CN109104639B - Live broadcast system, method and device for determining live broadcast video theme and electronic equipment - Google Patents
Live broadcast system, method and device for determining live broadcast video theme and electronic equipment Download PDFInfo
- Publication number
- CN109104639B CN109104639B CN201710470458.7A CN201710470458A CN109104639B CN 109104639 B CN109104639 B CN 109104639B CN 201710470458 A CN201710470458 A CN 201710470458A CN 109104639 B CN109104639 B CN 109104639B
- Authority
- CN
- China
- Prior art keywords
- live
- video
- live video
- information
- live broadcast
- 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.)
- Active
Links
- 238000000034 method Methods 0.000 title claims description 70
- 238000000605 extraction Methods 0.000 claims abstract description 11
- 238000012549 training Methods 0.000 claims description 35
- 230000008569 process Effects 0.000 claims description 30
- 230000006399 behavior Effects 0.000 claims description 19
- 238000010801 machine learning Methods 0.000 claims description 19
- 238000012545 processing Methods 0.000 claims description 14
- 238000006243 chemical reaction Methods 0.000 claims description 5
- 238000002790 cross-validation Methods 0.000 claims description 5
- 238000010586 diagram Methods 0.000 description 11
- 238000005516 engineering process Methods 0.000 description 9
- 238000003860 storage Methods 0.000 description 8
- 230000006870 function Effects 0.000 description 6
- 238000012706 support-vector machine Methods 0.000 description 5
- 230000003993 interaction Effects 0.000 description 4
- 230000005540 biological transmission Effects 0.000 description 3
- 230000000694 effects Effects 0.000 description 3
- 238000004590 computer program Methods 0.000 description 2
- 239000000284 extract Substances 0.000 description 2
- 239000012634 fragment Substances 0.000 description 2
- 238000002372 labelling Methods 0.000 description 2
- 230000003287 optical effect Effects 0.000 description 2
- 229920000742 Cotton Polymers 0.000 description 1
- 238000009825 accumulation Methods 0.000 description 1
- 230000004931 aggregating effect Effects 0.000 description 1
- 238000013398 bayesian method Methods 0.000 description 1
- 238000004422 calculation algorithm Methods 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 239000002537 cosmetic Substances 0.000 description 1
- 238000007418 data mining Methods 0.000 description 1
- 238000013500 data storage Methods 0.000 description 1
- 238000003066 decision tree Methods 0.000 description 1
- 238000009826 distribution Methods 0.000 description 1
- 238000005243 fluidization Methods 0.000 description 1
- 230000008676 import Effects 0.000 description 1
- 230000002452 interceptive effect Effects 0.000 description 1
- 238000007477 logistic regression Methods 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000000877 morphologic effect Effects 0.000 description 1
- 238000003062 neural network model Methods 0.000 description 1
- 238000004091 panning Methods 0.000 description 1
- 238000006116 polymerization reaction Methods 0.000 description 1
- 238000007781 pre-processing Methods 0.000 description 1
- 238000003672 processing method Methods 0.000 description 1
- 230000001737 promoting effect Effects 0.000 description 1
- 238000007637 random forest analysis Methods 0.000 description 1
- 238000012552 review Methods 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
- 230000001052 transient effect 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
- H04N21/47—End-user applications
- H04N21/478—Supplemental services, e.g. displaying phone caller identification, shopping application
- H04N21/47815—Electronic shopping
-
- 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
- H04N21/44—Processing of video elementary streams, e.g. splicing a video clip retrieved from local storage with an incoming video stream or rendering scenes according to encoded video stream scene graphs
- H04N21/44008—Processing of video elementary streams, e.g. splicing a video clip retrieved from local storage with an incoming video stream or rendering scenes according to encoded video stream scene graphs involving operations for analysing video streams, e.g. detecting features or characteristics in the video stream
-
- 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
-
- 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]
- H04N21/80—Generation or processing of content or additional data by content creator independently of the distribution process; Content per se
- H04N21/81—Monomedia components thereof
- H04N21/812—Monomedia components thereof involving advertisement data
Landscapes
- Engineering & Computer Science (AREA)
- Multimedia (AREA)
- Signal Processing (AREA)
- Databases & Information Systems (AREA)
- Human Computer Interaction (AREA)
- Business, Economics & Management (AREA)
- Marketing (AREA)
- Two-Way Televisions, Distribution Of Moving Picture Or The Like (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The application discloses live broadcast system includes: the live broadcast module is used for receiving a live broadcast video sent by a main broadcast client and sending at least part of the live broadcast video to at least one user client; the content service module is used for receiving text or picture information corresponding to the live video and sending at least part of the information to at least one other user client; the related commodity module is used for providing a list of commodities related to the live video; the extraction module is used for extracting the received live video and the text or picture information related to the live video; and the live broadcast theme determining module is used for matching the corresponding commodities aiming at the live broadcast video according to the extracted characteristic information and generating a live broadcast theme for identifying the live broadcast video according to a matching result, so that commodities corresponding to the live broadcast video or video clips are reflected at any time and a user is helped to select the corresponding live broadcast video according to interested commodities.
Description
Technical Field
The application relates to a video live broadcast technology, in particular to a live broadcast system and a method for determining a live broadcast video theme; the application also relates to a device for determining the theme of the live video and a related electronic device.
Background
Video live broadcast is increasingly becoming the key application mode of the internet. Through live video broadcast, the method can quickly and effectively transmit information, has the characteristics of site type, real-time property, intuition, entertainment and the like, and particularly accords with the interest of young people growing in the Internet era.
In the face of the general live video source of the forest, live subject identification is needed, so that a video receiver can accurately obtain the required live content.
In the prior art, a host of a live video generally determines a title of the live video in advance or tags the live video, and the like, so as to identify a live subject in advance. However, live video itself has features of live performance and real-time performance, and even a host of the live video cannot necessarily completely predict the content of the live video. For video programs related to commodities, the theme of the live video cannot be accurately calibrated according to actual commodities in the live video by adopting a pre-identification mode, so that the theme of the live video or the theme of playback of the live video cannot help a video user to accurately select the live video to be watched according to interested commodities in real time.
Disclosure of Invention
The application provides a live broadcast system which can generate a theme of live broadcast video in real time and identify the live broadcast video by using the obtained theme of the live broadcast video. The application also provides a method and a device for determining the live video theme and electronic equipment for executing the method for acquiring the live video theme.
The application provides a live system, includes:
the live broadcast module is used for receiving a live broadcast video sent by a main broadcast client and sending at least part of the live broadcast video to at least one user client;
the content service module is used for receiving text or picture information corresponding to the live video and sending at least part of the information to at least one other user client;
the related commodity module is used for providing a list of commodities related to the live video;
the extraction module is used for extracting the received live video and the text or picture information related to the live video;
and the live broadcast theme determining module is used for matching the corresponding commodities aiming at live broadcast video or live broadcast video clips obtained by intercepting the live broadcast video according to the extracted characteristic information and generating live broadcast themes used for identifying the live broadcast video or the live broadcast video clips according to matching results.
Optionally, the content service module includes a message service sub-module, configured to receive information for a live video sent by at least one user client, and send at least part of the information to at least one other user client.
Optionally, the content service module includes a commodity information publishing sub-module, configured to receive a commodity publishing request corresponding to the anchor client, obtain commodity information corresponding to the commodity publishing request, and send the commodity information to at least part of user clients accessing the live video.
Optionally, the content service module includes a commodity information list submodule configured to display a list of commodities related to the live video to the user client.
Optionally, the content service module includes a shopping information sub-module, configured to obtain purchasing behavior information of a behavior of purchasing a commodity, which occurs at the user client, and send the purchasing behavior information to at least one user client.
Optionally, the live topic determined by the live topic determination module includes at least one of the following information:
merchandise information related to the live video;
merchandise information related to a segment of the live video.
Optionally, the live topic determined by the live topic determination module is used in the following occasions or one of the following occasions:
a title of the live video, or a title of a segment of the live video;
a portion of a content vignette of the live video, or a portion of a content vignette of a segment of the live video;
a title of playback of the live video, or a title of playback of a segment of the live video;
a portion of a content vignette of the playback of the live video, or a portion of a content vignette of the playback of a segment of the live video.
Optionally, the system includes a live topic pushing module, configured to push the determined live topic to a user client that watches the live video or a segment of the live video, or put the determined live topic into a live video directory for the user client to browse.
Optionally, the system includes a live broadcast theme and video association module, configured to establish and store an association relationship between the live broadcast video or the live broadcast video clip and the live broadcast theme.
The application also provides a method for determining the theme of the live video, which comprises the following steps:
in the process of live video playing, acquiring the characteristic information of the live video in real time, and acquiring the characteristic information of picture information or text information related to the live video; and obtaining a list of commodities related to the live video;
substituting the characteristic information into a live broadcast topic identification model obtained through pre-training, and identifying commodities matched with the live broadcast video or each live broadcast video clip of the live broadcast video;
and determining the theme of the live video or the live video clip according to the determined commodity matched with the live video or the live video clip.
Optionally, the live topic identification model is obtained by training in the following manner:
selecting the characteristic information of the marked live videos and the picture information or text information related to the live videos as sample data; the marking means determining commodities related to the live video;
and training a machine learning model by using the sample data as training data, so that the machine learning model can be matched with the commodities related to the live video according to the characteristic information of the live video and the picture information or text information related to the live video.
Optionally, the machine learning model adopts a TreeLink model or an SVM model.
Optionally, before the step of substituting the feature information into the live topic recognition model obtained through pre-training, the step includes:
performing cross validation and/or manual marking on the characteristic information, and removing information which is actually irrelevant to the commodities related to the live video; and taking the characteristic information after the eliminating step as the characteristic information substituted into the live topic identification model.
Optionally, the theme of the live video or the live video clip is determined according to the determined commodity matched with the live video or the live video clip, and the theme is implemented by the following method:
acquiring commodity information related to the commodity according to the determined matched commodity;
and determining the theme of the live video or the live video clip by adopting a preset conversion method according to the commodity information related to the commodity.
Optionally, the commodity is represented by at least one of the following ways: item class, name of the item, SPU of the item, SKU of the item, item number of the item.
Optionally, after determining the theme of the live video or the live video clip, at least one of the following steps is performed:
when determining a title for the live video or the live video clip, taking the title as the title or a component of the title;
when a content introduction is made for the live video or the live video clip, the subject is used as a part of the content introduction;
when the live video or the playback of the live video clip is made, the theme is taken as a title or a part of the title;
and when the played back content introduction of the live video or the live video clip is made, taking the theme as a part of the played back content introduction.
The present application further provides a device for determining a live video topic, including:
the system comprises a characteristic information acquisition unit, a video processing unit and a video processing unit, wherein the characteristic information acquisition unit is used for acquiring the characteristic information of a live video in real time and acquiring the characteristic information of picture information or text information related to the live video in the process of carrying out the live video;
the matching unit is used for substituting the characteristic information into a live broadcast theme recognition model obtained through pre-training, and recognizing commodities matched with the live broadcast video or each live broadcast video clip of the live broadcast video;
and the live video theme determining unit is used for determining the theme of the live video or the live video clip according to the determined commodity matched with the live video or the live video clip.
The present application further provides an electronic device, comprising:
a processor; and
the memorizer is used for storing a program for determining the theme of the live video, and after the equipment is powered on and runs the program for determining the theme of the live video through the processor, the following steps are executed:
in the process of live video playing, acquiring the characteristic information of the live video in real time, and acquiring the characteristic information of picture information or text information related to the live video; the feature information of the live video comprises screenshot information and sound information acquired from the live video; the picture information or the text information related to the live video comprises commodity related information reflecting commodities related to the live video; and obtaining a list of commodities related to the live video;
substituting the characteristic information into a live broadcast topic identification model obtained through pre-training, and identifying commodities matched with the live broadcast video or each live broadcast video clip of the live broadcast video;
and determining the theme of the live video or the live video clip according to the determined commodity matched with the live video or the live video clip.
Compared with the prior art, the live broadcast system that this application provided includes: the live broadcast module is used for receiving a live broadcast video sent by a main broadcast client and sending at least part of the live broadcast video to at least one user client; the content service module is used for receiving text or picture information corresponding to the live video and sending at least part of the information to at least one other user client; the related commodity module is used for providing a list of commodities related to the live video; the extraction module is used for extracting the received live video and the text or picture information related to the live video; and the live broadcast theme determining module is used for matching the corresponding commodities aiming at live broadcast video or live broadcast video clips obtained by intercepting the live broadcast video according to the extracted characteristic information and generating live broadcast themes used for identifying the live broadcast video or the live broadcast video clips according to matching results. The technical scheme can acquire the characteristic information of the live video in real time and at multiple angles, and the characteristic information is matched with the commodity information contained in the live video, so that the commodities corresponding to the live video and the intercepted video clip are reflected at any time, the live theme which is most matched with the live video and the intercepted video clip is provided for a video receiver, and the video receiver is helped to effectively select the required live program according to the interested commodities.
Drawings
Fig. 1 is a schematic diagram of a first implementation manner of a live broadcast system provided in an embodiment of the present application;
fig. 2 is a schematic diagram of a content service module of a first implementation of a live broadcast system according to an embodiment of the present application;
fig. 3 is a schematic diagram of a second implementation manner of a live broadcast system provided in an embodiment of the present application;
FIG. 4 is a flowchart of a processing method for determining a topic of a live video according to an embodiment of the present application;
fig. 5 is a flowchart illustrating a training process of a live topic recognition model in a method for determining a live video topic according to an embodiment of the present application;
fig. 6 is a flowchart illustrating a process of determining a live video topic according to a determined commodity matching a live video or a live video segment in the method for determining a live video topic provided in the embodiment of the present application;
fig. 7 is a schematic diagram of an apparatus for determining a live video topic provided in an embodiment of the present application;
fig. 8 is a schematic view of an electronic device for determining a live video topic provided in an embodiment of the present application.
Detailed Description
In the following description, numerous specific details are set forth in order to provide a thorough understanding of the present application. This application is capable of implementation in many different ways than those herein set forth and of similar import by those skilled in the art without departing from the spirit of this application and is therefore not limited to the specific implementations disclosed below.
In the present application, a live broadcast system, a method for determining a live video topic, an apparatus for determining a live video topic, and a related electronic device are provided, and the following embodiments are provided to describe in detail one by one.
The application provides a live broadcast system includes: the live broadcast module is used for receiving a live broadcast video sent by a main broadcast client and sending at least part of the live broadcast video to at least one user client; the content service module is used for receiving text or picture information corresponding to the live video and sending at least part of the information to at least one other user client; the related commodity module is used for providing a list of commodities related to the live video; the extraction module is used for extracting the received live video and the text or picture information related to the live video; and the live broadcast theme determining module is used for matching the corresponding commodities aiming at live broadcast video or live broadcast video clips obtained by intercepting the live broadcast video according to the extracted characteristic information and generating live broadcast themes used for identifying the live broadcast video or the live broadcast video clips according to matching results. The live broadcast system is provided with a related commodity module, and provides a list of commodities related to live broadcast videos, so that the live broadcast videos (or live broadcast video clips) can be matched with specific commodities according to the feature information of the live broadcast videos, and further, a proper theme can be determined for the live broadcast videos according to the matching relation.
Fig. 1 to fig. 3 are schematic diagrams of a live broadcast system provided in an embodiment of the present application, where fig. 1 is a schematic diagram of a first implementation manner of the live broadcast system provided in the embodiment of the present application; fig. 2 is a schematic diagram of a content service module of a first implementation of a live broadcast system according to an embodiment of the present application; fig. 3 is a schematic diagram of a second implementation manner of a live broadcast system provided in an embodiment of the present application.
The live broadcast system comprises a live broadcast module 101, a content service module 102, an extraction module 103, an associated commodity module 104 and a live broadcast theme determination module 105.
In specific implementation, the live broadcast system needs to be combined with a main broadcast client and a user client which are respectively connected with the live broadcast system through a network to realize the live broadcast function. The anchor client with the anchor function is used for recording the video of the live broadcast room and sending the video to the live broadcast system, and the live broadcast system forwards the live broadcast video to at least one user client requesting to enter the live broadcast room and watch the live broadcast video. The anchor client and the user client can be realized by adopting a computer, a smart phone or a tablet computer and other equipment.
The live broadcast module 101 is configured to receive a live broadcast video sent by a main broadcast client, and send at least part of the live broadcast video to at least one user client.
The live video refers to a video program transmitted in a real-time manner, and is generally transmitted in a video streaming manner.
Because video information is abundant and information amount is large, and the transmission speed of the current network is slow, if the network video data information is processed according to the traditional processing mode of computer files, problems can occur. In a traditional computer file processing mode, a computer processing file is completely processed, namely the file is required to be a complete whole when being processed. If a portion of a file is corrupted or has only half the content, the computer will consider the file as bad and untreatable. If the set of principles is also applicable to network video files, the viewers need to wait at least tens of minutes to hours to see the video files after downloading is completed, which obviously does not meet the requirements of live video. Therefore, the transmission of live video usually employs a special fluidization technology to extract files, i.e. video streams.
Video streaming is a means of transmitting video data information in such a way that a user can process the received information without receiving the complete data information. In the video streaming mode, when a live broadcast system transmits a video file to a user, the file is not sent out in its entirety at a time, but the file is divided into small segments in the order of playing time, the segments are similar to frames in an image, and then the segments are sent to the user in sequence. After the network playing tool of the user receives the fragments, the fragments are continuously played, so that complete sound and images can be generated, and only a certain delay is caused in the beginning.
For live video, the generated video program realizes the basic real-time broadcasting and receiving of the video program through the medium of a live system, and the live effect is obtained.
In practical application, the live broadcast system can be used for consumer live broadcast, for example, for online shopping, and generally includes a live broadcast product list, directly shows or classically shows all live broadcast videos, and applies for a user (i.e., a main broadcast) with a main broadcast function to open, the title and corresponding link of the live broadcast video released by the live broadcast system are shown in the live broadcast product list, and the main broadcast also provides pictures which can reflect the video content theme and are attractive, and provides information such as main broadcast name and live broadcast content introduction, so that a user (i.e., audience) watching the live broadcast video can select an interested video to watch; and the audience applies for watching the corresponding live video by clicking the link. In addition, when the live video starts to be live, the anchor also marks the live video with a label reflecting the live content information of the live video. In consumer live broadcasting, a live video title and a live video tag of the consumer live broadcasting generally represent commodity information or service information provided by the live video.
The live video content released by the anchor in the live broadcasting system comprises the live broadcasting room which can intuitively display commodities to users and solve the living problems of the users, such as how to more effectively protect skin, how to match clothes in different occasions, how to make cate, how to carry out effective and appropriate outdoor exercises to achieve the purpose of body building, how to care infants and the like. When watching the live broadcast video through the user client, the audience clicks the live broadcast video in the home page list page of the live broadcast product to request to watch the live broadcast, and can buy while watching the live broadcast.
Besides, the live broadcast system can also be used for other scenes such as social live broadcast, show live broadcast, live entertainment live broadcast, game live broadcast, sports live broadcast, online education live broadcast, travel live broadcast and the like, the embodiment takes consumer live broadcast as an example for description, and the content of other existing multi-form scenes and live broadcast services is not repeated one by one and is all within the protection scope of the application.
Specific examples are as follows:
the anchor A issues a live video matched with clothes, a title of the live video is ' people's clothing ', an anchor wears the modeling picture of the anchor clothes and is attached with a video start button, and the name of the anchor A is ' anchor A ', and a viewer B clicks a corresponding link to request to watch the video. In the live broadcast process, the anchor A publishes and displays a series of clothes, the audience B can consult or comment commodities in the live broadcast process, and when seeing interested clothes, clicks the corresponding link to see the relevant information of the commodities and directly buys the commodities.
The live broadcast module 101 provides sound information and image information for identifying a live broadcast theme, and is a main information source for extracting feature information of live broadcast video.
The content service module 102 is configured to receive text or picture information corresponding to a live video, and send at least part of the information to at least one other user client.
After the anchor client accesses the live broadcast system provided by the embodiment, the anchor can publish video information through the live broadcast system and also publish corresponding video related information in a text form or a picture form so that the live broadcast system and the user client can manage, classify and retrieve the video related information, in addition, the anchor can also publish commodity information while displaying commodities in the live broadcast process of the video, the commodity information publishing form comprises pictures or word descriptions or links of commodity pages, and the information is related content information with live broadcast video; in addition to the information provided by the anchor, the administrator at the background may also perform some calibration on the live video content, for example, the administrator at the background may also determine the theme of the live video or set a tag for the theme according to the live video content, which is also related content information of the live video; in the live broadcast process, the anchor and the audience can carry out instant message interaction in a live broadcast room, the instant message interaction information comprises commodity related information, such as commodity consultation information, commented commodities and the like, and the information is also related content information of live broadcast video; in addition, in the live broadcasting process, the behavior of purchasing commodities generated by the audiences due to watching the live video is also the related content information of the live video.
The content information related to the live video, such as the title, tag, text introduction, picture, commodity consultation, commodity comment, commodity purchase, etc. of the live video, is received and processed by the content service module 102, and provides text information and picture information for determining the live topic, which is one of the information sources for extracting the feature information of the live video.
It should be noted that, the related content information of the live video, such as a title, a tag, a text introduction, a picture, a commodity consultation, a commodity comment, a commodity purchase, and the like of the live video, where at least part of the information is sent to at least one other user client; for example, the title and representative picture provided by the anchor client for the video stream of the anchor client are used as the title and still picture of the live video stream, and are used as the title bar content for the user to search through the user client, and are sent to the user client along with the live video stream.
In practical applications, the live content service module may include a plurality of parts according to the foregoing functions provided by the live content service module, and in a preferred scheme provided in this embodiment, the live content service module may include one of the following four different sub-modules: a message service sub-module 102-1, a commodity information publishing sub-module 102-2, a commodity information list sub-module 102-3 and a shopping information sub-module 102-4.
In this embodiment, the content service module 102 includes the four sub-modules, which are shown in fig. 2:
the message service sub-module 102-1 is configured to receive information for a live video sent by at least one user client, and send at least part of the information to at least one other user client.
In the live broadcast process, the audience uses the user client to interact with the anchor broadcast or other audiences, and the interactive instant messages comprise published text information, picture information, label information, title information and the like. The message service sub-module 102-1 extracts at least one instant message from a user client for live video from the live broadcast instant message, and sends at least part of the information to at least one other user client, thereby realizing interaction.
In the live broadcast process, the instant messages in the live broadcast room are various, for example, messages for calling, daily chat messages, messages for consulting the goods being displayed by the anchor, messages for consulting other goods of interest, messages for consulting life problems, and the like, wherein the messages include goods related information related to the goods related to the live broadcast video, for example, instant messages including contents such as goods information, goods comment information, goods consultation information, and the like.
The message service sub-module 102-1 sends information for live video as one of information sources for extracting feature information of the live video.
The commodity information publishing sub-module 102-2 is configured to receive a commodity publishing request corresponding to the anchor client, obtain commodity information corresponding to the commodity publishing request, and send the commodity information to at least some user clients accessing the live video.
In the live broadcast process, the anchor can release commodities in the live broadcast room at any time, the commodities are released in the live broadcast process in a mode of providing commodity related links, and audiences obtain commodity information of the commodities according to the commodity related links released by the anchor, wherein the commodity information comprises the following steps: commodity names, classifications, labels, prices, reviews, etc.; in addition, the anchor can also publish the commodities in other forms, for example, in the form of commodity pictures or pictures of commodity introduction pages or commodity text description information, and audiences can directly obtain commodity related information on a video barrage; or, the distribution may be in the form of an aggregate list or SPU list of all products involved in the live video. The commodity information publishing sub-module 102-2 receives the commodity publishing request, obtains the commodity information corresponding to the commodity publishing request, and sends the commodity information to at least part of user clients accessing the live broadcast video, so that viewers can conveniently obtain detailed information of interested products and purchase the products conveniently in the live broadcast watching process. The commodity information sent by the commodity information publishing sub-module 102-2 is used as one of the information sources for extracting the feature information of the live video.
The commodity information list sub-module 102-3 is configured to display a list of commodities related to the live video to a user client.
In practical application, in the live broadcasting process of the live broadcasting system, all commodities or part of commodities issued by a main broadcaster are displayed to a user client, and audiences can watch and purchase the commodities at the same time. The anchor releases the display forms of the goods, including the display in SPU units or in SKU units or in singleton units or in other forms.
The SPU (Standard Product Unit) is a minimum Unit for aggregating commodity information, and when a commodity is displayed in the SPU form, a basic Product is displayed first, and then the commodity of a merchant sold under the SPU is displayed after clicking. For example, iphone7plus is an SPU, N97 is also an SPU, this is independent of merchants, color, style, and package, and the merchandise information of iphone7plus displayed in the live video is first aggregated into an SPU, and a "iphone 7 plus" label is printed, and then a merchandise link of each specific merchant is displayed under the "iphone 7 plus" label.
The SKU (Stock Keeping Unit), Stock Keeping Unit of Stock of goods, for example, one SKU used in articles of clothing and footwear, generally represents: specification, colour, style, the exhibition of brand C dress that shows in the live broadcast video is as a SKU show, with the red XL of brand C dress polymerization to a SKU during the show commodity, take on "brand C red dress XL" label, under this label, show the red dress commodity of brand C of every size again.
The displaying of the merchandise in the form of the single item refers to independent displaying of each single item, for example, an information page with one link pointing to one single item.
The merchandise information list sub-module 102-3 displays a list of merchandise related to the live video to the user client. The list of these products or the product-related information corresponding to the products is one of the information sources from which the extraction module 103 extracts the feature information of the live video.
The shopping information sub-module 102-4 is configured to obtain purchasing behavior information of a behavior of purchasing a commodity, which occurs at the user client, and send the purchasing behavior information to at least one user client.
In the live broadcast process, if a viewer watching a live broadcast video generates a purchasing behavior for a commodity published in the video through a user client, purchasing behavior information of the viewer is acquired by the shopping information submodule 102-4, and meanwhile, the purchasing behavior information is sent to at least one user client by the module. Through the shopping information submodule 102-4, the purchasing behavior of the user client can be presented in the live broadcasting process, and the existence sense and the participation sense of the user are improved. Meanwhile, the purchasing behavior information can count the shipment quantity of SPUs or SKUs or single products of the commodities, provide reference for purchasing of other users, and can also be used as one of reference conditions of hot-selling products when determining a live broadcast theme.
The purchasing behavior information also serves as one of information sources for extracting characteristic information of the live video.
It should be noted that, in implementation, the four different sub-modules all include commodity-related information, and all of the four sub-modules may be used alone as a presentation mode of the content service module, or may be used as a presentation mode of the content service module by combining any several sub-modules, which is within the protection scope of the present application.
The extraction module 103 is configured to extract feature information of the received live video and text or picture information related to the live video.
In the process of live video, the related information of the live video is acquired in real time, and in order to comprehensively and accurately reflect the content of the live theme, the live video must be acquired from multiple dimensions. The dimension refers to the form of obtaining the source information, and one form is one dimension. The characteristic information refers to information valuable for identifying the live subject. The characteristic information of a plurality of preset dimensions is valuable information obtained by performing diversified acquisition on the same live video from different determined morphological aspects. In practical application, through a voice recognition technology, feature information of one preset dimension is extracted from sound information of a live video, and through an image recognition technology, feature information of the other dimension is extracted from picture information of the live video.
The voice recognition means that the computer automatically converts the voice content of the human into corresponding characters. In this embodiment, a sound extraction function is used to obtain sound, the collected sound is further recognized and processed by using an ali speech recognition technology, and the word key information of the lecture of the anchor is obtained from the sound of the anchor, where the word key information at least includes one keyword.
The image recognition refers to a technology for processing, analyzing and understanding images by using a computer to recognize various targets and objects in different modes. In this embodiment, the screenshot function is used to obtain screenshot information, and the obtained screenshot is further processed by using a shooting and panning technology in ali to obtain a picture scene classification of the live video.
In practical application, when recognizing the voice and the image of the live video, a segment with a preset duration may be extracted for recognition, for example, the preset duration may be 2 minutes, 5 minutes, 10 minutes, and the like. In this embodiment, the feature information of the live video is further identified from the extracted text key information and the extracted picture scene information, where the feature information of the live video is content that can reflect live video content or an intercepted live video clip, and text information related to commodity-related information included in the live video or the intercepted live video clip at least includes one keyword.
In addition, the extraction module 103 further obtains text or picture information related to the live video from the content service module 102, and identifies feature information of the live video or the captured video clip from the text or picture information, where the feature information includes commodity information or commodity list information corresponding to the content of the live video or the captured video clip. The acquisition may be directly sent by the content service module 102, or may be directly read by the extraction module 103.
Specific examples are as follows:
the anchor A carries out the live broadcast of the video of the title "people lean on clothes", the anchor A shows C commodity D style T-shirt directly perceivedly through the video in the live broadcast process, the style of the T-shirt is introduced to the pronunciation, simultaneously declare that promotion activity and lottery drawing rule need to be carried out, audience B gets into this live broadcast room and consults a white commodity price, material, whether participate in the promotion, anchor A pronunciation answer price this clothes is pure cotton with elasticity 80 yuan RMB, can join in the graceful street and clap the wind with skirt or broken hole nine minutes trousers constitution, and directly perceived the object collocation of two kinds of modes, participate in promotion and whole point lottery drawing activity.
In this embodiment, "light street wind" and "C merchandise D-shirt" are recognized from the voice of the anchor of the live broadcast video stream, "clothing" is recognized from the title information, "sales promotion and lottery drawing" are recognized from the instant message and the voice interaction of the anchor, and these feature information are used as one of the information sources for determining the live broadcast theme.
The associated goods module 104 is configured to provide a list of goods related to the live video.
The live broadcast video theme determined by the live broadcast system provided by the application comprises commodity information corresponding to a live broadcast video (or a live broadcast video clip), and the live broadcast video (or the live broadcast video clip) can be matched with a specific commodity according to the live broadcast video theme, so that the live broadcast video theme can be used for consumption live broadcast related to online shopping; and the set of these specific commodities is the commodity information list provided by the commodity association module 103.
The associated goods module 104 provides the list of the goods related to the live video to the live subject determination 105 for determining the live subject.
The associated item module 104 is different from the item information list sub-module 102-3, and the associated item module 104 provides a list of items related to the live video, that is, the live video and the items to which each segment of the live video may be related; the goods information list sub-module 102-3 is responsible for displaying the goods to the user client, the displayed goods are provided by the associated goods module 104, but the specific display mode is realized by the goods information list sub-module 102-3.
In addition, the associated commodities provided by the associated commodity module 104 may be physical commodities such as ornaments, cosmetics, clothes and the like; but also service goods such as travel goods and the like.
The live broadcast theme determining module 105 is configured to match the corresponding commodity for a live broadcast video clip obtained by a live broadcast video or a captured live broadcast video according to the extracted feature information, and generate a live broadcast theme used for identifying the live broadcast video or the live broadcast video clip according to a matching result.
Identifying the contained commodities and the characteristic information of the related information thereof according to the live video and the live video clip, matching the commodities with the list of the commodities related to the live video provided by the related commodity module 104, and determining the live theme according to the matching result.
In practical application, the acquired feature information of the live video or the live video clip comprises a plurality of key fields, a plurality of data storage processing modes and a plurality of machine learning models suitable for classification can be selected to establish a video-commodity identification model, and the feature information is matched with specific commodities in the commodity list.
In the embodiment, in consideration of large data volume and convenience of data processing, an ODPS (aerial ladder) data system in ali is used and stored in the form of a database table; selecting a training sample from the characteristic information, performing cross validation and/or manual labeling on the sample data, removing information which is practically irrelevant to the live video content, and labeling expected results corresponding to all elements in the sample data; and training the sample data by using a machine learning model TreeLink so as to establish the video-commodity identification model. In addition, in this embodiment, the feature information at least includes a keyword field that can represent the commodity SPU, such as clothing, mother and infant products, mobile phones, and the like.
In this embodiment, after matching the feature information with a specific commodity in the list of commodities, a live theme for identifying the live video or the live video clip is generated according to a matching result; the live theme comprises at least one of the following information: merchandise information related to the live video; merchandise information related to a segment of the live video; for example, the name of the item, the promotional means are embodied in the live topic as part of the live topic.
The live topic determined by the live topic determination module 105 in the above manner necessarily includes: the video processing device comprises commodity information related to the live video and commodity information related to a segment of the live video.
In practical application, the live theme can be used in a plurality of scenes, specifically, the live theme is used as a title of the live video and a title of a section of the live video, the titles are shown in a live product catalog and can also be used as a retrieval clue of commodities or videos, and when a user searches the commodities or videos or the section of the videos, the titles are searched; the live theme can also be used as a part of the content introduction of the live video or a part of the content introduction of the segment of the live video, so that the theme is more fit with the content of the video or the video segment; in addition, the live topic may also be a title of playback of the live video, or a title of playback of a segment of the live video; the live topic may also be part of a content vignette of the playback of the live video or of a segment of the live video. Because the live broadcast theme can provide more accurate correspondence for commodities and videos or video clips, a user can watch live broadcast or playback of videos and live broadcast or playback of video clips for interested commodities conveniently, and time does not need to be wasted for waiting for the interested commodities.
Fig. 3 is a schematic diagram of a second implementation manner of a live broadcast system provided in an embodiment of the present application. The structure of this embodiment is that two modules are added to the implementation shown in fig. 1, including: a live theme and video association module 306 and a live theme push module 307.
The live broadcast theme and video association module 306 is configured to establish and store an association relationship between the live broadcast video or the live broadcast video clip and the live broadcast theme.
In the embodiment provided by the application, the live topic determining module obtains the live topic corresponding to the live video, inputs the live topic and video association module 306, establishes the association relationship between the live video (or the live video clip) and the live topic, and stores the association relationship, so that the retrieval and query of the live video or the live video clip can be conveniently realized. For example, the matching live subjects of a certain live video and the intercepted 5 video clips are all the commodity SPU a, so that the live video and the live video clips are associated with the commodity SPU a, and thus, when a user searches the commodity a, all the corresponding live videos or video clips under the commodity SPU a can be queried.
The live topic pushing module 307 is configured to push the determined live topic to a user client that watches the live video or a segment of the live video, or correspondingly put the determined live topic into a live video directory for the user client to browse.
The live topic and video association module 306 obtains the association relationship between the live topic and the live video (or the live video clip), and inputs the association relationship to the live topic push module 307, which pushes the live topic of the live video to the user client that watches the live video or the live video clip, or correspondingly puts the determined live topic into the directory information for the user client to browse, and the user client clicks the live topic responded in the directory information to obtain and watch the corresponding live video or the live video clip.
In addition, the live topic pushing module 307 may also push the live topic determined by the live topic determining module to a user client that watches the live video or a live video clip, or correspondingly put the determined live topic into directory information for the user client to browse.
In this embodiment, the live theme and video association module and the live theme push module are added, when a user client searches for a live video related to a certain commodity, the live theme push module 307 pushes the related live video or live video clip information to the user client by using an SPU, and a user clicks and displays each specific related live video and its corresponding introduction information or live video clip and its corresponding introduction information, so that the user can select and view the live video. The introduction information corresponding to the video or the received information corresponding to the video clip is the content of the live topic information generated by the live topic determination module 105, and includes commodity-related information, such as commodity purchase amount information. In addition, in the embodiment, the corresponding relationship between the live broadcast theme and the live broadcast video or the live broadcast video clip is added to the live broadcast video classification directory in real time, and meanwhile, the corresponding relationship between the live broadcast theme and the live broadcast video playback or the live broadcast video clip playback is added to the live broadcast video playback classification directory in real time, so that a user can conveniently and visually select the interested live broadcast video or live broadcast video clip or live broadcast video playback or live broadcast video clip playback, more convenient and more precise service is provided for shopping of the user, the user can pertinently select the interested content instead of blindly watching the video, and time is saved.
It should be noted that, both the live subject and video association module 306 and the live subject pushing module 307 may include one or both, and are within the scope of the present application.
In the foregoing embodiment, a live broadcast system is provided, and based on the live broadcast system provided in the embodiment of the present application, an embodiment of the present application further provides a method for determining a live broadcast video topic, please refer to fig. 4 to 6, where fig. 4 is a processing flow chart of the method for determining a live broadcast video topic provided in the embodiment of the present application; fig. 5 is a flowchart illustrating a training process of a live topic recognition model in a method for determining a live video topic according to an embodiment of the present application; fig. 6 is a processing flow chart of determining a live video topic according to a determined commodity matched with a live video or a live video segment in the method for determining a live video topic provided in the embodiment of the present application.
An embodiment of the present application further provides a method for determining a live video topic, as shown in fig. 4, including:
step S401, in the process of live video playing, acquiring the characteristic information of the live video in real time, and acquiring the characteristic information of picture information or text information related to the live video; and obtaining a list of commodities related to the live video.
The live video, the feature information of the picture information or the text information related to the live video and the list of the commodities related to the live video are acquired in real time, and detailed explanation is already carried out in the live system provided by the embodiment of the application.
Optionally, receiving information for live video sent from at least one user client; and sending at least part of the information to at least one other user client.
Optionally, a commodity release request corresponding to the anchor client is received, and commodity information corresponding to the commodity release request is obtained; and sending the commodity information to a user client side which at least partially accesses the live video.
Optionally, a list of the goods related to the live video is presented to the user client.
Optionally, purchasing behavior information of a behavior of purchasing a commodity, which occurs at the user client, is acquired; and transmitting the purchasing behavior information to at least one user client.
And S402, substituting the characteristic information into a live broadcast topic identification model obtained through pre-training, and identifying commodities matched with the live broadcast video or each live broadcast video clip of the live broadcast video.
In this step, according to the feature information of the live video or the live video clip acquired in step 401, the commodities matched with the live video or each live video clip of the live video are identified through the live topic identification model.
The task executed by the live broadcast topic identification model is matching, namely under the condition that commodities possibly related to a live broadcast video are determined in advance, the commodities are taken as candidate items, and the commodity or the commodities are actually matched with the live broadcast video through collecting characteristic information of various dimensions of the live broadcast video. As the related commodities are selected from a limited candidate set, the training difficulty of the live topic recognition module is low, and the recognition accuracy is high.
In this embodiment, a training process for obtaining the live topic recognition model is shown in fig. 5, and includes:
step S402-1, selecting the characteristic information of the marked live videos and the picture information or text information related to the live videos as sample data; and the marking refers to determining the commodities related to the live video.
In this step, sample data for establishing the live topic identification model is selected. In the embodiment, live videos of commodities related to the live videos are selected firstly, and feature information of the live videos and related picture information or text information are used for establishing the live topic identification model and commodities corresponding to the live videos are used as sample data; the characteristic information, the related picture information or text information, and the commodity have been described in detail in the live broadcast system provided in the above embodiment of the present application.
And S402-2, taking the sample data as training data, training a machine learning model, and enabling the machine learning model to be capable of matching commodities related to the live video according to the feature information of the live video and the picture information or text information related to the live video.
In the step, training is achieved to obtain a machine learning model, and matching with the commodities related to the videos is achieved according to the live video characteristic information and the picture information or the text information related to the live videos through the machine learning model. Specifically, the training sample obtained in step S402-1 is used to train a current live topic recognition model, and the live topic recognition model obtained after training is used as the live topic recognition model obtained by pre-training and used to identify a commodity matched with the live video or each live video segment of the live video.
The Machine learning models are various, such as a logistic regression model, a random forest model, a bayesian method model, a neural network model, a TreeLink (decision tree), an SVM (support Vector Machine), and the like. The model can be selected through processes of iteration, trial and error and the like according to business characteristics and experience accumulation, and can also be selected by other existing tools and methods, such as a data mining map, a method provided by Sciket-lean (machine learning module based on BSD open source licensing) and the like.
In this embodiment, the characteristic information of the marked live videos and the picture information or text information related to the live videos are used as sample data and the determined commodities related to the live videos, the sample data and the determined commodities are processed by an Aliyun ODPS data processing service system to be used as training data, and a live topic identification model is preferably established by adopting machine learning TreeLink algorithm training. In addition, the machine learning model in this embodiment may also select an SVM model.
In practical applications, before the step is executed, the method generally further includes preprocessing the feature information, including: performing cross validation and/or manual marking on the characteristic information, and removing information which is actually irrelevant to the commodities related to the live video; and the characteristic information after the eliminating step is taken as the characteristic information substituted into the live broadcast topic identification model, so that the live broadcast video is matched with the related commodities more quickly and accurately.
Step S403, determining the theme of the live video or the live video clip according to the determined commodity matched with the live video or the live video clip.
The purpose of this step is to determine the theme of the live video or the live video clip, and specifically includes steps S403-1 to S403-2, as shown in fig. 6:
step S403-1, according to the matched commodity, acquiring commodity information related to the commodity;
and S403-2, determining the theme of the live video or the live video clip by adopting a preset conversion method according to the commodity information related to the commodity.
In this embodiment, the theme of the live video or the live video clip is re-calibrated by inserting the commodity name and the anchor name into the live title. In addition, other conversion methods can be used to determine the live video topic.
Optionally, the commodity is represented by at least one of the following ways: item class, name of the item, SPU of the item, SKU of the item, item number of the item.
The SPU, SKU of the merchandise has been described in detail in the embodiment of a live broadcast system provided herein.
After determining the theme of the live video or the live video segment, the live video theme has a plurality of application scenes, for example, at least one of the following steps is performed: when determining a title for the live video or the live video clip, taking the title as the title or a component of the title; when a content introduction is made for the live video or the live video clip, the subject is used as a part of the content introduction; when the live video or the playback of the live video clip is made, the theme is taken as a title or a part of the title; and when the played back content introduction of the live video or the live video clip is made, taking the theme as a part of the played back content introduction.
Corresponding to the method for determining the live video theme provided by the embodiment of the present application, an embodiment of the present application further provides a device for determining the live video theme, which is described below with reference to the accompanying drawings.
Referring to fig. 7, a schematic diagram of an apparatus for determining a live video topic provided by an embodiment of the present application is shown. Since the apparatus embodiments are substantially similar to the method embodiments, they are described in a relatively simple manner, and reference may be made to some of the descriptions of the method embodiments for relevant points. The device embodiments described below are merely illustrative.
The application provides a device for confirming live video theme, includes:
a feature information obtaining unit 701, configured to obtain, in real time, feature information of a live video during a live video playing process, and obtain feature information of picture information or text information related to the live video;
a matching unit 702, configured to substitute the feature information into a live topic identification model obtained through pre-training, and identify a commodity that matches the live video or each live video segment of the live video;
a live video topic determining unit 703, configured to determine a topic of the live video or the live video segment according to the determined commodity matching the live video or the live video segment.
Optionally, the live topic identification model is obtained by training in the following manner:
selecting the characteristic information of the marked live videos and the picture information or text information related to the live videos as sample data; the marking means determining commodities related to the live video;
and training a machine learning model by using the sample data as training data, so that the machine learning model can be matched with the commodities related to the live video according to the characteristic information of the live video and the picture information or text information related to the live video.
Optionally, the machine learning model adopts a TreeLink model or an SVM model.
Optionally, before the step of substituting the feature information into the live topic recognition model obtained through pre-training, the step includes:
performing cross validation and/or manual marking on the characteristic information, and removing information which is actually irrelevant to the commodities related to the live video; and taking the characteristic information after the eliminating step as the characteristic information substituted into the live topic identification model.
Optionally, the theme of the live video or the live video clip is determined according to the determined commodity matched with the live video or the live video clip, and the theme is implemented by the following method:
acquiring commodity information related to the commodity according to the determined matched commodity;
and determining the theme of the live video or the live video clip by adopting a preset conversion method according to the commodity information related to the commodity.
Optionally, the commodity is represented by at least one of the following ways: item class, name of the item, SPU of the item, SKU of the item, item number of the item.
Optionally, after determining the theme of the live video or the live video clip, at least one of the following steps is performed:
when determining a title for the live video or the live video clip, taking the title as the title or a component of the title;
when a content introduction is made for the live video or the live video clip, the subject is used as a part of the content introduction;
when the live video or the playback of the live video clip is made, the theme is taken as a title or a part of the title;
and when the played back content introduction of the live video or the live video clip is made, taking the theme as a part of the played back content introduction.
An embodiment of the present application further provides an electronic device, please refer to fig. 7, which is a schematic diagram of an embodiment of an electronic device for determining a live video topic provided in the present application. Since the apparatus embodiments are substantially similar to the method embodiments, they are described in a relatively simple manner, and reference may be made to some of the descriptions of the method embodiments for relevant points. The device embodiments described below are merely illustrative.
An electronic device of the present embodiment includes: a processor 801; and a memory 802, where the memory 802 is used to store a program for determining a live video theme, and after the device is powered on and the program for determining a live video theme is run by the processor, the following steps are performed:
in the process of live video playing, acquiring the characteristic information of the live video in real time, and acquiring the characteristic information of picture information or text information related to the live video; the feature information of the live video comprises screenshot information and sound information acquired from the live video; the picture information or the text information related to the live video comprises commodity related information reflecting commodities related to the live video; and obtaining a list of commodities related to the live video;
substituting the characteristic information into a live broadcast topic identification model obtained through pre-training, and identifying commodities matched with the live broadcast video or each live broadcast video clip of the live broadcast video;
and determining the theme of the live video or the live video clip according to the determined commodity matched with the live video or the live video clip.
Although the present application has been described with reference to the preferred embodiments, it is not intended to limit the present application, and those skilled in the art can make variations and modifications without departing from the spirit and scope of the present application, therefore, the scope of the present application should be determined by the claims that follow.
In a typical configuration, a computing device includes one or more processors (CPUs), input/output interfaces, network interfaces, and memory.
The memory may include forms of volatile memory in a computer readable medium, Random Access Memory (RAM) and/or non-volatile memory, such as Read Only Memory (ROM) or flash memory (flash RAM). Memory is an example of a computer-readable medium.
1. Computer-readable media, including both non-transitory and non-transitory, removable and non-removable media, may implement information storage by any method or technology. The information may be computer readable instructions, data structures, modules of a program, or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), Static Random Access Memory (SRAM), Dynamic Random Access Memory (DRAM), other types of Random Access Memory (RAM), Read Only Memory (ROM), Electrically Erasable Programmable Read Only Memory (EEPROM), flash memory or other memory technology, compact disc read only memory (CD-ROM), Digital Versatile Discs (DVD) or other optical storage, magnetic cassettes, magnetic tape magnetic disk storage or other magnetic storage devices, or any other non-transmission medium that can be used to store information that can be accessed by a computing device. As defined herein, computer readable media does not include non-transitory computer readable media (transient media), such as modulated data signals and carrier waves.
2. As will be appreciated by one skilled in the art, embodiments of the present application may be provided as a method, system, or computer program product. Accordingly, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present application may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.
Claims (18)
1. A live broadcast system, comprising:
the live broadcast module is used for receiving a live broadcast video sent by a main broadcast client and sending at least part of the live broadcast video to at least one user client;
the content service module is used for receiving text or picture information corresponding to the live video and sending at least part of the information to at least one other user client;
the related commodity module is used for providing a list of commodities related to the live video;
the extraction module is used for extracting the received live video and the text or picture information related to the live video;
and the live broadcast theme determining module is used for matching the corresponding commodities aiming at live broadcast video or live broadcast video clips obtained by intercepting the live broadcast video according to the extracted characteristic information and generating live broadcast themes used for identifying the live broadcast video or the live broadcast video clips according to matching results.
2. The live system of claim 1, wherein the content service module comprises a message service sub-module configured to receive information for live video from at least one of the user clients and to send at least part of the information to at least one other user client.
3. The live broadcasting system of claim 1, wherein the content service module includes a commodity information publishing sub-module, configured to receive a commodity publishing request corresponding to the anchor client, obtain commodity information corresponding to the commodity publishing request, and send the commodity information to at least some user clients accessing the live video.
4. The live system of claim 1, wherein the content service module comprises a merchandise information listing sub-module configured to present a listing of merchandise associated with the live video to a user client.
5. The live broadcast system of claim 1, wherein the content service module comprises a shopping information sub-module configured to obtain purchasing behavior information of a behavior of purchasing goods occurring at the user client, and send the purchasing behavior information to at least one user client.
6. The live broadcast system of claim 1, wherein the live broadcast topic determined by the live broadcast topic determination module comprises at least one of the following information:
merchandise information related to the live video;
merchandise information related to a segment of the live video.
7. A live system according to claim 6, wherein the live topic determined by the live topic determination module is for use in one or more of the following:
a title of the live video, or a title of a segment of the live video;
a portion of a content vignette of the live video, or a portion of a content vignette of a segment of the live video;
a title of playback of the live video, or a title of playback of a segment of the live video;
a portion of a content vignette of the playback of the live video, or a portion of a content vignette of the playback of a segment of the live video.
8. The live broadcasting system of claim 1, comprising a live broadcasting theme pushing module, configured to push the determined live broadcasting theme to a user client that watches the live broadcasting video or a segment of the live broadcasting video, or put the determined live broadcasting theme in a live broadcasting video directory for the user client to browse.
9. The live broadcasting system of claim 1, comprising a live broadcasting theme and video associating module, configured to establish and store an association relationship between the live broadcasting video or the live broadcasting video clip and the live broadcasting theme.
10. A method of determining a live video topic, comprising:
in the process of live video playing, acquiring the characteristic information of the live video in real time, and acquiring the characteristic information of picture information or text information related to the live video; and obtaining a list of commodities related to the live video;
substituting the characteristic information into a live broadcast topic identification model obtained through pre-training, and identifying commodities matched with the live broadcast video or each live broadcast video clip of the live broadcast video;
and determining the theme of the live video or the live video clip according to the determined commodity matched with the live video or the live video clip.
11. The method for determining the theme of the live video according to claim 10, wherein the live theme recognition model is obtained by training as follows:
selecting the characteristic information of the marked live videos and the picture information or text information related to the live videos as sample data; the marking means determining commodities related to the live video;
and training a machine learning model by using the sample data as training data, so that the machine learning model can be matched with the commodities related to the live video according to the characteristic information of the live video and the picture information or text information related to the live video.
12. The method of determining live video topics as claimed in claim 11, wherein said machine learning model employs a TreeLink model or an SVM model.
13. The method for determining the theme of the live video, as claimed in claim 10, wherein before the step of substituting the feature information into the live theme recognition model obtained by pre-training, the step of:
performing cross validation and/or manual marking on the characteristic information, and removing information which is actually irrelevant to the commodities related to the live video; and taking the characteristic information after the eliminating step as the characteristic information substituted into the live topic identification model.
14. The method for determining the theme of the live video, as claimed in claim 10, wherein the theme of the live video or the live video segment is determined according to the determined commodities matched with the live video or the live video segment, and the method is implemented as follows:
acquiring commodity information related to the commodity according to the determined matched commodity;
and determining the theme of the live video or the live video clip by adopting a preset conversion method according to the commodity information related to the commodity.
15. The method of claim 14, wherein the item is represented by at least one of: item class, name of the item, SPU of the item, SKU of the item, item number of the item.
16. The method of claim 10, wherein determining the topic of the live video or live video segment is performed by at least one of:
when determining a title for the live video or the live video clip, taking the title as the title or a component of the title;
when a content introduction is made for the live video or the live video clip, the subject is used as a part of the content introduction;
when the live video or the playback of the live video clip is made, the theme is taken as a title or a part of the title;
and when the played back content introduction of the live video or the live video clip is made, taking the theme as a part of the played back content introduction.
17. An apparatus for determining a topic of a live video, comprising:
the system comprises a characteristic information acquisition unit, a video processing unit and a video processing unit, wherein the characteristic information acquisition unit is used for acquiring the characteristic information of a live video in real time and acquiring the characteristic information of picture information or text information related to the live video in the process of carrying out the live video;
the matching unit is used for substituting the characteristic information into a live broadcast theme recognition model obtained through pre-training, and recognizing commodities matched with the live broadcast video or each live broadcast video clip of the live broadcast video;
and the live video theme determining unit is used for determining the theme of the live video or the live video clip according to the determined commodity matched with the live video or the live video clip.
18. An electronic device, comprising:
a processor; and
the memorizer is used for storing a program for determining the theme of the live video, and after the equipment is powered on and runs the program for determining the theme of the live video through the processor, the following steps are executed:
in the process of live video playing, acquiring the characteristic information of the live video in real time, and acquiring the characteristic information of picture information or text information related to the live video; the feature information of the live video comprises screenshot information and sound information acquired from the live video; the picture information or the text information related to the live video comprises commodity related information reflecting commodities related to the live video; and obtaining a list of commodities related to the live video;
substituting the characteristic information into a live broadcast topic identification model obtained through pre-training, and identifying commodities matched with the live broadcast video or each live broadcast video clip of the live broadcast video;
and determining the theme of the live video or the live video clip according to the determined commodity matched with the live video or the live video clip.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710470458.7A CN109104639B (en) | 2017-06-20 | 2017-06-20 | Live broadcast system, method and device for determining live broadcast video theme and electronic equipment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710470458.7A CN109104639B (en) | 2017-06-20 | 2017-06-20 | Live broadcast system, method and device for determining live broadcast video theme and electronic equipment |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109104639A CN109104639A (en) | 2018-12-28 |
| CN109104639B true CN109104639B (en) | 2021-11-02 |
Family
ID=64795683
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710470458.7A Active CN109104639B (en) | 2017-06-20 | 2017-06-20 | Live broadcast system, method and device for determining live broadcast video theme and electronic equipment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109104639B (en) |
Families Citing this family (27)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111385642A (en) * | 2018-12-29 | 2020-07-07 | 阿里巴巴集团控股有限公司 | Media information processing method, device, server, equipment and storage medium |
| CN110704610A (en) * | 2019-05-15 | 2020-01-17 | 北京信息科技大学 | Sports news war topic classification method |
| CN112055225B (en) * | 2019-06-06 | 2023-02-17 | 阿里巴巴集团控股有限公司 | Live broadcast video interception, commodity information generation and object information generation methods and devices |
| CN110381328A (en) * | 2019-06-11 | 2019-10-25 | 拉扎斯网络科技（上海）有限公司 | Commodity operation method, apparatus, server based on video and can storage medium |
| CN111757130A (en) * | 2019-06-27 | 2020-10-09 | 上海妃鱼网络科技有限公司 | Broadcast selection method and system for Internet live broadcast |
| CN112351288B (en) * | 2019-08-07 | 2023-04-28 | 阿里巴巴集团控股有限公司 | Live broadcast information processing method, device, server, terminal and storage medium |
| CN112399258B (en) * | 2019-08-13 | 2022-06-07 | 腾讯科技（深圳）有限公司 | Live playback video generation playing method and device, storage medium and electronic equipment |
| CN110418156A (en) * | 2019-08-27 | 2019-11-05 | 上海掌门科技有限公司 | Information processing method and device |
| CN111127079A (en) * | 2019-12-04 | 2020-05-08 | 北京奇艺世纪科技有限公司 | Method, device, computer equipment and storage medium for issuing commodity resources |
| CN111131875A (en) * | 2019-12-06 | 2020-05-08 | 北京达佳互联信息技术有限公司 | Information display method, device and system, electronic equipment and storage medium |
| CN113327121A (en) * | 2020-02-28 | 2021-08-31 | 阿里巴巴集团控股有限公司 | Anchor selection related system, method, device and equipment |
| CN111429217A (en) * | 2020-03-20 | 2020-07-17 | 青岛聚好联科技有限公司 | Commodity display method and device based on smart community live broadcast |
| CN111754302B (en) * | 2020-06-24 | 2021-04-27 | 广西壹农乐信息技术有限公司 | Video live broadcast interface commodity display intelligent management system based on big data |
| CN113763072B (en) * | 2020-07-01 | 2023-09-22 | 北京京东振世信息技术有限公司 | Method and device for analyzing information |
| CN111970523B (en) * | 2020-08-13 | 2022-09-09 | 北京达佳互联信息技术有限公司 | Information display method, device, terminal, server and storage medium |
| CN111954077A (en) * | 2020-08-24 | 2020-11-17 | 上海连尚网络科技有限公司 | Video stream processing method and device for live broadcast |
| JP7473930B2 (en) | 2020-10-28 | 2024-04-24 | ビゴ テクノロジー ピーティーイー． リミテッド | Video playback method, device, terminal, and storage medium |
| CN113852832B (en) * | 2020-11-26 | 2022-09-20 | 阿里巴巴集团控股有限公司 | Video processing method, device, equipment and storage medium |
| CN112465596B (en) * | 2020-12-01 | 2021-07-23 | 山西易通天下网络科技有限公司 | Image information processing cloud computing platform based on electronic commerce live broadcast |
| CN112348649B (en) * | 2020-12-07 | 2022-07-29 | 福州灵和晞科技有限公司 | Live broadcast category recommendation method |
| CN112785381A (en) * | 2021-01-28 | 2021-05-11 | 维沃移动通信有限公司 | Information display method, device and equipment |
| CN112995691B (en) * | 2021-02-26 | 2023-02-28 | 北京达佳互联信息技术有限公司 | Live broadcast data processing method and device, electronic equipment and storage medium |
| CN115079878B (en) * | 2021-03-15 | 2024-04-16 | 北京字节跳动网络技术有限公司 | Object display method, device, electronic equipment and storage medium |
| CN113438531B (en) * | 2021-05-18 | 2023-09-05 | 北京达佳互联信息技术有限公司 | Object display method and device, electronic equipment and storage medium |
| CN113315989B (en) * | 2021-05-28 | 2022-12-27 | 北京沃东天骏信息技术有限公司 | Live broadcast processing method, live broadcast platform, device, system, medium and equipment |
| CN113613054B (en) * | 2021-07-30 | 2023-05-09 | 北京达佳互联信息技术有限公司 | Information prompting method, device, electronic equipment and computer readable storage medium |
| CN115115994A (en) * | 2022-08-25 | 2022-09-27 | 北京凯利时科技有限公司 | Multi-modality based director method and system and computer program product |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2015078255A1 (en) * | 2013-11-28 | 2015-06-04 | 乐视网信息技术（北京）股份有限公司 | Video title display method and system, and browser |
| CN104994404A (en) * | 2015-07-06 | 2015-10-21 | 无锡天脉聚源传媒科技有限公司 | Method and device for obtaining keywords for video |
| CN106791907A (en) * | 2017-01-13 | 2017-05-31 | 四川鑫义堂文化传播有限公司 | Auction net cast method, apparatus and system |
Family Cites Families (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104715023B (en) * | 2015-03-02 | 2018-08-03 | 北京奇艺世纪科技有限公司 | Method of Commodity Recommendation based on video content and system |
| US20170171622A1 (en) * | 2015-12-15 | 2017-06-15 | Le Holdings (Beijing) Co., Ltd. | Methods and content systems, servers, terminals and communication systems |
| CN105656917B (en) * | 2016-02-02 | 2018-11-16 | 广州华多网络科技有限公司 | Control the method and device that online direct broadcast band starts broadcasting |
| CN106202317A (en) * | 2016-07-01 | 2016-12-07 | 传线网络科技（上海）有限公司 | Method of Commodity Recommendation based on video and device |
| CN106412621B (en) * | 2016-09-28 | 2019-11-26 | 广州华多网络科技有限公司 | Image display method and device, control method and relevant device between network direct broadcasting |
| CN106792081B (en) * | 2016-12-07 | 2020-01-10 | 腾讯科技（深圳）有限公司 | Live video pushing method and device |
| CN106604048A (en) * | 2016-12-15 | 2017-04-26 | 深圳市智联宝生态科技有限公司 | Direct broadcasting method and system with shopping module embedded in direct broadcasting room |
-
2017
- 2017-06-20 CN CN201710470458.7A patent/CN109104639B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2015078255A1 (en) * | 2013-11-28 | 2015-06-04 | 乐视网信息技术（北京）股份有限公司 | Video title display method and system, and browser |
| CN104994404A (en) * | 2015-07-06 | 2015-10-21 | 无锡天脉聚源传媒科技有限公司 | Method and device for obtaining keywords for video |
| CN106791907A (en) * | 2017-01-13 | 2017-05-31 | 四川鑫义堂文化传播有限公司 | Auction net cast method, apparatus and system |
Non-Patent Citations (1)
| Title |
|---|
| 贷齐乐直播***发布;马建堂;《计算机与网络》;20160926(第18期);全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109104639A (en) | 2018-12-28 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109104639B (en) | Live broadcast system, method and device for determining live broadcast video theme and electronic equipment | |
| TWI684875B (en) | Method for determining user behavior preference, method and device for displaying recommended information | |
| KR102315474B1 (en) | A computer-implemented method and non-transitory computer-readable storage medium for presentation of a content item synchronized with a media display | |
| US10299010B2 (en) | Method of displaying advertising during a video pause | |
| US10311103B2 (en) | Information search method, device, server and storage medium | |
| WO2018036456A1 (en) | Method and device for tracking and recognizing commodity in video image and displaying commodity information | |
| US20150256858A1 (en) | Method and device for providing information | |
| US20130282532A1 (en) | Systems and methods for presentation and analysis of media content | |
| CN112055225A (en) | Live broadcast video interception, commodity information generation and object information generation methods and devices | |
| US10854014B2 (en) | Intelligent object recognizer | |
| CN103190146A (en) | Content capture device and methods for automatically tagging content | |
| CN113301360B (en) | Information prompting method, computing device and storage medium | |
| WO2012092240A2 (en) | Method and apparatus for providing or utilizing interactive video with tagged objects | |
| CN105657558B (en) | A kind of information providing method, methods of exhibiting, device and smart television | |
| US10440435B1 (en) | Performing searches while viewing video content | |
| CN105893404A (en) | Natural information identification based pushing system and method, and client | |
| CN112396456A (en) | Advertisement pushing method and device, storage medium and terminal | |
| KR20120128511A (en) | System and method for providing information on product exposed | |
| CN110324676A (en) | Data processing method, media content put-on method, device and storage medium | |
| CN112714349A (en) | Data processing method, commodity display method and video playing method | |
| KR20160027486A (en) | Apparatus and method of providing advertisement, and apparatus and method of displaying advertisement | |
| US20180204270A1 (en) | Method, Apparatus, and System for Data Object Information Processing | |
| WO2014076442A1 (en) | A self-service facility for content providers | |
| US11375290B2 (en) | Rules-based ancillary data | |
| US20160132593A1 (en) | Method and system for audio signal recognition and cloud search architecure utilizing same |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |