CN111107423A - Video service playing card pause identification method and device - Google Patents
Video service playing card pause identification method and device Download PDFInfo
- Publication number
- CN111107423A CN111107423A CN201811270250.1A CN201811270250A CN111107423A CN 111107423 A CN111107423 A CN 111107423A CN 201811270250 A CN201811270250 A CN 201811270250A CN 111107423 A CN111107423 A CN 111107423A
- Authority
- CN
- China
- Prior art keywords
- xdr
- behavior
- video service
- stuck
- index
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 73
- 230000006399 behavior Effects 0.000 claims abstract description 117
- 238000003062 neural network model Methods 0.000 claims abstract description 23
- 238000000605 extraction Methods 0.000 claims abstract description 11
- 238000012360 testing method Methods 0.000 claims abstract description 11
- 239000000523 sample Substances 0.000 claims description 23
- 238000012549 training Methods 0.000 claims description 22
- 238000004590 computer program Methods 0.000 claims description 13
- 238000013528 artificial neural network Methods 0.000 claims description 7
- 238000012545 processing Methods 0.000 claims description 5
- 238000013526 transfer learning Methods 0.000 claims description 5
- 238000007689 inspection Methods 0.000 claims description 3
- 238000010606 normalization Methods 0.000 claims description 3
- 230000008569 process Effects 0.000 abstract description 26
- 238000013507 mapping Methods 0.000 abstract description 7
- 238000011156 evaluation Methods 0.000 description 12
- 230000011664 signaling Effects 0.000 description 10
- 238000004458 analytical method Methods 0.000 description 8
- 238000010586 diagram Methods 0.000 description 7
- 230000008447 perception Effects 0.000 description 6
- 238000004891 communication Methods 0.000 description 4
- 230000006870 function Effects 0.000 description 4
- 239000000284 extract Substances 0.000 description 3
- 238000007781 pre-processing Methods 0.000 description 3
- 238000002360 preparation method Methods 0.000 description 3
- 238000013473 artificial intelligence Methods 0.000 description 2
- 238000010276 construction Methods 0.000 description 2
- 238000013135 deep learning Methods 0.000 description 2
- 238000005065 mining Methods 0.000 description 2
- 238000012544 monitoring process Methods 0.000 description 2
- 230000003287 optical effect Effects 0.000 description 2
- 238000005457 optimization Methods 0.000 description 2
- 230000004044 response Effects 0.000 description 2
- 235000011464 Pachycereus pringlei Nutrition 0.000 description 1
- 240000006939 Pachycereus weberi Species 0.000 description 1
- 235000011466 Pachycereus weberi Nutrition 0.000 description 1
- 230000004913 activation Effects 0.000 description 1
- 230000003139 buffering effect Effects 0.000 description 1
- 238000011217 control strategy Methods 0.000 description 1
- 238000013136 deep learning model Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 238000013210 evaluation model Methods 0.000 description 1
- 238000012423 maintenance Methods 0.000 description 1
- 238000010295 mobile communication Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 238000010998 test method Methods 0.000 description 1
- 238000012546 transfer Methods 0.000 description 1
- 230000007704 transition Effects 0.000 description 1
Images
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N21/00—Selective content distribution, e.g. interactive television or video on demand [VOD]
- H04N21/40—Client devices specifically adapted for the reception of or interaction with content, e.g. set-top-box [STB]; Operations thereof
- H04N21/43—Processing of content or additional data, e.g. demultiplexing additional data from a digital video stream; Elementary client operations, e.g. monitoring of home network or synchronising decoder's clock; Client middleware
- H04N21/436—Interfacing a local distribution network, e.g. communicating with another STB or one or more peripheral devices inside the home
- H04N21/4363—Adapting the video stream to a specific local network, e.g. a Bluetooth® network
- H04N21/43637—Adapting the video stream to a specific local network, e.g. a Bluetooth® network involving a wireless protocol, e.g. Bluetooth, RF or wireless LAN [IEEE 802.11]
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
- H04L41/14—Network analysis or design
- H04L41/145—Network analysis or design involving simulating, designing, planning or modelling of a network
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N21/00—Selective content distribution, e.g. interactive television or video on demand [VOD]
- H04N21/40—Client devices specifically adapted for the reception of or interaction with content, e.g. set-top-box [STB]; Operations thereof
- H04N21/43—Processing of content or additional data, e.g. demultiplexing additional data from a digital video stream; Elementary client operations, e.g. monitoring of home network or synchronising decoder's clock; Client middleware
- H04N21/442—Monitoring of processes or resources, e.g. detecting the failure of a recording device, monitoring the downstream bandwidth, the number of times a movie has been viewed, the storage space available from the internal hard disk
- H04N21/44204—Monitoring of content usage, e.g. the number of times a movie has been viewed, copied or the amount which has been watched
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N21/00—Selective content distribution, e.g. interactive television or video on demand [VOD]
- H04N21/40—Client devices specifically adapted for the reception of or interaction with content, e.g. set-top-box [STB]; Operations thereof
- H04N21/43—Processing of content or additional data, e.g. demultiplexing additional data from a digital video stream; Elementary client operations, e.g. monitoring of home network or synchronising decoder's clock; Client middleware
- H04N21/442—Monitoring of processes or resources, e.g. detecting the failure of a recording device, monitoring the downstream bandwidth, the number of times a movie has been viewed, the storage space available from the internal hard disk
- H04N21/44213—Monitoring of end-user related data
- H04N21/44218—Detecting physical presence or behaviour of the user, e.g. using sensors to detect if the user is leaving the room or changes his face expression during a TV program
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N21/00—Selective content distribution, e.g. interactive television or video on demand [VOD]
- H04N21/40—Client devices specifically adapted for the reception of or interaction with content, e.g. set-top-box [STB]; Operations thereof
- H04N21/47—End-user applications
- H04N21/472—End-user interface for requesting content, additional data or services; End-user interface for interacting with content, e.g. for content reservation or setting reminders, for requesting event notification, for manipulating displayed content
- H04N21/47202—End-user interface for requesting content, additional data or services; End-user interface for interacting with content, e.g. for content reservation or setting reminders, for requesting event notification, for manipulating displayed content for requesting content on demand, e.g. video on demand
Landscapes
- Engineering & Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Databases & Information Systems (AREA)
- Multimedia (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Social Psychology (AREA)
- Human Computer Interaction (AREA)
- Two-Way Televisions, Distribution Of Moving Picture Or The Like (AREA)
Abstract
The embodiment of the invention provides a video service playing stuck identification method and device, through the incidence relation between the indexes related to user experience or service behaviors in an XDR and stuck behaviors, the stuck behaviors in the playing process are accurately identified, through the extraction of mathematical relations among the indexes, the incidence mapping expression of multi-dimension and multi-index is realized, through constructing a deep neural network model, the stuck identification of a new service can be completed only by a small amount of new characteristic samples, the difficulty and the cost of deployment and implementation of the proposal are reduced, and the problems of few stuck samples and incomplete coverage under the traditional service dial testing method are solved.
Description
Technical Field
The embodiment of the invention relates to the technical field of mobile communication, in particular to a video service playing card pause identification method and device.
Background
At present, the proportion of video service traffic under a mobile network exceeds a webpage browsing class, and the importance of the service is increasingly highlighted. In order to guarantee customer perception and improve the network optimization and control capability, operators urgently need to establish an accurate video service perception evaluation means. The Kadun is the most important evaluation index in the video playing process, and how to utilize the pipeline data of the operator has important significance on accurate modeling.
In the prior art, an operator mainly uses DPI to collect an XDR document of a user plane of an S1-U port, and approximately evaluates the quality of video service for indexes such as download rate, play code rate, video initial delay time, TCP handshake delay, GET response delay, etc. in the XDR, and mainly aims at index evaluation of play preparation stage or overall play quality, such as play success rate, initial delay time, average download rate, etc.
In the prior art, although the XDR analysis of the user plane of the S1-U port has the capability of analyzing user behaviors such as KQI indexes including download rate and time delay, and playback code rate, the XDR analysis does not have the capability of accurately evaluating whether a stuck behavior occurs in the playback process. If the video playing and downloading rate is related to the film source code rate, the video playing and downloading rate is also influenced by the storage space of the terminal, the stepping control strategy of the server and the like. The evaluation of the initial delay time index is mainly the problem of the waiting time of the playing preparation stage. Namely, the existing method for evaluating the video service quality through XDR network big data of DPI lacks the identification capability of playing card pause behaviors, and cannot accurately insights the quality difference behaviors occurring in the video playing process of a user.
Disclosure of Invention
Embodiments of the present invention provide a method and an apparatus for identifying video service playing card pause, which overcome the above problems or at least partially solve the above problems.
In a first aspect, an embodiment of the present invention provides a method for identifying a video service playing card pause, including:
the method comprises the steps of extracting indexes related to user experience or business behaviors in an XDR (transaction detail document) of the video business, and identifying whether the XDR has the stuck behaviors or not based on the pre-obtained incidence relation between the indexes and the stuck behaviors.
In a second aspect, an embodiment of the present invention provides an apparatus for identifying a video service playing card pause, including:
the extraction module is used for extracting indexes related to user experience or service behaviors in the transaction detail document XDR of the video service;
and the identification module is used for identifying whether the card pause behavior exists in the XDR based on the pre-obtained incidence relation between the index and the card pause behavior.
In a third aspect, an embodiment of the present invention provides an electronic device, which includes a memory, a processor, and a computer program stored in the memory and executable on the processor, where the processor implements the steps of the video service playing card identification method provided in the first aspect when executing the program.
In a fourth aspect, an embodiment of the present invention provides a non-transitory computer-readable storage medium, on which a computer program is stored, where the computer program, when executed by a processor, implements the steps of the video service playing stuck identification method as provided in the first aspect.
The embodiment of the invention provides a video service playing stuck identification method and device, through the incidence relation between the indexes related to user experience or service behaviors in an XDR and stuck behaviors, the stuck behaviors in the playing process are accurately identified, through the extraction of mathematical relations among the indexes, the incidence mapping expression of multi-dimension and multi-index is realized, through the construction of a deep neural network model, the stuck identification of a new service can be completed only by a small number of new characteristic samples, the difficulty and the cost of deployment and implementation of the proposal are reduced, and the problems of few stuck samples and incomplete coverage under the traditional service dial testing method are solved.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, and it is obvious that the drawings in the following description are some embodiments of the present invention, and those skilled in the art can also obtain other drawings according to the drawings without creative efforts.
Fig. 1 is a schematic diagram illustrating a video service playing card pause identification method according to an embodiment of the present invention;
FIG. 2 is a diagram illustrating a learning method for DNN training according to an embodiment of the present invention;
FIG. 3 is a schematic diagram of a DNN network topology according to an embodiment of the present invention;
FIG. 4 is a schematic diagram of DNN network transition learning according to an embodiment of the present invention
Fig. 5 is a schematic diagram of an apparatus for identifying video service playing card pause according to an embodiment of the present invention;
fig. 6 is a schematic physical structure diagram of an electronic device according to an embodiment of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are some, but not all, embodiments of the present invention. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
The video service is a mobile service with high code stream and high concurrency, the service quality requirement is high, and the video service perception becomes an important evaluation dimension of customer perception. Currently, mobile operators basically adopt a streaming media mode to realize video on demand services, and domestic mainstream video APPs deliver media data based on a TCP/HTTP protocol. When the user mobile phone terminal plays the video on line, the mobile phone terminal can request the server for corresponding video information, the server responds to the request to send the relevant information of the video, the user side initiates a resource request according to the obtained video downloading address, and the server responds to the resource request message to send corresponding video data. The client only needs to wait for a short period of time before starting playing to download and buffer the foremost part of the data of the media file, and the client can download and play the video data while playing the video data after the video data received by the client exceeds the initial buffering threshold. In the video service playing phase, katton is a main factor affecting the perception of the user. Video traffic stuck is defined as a freeze of a picture due to low network throughput, independent of a stop of the picture due to a user's pause operation. Stuck occurs when the transfer rate is less than the play-out rate and the buffer is empty.
The existing morton monitoring technical scheme mainly aims at index evaluation of a play preparation stage or overall play quality, such as play success rate, initial delay time, average download rate and the like, and approximately evaluates video service quality for indexes such as download rate, play code rate, video initial delay time, TCP handshake delay, GET response delay and the like in an XDR by utilizing a DPI to acquire an XDR document of a user plane of an S1-U port; the method lacks the identification capability of playing the pause behavior, and cannot accurately observe the poor quality behavior occurring in the video playing process of the user.
Because the jamming monitoring method in the prior art lacks the capability of identifying the playing jamming behavior and cannot accurately observe the quality difference behavior in the video playing process of the user, the jamming behavior in the video playing process can be identified by utilizing the network signaling data through the combination of the network big data and the artificial intelligence algorithm in each embodiment of the invention. The following description and description will proceed with reference being made to various embodiments.
Fig. 1 is a method for identifying a video service playing card pause according to an embodiment of the present invention, including:
s1, extracting indexes related to user experience or service behavior in the transaction detail document XDR of the video service;
s2, based on the correlation between the index and the katton behavior, identifying whether the XDR has the katton behavior.
In this embodiment, the video service in the video service playing process in the above description, and then the embodiment can identify the occurrence of the stuck behavior in the video playing process, and by obtaining the Transaction Detail document XDR in the mobile internet video service, extract the indexes related to the user experience or the service behavior in the XDR, the association relationship between each index related to the stuck behavior in the XDR and the stuck behavior is mined, and the occurrence of the stuck behavior in the video playing process is identified, where the Transaction Detail document (XDR) is a detailed Record of the signaling and the service for the signaling detection platform and the signaling application generated after processing based on the full data in the playing process.
On the basis of the above embodiment, identifying whether there is a stuck behavior in the XDR based on the pre-obtained association relationship between the index and the stuck behavior specifically includes:
and based on a trained deep neural network model, carrying out recognition processing on the index, and judging whether the XDR has a stuck behavior or not.
In the embodiment, in the process of mining the XDR, a Deep neural network (Deep neural network, DNN) is introduced, through extraction of mathematical relationships between indexes, the associative mapping expression of multi-dimensional and multi-index is realized, a Deep neural network model capable of identifying the associative relationships between each index and the stuck behavior is constructed in advance, in the identification process, only each index related to user experience or business behavior in the transaction detailed document XDR of the video service needs to be obtained, and direct identification is performed through the Deep neural network model, so that the stuck behavior occurring in the video playing process can be identified, and the problem that the stuck behavior occurring in the playing process cannot be accurately observed in the traditional single index evaluation mode is solved.
On the basis of the foregoing embodiments, as shown in fig. 2, before the training-based deep neural network model, i.e. before step S1, the method further includes S0:
s01, obtaining the XDR containing the stuck behavior and the XDR not containing the stuck behavior in the video service, marking the XDR containing the stuck behavior as a stuck document, and marking the XDR not containing the stuck behavior as a normal document;
s02, extracting indexes related to user experience or business behavior in an XDR of the video business, wherein the XDR comprises a card document and a normal document;
and S03, training and learning by taking the indexes as nodes of the input layer of the deep neural network DNN and taking the Kanton labels as nodes of the corresponding output layer to obtain a deep neural network model for identifying whether the Kanton behavior exists in the XDR.
In this embodiment, before training and learning of the deep neural network model, a stuck condition and a corresponding XDR in a video playing process need to be acquired, and specifically, in this embodiment, the method includes acquiring an XDR with a stuck behavior and an XDR without a stuck behavior in a video service, marking the XDR with the stuck behavior as a stuck document, and marking the XDR without the stuck behavior as a normal document.
Using the XDR (including normal documents and cardon documents) as a sample, designating fields related to user experience or service behavior, in this embodiment, including network access type, uplink and downlink traffic, failure reason code, time delay, jitter, packet loss, and extracting corresponding index data; in this embodiment, the indexes include 63 indexes, which are specifically shown in table 1 below:
TABLE 1 XDR sample extraction field
A DNN deep learning model is constructed, and the speed, the code rate, the packet loss, the disorder and other multi-dimensional 60 indexes are expressed by the incidence relation between the stuck behavior and the previous 60 indexes, so that the stuck behavior can be quantitatively evaluated, and the problems that the stuck index is lacked in single index evaluation and the analysis accuracy is low in the traditional method are solved.
On the basis of the above embodiments, extracting the index related to the user experience or the service behavior in the transaction detail document XDR of the video service specifically includes:
extracting an index field related to user experience or service behavior in the XDR, extracting corresponding index data based on the index field, and establishing an initial index set;
and performing zero filling pretreatment on the missing values in the initial index set, and performing data normalization treatment by taking each index as a dimension to obtain an index data set.
In this embodiment, as shown in table 1, an initial index set is constructed by a total of 63 indexes, and then the missing values in the initial index set are preprocessed in a "zero padding" manner. And finally, carrying out data normalization processing by taking each index as a dimension, wherein a stuck label represents stuck through one-hot coding, and 1, and finally obtaining an initialized index data set.
On the basis of the above embodiments, acquiring an XDR with a stuck behavior and an XDR without a stuck behavior in a video service includes:
for the own video service, acquiring a playing index through an embedded video quality acquisition kit (SDK), and if the SDK acquires the stuck behavior, sending an invalid service request to a set Uniform Resource Locator (URL);
and after the deep packet inspection DPI probe of the network side S1-U port is acquired and analyzed to the service request, considering that the card pause behavior occurs, adding a card pause label in the corresponding XDR, and otherwise, considering that the card pause behavior does not occur, adding a normal label in the corresponding XDR.
Each operator is provided with a DPI signaling collection system at a core network interface, and user-level, cell/base station-level and OTT-level video service user perception can be obtained through whole-network whole-signaling collection, analysis and summarization. The XDR signaling can be used for directly dissecting the video service of the whole network to obtain the video recording log of the whole amount of users.
In this embodiment, for a self-owned video service including a self-owned video APP, player indexes are collected through an embedded video quality collection Kit SDK (Software Development Kit), after a card-pause behavior is monitored, an invalid service request is sent once to a set URL (Uniform Resource Locator) (an identification field of 32 bits is added after the URL is originally requested), and after a corresponding DPI (Deep packet inspection) probe at an S1-U port on a network side collects and analyzes the URL request, it is considered that the card-pause behavior occurs in the playing process, a tag is added to a synthesized XDR, and the XDR is marked as a card-pause document, otherwise, the XDR is marked as a normal document. The video playing process can be monitored and analyzed by DPI equipment of an operator and the card pause document can be obtained, and the defects of few card pause samples and incomplete coverage in the traditional business dial testing method are overcome by the automatic acquisition and analysis method.
By formulating a rule for sending a specific URL invalid request after the video APP is blocked, the DPI on the network pipeline side can analyze the method for obtaining the marked sample, and the problems of high cost and poor coverage of the traditional method for obtaining the video blocked sample through manual dial-testing are solved. Therefore, the characteristics of the pipeline data after the occurrence of the stuck business behavior can be analyzed more comprehensively and more efficiently.
On the basis of the above embodiments, acquiring an XDR with a stuck behavior and an XDR without a stuck behavior in a video service further includes:
for the non-owned video service, the time mode of occurrence of the pause behavior of the card is recorded by a dial testing method, the corresponding XDR is extracted from the background of the network big data, and the pause label or the normal label is added.
In this embodiment, the automatic acquisition and analysis method cannot be adopted for non-self videos, so that a dial-up test method is adopted to simulate the playing process of a video service, obtain a card-pause document in the playing process of the video service, and extract an XDR record of a corresponding dial-up test number from a network big data background for marking.
Through the two modes in the embodiment, the training sample set is obtained, wherein the scale of the self-owned video sample is large and is used for the training of the subsequent basic model, and the dial testing sample of the non-self-owned video is used for the transfer learning training, and only a small amount of samples are needed.
In this embodiment, an artificial intelligence algorithm is introduced, a stuck document reported by a self-owned video APP is used as a sample, indexes such as uplink and downlink flow, failure reason codes, code rate, RTT delay, TCP delay, packet loss and the like in an XDR are used as input, a stuck state is used as output, a DNN deep learning network is constructed, an association relation of stuck behaviors such as rate, code rate, delay, packet loss, disorder and the like in an XDR signaling is mined, and an evaluation model of the stuck behavior is constructed. And through transfer learning, a small number of stuck samples are obtained according to the hot video APP dialing and testing mode, and then model training can be completed. And further, the purpose of evaluating the card pause behavior of the whole network video service through the network XDR big data of the operator is achieved. By the method in the embodiment, the total identification accuracy of the stuck behavior reaches 82%, a video stuck behavior identification method based on network signaling data is constructed, and the value of network operation and maintenance support is highlighted.
On the basis of the above embodiments, after obtaining a corresponding XDR sample and preprocessing the XDR sample to obtain an index data set, each index in the index data set is used as a node of the DNN learning network input layer, and the katton label corresponds to a node of the output layer. And further defining a 63 × 100 × 50 × 2-dimensional DNN deep neural network comprising an input layer, two hidden layers and an output layer, and defining a Linear rectification function (ReLU) as an activation function to complete a multi-dimensional correlation mapping model from the index recorded in the XDR to the playing card. The topology of the DNN network is schematically shown in fig. 3.
Specifically, for training the learning model, a training sample set is formed by recording the N video services preprocessed in the above embodiments. Defining the cross entropy as a loss function, adjusting the model updating weight and deviation parameters through a gradient descent algorithm, finishing training when the loss function does not descend or the judgment accuracy reaches a target, obtaining model parameters of a DNN (deep neural network), and finally obtaining a deep neural network model for judging whether the XDR has the Kanton behavior.
In the embodiment, a method for mining the association relationship between indexes such as speed, code rate, time delay, packet loss and disorder in operator network signaling data and playing card pause is based on DNN deep learning, a complete index system including network access types and the like is defined, the association mapping expression of multi-dimensional and multi-index is realized through the extraction of mathematical relationships among the indexes, and the problem that the card pause behavior in the playing process cannot be accurately known in the traditional single index evaluation mode is solved.
On the basis of the above embodiments, as shown in fig. 4, the training and learning specifically includes:
s031, training and learning by taking XDR in the free video service as a sample to obtain an initial deep neural network model;
and S032, taking the XDR in the non-self video service as a sample, and further training the initial deep neural network model by a transfer learning method to obtain the deep neural network model corresponding to the non-self video service.
In this embodiment, for some popular video services APPs or target APPs, a dial-up test sample is obtained through the steps in the above embodiments, and a new training sample is constructed after preprocessing; and by utilizing the network structure constructed in each embodiment, inheriting the model parameters obtained by learning of self-video morton evaluation, inputting the new training sample into the DNN network model by dividing the APP for further training, and completing parameter model tuning of the hotspot APP one by one.
The embodiment fully utilizes the characteristic that the video morton sample is convenient to obtain, completes the training of the basic model, obtains the basic incidence relation between the morton sample and the service index, and realizes the purpose of completing the model optimization only by a small amount of non-video APP samples through the transfer learning.
And during video service quality analysis, extracting the XDR corresponding to the video service, extracting index data according to the index field extraction method in each embodiment, preprocessing the index data, and inputting the preprocessed index data into the corresponding trained DNN deep neural network model, namely, identifying and judging whether each XDR is stuck or not.
The embodiment further provides an identification apparatus for video service playing stuck, based on the identification method for video service playing stuck described in the foregoing embodiments, as shown in fig. 5, including an extraction module 40 and an identification module 50, where:
the extraction module 40 extracts indexes related to user experience or service behaviors in the transaction detail document XDR of the video service; through the expression of the incidence relation between 60 multidimensional indexes such as rate, code rate, packet loss, disorder and the like and before the stuck behavior, the stuck behavior can be quantitatively evaluated, and the problems that the stuck index is lacked in single index evaluation and the analysis accuracy is low in the traditional method are solved.
The recognition module 50 recognizes whether the XDR has the stuck behavior based on a correlation between the index and the stuck behavior obtained in advance, performs recognition processing on the index based on a trained deep neural network model, and determines whether the XDR has the stuck behavior. By introducing Deep Neural Networks (DNN) and extracting mathematical relations among indexes, the relevance mapping expression of multi-dimensional and multi-index is realized, a Deep Neural Network model capable of identifying the relevance relations between each index and a stuck behavior is constructed in advance, in the identification process, only each index related to user experience or business behavior in an affair detailed document XDR of a video business needs to be obtained, direct identification is carried out through the Deep Neural Network model, the stuck behavior in the video playing process can be identified, and the problem that the stuck behavior in the playing process cannot be accurately observed in the traditional single index evaluation mode is solved.
Fig. 6 is a schematic entity structure diagram of an electronic device according to an embodiment of the present invention, and as shown in fig. 6, the electronic device may include: a processor (processor)810, a communication Interface 820, a memory 830 and a communication bus 840, wherein the processor 810, the communication Interface 820 and the memory 830 communicate with each other via the communication bus 840. The processor 810 may call a computer program stored in the memory 830 and operable on the processor 810 to execute the video service playing card identification method provided by the above embodiments, for example, including:
s1, extracting indexes related to user experience or service behavior in the transaction detail document XDR of the video service;
s2, based on the correlation between the index and the katton behavior, identifying whether the XDR has the katton behavior.
In addition, the logic instructions in the memory 830 may be implemented in software functional units and stored in a computer readable storage medium when the logic instructions are sold or used as independent products. Based on such understanding, the technical solutions of the embodiments of the present invention may be essentially implemented or make a contribution to the prior art, or may be implemented in the form of a software product stored in a storage medium and including instructions for causing a computer device (which may be a personal computer, a server, or a network device) to execute all or part of the steps of the methods described in the embodiments of the present invention. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a magnetic disk or an optical disk, and other various media capable of storing program codes.
An embodiment of the present invention further provides a non-transitory computer-readable storage medium, on which a computer program is stored, where the computer program is implemented to perform the video service playing card identification method provided in the foregoing embodiments when executed by a processor, for example, the method includes:
s1, extracting indexes related to user experience or service behavior in the transaction detail document XDR of the video service;
s2, based on the correlation between the index and the katton behavior, identifying whether the XDR has the katton behavior.
An embodiment of the present invention further provides a computer program product, where the computer program product includes a computer program stored on a non-transitory computer readable storage medium, where the computer program includes program instructions, and when the program instructions are executed by a computer, the computer can execute the video service playing card identification method as described above, for example, the method includes:
s1, extracting indexes related to user experience or service behavior in the transaction detail document XDR of the video service;
s2, based on the correlation between the index and the katton behavior, identifying whether the XDR has the katton behavior.
In summary, according to the identification method and apparatus for video service playing stuck behavior provided in the embodiments of the present invention, through the association relationship between the index related to the user experience or service behavior in the XDR and the stuck behavior, the stuck behavior in the playing process is accurately identified, through the extraction of the mathematical relationship between the indexes, the association mapping expression of multiple dimensions and multiple indexes is realized, and through the construction of the deep neural network model, the stuck identification of a new service can be completed only by a small number of new feature samples, so that the difficulty and cost of deployment and implementation of the present proposal are reduced, and the problems of few stuck samples and incomplete coverage in the conventional service dial testing method are solved.
The above-described embodiments of the apparatus are merely illustrative, and the units described as separate parts may or may not be physically separate, and parts displayed as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of the present embodiment. One of ordinary skill in the art can understand and implement it without inventive effort.
Through the above description of the embodiments, those skilled in the art will clearly understand that each embodiment can be implemented by software plus a necessary general hardware platform, and certainly can also be implemented by hardware. With this understanding in mind, the above-described technical solutions may be embodied in the form of a software product, which can be stored in a computer-readable storage medium such as ROM/RAM, magnetic disk, optical disk, etc., and includes instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute the methods described in the embodiments or some parts of the embodiments.
Finally, it should be noted that: the above examples are only intended to illustrate the technical solution of the present invention, but not to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some technical features may be equivalently replaced; and such modifications or substitutions do not depart from the spirit and scope of the corresponding technical solutions of the embodiments of the present invention.
Claims (10)
1. A video service playing card pause identification method is characterized by comprising the following steps:
the method comprises the steps of extracting indexes related to user experience or business behaviors in an XDR (transaction detail document) of the video business, and identifying whether the XDR has the stuck behaviors or not based on the pre-obtained incidence relation between the indexes and the stuck behaviors.
2. The method for identifying video service playing card pause according to claim 1, wherein identifying whether card pause behavior exists in the XDR specifically comprises, based on the pre-obtained association relationship between the indicator and card pause behavior:
and based on a trained deep neural network model, carrying out recognition processing on the index, and judging whether the XDR has a Caton behavior.
3. The video service playing card recognition method of claim 2, wherein before being based on the trained deep neural network model, the video service playing card recognition method further comprises:
obtaining an XDR containing a stuck behavior and an XDR not containing the stuck behavior in the video service, marking the XDR containing the stuck behavior as a stuck document, and marking the XDR not containing the stuck behavior as a normal document;
extracting indexes related to user experience or service behaviors in an XDR of the video service, wherein the XDR comprises a card document and a normal document;
and training and learning by taking the index as a node of a DNN input layer of the deep neural network and taking the Kanton label as a node of a corresponding output layer to obtain a deep neural network model for identifying whether the Kanton behavior exists in the XDR.
4. The method for identifying the video service playing card pause according to claim 1 or 3, characterized in that the extracting of the index related to the user experience or the service behavior in the transaction detail document XDR of the video service specifically comprises:
extracting an index field related to user experience or service behavior in the XDR, extracting corresponding index data based on the index field, and establishing an initial index set;
and performing zero filling pretreatment on the missing values in the initial index set, and performing data normalization treatment by taking each index as a dimension to obtain an index data set.
5. The method for identifying video service playing card pause as claimed in claim 3, wherein the step of obtaining the XDR with card pause behavior and the XDR without card pause behavior in the video service comprises:
for the own video service, acquiring a playing index through an embedded video quality acquisition kit (SDK), and if the SDK acquires the stuck behavior, sending an invalid service request to a set Uniform Resource Locator (URL);
and after the deep packet inspection DPI probe of the network side S1-U port is acquired and analyzed to the service request, considering that the card pause behavior occurs, adding a card pause label in the corresponding XDR, and otherwise, considering that the card pause behavior does not occur, adding a normal label in the corresponding XDR.
6. The method for identifying video service playing card pause as claimed in claim 5, wherein the obtaining of the XDR with card pause behavior and the XDR without card pause behavior in the video service further comprises:
for the non-owned video service, the time mode of occurrence of the pause behavior of the card is recorded by a dial testing method, the corresponding XDR is extracted from the background of the network big data, and the pause label or the normal label is added.
7. The video service playing card recognition method of claim 6, wherein the training learning specifically comprises:
training and learning by taking an XDR in free video service as a sample to obtain an initial deep neural network model;
and taking the XDR in the non-self video service as a sample, and further training the initial deep neural network model by a transfer learning method to obtain a deep neural network model corresponding to the non-self video service.
8. An apparatus for identifying video service playing card pause, comprising:
the extraction module is used for extracting indexes related to user experience or service behaviors in the transaction detail document XDR of the video service;
and the identification module is used for identifying whether the card pause behavior exists in the XDR based on the pre-obtained incidence relation between the index and the card pause behavior.
9. An electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, wherein the processor when executing the program implements the steps of the video service playback stuck identification method according to any one of claims 1 to 7.
10. A non-transitory computer readable storage medium, on which a computer program is stored, wherein the computer program, when being executed by a processor, implements the steps of the video service playback stuck identification method according to any one of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811270250.1A CN111107423A (en) | 2018-10-29 | 2018-10-29 | Video service playing card pause identification method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811270250.1A CN111107423A (en) | 2018-10-29 | 2018-10-29 | Video service playing card pause identification method and device |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN111107423A true CN111107423A (en) | 2020-05-05 |
Family
ID=70420069
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811270250.1A Pending CN111107423A (en) | 2018-10-29 | 2018-10-29 | Video service playing card pause identification method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111107423A (en) |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111586475A (en) * | 2020-05-27 | 2020-08-25 | 飞思达技术（北京）有限公司 | IPTV and OTT live broadcast audio and video quality and perception evaluation system |
| CN112333756A (en) * | 2020-09-14 | 2021-02-05 | 咪咕文化科技有限公司 | Method, system, electronic device and storage medium for monitoring regional network quality |
| CN112423096A (en) * | 2020-11-03 | 2021-02-26 | 上海哔哩哔哩科技有限公司 | Playing optimization method and system |
| CN112511818A (en) * | 2020-11-24 | 2021-03-16 | 上海哔哩哔哩科技有限公司 | Video playing quality detection method and device |
| CN115412762A (en) * | 2021-05-26 | 2022-11-29 | 中国移动通信有限公司研究院 | Data processing method, device and equipment of video service and storage medium |
| CN117241071A (en) * | 2023-11-15 | 2023-12-15 | 北京浩瀚深度信息技术股份有限公司 | Method for sensing video katon quality difference based on machine learning algorithm |
| CN117692948A (en) * | 2023-11-10 | 2024-03-12 | 广州丰石科技有限公司 | Quality difference IP identification method and device |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102340789A (en) * | 2010-07-20 | 2012-02-01 | 中国联合网络通信集团有限公司 | Dialing test terminal and dialing test method |
| FR2992809A1 (en) * | 2012-06-28 | 2014-01-03 | France Telecom | Method for estimating quality of experience of user with respect to e.g. watching of real-time video, involves providing parameters related to context and quality of service of connection of terminal to network independent of service |
| CN105554782A (en) * | 2015-12-09 | 2016-05-04 | 中国联合网络通信集团有限公司 | Prediction method and device for user perception index |
| US20170093648A1 (en) * | 2015-09-28 | 2017-03-30 | Wi-Lan Labs, Inc. | System and method for assessing streaming video quality of experience in the presence of end-to-end encryption |
| CN107493509A (en) * | 2017-09-25 | 2017-12-19 | 中国联合网络通信集团有限公司 | Video quality monitoring method and device |
| CN107888579A (en) * | 2017-11-06 | 2018-04-06 | 浙江大学 | A kind of mobile video user experience quality index modeling method of non-interfering type |
| CN108600790A (en) * | 2018-05-17 | 2018-09-28 | 北京奇艺世纪科技有限公司 | A kind of detection method and device of interim card failure |
| CN108696403A (en) * | 2018-03-23 | 2018-10-23 | 中国科学技术大学 | A kind of encrypted video QoE evaluating methods based on the study of network flow latent structure |
-
2018
- 2018-10-29 CN CN201811270250.1A patent/CN111107423A/en active Pending
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102340789A (en) * | 2010-07-20 | 2012-02-01 | 中国联合网络通信集团有限公司 | Dialing test terminal and dialing test method |
| FR2992809A1 (en) * | 2012-06-28 | 2014-01-03 | France Telecom | Method for estimating quality of experience of user with respect to e.g. watching of real-time video, involves providing parameters related to context and quality of service of connection of terminal to network independent of service |
| US20170093648A1 (en) * | 2015-09-28 | 2017-03-30 | Wi-Lan Labs, Inc. | System and method for assessing streaming video quality of experience in the presence of end-to-end encryption |
| CN105554782A (en) * | 2015-12-09 | 2016-05-04 | 中国联合网络通信集团有限公司 | Prediction method and device for user perception index |
| CN107493509A (en) * | 2017-09-25 | 2017-12-19 | 中国联合网络通信集团有限公司 | Video quality monitoring method and device |
| CN107888579A (en) * | 2017-11-06 | 2018-04-06 | 浙江大学 | A kind of mobile video user experience quality index modeling method of non-interfering type |
| CN108696403A (en) * | 2018-03-23 | 2018-10-23 | 中国科学技术大学 | A kind of encrypted video QoE evaluating methods based on the study of network flow latent structure |
| CN108600790A (en) * | 2018-05-17 | 2018-09-28 | 北京奇艺世纪科技有限公司 | A kind of detection method and device of interim card failure |
Non-Patent Citations (1)
| Title |
|---|
| 程婕等: "移动视频业务卡顿现象评估方法研究与演进", 《邮电设计技术》 * |
Cited By (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111586475A (en) * | 2020-05-27 | 2020-08-25 | 飞思达技术（北京）有限公司 | IPTV and OTT live broadcast audio and video quality and perception evaluation system |
| CN111586475B (en) * | 2020-05-27 | 2022-05-06 | 飞思达技术（北京）有限公司 | IPTV and OTT live broadcast audio and video quality and perception evaluation system |
| CN112333756A (en) * | 2020-09-14 | 2021-02-05 | 咪咕文化科技有限公司 | Method, system, electronic device and storage medium for monitoring regional network quality |
| CN112333756B (en) * | 2020-09-14 | 2024-02-27 | 咪咕文化科技有限公司 | Regional network quality monitoring method, system, electronic equipment and storage medium |
| CN112423096A (en) * | 2020-11-03 | 2021-02-26 | 上海哔哩哔哩科技有限公司 | Playing optimization method and system |
| CN112511818A (en) * | 2020-11-24 | 2021-03-16 | 上海哔哩哔哩科技有限公司 | Video playing quality detection method and device |
| CN112511818B (en) * | 2020-11-24 | 2022-08-19 | 上海哔哩哔哩科技有限公司 | Video playing quality detection method and device |
| CN115412762A (en) * | 2021-05-26 | 2022-11-29 | 中国移动通信有限公司研究院 | Data processing method, device and equipment of video service and storage medium |
| CN117692948A (en) * | 2023-11-10 | 2024-03-12 | 广州丰石科技有限公司 | Quality difference IP identification method and device |
| CN117241071A (en) * | 2023-11-15 | 2023-12-15 | 北京浩瀚深度信息技术股份有限公司 | Method for sensing video katon quality difference based on machine learning algorithm |
| CN117241071B (en) * | 2023-11-15 | 2024-02-06 | 北京浩瀚深度信息技术股份有限公司 | Method for sensing video katon quality difference based on machine learning algorithm |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111107423A (en) | Video service playing card pause identification method and device | |
| CN111382623B (en) | Live broadcast auditing method, device, server and storage medium | |
| CN109936512B (en) | Flow analysis method, public service flow attribution method and corresponding computer system | |
| CN111526119B (en) | Abnormal flow detection method and device, electronic equipment and computer readable medium | |
| CN108206769B (en) | Method, apparatus, device and medium for filtering network quality alarms | |
| CN109426700B (en) | Data processing method, data processing device, storage medium and electronic device | |
| CN115967504A (en) | Encrypted malicious traffic detection method and device, storage medium and electronic device | |
| CN113469298A (en) | Model training method and resource recommendation method | |
| CN110020161B (en) | Data processing method, log processing method and terminal | |
| CN114422211A (en) | HTTP malicious traffic detection method and device based on graph attention network | |
| CN110457601B (en) | Social account identification method and device, storage medium and electronic device | |
| CN112822121A (en) | Traffic identification method, traffic determination method and knowledge graph establishment method | |
| CN112995690B (en) | Live content category identification method, device, electronic equipment and readable storage medium | |
| CN110225009A (en) | It is a kind of that user's detection method is acted on behalf of based on communication behavior portrait | |
| CN116939661A (en) | SIM card abnormality detection method and system, electronic equipment and storage medium | |
| CN108985059B (en) | Webpage backdoor detection method, device, equipment and storage medium | |
| WO2023045365A1 (en) | Video quality evaluation method and apparatus, electronic device, and storage medium | |
| CN116915630A (en) | Network stuck prediction method, device, electronic equipment, medium and program product | |
| CN111143688B (en) | Evaluation method and system based on mobile news client | |
| CN113453076B (en) | User video service quality evaluation method, device, computing equipment and storage medium | |
| CN113297417B (en) | Video pushing method, device, electronic equipment and storage medium | |
| CN114422168A (en) | Malicious machine traffic identification method and system | |
| CN115767069A (en) | Camera fault determination method and device and computer readable storage medium | |
| CN116915767B (en) | Document transmission method and device | |
| CN111385136B (en) | Method and device for determining user communication identifier |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20200505 |
|
| RJ01 | Rejection of invention patent application after publication |