth feature vector used to represent the alarm information associated with the trouble ticket;

a sixth feature vector for characterizing the fault cause sub-category corresponding to the alarm title; and

a seventh feature vector used to characterize the fault cause sub-category corresponding to the network element type;

Wherein, the first type of eigenvectors include: the first eigenvector, the second eigenvector, the third eigenvector, the fourth eigenvector, and the fifth eigenvector;

The second type of eigenvectors include: the first eigenvector, the second eigenvector, the third eigenvector, the fourth eigenvector, the fifth eigenvector, and the sixth eigenvector and the seventh eigenvector.

The first feature vector, the second feature vector, and the sixth feature vector are all obtained according to the correspondence between the alarm title of the fault ticket and the feature vector.

That is to say, the alarm title has a first correspondence with the feature vector representing itself. Through the first correspondence, the first feature vector representing the alarm title of the trouble ticket is obtained; the alarm title corresponds to the alarm title. The feature vector of the fault cause category has a second corresponding relationship, and through the second corresponding relationship, the second feature vector of the fault work order representing the fault cause category corresponding to its alarm title is obtained; the alarm title and the corresponding alarm title are obtained. The feature vector of the fault cause subcategory has a third correspondence, and through the third correspondence, a sixth feature vector of the fault work order is obtained, which represents the fault cause subcategory corresponding to the alarm title of the fault work order.

The third eigenvector, the fourth eigenvector, and the seventh eigenvector are all obtained according to the corresponding relationship between the network element type of the fault work order and the eigenvectors.

That is to say, the network element type has a fourth correspondence with the feature vector that characterizes itself. Through the fourth correspondence, the third feature vector representing the network element type of the fault work order is obtained; The feature vector of the fault cause category corresponding to the network element type has a fifth correspondence, and through the fifth correspondence, the fourth feature vector of the fault work order that represents the fault cause category corresponding to the network element type is obtained; The feature vector representing the fault cause sub-category corresponding to the network element type has a sixth correspondence, and through the sixth correspondence, the seventh feature vector of the fault work order representing the fault cause sub-category corresponding to the network element type is obtained.

It should be specially pointed out that the fifth feature vector is obtained based on the feature extraction model. Specifically, first, extract the alarm information associated with the fault ticket to obtain m alarms, and use each alarm information to obtain a corresponding vector through the first feature extraction model (such as the CBOW model of word2vec); then, obtain m The first mean vec_cbow of the vector; then, the corresponding vector is obtained for each alarm message through the second feature extraction model (such as the Skip-grams model of word2vec); after that, the second mean vec_sg of the m vectors is obtained; finally, the The first mean vec_cbow and the second mean vec_sg are concatenated to obtain the fifth feature vector.

It should be noted that, when the alarm information satisfies the first condition and the second condition, it is determined that the alarm information is the alarm information associated with the fault work order.

Here, the first condition is that the alarm start time t ₂ of the alarm information is between (t ₁ -t _A , t ₁ +t _B ), where t ₁ represents the fault occurrence time of the fault work order, t _A and t _B is the preset time value.

That is, the alarm start time of the alarm information is between a period of time before and a period of time after the fault occurrence time of the fault ticket.

The second condition is that the network element name in the alarm information is the same as the network element name in the trouble ticket.

Here, word2vec can quickly and efficiently express a word into a vector form through an optimized training model according to a given corpus, which provides a new tool for applied research in the field of natural language processing. word2vec relies on skip-grams models or continuous bag-of-words CBOW models to build neural word embeddings.

As an optional implementation manner, in step 102 of the method in this embodiment of the present invention, according to the first type feature vector and the first classification prediction model, the target failure cause category to which the fault work order belongs is obtained, which may include:

Classify the first-class feature vector by using the first classification prediction model to obtain the probability value of each fault cause category;

Here, each fault cause category refers to each fault cause category to which the fault work order belongs, that is, each fault cause category to which the fault work order belongs.

It should be noted that, based on the feature vector of the first type, through the classification of the first classification prediction model, it can be obtained that the fault cause category to which the fault work order belongs may correspond to many kinds, and the most likely fault cause category can pass the probability value. Compare and measure to determine.

The failure cause category corresponding to the maximum probability value among the probability values of each failure cause category is determined as the target failure cause category to which the failure work order belongs.

Through this implementation, the target fault cause category to which the fault work order belongs can be predicted.

As an optional implementation manner, in step 103 of the method in this embodiment of the present invention, according to the second type feature vector and the second classification prediction model corresponding to the target fault cause type, obtain the location where the fault work order is located. target failure cause subcategories in the target failure cause category, including:

Classify the feature vector of the second type by using the second classification prediction model to obtain the probability value of each fault cause sub-category of the fault work order in the target fault cause category;

In this step, it should be noted that the second classification prediction model is determined according to the target fault cause category to which the fault work order belongs. That is to say, different failure cause categories correspond to different second classification prediction models.

The fault cause sub-category corresponding to the maximum probability value among the probability values of each fault cause sub-category is determined as the target fault cause sub-category.

This implementation method is similar to the above-mentioned determination of the target fault cause category of the fault work order, and the most likely sub-category of the fault cause can be determined by comparing the probability values.

It can be seen from the above description that the key to the accuracy of predicting the failure cause of a fault work order lies in the classification prediction model. How to train a good classification prediction model, as an optional implementation manner, the method in the embodiment of the present invention may further include:

Obtain multiple historical fault tickets and multiple historical alarm information. The fields of each historical fault ticket include the alarm title, NE name, NE type, fault occurrence time, fault cause category, and the fault corresponding to the fault cause category. Cause sub-category, each field of the historical alarm information includes the alarm title, network element name and alarm start time;

Here, "multiple" among the multiple historical fault work orders and the multiple historical alarm information can be understood as a large number here. It should be noted that these historical fault tickets and historical alarm information are valid data, that is, there are no empty fields in the above historical fault tickets and historical alarm information.

In a large number of historical fault tickets and a large amount of historical alarm information, first, remove the data containing empty fields; then, perform regularization matching on the filtered historical fault tickets and the alarm titles in the two alarm information, and extract the alarms The content, such as "yyy alarm" for "xxx", extract the "yyy alarm" part; finally, format the fault occurrence time and alarm start time, and convert them into a unified time format, such as datetime64 format.

According to the field of the historical trouble ticket and the field of the historical alarm information, a classification feature vector is obtained, and the classification feature vector includes: a first feature vector used to represent the alarm title, used to represent the alarm title The second eigenvector of the corresponding fault cause category, used to characterize the third eigenvector of the network element type, used to characterize the fourth eigenvector of the fault cause category corresponding to the NE type, used to characterize the fault The fifth eigenvector of the alarm information associated with the work order, the sixth eigenvector used to represent the fault cause subcategory corresponding to the alarm title, and the seventh eigenvector used to represent the fault cause subcategory corresponding to the network element type;

In this step, the first feature vector, the second feature vector and the sixth feature vector are obtained according to the alarm title field of the historical trouble ticket.

Specifically, the historical fault tickets are processed as follows:

1) One-hot encoding is performed on the alarm title of the historical fault ticket to obtain a one-hot vector; the alarm title and the corresponding one-hot vector are stored as the key and value of the dictionary, and the dictionary is recorded as dict_1.

The one-hot vector here is the first feature vector used to represent the alarm title.

2) Obtain the number of times of each fault cause category corresponding to the alarm title in the historical fault work order, and perform normalization processing to obtain the vector corresponding to the alarm title; use the alarm title and the corresponding vector as the key of the dictionary and value are stored, and the dictionary is recorded as dict_2.

Here, the vector is the second feature vector used to represent the fault cause category corresponding to the alarm title.

For example, in all work orders, there are a total of 4 work orders with alarm title A. The alarm titles and fault reasons in these four work orders are as follows:

Alarm Title A Fault Cause Category 1

Alarm Title A Fault Cause Category 1

Alarm Title A Fault Cause Category 3

Alarm Title A Fault Cause Category 4

Then the vector corresponding to the alarm title A is [2/4, 0, 1/4, 1/4, 0, . . . ], and the dimension of the vector is the number of categories of fault cause categories.

3) One-hot encoding is performed on the network element type of the historical fault work order to obtain a one-hot vector; the network element type and the corresponding one-hot vector are stored as the key and value of the dictionary, and the dictionary is recorded as dict_3.

The one-hot vector here is the third feature vector used to represent the alarm title.

4) Obtain the number of times of each failure cause category corresponding to the network element type in the historical fault work order, and perform normalization processing to obtain the vector corresponding to the network element type; the network element type and its corresponding vector It is stored as the key and value of the dictionary, and the dictionary is recorded as dict_4.

Here, the vector is the fourth feature vector used to represent the failure cause category corresponding to the network element type.

For example, in all work orders, there are a total of 4 work orders with NE type A. The network element types and fault causes in these four work orders are as follows:

NE Type A Fault Cause Category 1

NE Type A Fault Cause Category 1

NE Type A Fault Cause Category 3

NE Type A Fault Causes Category 4

The vector corresponding to network element type A is [2/4,0,1/4,1/4,0,...], and the dimension of the vector is the number of categories of fault cause categories.

5) For the fifth feature vector used to represent the alarm information associated with the fault ticket

The trouble ticket here refers to the historical trouble ticket.

First, for each historical fault ticket, extract the associated alarm information.

Here, when the alarm information satisfies the first condition and the second condition, it is determined that the alarm information is the alarm information associated with the historical trouble ticket.

Here, the first condition is that the alarm start time t ₂ of the alarm information is between (t ₁ -t _A , t ₁ +t _B ), where t ₁ represents the fault occurrence time of the historical fault work order, t _A and t _B is the preset time value.

That is, the alarm start time of the alarm information is between a period of time before and a period of time after the fault occurrence time of the historical fault work order.

The second condition is that the network element name in the alarm information is the same as the network element name in the historical trouble ticket.

After that, sort the alarm information associated with each historical fault ticket according to the order of the alarm start time, extract the alarm title of each alarm information, and use it as a word to form an alarm sentence.

Here, the alarm statement is used to describe a series of ordered alarm information generated on the network element within a period of time when the historical fault work order fault occurs.

Then, the alarm statement corresponding to each historical fault ticket is formed into a document, which is used as the corpus, and the two wrod2vec models, the CBOW model and the Skip-gram model, are respectively trained with the corpus, and two vector representations of each alarm information are obtained; Save the trained CBOW model model_cbow and Skip-gram model model_sg;

Finally, query the alarm vector of the alarm title in the alarm statement corresponding to each historical fault ticket, and then average the alarm vectors. The CBOW model and Skip-gram model vectorize and average the alert titles, respectively. The two vectors obtained by the two models are spliced together as the feature vector of the historical fault ticket matching the alarm information, that is, the fifth feature vector.

6) Obtain the number of times of each fault cause sub-category (that is, the fault cause subcategory) corresponding to the alarm title in the historical fault work order, and perform normalization processing to obtain the vector corresponding to the alarm title; The title and the corresponding vector are stored as the key and value, and the dictionary is recorded as dict_5.

Here, the vector is the sixth feature vector used to represent the fault cause sub-category corresponding to the alarm title.

7) Obtain the number of times of each fault cause sub-category (that is, the sub-category of fault causes) corresponding to the network element type in the historical fault work order, and perform normalization processing to obtain the vector corresponding to the network element type; The network element type and the corresponding vector are stored as the key and value, and the dictionary is recorded as dict_6.

Here, the vector is the seventh feature vector used to represent the fault cause sub-category corresponding to the network element type.

According to the first eigenvector, the second eigenvector, the third eigenvector, the fourth eigenvector, the fifth eigenvector, and the category label of the fault cause category, perform model training to obtain the first Classification prediction model.

It should be noted that, optionally, the category label of the fault cause category is obtained by digitally encoding the fault cause category of the historical fault work order. That is, numbers 1 to N are used to identify N types of fault cause categories as category labels.

Here, the first eigenvector, the second eigenvector, the third eigenvector, the fourth eigenvector, and the fifth eigenvector are used as input, and are input into the preset classification model, and the classification result is output, that is, the output category label, and the classification result is Compare the category labels of the classification results with the corresponding actual category labels, continuously adjust the parameters in the preset classification model, reduce the difference between the category labels in the classification results and the corresponding actual category labels, and reduce the difference to the preset range, or Brings the preset classification model to the minimum convergence position.

Here, the preset classification model used in model training is the GBDT model or the XGBoost model.

Here, the GBDT (Gradient Boosting Decision Tree) model is an additive model that trains a set of CART (Classification and Regression Trees) serially, and finally sums the prediction results of all regression trees , resulting in a strong learner, where each new tree fits the negative gradient direction of the current loss function.

The XGBoost (Extreme Gradient Boosting, gradient boosting tree) model also generates models serially, and takes the sum of all models as the output.

Here, save the trained first classification prediction model model_1.

Further, after obtaining the classification feature vector according to the field of the historical trouble ticket and the field of the historical alarm information, the method further includes:

Grouping a plurality of the historical fault work orders according to the fault cause category to obtain multiple sets of historical fault work order data;

Here, all data, that is, all historical fault work orders, are grouped according to fault cause categories to obtain multiple sets of historical fault work order data.

Here, different groups of historical trouble ticket data correspond to different trouble cause categories.

According to the category label and classification feature vector of the fault cause sub-category corresponding to each group of historical fault work order data, model training is performed on each group of historical fault work order data to obtain a plurality of second classification prediction models.

到

共n_1种标签。It should be noted that, assuming that there are N kinds of failure cause categories, A ¹ , ···, A ^N , then in the first group of data, the failure cause categories, that is, the major categories of failure causes, are all A ¹ , and the labels of the first group of data It is its fault cause sub-category, assuming the fault cause sub-category corresponding to the fault cause category A ¹ , that is, there are n_1 types of fault cause sub-categories, then the labels of the first group of data are:

arrive

A total of n_1 kinds of labels.

Here, the classification feature vector specifically refers to the first feature vector, the second feature vector, the third feature vector, the fourth feature vector, the fifth feature vector, the sixth feature vector, and the seventh feature vector.

A model is trained by each group of historical fault ticket data, that is, the classification feature vector in each group of historical fault tickets. The preset classification model used in model training is the GBDT model or the XGBoost model.

Here, the trained N second classification prediction models model_2_1 to model_2_N are saved.

Here, for the specific training process of the first classification prediction model and the second classification prediction model, reference may be made to FIG. 2 .

The following is an example, as shown in FIG. 3 , to specifically describe the implementation process of the method in the embodiment of the present invention.

S1: The work order to be predicted is received.

Here, the work order to be predicted is a new work order, and the work order to be predicted includes four fields: alarm title, network element name, network element type, and fault occurrence time.

It should be noted that the field of the fault cause sub-category (ie, the fault cause sub-category) of the work order to be predicted is predicted based on the above four fields and the alarm information associated with the work order.

S2: Extract the feature vector of the first-level classification of the work order.

Specifically, 1) the one-hot feature of the alarm title: the alarm title in the work order is used as the key, and the feature vector vec_1 corresponding to the alarm title is queried in the dictionary dict_1, that is, the first feature vector.

2) Distributed characteristics of the major categories of fault causes of the alarm title: The alarm title in the work order is used as the key, and the feature vector vec_2 corresponding to the alarm title is queried in the dictionary dict_2, that is, the second feature vector.

3) One-hot feature of network element type: The network element type of the work order is used as the key, and the feature vector vec_3 corresponding to the network element type is queried in the dictionary dict_3, that is, the third feature vector.

4) The distribution characteristics of the fault causes of the network element type: the network element type of the work order is used as the key, and the feature vector vec_4 corresponding to the network element type is queried in the dictionary dict_4, that is, the fourth feature vector.

5) The word2vec feature associated with the work order to the alarm:

First, extract the alarm information associated with the work order. The extraction method is the same as the extraction method in the training phase. Assuming that h alarms are associated, for each alarm, use the wrod2vec model CBOW model to get its vector, and then find the mean vec_cbow of the h vectors; then, for each alarm, use the Skip-grams model to get its vector, and find h The mean vec_sg of the strip vectors; finally, the mean vec_cbow and the mean vec_sg are spliced to obtain the feature vector vec_5, which is the fifth feature vector.

6) Distribution characteristics of fault cause subdivision on the alarm title: The alarm title in the work order is used as a key, and the feature vector vec_6 corresponding to the alarm title is queried in the dictionary dict_5.

7) The distribution characteristics of the breakdown of the fault causes on the network element type: The network element type in the work order is used as the key, and the feature vector vec_7 corresponding to the network element type is queried in the dictionary dict_6.

Here, the above-mentioned vec_1, vec_2, vec_3, vec_4, and vec_5 belong to the first-class classification feature vectors, that is, the first-class feature vectors in the foregoing embodiment.

S3: Input the feature vector of the first-level classification into model_1 for classification.

Here, the extracted feature vectors vec_1, vec_2, vec_3, vec_4, and vec_5 are spliced, and then input to model_1 for classification, and the probability p_1, ···, p_N of each fault cause category is obtained.

S4: Select the fault cause category i with the highest probability in the prediction result.

Among them, the prediction result is the probability p_1, ···, p_N belonging to each fault cause category, and the category i with the largest probability value (1≤i≤N) is the category to which the work order belongs in the first step of classification.

S5: Extract the feature vector of the secondary classification of the work order.

The above-mentioned vec_1, vec_2, vec_3, vec_4, vec_5, vec_6, and vec_7 belong to the feature vectors of the secondary classification, that is, the second type of feature vectors in the foregoing embodiment. For the specific extraction process, please refer to the description of the S2 part, which will not be repeated here.

S6: Input the feature vector of the secondary classification into model_2_i for classification.

Here, the extracted feature vectors vec_1, vec_2, vec_3, vec_4, vec_5, vec_6 and vec_7 are spliced, and then input into model_2_i for classification, and the probability of each fault cause subdivision belonging to the fault cause category i is obtained.

S7: Select the fault cause subdivision j with the highest probability in the prediction result.

其中概率值最大的类别j(1≤j≤n_i)即为该工单在第二步分类时所属的类别，即最终的故障原因细分类别。Among them, the prediction result is the probability of each fault cause subdivision in the fault cause category i

The category j with the largest probability value (1≤j≤n_i) is the category to which the work order belongs in the second step of classification, that is, the final breakdown cause classification category.

In the method for predicting the cause of a network fault according to the embodiment of the present invention, the classification feature vector in the fault work order is obtained, and the classification feature vector includes a first-class feature vector and a second-class feature vector; according to the first-class feature vector and the second-class feature vector a classification prediction model, to obtain the target failure cause category to which the fault work order belongs; The target fault cause sub-category in the target fault cause category. In this way, through a two-step prediction method, that is, first predict the fault cause category, and then predict the sub-categories of the fault cause in the fault cause category. It can effectively reduce the number of categories predicted at each step and improve the accuracy of the prediction results.

As shown in FIG. 4 , an embodiment of the present invention further provides an apparatus for predicting the cause of a network fault, and the apparatus includes:

The first obtaining module 401 is configured to obtain a classification feature vector in the trouble ticket, where the classification feature vector includes a first-type feature vector and a second-class feature vector;

A first fault cause prediction module 402, configured to obtain the target fault cause category to which the fault work order belongs according to the first type of feature vector and the first classification prediction model;

The second fault cause prediction module 403 is configured to obtain, according to the second type feature vector and the second classification prediction model corresponding to the target fault cause category, the fault work order in the target fault cause category Target failure cause subcategory.

Optionally, the first obtaining module 401 includes:

a first acquiring unit, configured to acquire a fault work order to be processed, the fields of the fault work order include an alarm title, a network element name, a network element type, and a fault occurrence time;

A feature extraction unit, configured to extract the classification feature vector in the trouble ticket based on the corresponding relationship between the fields of the trouble ticket and the feature vector and/or the feature extraction model.

Optionally, the classification feature vector includes:

a first feature vector for characterizing the alert title;

a second feature vector used to characterize the fault cause category corresponding to the alarm title;

a third feature vector for characterizing the network element type;

a fourth feature vector used to characterize the failure cause category corresponding to the network element type;

a fifth feature vector used to represent the alarm information associated with the trouble ticket;

a sixth feature vector for characterizing the fault cause sub-category corresponding to the alarm title; and

a seventh feature vector used to characterize the fault cause sub-category corresponding to the network element type;

Wherein, the first type of eigenvectors include: the first eigenvector, the second eigenvector, the third eigenvector, the fourth eigenvector, and the fifth eigenvector;

The second type of eigenvectors include: the first eigenvector, the second eigenvector, the third eigenvector, the fourth eigenvector, the fifth eigenvector, and the sixth eigenvector and the seventh eigenvector.

Optionally, the first fault cause prediction module 402 includes:

a first processing unit, configured to classify the first-type feature vector by using the first classification prediction model to obtain the probability value of each fault cause category;

The second processing unit is configured to determine the failure cause category corresponding to the maximum probability value among the probability values of each failure cause category as the target failure cause category to which the failure work order belongs.

Optionally, the second fault cause prediction module 403 includes:

a third processing unit, configured to classify the feature vector of the second type by using the second classification prediction model, and obtain the probability value of each fault cause sub-category of the fault work order in the target fault cause category;

The fourth processing unit is configured to determine the fault cause sub-category corresponding to the maximum probability value among the probability values of each fault cause sub-category as the target fault cause sub-category.

Optionally, the device further includes:

The second acquisition module is used to acquire multiple historical fault work orders and multiple historical alarm information. The fields of each historical fault work order include the alarm title, network element name, network element type, fault occurrence time, and fault cause category. and the fault cause sub-category corresponding to the fault cause category, and the fields of each piece of historical alarm information include the alarm title, the name of the network element, and the alarm start time;

a first processing module, configured to obtain a classification feature vector according to the field of the historical fault ticket and the field of the historical alarm information, where the classification feature vector includes: a first feature vector used to represent the alarm title, a second eigenvector used to characterize the fault cause category corresponding to the alarm title, a third eigenvector used to characterize the network element type, and a fourth eigenvector used to characterize the fault cause category corresponding to the network element type , used to represent the fifth feature vector of the alarm information associated with the fault ticket, used to represent the sixth feature vector of the fault cause sub-category corresponding to the alarm title, and used to represent the fault cause sub-category corresponding to the network element type the seventh eigenvector of the category;

A first model training module, configured to classify labels according to the first feature vector, the second feature vector, the third feature vector, the fourth feature vector, the fifth feature vector and the fault cause category , and perform model training to obtain the first classification prediction model.

Optionally, the device further includes:

The second processing module is configured to group a plurality of the historical fault work orders according to the fault cause category to obtain multiple sets of historical fault work order data;

The second model training module is used to perform model training on each group of historical fault work order data according to the category label and classification feature vector of the fault cause sub-category corresponding to each group of historical fault work order data to obtain multiple second classification predictions Model.

The apparatus for predicting the cause of a network fault in the embodiment of the present invention obtains the classification feature vector in the failure work order, where the classification feature vector includes a first-class feature vector and a second-class feature vector; according to the first-class feature vector and the second-class feature vector a classification prediction model, to obtain the target failure cause category to which the fault work order belongs; The target fault cause sub-category in the target fault cause category. In this way, through a two-step prediction method, that is, first predict the fault cause category, and then predict the sub-categories of the fault cause in the fault cause category. It can effectively reduce the number of categories predicted at each step and improve the accuracy of the prediction results.

It should be noted here that the above-mentioned device provided by the embodiment of the present invention can realize all the method steps realized by the above-mentioned method embodiment, and can achieve the same technical effect, and the same as the method embodiment in this embodiment is not repeated here. The parts and beneficial effects will be described in detail.

In order to better achieve the above purpose, as shown in FIG. 5 , an embodiment of the present invention further provides an electronic device, including a processor 500 and a transceiver 510, the transceiver 510 receives and sends data under the control of the processor, The processor 500 is configured to perform the following processes:

obtaining a classification feature vector in the trouble ticket, where the classification feature vector includes a first-class feature vector and a second-class feature vector;

According to the first type of feature vector and the first classification prediction model, obtain the target fault cause category to which the fault work order belongs;

According to the second type feature vector and the second classification prediction model corresponding to the target fault cause category, the target fault cause subcategory of the fault work order in the target fault cause category is obtained.

Optionally, the processor 500 is further configured to:

Acquiring a to-be-processed fault ticket, where the fields of the fault ticket include an alarm title, a network element name, a network element type, and a fault occurrence time;

Based on the correspondence between the fields of the trouble ticket and the feature vector and/or the feature extraction model, the classification feature vector in the trouble ticket is extracted.

Optionally, the classification feature vector includes:

a first feature vector for characterizing the alert title;

a second feature vector used to characterize the fault cause category corresponding to the alarm title;

a third feature vector for characterizing the network element type;

a fourth feature vector used to characterize the failure cause category corresponding to the network element type;

a fifth feature vector used to represent the alarm information associated with the trouble ticket;

a sixth feature vector for characterizing the fault cause sub-category corresponding to the alarm title; and

a seventh feature vector used to characterize the fault cause sub-category corresponding to the network element type;

Wherein, the first type of eigenvectors include: the first eigenvector, the second eigenvector, the third eigenvector, the fourth eigenvector, and the fifth eigenvector;

The second type of eigenvectors include: the first eigenvector, the second eigenvector, the third eigenvector, the fourth eigenvector, the fifth eigenvector, and the sixth eigenvector and the seventh eigenvector.

Optionally, the processor 500 is further configured to:

Classify the first-class feature vector by using the first classification prediction model to obtain the probability value of each fault cause category;

The failure cause category corresponding to the maximum probability value among the probability values of each failure cause category is determined as the target failure cause category to which the failure work order belongs.

Optionally, the processor 500 is further configured to:

Classify the feature vector of the second type by using the second classification prediction model to obtain the probability value of each fault cause sub-category of the fault work order in the target fault cause category;

The fault cause sub-category corresponding to the maximum probability value among the probability values of each fault cause sub-category is determined as the target fault cause sub-category.

Optionally, the processor 500 is further configured to:

Obtain multiple historical fault tickets and multiple historical alarm information. The fields of each historical fault ticket include the alarm title, NE name, NE type, fault occurrence time, fault cause category, and the fault corresponding to the fault cause category. Cause sub-category, each field of the historical alarm information includes the alarm title, network element name and alarm start time;

According to the field of the historical trouble ticket and the field of the historical alarm information, a classification feature vector is obtained, and the classification feature vector includes: a first feature vector used to represent the alarm title, used to represent the alarm title The second eigenvector of the corresponding fault cause category, used to characterize the third eigenvector of the network element type, used to characterize the fourth eigenvector of the fault cause category corresponding to the NE type, used to characterize the fault The fifth eigenvector of the alarm information associated with the work order, the sixth eigenvector used to represent the fault cause subcategory corresponding to the alarm title, and the seventh eigenvector used to represent the fault cause subcategory corresponding to the network element type;

According to the first eigenvector, the second eigenvector, the third eigenvector, the fourth eigenvector, the fifth eigenvector, and the category label of the fault cause category, perform model training to obtain the first Classification prediction model.

Optionally, the processor 500 is further configured to:

Grouping a plurality of the historical fault work orders according to the fault cause category to obtain multiple sets of historical fault work order data;

According to the category label and classification feature vector of the fault cause sub-category corresponding to each group of historical fault work order data, model training is performed on each group of historical fault work order data to obtain a plurality of second classification prediction models.

In the electronic device according to the embodiment of the present invention, by acquiring the classification feature vector in the fault work order, the classification feature vector includes a first-class feature vector and a second-class feature vector; predicting based on the first-class feature vector and the first classification feature vector model to obtain the target failure cause category to which the fault work order belongs; according to the second type feature vector and the second classification prediction model, obtain the target failure cause sub-category of the fault work order in the target failure cause category , so, through the two-step prediction method, that is, first predicting the major categories of fault causes, and then predicting the sub-categories of the fault causes in the major categories of fault causes, which can effectively reduce the number of categories predicted in each step and improve the prediction. accuracy of the results.

An embodiment of the present invention further provides an electronic device, including a memory, a processor, and a computer program stored on the memory and running on the processor, where the processor implements the above-mentioned program when the processor executes the program The various processes in the embodiments of the method for predicting the cause of a network fault can achieve the same technical effect, and are not repeated here to avoid repetition.

Embodiments of the present invention also provide a computer-readable storage medium on which a computer program is stored, and when the program is executed by a processor, implements each process in the above-mentioned embodiment of the method for predicting a cause of a network failure, and can achieve the same The technical effect, in order to avoid repetition, will not be repeated here. The computer-readable storage medium is, for example, a read-only memory (Read-Only Memory, ROM for short), a random access memory (Random Access Memory, RAM for short), a magnetic disk, or an optical disk.

As will be appreciated by one skilled in the art, the embodiments of the present application may be provided as a method, a system or a computer program product. Accordingly, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present application may take the form of a computer program product embodied on one or more computer-readable storage media having computer-usable program code embodied therein, including but not limited to disk storage, optical storage, and the like.

The present application is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems) and computer program products according to embodiments of the present application. It will be understood that each flow and/or block in the flowchart illustrations and/or block diagrams, and combinations of flows and/or blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to the processor of a general purpose computer, special purpose computer, embedded processor or other programmable data processing device to produce a machine such that the instructions executed by the processor of the computer or other programmable data processing device produce Means for implementing the functions specified in a flow or flow and/or a block or blocks of the flowchart.

These computer program instructions may also be stored in a computer-readable storage medium capable of directing a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce a paper product comprising the instruction means, The instruction means implements the functions specified in the flow or flows of the flowcharts and/or the block or blocks of the block diagrams.

These computer program instructions can also be loaded onto a computer or other programmable data processing device to cause the computer or other programmable device to perform a series of operational steps to produce a computer-implemented process, whereby the instructions to be executed on the computer or other programmable device Steps are provided for implementing the functions specified in a flow or flows of the flowcharts and/or a block or blocks of the block diagrams.

The above are the preferred embodiments of the present invention. It should be pointed out that for those skilled in the art, without departing from the principles of the present invention, several improvements and modifications can be made. These improvements and modifications It should also be regarded as the protection scope of the present invention.

Claims (17)

1. A method for predicting a cause of a network fault includes:

obtaining a classification feature vector in a fault work order, wherein the classification feature vector comprises a first class feature vector and a second class feature vector;

obtaining the target fault reason category to which the fault work order belongs according to the first class feature vector and a first classification prediction model;

and obtaining a target fault cause sub-category of the fault work order in the target fault cause category according to the second class of feature vectors and a second classification prediction model corresponding to the target fault cause category.

2. The method of claim 1, wherein the obtaining classified feature vectors in the faulty work order comprises:

acquiring a fault work order to be processed, wherein fields of the fault work order comprise an alarm title, a network element name, a network element type and fault occurrence time;

and extracting the classification characteristic vector in the fault work order based on the corresponding relation between the field of the fault work order and the characteristic vector and/or a characteristic extraction model.

3. The method of claim 2, wherein the classifying the feature vector comprises:

a first feature vector for characterizing the alert caption;

the second characteristic vector is used for representing the fault reason category corresponding to the alarm title;

a third feature vector for characterizing the network element type;

a fourth feature vector for characterizing a fault cause category corresponding to the network element type;

a fifth feature vector used for representing the alarm information associated with the fault work order;

a sixth feature vector used for characterizing the fault cause subcategory corresponding to the alarm title; and

a seventh feature vector for characterizing a fault cause subcategory corresponding to the network element type;

wherein the first class of feature vectors comprises: the first feature vector, the second feature vector, the third feature vector, the fourth feature vector, and the fifth feature vector;

the second class of feature vectors comprises: the first feature vector, the second feature vector, the third feature vector, the fourth feature vector, the fifth feature vector, the sixth feature vector, and the seventh feature vector.

4. The method according to claim 1, wherein the obtaining a target fault cause category to which the fault work order belongs according to the first class eigenvector and the first classification prediction model comprises:

classifying the first class of feature vectors through the first classification prediction model to obtain probability values of all fault reason classes;

and determining the fault reason category corresponding to the maximum probability value in the probability values of all fault reason categories as a target fault reason category to which the fault work order belongs.

5. The method of claim 1, wherein obtaining the target fault cause subcategory of the fault work order in the target fault cause category according to the second class feature vector and a second classification prediction model corresponding to the target fault cause category comprises:

classifying the second class of feature vectors through the second classification prediction model to obtain probability values of fault cause sub-categories of the fault work order in the target fault cause category;

and determining the fault reason subcategory corresponding to the maximum probability value in the probability values of the fault reason subcategories as a target fault reason subcategory.

6. The method of claim 1, further comprising:

acquiring a plurality of historical fault work orders and a plurality of historical alarm information, wherein the field of each historical fault work order comprises an alarm title, a network element name, a network element type, fault occurrence time, a fault reason category and a fault reason subcategory corresponding to the fault reason category, and the field of each historical alarm information comprises an alarm title, a network element name and alarm starting time;

obtaining a classification feature vector according to the field of the historical fault work order and the field of the historical alarm information, wherein the classification feature vector comprises: a first feature vector used for characterizing the alarm title, a second feature vector used for characterizing a fault cause category corresponding to the alarm title, a third feature vector used for characterizing the network element type, a fourth feature vector used for characterizing the fault cause category corresponding to the network element type, a fifth feature vector used for characterizing alarm information associated with the fault work order, a sixth feature vector used for characterizing a fault cause sub-category corresponding to the alarm title, and a seventh feature vector used for characterizing the fault cause sub-category corresponding to the network element type;

and performing model training according to the first feature vector, the second feature vector, the third feature vector, the fourth feature vector, the fifth feature vector and the class label of the fault reason class to obtain a first class prediction model.

7. The method of claim 6, wherein after obtaining the classification feature vector according to the fields of the historical fault work order and the fields of the historical alarm information, the method further comprises:

according to the fault reason category, grouping a plurality of historical fault work orders to obtain a plurality of groups of historical fault work order data;

and respectively carrying out model training on each group of historical fault work order data according to the class label and the classification characteristic vector of the fault reason subclass corresponding to each group of historical fault work order data to obtain a plurality of second classification prediction models.

8. A network failure cause prediction apparatus, comprising:

the first obtaining module is used for obtaining a classification feature vector in the fault work order, wherein the classification feature vector comprises a first class feature vector and a second class feature vector;

the first fault cause prediction module is used for obtaining the target fault cause category to which the fault work order belongs according to the first class feature vector and a first classification prediction model;

and the second fault cause prediction module is used for obtaining a target fault cause sub-category of the fault work order in the target fault cause category according to the second class of feature vectors and a second classification prediction model corresponding to the target fault cause category.

9. An electronic device comprising a processor and a transceiver, the transceiver receiving and transmitting data under control of the processor, characterized in that the processor is adapted to:

obtaining a classification feature vector in a fault work order, wherein the classification feature vector comprises a first class feature vector and a second class feature vector;

obtaining the target fault reason category to which the fault work order belongs according to the first class feature vector and a first classification prediction model;

and obtaining a target fault cause sub-category of the fault work order in the target fault cause category according to the second class feature vector and a second classification prediction model.

10. The electronic device of claim 9, wherein the processor is further configured to:

acquiring a fault work order to be processed, wherein fields of the fault work order comprise an alarm title, a network element name, a network element type and fault occurrence time;

and extracting the classification characteristic vector in the fault work order based on the corresponding relation between the field of the fault work order and the characteristic vector and/or a characteristic extraction model.

11. The electronic device of claim 10, wherein the classification feature vector comprises:

a first feature vector for characterizing the alert caption;

the second characteristic vector is used for representing the fault reason category corresponding to the alarm title;

a third feature vector for characterizing the network element type;

a fourth feature vector used for characterizing the fault reason category corresponding to the network element type;

a fifth feature vector used for representing the alarm information associated with the fault work order;

a sixth feature vector used for characterizing the fault cause subcategory corresponding to the alarm title; and

a seventh feature vector for characterizing a fault cause subcategory corresponding to the network element type;

wherein the first class of feature vectors comprises: the first, second, third, fourth, and fifth feature vectors;

the second class of feature vectors includes: the first feature vector, the second feature vector, the third feature vector, the fourth feature vector, the fifth feature vector, the sixth feature vector, and the seventh feature vector.

12. The electronic device of claim 9, wherein the processor is further configured to:

classifying the first class of feature vectors through the first classification prediction model to obtain probability values of all fault reason classes;

and determining the fault reason category corresponding to the maximum probability value in the probability values of all fault reason categories as a target fault reason category to which the fault work order belongs.

13. The electronic device of claim 9, wherein the processor is further configured to:

classifying the second class of feature vectors through the second classification prediction model to obtain probability values of fault cause sub-categories of the fault work order in the target fault cause category;

and determining the fault reason subcategory corresponding to the maximum probability value in the probability values of the fault reason subcategories as a target fault reason subcategory.

14. The electronic device of claim 9, wherein the processor is further configured to:

acquiring a plurality of historical fault work orders and a plurality of historical alarm information, wherein the field of each historical fault work order comprises an alarm title, a network element name, a network element type, fault occurrence time, a fault reason category and a fault reason subcategory corresponding to the fault reason category, and the field of each historical alarm information comprises an alarm title, a network element name and alarm starting time;

obtaining a classification feature vector according to the field of the historical fault work order and the field of the historical alarm information, wherein the classification feature vector comprises: the first feature vector is used for representing the alarm title, the second feature vector is used for representing the fault cause category corresponding to the alarm title, the third feature vector is used for representing the network element type, the fourth feature vector is used for representing the fault cause category corresponding to the network element type, the fifth feature vector is used for representing the alarm information associated with the fault work order, the sixth feature vector is used for representing the fault cause sub-category corresponding to the alarm title, and the seventh feature vector is used for representing the fault cause sub-category corresponding to the network element type;

and performing model training according to the first feature vector, the second feature vector, the third feature vector, the fourth feature vector, the fifth feature vector and the class label of the fault reason class to obtain a first classification prediction model.

15. The electronic device of claim 14, wherein the processor is further configured to:

according to the fault reason category, grouping a plurality of historical fault work orders to obtain a plurality of groups of historical fault work order data;

and respectively carrying out model training on each group of historical fault work order data according to the category label and the classification characteristic vector of the fault reason sub-category corresponding to each group of historical fault work order data to obtain a plurality of second classification prediction models.

16. An electronic device comprising a memory, a processor, and a program stored on the memory and executable on the processor; characterized in that the processor implements the method for predicting the cause of network failure according to any one of claims 1 to 7 when executing the program.

17. A computer-readable storage medium, on which a computer program is stored, which program, when being executed by a processor, carries out the steps of the method for predicting a cause of a network failure as set forth in any one of claims 1 to 7.

Images