CN116915630A - Network jam prediction methods, devices, electronic equipment, media and program products - Google Patents

Abstract

The application discloses a network katon prediction method, a network katon prediction device, electronic equipment, a medium and a program product. The network jamming prediction method comprises the following steps: acquiring historical internet surfing record data corresponding to each historical preset time period of a user; for each history preset time period, determining whether a history label of network blocking occurs in the history preset time period and network quality index data in the history preset time period based on the history internet log data corresponding to the history preset time period; taking each history label and network quality index data corresponding to each history label as training samples, and training a network katon prediction model to obtain a trained network katon prediction model; and obtaining a target label of the occurrence of the jamming in the network in the time period to be predicted based on the trained network jamming prediction model and the network quality index data in the time period to be predicted. The method can realize the effect of accurately predicting the user's stuck behavior in any time period to be predicted.

Description

Network jam prediction methods, devices, electronic equipment, media and program products

Technical field

This application relates to the field of data processing technology, and specifically relates to a network jam prediction method, device, electronic equipment, media and program products.

Background technique

In order to ensure the development of video services and user perception of video service access, detecting and evaluating the network quality of video services is also an indispensable step in the development process of video services. In order to monitor the quality of user access to video services, there are currently two solutions: One is to use dial detection technology, and the other is to detect based on deep packet inspection (Deep Packet Inspection, DPI) technology.

The inventor found that although the indicators obtained by the dial test technology can evaluate the user's perception of accessing the video service from the user side, the dial test is always conducted under specific conditions and cannot comprehensively and truly evaluate the user's perception. In addition, DPI technology cannot process and adjust related network links in advance to avoid possible lags and affect the user's online experience.

Contents of the invention

The purpose of the embodiments of the present application is to provide a network jam prediction method, device, electronic equipment, media and program products to achieve a comprehensive prediction effect of whether the network is stuck.

The technical solution of this application is as follows:

The first aspect provides a network jam prediction method, which includes:

Obtain the user's historical Internet access record data corresponding to each historical preset time period;

For each historical preset time period, based on the historical Internet access record data corresponding to the historical preset time period, determine the historical tag of whether network lag occurred during the historical preset time period, and the network quality indicator data within the historical preset time period. ;

Use each historical tag and the network quality index data corresponding to each historical tag as training samples to train the network stuck prediction model and obtain the trained network stuck prediction model;

Based on the trained network lag prediction model and the network quality indicator data in the period to be predicted, the target label of network lag in the period to be predicted is obtained.

In a second aspect, a network jam prediction device is provided, which device includes:

The first acquisition module is used to acquire the user's historical Internet access record data corresponding to each historical preset time period;

The first determination module is used for each historical preset time period, based on the historical Internet access record data corresponding to the historical preset time period, to determine whether the historical tag of network lag occurs within the historical preset time period, and the historical preset time Network quality indicator data within the segment;

The second determination module is used to use each historical label and the network quality indicator data corresponding to each historical label as training samples to train the network stuck prediction model and obtain the trained network stuck prediction model;

The third determination module is used to obtain the target label of network lag in the time period to be predicted based on the trained network lag prediction model and the network quality indicator data in the time period to be predicted.

In a third aspect, embodiments of the present application provide an electronic device. The electronic device includes a processor, a memory, and a program or instruction stored in the memory and executable on the processor. When the program or instruction is executed by the processor, the present invention is implemented. Apply the steps of the network jam prediction method in any embodiment.

In the fourth aspect, embodiments of the present application provide a readable storage medium. The readable storage medium stores programs or instructions. When the program or instructions are executed by a processor, the steps of implementing any of the network jam prediction methods in the embodiments of the present application are implemented. .

In a fifth aspect, embodiments of the present application provide a computer program product. When the instructions in the computer program product are executed by a processor of an electronic device, the electronic device can perform any of the steps of the network jam prediction method in the embodiments of the present application. .

The technical solutions provided by the embodiments of the present application at least bring the following beneficial effects:

In the embodiment of this application, by obtaining the historical Internet access record data corresponding to each historical preset time period of the user, for each historical preset time period, the historical preset time period is determined based on the historical Internet access record data corresponding to the historical preset time period. The historical labels of whether network lag occurs within the period, and the network quality indicator data within the historical preset time period, and then use each historical label and the network quality indicator data corresponding to each historical label as training samples to train the network lag prediction model. Obtain the trained network lag prediction model, so that based on the trained network lag prediction model and the network quality indicator data in the time period to be predicted, the target label of network lag in the time period to be predicted can be obtained, so through Correlate user stuck characteristics and network quality indicator data, establish characteristic equations, train network stuck prediction models, and accurately predict whether the network is stuck based on the trained network stuck prediction model. This way, it is possible to predict whether the network is stuck in any time period to be predicted. Both can accurately predict user freezing behavior.

It should be understood that the above general description and the following detailed description are only exemplary and explanatory, and do not limit the present application.

Description of the drawings

The drawings herein are incorporated into the specification and constitute a part of the specification, illustrate embodiments consistent with the present application, and are used together with the description to explain the principles of the present application, and do not constitute undue limitations on the present application.

Figure 1 is a schematic flowchart of a network jam prediction method provided by an embodiment of the present application;

Figure 2 is a system architecture diagram of a network jam prediction method provided by an embodiment of the present application;

Figure 3 is another implementable manner of the network jam prediction method provided by the embodiment of the present application;

Figure 4 is a schematic structural diagram of a network jam prediction device provided by the second embodiment of the present application;

FIG. 5 is a schematic structural diagram of an electronic device provided by the third embodiment of the present application.

Detailed ways

In order to enable ordinary people in the art to better understand the technical solutions of the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below in conjunction with the accompanying drawings. It should be understood that the specific embodiments described here are only intended to explain the application, but not to limit the application. It will be apparent to one skilled in the art that the present application may be practiced without some of these specific details. The following description of embodiments is merely intended to provide a better understanding of the present application by illustrating examples thereof.

It should be noted that the terms "first", "second", etc. in the description and claims of this application and the above-mentioned drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It is to be understood that the data so used are interchangeable under appropriate circumstances so that the embodiments of the application described herein can be practiced in sequences other than those illustrated or described herein. The implementations described in the following exemplary embodiments do not represent all implementations consistent with this application. Rather, they are merely examples consistent with aspects of the application as detailed in the appended claims.

As shown in the background technology section, in order to ensure the development of video services and user perception of video service access, detecting and evaluating the network quality of video services is also an indispensable step in the development process of video services. In order to monitor the quality of user access to video services, the current There are two solutions: one is to use dial detection technology, and the other is to detect based on deep packet inspection (Deep Packet Inspection, DPI) technology.

The inventor found that although the indicators obtained by the dial test technology can evaluate the user's perception of accessing the video service from the user side, the dial test is always conducted under specific conditions and cannot comprehensively and truly evaluate the user's perception. In addition, DPI technology cannot process and adjust related network links in advance to avoid possible lags and affect the user's online experience.

In order to solve the above problems, embodiments of the present application provide a network jam prediction method, device, equipment, media and computer program products. By obtaining the user's historical Internet access record data corresponding to each historical preset time period, for each historical prediction Set a time period, based on the historical Internet access record data corresponding to the historical preset time period, determine the historical tag of whether network lag occurs within the historical preset time period, and the network quality indicator data within the historical preset time period, and then combine each historical tags, and the network quality index data corresponding to each historical tag are used as training samples to train the network jam prediction model and obtain the trained network jam prediction model. This can be based on the trained network jam prediction model and the time period to be predicted. Based on the network quality index data within the time period to be predicted, the target label of network lag in the period to be predicted is obtained. In this way, by correlating the user lag characteristics with the network quality index data, a characteristic equation is established to train the network lag prediction model. Based on the well-trained The network stuck prediction model accurately predicts whether the network is stuck, so that user stuck behavior can be accurately predicted in any time period to be predicted.

Embodiments of the present application provide a network jam prediction method, device, equipment, storage medium and product. The following first introduces the network jam prediction method provided by the embodiment of the present application.

Figure 1 is a schematic flowchart of a network jam prediction method provided by an embodiment of the present application. As shown in Figure 1, the method includes:

S110. Obtain the user's historical Internet access record data corresponding to each historical preset time period.

The historical preset time period may be a time period during which the user accesses resources, for example, it may be a time period during which the user watches a video.

Historical Internet access record data can include any one or more of the user's Internet Protocol (IP), resource IP, domain name information or Universal Resource Locator (URL) information when the user accesses resources. .

Obtain the user's historical Internet access record data corresponding to each historical preset time period. Specifically, you can use Deep Packet Inspection (DPI) probes to send the collected packets to the server, or you can use other methods to obtain the above data. , no limitation is made here.

As an example, use the deep packet parsing capability of the DPI probe and the synthesis function of the terminal detection and response (Extended Detection and Response, By capturing the packets generated when watching a website, the corresponding historical Internet record data generated when watching a certain video website can be obtained.

S120. For each historical preset time period, based on the historical Internet access record data corresponding to the historical preset time period, determine the historical tag of whether network lag occurs within the historical preset time period, and the network quality within the historical preset time period. indicator data.

The historical label may be a label used to indicate whether network lag occurs within a preset historical time period, for example, it may be "stuck" or "no lag".

In some embodiments, in order to more accurately determine the historical label of whether network lag occurs within the historical preset time period, the network quality indicator data may at least include: uplink rate, downlink rate, server delay, client delay, response Delay, uplink retransmission rate, downlink retransmission rate, access success rate, client failure rate, and server failure rate.

S130. Use each historical label and the network quality index data corresponding to each historical label as a training sample to train the network stuck prediction model to obtain the trained network stuck prediction model.

The network lag prediction model can be a model used to predict whether lag will occur within a certain period of time. Specifically, it can be one of any machine learning models, or it can be a prediction model preset based on the actual situation. This is not the case here. Make limitations.

S140. Based on the trained network jam prediction model and the network quality indicator data in the time period to be predicted, obtain the target label for network jams in the time period to be predicted.

The target label may be a label used to characterize whether the network is stuck in the time period to be predicted, for example, it may be "stuck" or "non-stuck". Specifically, the label can be obtained by using the network quality index data in the time period to be predicted as the input of the network jam prediction model and using the target label as the output, thereby obtaining the label.

In this way, by obtaining the historical Internet access record data corresponding to each historical preset time period of the user, for each historical preset time period, based on the historical Internet access record data corresponding to the historical preset time period, it is determined whether a network incident occurred within the historical preset time period. The historical tags of lags and the network quality indicator data within the historical preset time period are then used as training samples to train the network lag prediction model and obtain the trained Network jam prediction model, so that based on the trained network jam prediction model and the network quality indicator data in the time period to be predicted, the target label of network jams in the time period to be predicted can be obtained, so as to prevent users from being stuck. Features are associated with network quality index data, feature equations are established, network stuck prediction models are trained, and network stuck prediction models are accurately predicted based on the trained network stuck prediction model. In this way, users can be predicted in any time period to be predicted. Accurate prediction of stuck behavior.

In some embodiments, in order to more accurately determine the historical tag of whether network lag occurs within the historical preset time period, the historical Internet access record data may include the unified resource identifier of the user's Internet access. The above S120 may specifically include:

Delete the unified resource identifiers that do not contain preset information among the unified resource identifiers corresponding to the historical preset time period to obtain the first unified resource identifier; where the preset information at least includes files, path information and query parameters; query parameters Characteristic information used to characterize network lag;

Split the first unified resource identifier according to the domain name and the key-value pairs corresponding to the query parameters to obtain a set of alternative features;

Based on the set of alternative features, determine the historical label of whether network stalling occurred within the historical preset time period.

In some embodiments, the Uniform Resource Identifier (URI) may be a string used to identify a certain Internet resource name. This identification allows users to interact with any resource (including local and Internet) through specific protocols. URI is one of the important fields that needs to be included in the XDR bill.

The first uniform resource identifier may be the same resource identifier obtained by deleting the uniform resource identifiers that do not contain preset information among the uniform resource identifiers corresponding to the historical preset time period.

The path information may be the path information of the network resources viewed by the user when surfing the Internet.

The candidate feature set may be a set of split key-value pairs obtained by splitting the first uniform resource identifier according to the domain name and the key-value pairs corresponding to the query parameters.

In some embodiments, the query parameters may be, for example, path and query parameters, and the key-value pair may be the key and value in the parameters.

Delete the uniform resource identifiers that do not contain preset information among the uniform resource identifiers corresponding to the historical preset time period to obtain the first uniform resource identifier. Specifically, after obtaining multiple strings of URIs, delete the uniform resource identifiers that do not contain preset information. URI to delete.

In some embodiments, the first uniform resource identifier is divided according to the domain name and the key-value pairs corresponding to the query parameters to obtain the candidate feature set. Specifically, specific symbols may be used to segment each string of URIs and extract the domain name, Query URIs with the same parameters to obtain a set of alternative features.

In some embodiments, based on the alternative feature set, determining the historical label of whether network lag occurs within the historical preset time period may be to use the key value representing the lag in the set as the historical label.

As an example, based on the above description of URIs, first filter out URIs that do not contain files, paths, or query parameters, and then use ‘:’, ‘/’, ‘.’, ‘-’, ‘? ', '&' and other symbols to split the remaining URIs, and extract the key and value values in the domain name, path and query parameters of the URI, as shown in the following table (1):

Table 1 Example table of first uniform resource identifiers

In one example, you can create lags when watching videos through methods such as artificial speed limits, watch videos and capture packets for analysis. Assume that user U experiences lags when watching video episode A and video episode B, and obtains a total of 5 A domain name and 5 URLs with query parameters were split into a total of 15 keys and values. The domain names, keys and values may be the same or different. Select the combination of domain name, key and value with the same value as the candidate feature set.

In this way, by obtaining the uniform resource identifiers within the preset time period of the user's history, filtering the identifiers, and dividing the obtained uniform resource identifiers according to domain names and key-value pairs corresponding to the query parameters, an alternative feature set is obtained , the candidate feature set obtained in this way contains feature information used to characterize network jams. Therefore, it can be determined based on this set whether network jams have occurred within the historical preset time period, which improves accuracy.

In some embodiments, in order to further improve the accuracy of historical tags for determining whether network stalling occurs within a historical preset time period, based on a set of alternative features, determine historical tags for whether network stalling occurs within a historical preset time period, Specifics may include:

Obtain respectively the first domain name and the first key-value pair of the query parameter when the network is stuck within the historical preset time period, and the second key-value pair of the second domain name and the query parameter when the network is not stuck;

Compare the first domain name, the first key-value pair, the second domain name and the second key-value pair with the domain names and key-value pairs in the alternative feature set, and select the domain name that is the same as the domain name in the alternative feature set and is the same as the domain name in the alternative feature set. The key attribute in the key-value pair in the candidate feature set is the same and the value attribute that is different from the value attribute in the key-value pair in the candidate feature set is determined as the target value attribute;

Based on the target value attribute, determine the historical tag of whether network lag occurs within the historical preset time period.

The first domain name may be the domain name when the network was stuck during a preset historical time period.

The first key-value pair may be a key-value pair of query parameters when the network is stuck within a historical preset time period.

The second domain name may be the domain name when the network is not stuck within a preset historical time period.

The second key-value pair may be a key-value pair of query parameters when the network is not stuck within a historical preset time period.

The selected target value attribute can be the same as the domain name in the candidate feature set, the same as the key attribute in the key-value pair in the candidate feature set, and the same as the value attribute in the key-value pair in the candidate feature set. Different value attributes.

As an example, in the above example, for user U, by observing the domain name, key, and value generated when user U watches video episode A, the domain name, key, and value generated when the user U is stuck and not stuck, and the domain name, key, and value generated when watching video episode B. Key and value values, compare the domain name, key and value value in the candidate feature set mentioned above, and select the combination of the domain name, key value and different value value in the candidate feature set as the target value attribute.

In this way, by obtaining the first key-value pair of the first domain name and the query parameter when the network is stuck within the historical preset time period, and the second key-value pair of the second domain name and the query parameter when the network is not stuck, and the alternative Compare the domain names and key-value pairs in the feature set, and select the domain name that is the same as the domain name in the candidate feature set, the same key attribute as the key-value pair in the candidate feature set, and the same as the key-value pair in the candidate feature set. Value attributes with different value attributes in the key-value pair are determined as target value attributes. This way, the attribute value when the network is stuck can be matched with the value in the alternative feature set, and further confirmed as the attribute value characterizing the stuck, narrowing the scope. , making the obtained historical labels more accurate.

In some embodiments, in order to more accurately obtain the network quality index data within the historical preset time period, the historical Internet access record data may also include: network feature information; the above S120 may include:

Based on the network characteristic information corresponding to the historical preset time period, the network quality indicator data within the historical preset time period is calculated.

In some embodiments, the network characteristic information may specifically be characteristics used to calculate network quality indicator data, such as download traffic, download duration, upload traffic, upload duration, or client-initiated synchronization sequence numbers (SynchronizeSequenceNumbers, SYN) to the server. Returns information such as the delay of response confirmation (SYN ACK).

As an example, the network quality indicator data can be a quality indicator that reflects the user's perception of Internet access. Specifically, it can include uplink rate, downlink rate, server delay, client delay, response delay, uplink retransmission rate, downlink retransmission rate, Access success rate, client failure rate, server failure rate.

The calculation method for each network quality indicator can be shown in Table 2:

Table 2 Example of calculation method for network quality indicators

In this way, the network quality index data within the historical preset time period is calculated by using the network feature information corresponding to the historical preset time period. Since the network feature information contains a variety of indicator information used to characterize the user's Internet access, and through the preset formula After calculation, the obtained network quality indicator data can include the user's lag information, thereby improving the accuracy of the network quality indicator data within the historical preset time period.

In some embodiments, in order to accurately obtain network quality index data, the network quality index data within the historical preset time period is calculated based on the network feature information corresponding to the historical preset time period, which may specifically include:

Clean the network feature information corresponding to the historical preset time period to obtain the target network feature information;

Based on the target network characteristic information, the network quality indicator data within the historical preset time period is calculated.

The target network feature information may be network feature information obtained by cleaning the network feature information corresponding to the historical preset time period.

In some embodiments of the present application, cleaning the network feature information corresponding to the historical preset time period may, but is not limited to, include deduplication processing, feature encoding processing, etc. on the network feature information corresponding to the historical preset time period.

In the embodiment of the present application, the target network characteristic information is obtained by cleaning the network characteristic information corresponding to the historical preset time period, and then based on the target network characteristic information, the network quality index data within the historical preset time period is calculated, so Better quality network feature information can be obtained, so that accurate network quality index data can be obtained based on the better quality network feature information.

In some embodiments, in order to more accurately obtain the target label of network jamming in the time period to be predicted, the network jamming prediction model may be a logistic regression model.

As an example, using the logistic regression model as the network jam prediction model, the training process can be: obtaining the user's real online records U ₀ in the past month from the big data platform distributed system infrastructure (Hadoop), including a total of m pieces of history Go online and record data. The data set U ₀ is processed to obtain the historical label of whether the user is stuck on the Internet as the dependent variable y (stuck occurs: 1; lag does not occur: 0). The data set U ₀ is processed to obtain the network quality index data of the user when surfing the Internet as the independent variable. X ₍ including _{x 1 ,} . Take _{out the m X<x 1 , x 2} ...

h(x)＝w ₀ +w ₁ x ₁ +w ₂ x ₂ +...+w _n x _n (1)

Among them, w is a parameter of the model, w ₀ is the intercept, and w ₁ to w _n are coefficients.

It can be understood that the task of the linear regression function is to construct the prediction function h _(x) to map the linear relationship between the input variable X and the label value y, so that h (x) = w ₀ +w ₁ x ₁ +w ₂ x ₂ +...+w _n x _n =0 is used as the decision boundary for whether stuck occurs, then when function h(x)>=0, y=1; when function h(x)<0, y=0.

Since {stuck: 1, not stuttered: 0} is an obvious two-classification problem, and the values of this model are continuous, a link function is introduced to transform the linear regression equation h(x) into g (h), and let the value of g(h) be distributed between (0, 1), and when g(z) is close to 0, the label of the sample is category 0 without freezing, and when g(z) is close to 1, the sample The label is Category 1 Caton, so a classification model is obtained. For logistic regression, this link function is the Sigmoid function, as shown in formula (2):

Put the linear regression function, that is, formula (1) into the Sigmoid function to get the final logistic regression function, as shown in formula (3):

Among them, x ₁ , x _{2 ...} transmission rate, access success rate, client failure rate, and server failure rate. Y(x) is the label value returned by logistic regression.

It can be understood that the values of y(x) are all between [0,1]. When y(x) tends to 1, the user is stuck. When y(x) tends to 0, then The user is not stuck.

In this way, by introducing the regression model as the network jam prediction model, and using each historical label and the network quality index data corresponding to each historical label as training samples, it can fit the actual network jam prediction situation and obtain the prediction more accurately. Target tags where the network is stuck during a certain period of time.

In order to understand the solution of the present application more clearly, the embodiment of the present application also provides a system architecture diagram of a network jam prediction method, as shown in Figure 2.

In some embodiments of the present application, in order to understand the technical solutions of the present application more clearly, the technical solutions of the present application are introduced below using specific scenarios.

The embodiment of the present application also provides another implementation method of the network jam prediction method. As shown in Figure 3, the network jam prediction method provided by the embodiment of the present application may include the following steps:

S310: Capture the packet to obtain the URI, and separate the domain name and query parameters of the URI.

S320: Obtain network characteristic information from the user's historical Internet access data.

S330: Filter the key key and value values in the URI and determine the target attribute value.

S340: Filter network feature information and calculate network quality index data.

S350: Establish a stuck prediction model.

S360, predicts user lag.

It should be noted that the execution subject of the network jam prediction method provided by the embodiments of the present application may be a network jam prediction device, or a control module in the network jam prediction device for executing the network jam prediction method.

Based on the same inventive concept as the above-mentioned network jam prediction method, this application also provides a network jam prediction device. The network jam prediction provided by the embodiment of the present application will be described in detail below with reference to Figure 4 .

Figure 4 is a schematic structural diagram of a network jam prediction device according to an exemplary embodiment.

As shown in Figure 4, the network jam prediction device 400 may include:

The first acquisition module 401 is used to acquire the user's historical Internet access record data corresponding to each historical preset time period;

The first determination module 402 is used to determine, for each historical preset time period, based on the historical Internet access record data corresponding to the historical preset time period, a historical tag of whether network lag occurs within the historical preset time period, and the historical preset Network quality indicator data within a time period;

The second determination module 403 is used to use each historical tag and the network quality index data corresponding to each historical tag as training samples to train the network stuck prediction model and obtain the trained network stuck prediction model;

The third determination module 404 is used to obtain the target label of network lag in the time period to be predicted based on the trained network lag prediction model and the network quality indicator data in the time period to be predicted.

In this way, by obtaining the historical Internet access record data corresponding to each historical preset time period of the user, for each historical preset time period, based on the historical Internet access record data corresponding to the historical preset time period, it is determined whether a network incident occurred within the historical preset time period. The historical tags of lags and the network quality indicator data within the historical preset time period are then used as training samples to train the network lag prediction model and obtain the trained Network jam prediction model, so that based on the trained network jam prediction model and the network quality indicator data in the time period to be predicted, the target label of network jams in the time period to be predicted can be obtained, so as to prevent users from being stuck. Features are associated with network quality index data, feature equations are established, network stuck prediction models are trained, and network stuck prediction models are accurately predicted based on the trained network stuck prediction model. In this way, users can be predicted in any time period to be predicted. Accurate prediction of stuck behavior.

In some embodiments, in order to more accurately determine the historical tag of whether network lag occurs within the historical preset time period, the historical Internet access record data may include the unified resource identifier of the user's Internet access. The above-mentioned first determination module 402 may specifically include the following: unit:

The deletion unit is used to delete the uniform resource identifiers that do not contain preset information among the uniform resource identifiers corresponding to the historical preset time period to obtain the first uniform resource identifier; wherein the preset information at least includes files, path information and Query parameters; query parameters are used to characterize the characteristic information of network lag;

A splitting unit used to split the first unified resource identifier according to the domain name and the key-value pairs corresponding to the query parameters to obtain a set of alternative features;

The determination unit is used to determine the historical label of whether network stalling occurs within the historical preset time period based on the candidate feature set.

In some embodiments, in order to further improve the accuracy of historical tags for determining whether network lag occurs within a historical preset time period, the above-mentioned determination unit may specifically include the following sub-units:

The acquisition subunit is used to respectively obtain the first key-value pair of the first domain name and the query parameter when the network is stuck within the historical preset time period, and the second key-value pair of the second domain name and the query parameter when the network is not stuck. ;

The comparison subunit is used to compare the first domain name, the first key-value pair, the second domain name and the second key-value pair with the domain names and key-value pairs in the alternative feature set, and select the ones that match the alternative feature set. The domain name in is the same, and the key attribute is the same as the key-value pair in the candidate feature set, and the value attribute that is different from the value attribute in the key-value pair in the candidate feature set is determined as the target value attribute;

Determine the subunit, which is used to determine the historical label of whether network stalling occurs within the historical preset time period based on the target value attribute.

In some embodiments, in order to more accurately obtain the network quality index data within the historical preset time period, the historical Internet access record data may also include network feature information. The above-mentioned first determination module 402 may specifically include the following units:

The calculation unit is used to calculate the network quality index data within the historical preset time period based on the network characteristic information corresponding to the historical preset time period.

The network jam prediction device provided by the embodiments of the present application can be used to execute the network jam prediction method provided by the above method embodiments. Its implementation principles and technical effects are similar, and for the sake of brief introduction, they will not be described again here.

Based on the same inventive concept, embodiments of the present application also provide an electronic device.

FIG. 5 is a schematic structural diagram of an electronic device provided by an embodiment of the present application. As shown in Figure 5, the electronic device may include a processor 501 and a memory 502 storing computer programs or instructions.

Specifically, the above-mentioned processor 501 may include a central processing unit (CPU), or an application specific integrated circuit (Application Specific Integrated Circuit, ASIC), or may be configured to implement one or more integrated circuits according to embodiments of the present invention.

Memory 502 may include bulk storage for data or instructions. By way of example and not limitation, the memory 502 may include a hard disk drive (HDD), a floppy disk drive, flash memory, an optical disk, a magneto-optical disk, a magnetic tape, or a Universal Serial Bus (USB) drive or two or more A combination of many of the above. Memory 502 may include removable or non-removable (or fixed) media, where appropriate. Where appropriate, the memory 502 may be internal or external to the integrated gateway disaster recovery device. In certain embodiments, memory 502 is non-volatile solid-state memory. The memory may include read-only memory (Read Only Memory image, ROM), random access memory (Random-Access Memory, RAM), magnetic disk storage media device, optical storage media device, flash memory device, electrical, optical or other physical/tangible Memory storage device. Thus, generally, memory includes one or more tangible (non-transitory) computer-readable storage media (e.g., memory devices) encoded with software including computer-executable instructions, and when the software is executed (e.g., by one or multiple processors), it may be operable to perform the operations described in the network jam prediction method provided in the above embodiment.

The processor 501 reads and executes the computer program instructions stored in the memory 502 to implement any of the network jam prediction methods in the above embodiments.

In one example, the electronic device may also include communication interface 503 and bus 510 . Among them, as shown in Figure 5, the processor 501, the memory 502, and the communication interface 503 are connected through the bus 510 and complete communication with each other.

The communication interface 503 is mainly used to implement communication between modules, devices, units and/or devices in the embodiment of the present invention.

Bus 510 includes hardware, software, or both, coupling the components of the electronic device to each other. By way of example, and not limitation, the bus may include Accelerated Graphics Port (AGP) or other graphics bus, Enhanced Industry Standard Architecture (EISA) bus, Front Side Bus (FSB), HyperTransport (HT) interconnect, Industry Standard Architecture (ISA) Bus, Infinite Bandwidth Interconnect, Low Pin Count (LPC) Bus, Memory Bus, Micro Channel Architecture (MCA) Bus, Peripheral Component Interconnect (PCI) Bus, PCI-Express (PCI-X) Bus, Serial Advanced Technology Attachment (SATA) bus, Video Electronics Standards Association Local (VLB) bus or other suitable bus or a combination of two or more of these. Where appropriate, bus 510 may include one or more buses. Although embodiments of the invention describe and illustrate a particular bus, the invention contemplates any suitable bus or interconnection.

The electronic device can execute the network jam prediction method in the embodiment of the present invention, thereby realizing the network jam prediction method described in Figures 1-3.

In addition, combined with the network jam prediction method in the above embodiment, embodiments of the present invention can provide a readable storage medium for implementation. The readable storage medium stores program instructions; when the program instructions are executed by the processor, any one of the network jam prediction methods in the above embodiments is implemented.

It is to be understood that this invention is not limited to the specific arrangements and processes described above and illustrated in the drawings. For the sake of brevity, detailed descriptions of known methods are omitted here. In the above embodiments, several specific steps are described and shown as examples. However, the method process of the present invention is not limited to the specific steps described and shown. Those skilled in the art can make various changes, modifications and additions, or change the order between steps after understanding the spirit of the present invention.

The functional blocks shown in the above structural block diagram can be implemented as hardware, software, firmware or a combination thereof. When implemented in hardware, it may be, for example, an electronic circuit, an application specific integrated circuit (ASIC), appropriate firmware, a plug-in, a function card, or the like. When implemented in software, elements of the invention are programs or code segments used to perform the required tasks. The program or code segments may be stored in a machine-readable medium or transmitted over a transmission medium or communications link via a data signal carried in a carrier wave. "Machine-readable medium" may include any medium capable of storing or transmitting information. Examples of machine-readable media include electronic circuits, semiconductor memory devices, ROM, flash memory, erasable ROM (EROM), floppy disks, CD-ROMs, optical disks, hard disks, fiber optic media, radio frequency (RF) links, and the like. Code segments may be downloaded via computer networks such as the Internet, intranets, and the like.

It should also be noted that the exemplary embodiments mentioned in the present invention describe some methods or systems based on a series of steps or devices. However, the present invention is not limited to the order of the above steps. That is to say, the steps may be performed in the order mentioned in the embodiments, or may be different from the order in the embodiments, or several steps may be performed simultaneously.

Aspects of the present application are described above with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems) and computer program products according to embodiments of the application. It will be understood that each block of the flowchart illustrations and/or block diagrams, and combinations of blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, or other programmable data processing apparatus to produce a machine such that execution of the instructions via the processor of the computer or other programmable data processing apparatus enables Implementation of the functions/actions specified in one or more blocks of a flowchart and/or block diagram. Such a processor may be, but is not limited to, a general-purpose processor, a special-purpose processor, a special application processor, or a field-programmable logic circuit. It will also be understood that each block in the block diagrams and/or flowchart illustrations, and combinations of blocks in the block diagrams and/or flowchart illustrations, can also be implemented by special purpose hardware that performs the specified functions or actions, or can be implemented by special purpose hardware and A combination of computer instructions.

The above are only specific implementations of the present invention. Those skilled in the art can clearly understand that for the convenience and simplicity of description, the specific working processes of the above-described systems, modules and units can be referred to the foregoing method embodiments. The corresponding process will not be described again here. It should be understood that the protection scope of the present invention is not limited thereto. Any person familiar with the technical field can easily think of various equivalent modifications or substitutions within the technical scope disclosed in the present invention, and these modifications or substitutions should be covered. within the protection scope of the present invention.

Claims (10)

1. A network jam prediction method, characterized in that the method includes: Obtain the user's historical Internet access record data corresponding to each historical preset time period; For each historical preset time period, based on the historical Internet access record data corresponding to the historical preset time period, determine the historical tag of whether network lag occurs within the historical preset time period, and the historical preset time period Network quality indicator data within; Use each of the historical tags and the network quality indicator data corresponding to each historical tag as training samples to train a network jam prediction model to obtain a trained network jam prediction model; Based on the trained network lag prediction model and the network quality indicator data in the time period to be predicted, the target label of network lag in the time period to be predicted is obtained. 2. The method according to claim 1, characterized in that the historical Internet access record data includes a unified resource identifier for users to access the Internet; The historical tag for determining whether network lag occurs within the historical preset time period based on the historical Internet access record data corresponding to the historical preset time period includes: Delete the unified resource identifiers that do not contain preset information among the unified resource identifiers corresponding to the historical preset time period to obtain the first unified resource identifier; wherein the preset information at least includes files, path information, and queries. Parameters; the query parameters are used to characterize the characteristic information of network lag; Split the first unified resource identifier according to domain names and key-value pairs corresponding to the query parameters to obtain a set of alternative features; Based on the alternative feature set, determine a historical label of whether network stalling occurs within the historical preset time period. 3. The method according to claim 2, characterized in that, based on the alternative feature set, determining the historical label of whether network lag occurs within the historical preset time period includes: Obtain respectively the first key-value pair of the first domain name and the query parameter when the network is stuck within the historical preset time period, and the second key-value pair of the second domain name and the query parameter when the network is not stuck; Compare the first domain name, the first key-value pair, the second domain name and the second key-value pair with the domain name and key-value pairs in the alternative feature set, and select the ones that match the The domain name in the alternative feature set is the same, the key attribute in the key-value pair in the alternative feature set is the same, and the value attribute is different from the value attribute in the key-value pair in the alternative feature set , determined as the target value attribute; Based on the target value attribute, a historical tag of whether network stalling occurs within the historical preset time period is determined. 4. The method according to claim 1, characterized in that the historical Internet access record data further includes: network feature information; Based on the historical Internet access record data corresponding to the historical preset time period, determine the network quality indicator data within the historical preset time period, including: Based on the network characteristic information corresponding to the historical preset time period, the network quality index data within the historical preset time period is calculated. 5. The method according to claim 1 or 4, characterized in that the network quality indicator data at least includes: uplink rate, downlink rate, server delay, client delay, response delay, uplink retransmission rate, Downlink retransmission rate, access success rate, client failure rate, and server failure rate. 6. The method of claim 1, wherein the network jam prediction model is a logistic regression model. 7. A network jam prediction device, characterized in that the device includes: The first acquisition module is used to acquire the user's historical Internet access record data corresponding to each historical preset time period; The first determination module is configured to determine, for each historical preset time period, based on the historical Internet access record data corresponding to the historical preset time period, a historical tag of whether network lag occurs within the historical preset time period, and Network quality indicator data within the historical preset time period; The second determination module is used to use each of the historical tags and the network quality indicator data corresponding to each historical tag as training samples to train a network jam prediction model and obtain a trained network jam prediction model; The third determination module is used to obtain the target label of network lag in the to-be-predicted time period based on the trained network jam prediction model and the network quality indicator data in the to-be-predicted time period. 8. An electronic device, characterized in that it includes a processor, a memory and a program or instructions stored on the memory and executable on the processor. The program or instructions are implemented when executed by the processor. The steps of the network jam prediction method according to any one of claims 1-6. 9. A readable storage medium, characterized in that the readable storage medium stores programs or instructions, and when the programs or instructions are executed by a processor, the network jam described in any one of claims 1-6 is achieved. Steps in the forecasting method. 10. A computer program product, characterized in that, when the instructions in the computer program product are executed by a processor of an electronic device, the electronic device causes the electronic device to perform the network jam prediction according to any one of claims 1-6. Method steps.