CN114443943A - Information scheduling method, device and equipment and computer readable storage medium - Google Patents

Information scheduling method, device and equipment and computer readable storage medium Download PDF

Info

Publication number
CN114443943A
CN114443943A CN202011216286.9A CN202011216286A CN114443943A CN 114443943 A CN114443943 A CN 114443943A CN 202011216286 A CN202011216286 A CN 202011216286A CN 114443943 A CN114443943 A CN 114443943A
Authority
CN
China
Prior art keywords
information
scheduling
audited
real
index
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
CN202011216286.9A
Other languages
Chinese (zh)
Inventor
朱昱锦
李建勇
朱新宇
茅克盾
袁易之
司超
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Tencent Technology Shenzhen Co Ltd
Original Assignee
Tencent Technology Shenzhen Co Ltd
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Tencent Technology Shenzhen Co Ltd filed Critical Tencent Technology Shenzhen Co Ltd
Priority to CN202011216286.9A priority Critical patent/CN114443943A/en
Publication of CN114443943A publication Critical patent/CN114443943A/en
Pending legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/90—Details of database functions independent of the retrieved data types
    • G06F16/95—Retrieval from the web
    • G06F16/953—Querying, e.g. by the use of web search engines
    • G06F16/9535—Search customisation based on user profiles and personalisation
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/10—Office automation; Time management
    • G06Q10/103—Workflow collaboration or project management

Landscapes

  • Engineering & Computer Science (AREA)
  • Business, Economics & Management (AREA)
  • Databases & Information Systems (AREA)
  • Theoretical Computer Science (AREA)
  • Human Resources & Organizations (AREA)
  • Strategic Management (AREA)
  • Entrepreneurship & Innovation (AREA)
  • Data Mining & Analysis (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • General Engineering & Computer Science (AREA)
  • Economics (AREA)
  • Marketing (AREA)
  • Operations Research (AREA)
  • Quality & Reliability (AREA)
  • Tourism & Hospitality (AREA)
  • General Business, Economics & Management (AREA)
  • Debugging And Monitoring (AREA)

Abstract

The application provides an information scheduling method, device, equipment and a computer readable storage medium; the method comprises the following steps: when push information pushed to a user is received, extracting information to be audited from the push information, and extracting real-time feedback data corresponding to the information to be audited from real-time behavior data corresponding to the push information; determining a scheduling index corresponding to the information to be audited based on the real-time feedback data; the scheduling index represents the abnormal influence degree of the information to be audited; determining a real-time scheduling sequence of the information to be audited by using a scheduling index corresponding to the information to be audited; and according to the real-time scheduling sequence, performing auditing scheduling on the information to be audited. Through the application, the stability of the online environment can be improved.

Description

Information scheduling method, device and equipment and computer readable storage medium
Technical Field
The present application relates to information processing technologies, and in particular, to an information scheduling method, apparatus, device, and computer readable storage medium.
Background
When using social applications, users often subscribe to some information sources actively. When an information source publishes new push information, the social application can quickly acquire the push information, and push the push information to a user subscribing the information source in a push information stream aggregated by the push information, so that the user can acquire the push information published by the information source concerned by the user in time.
The push information brings certain influence to users, and in order to maintain the online environment, the push information with adverse effect needs to be searched out and off-shelf by means of posterior audit. In the related art, the pushed information to be audited is audited mainly according to the time sequence. Therefore, the audit sequence for issuing late abnormal push information is bound to be backward, so that the possibility that the abnormal push information is audited in time is low, the efficiency of auditing the abnormal push information is low, and the stability of the online environment is low finally.
Disclosure of Invention
The embodiment of the application provides an information scheduling method, an information scheduling device, information scheduling equipment and a computer readable storage medium, which can improve the stability of an online environment.
The technical scheme of the embodiment of the application is realized as follows:
an embodiment of the present application provides an information scheduling method, including:
when push information pushed to a user is received, extracting information to be audited from the push information, and extracting real-time feedback data corresponding to the information to be audited from real-time behavior data corresponding to the push information;
determining a scheduling index corresponding to the information to be audited based on the real-time feedback data; the scheduling index represents the abnormal influence degree of the information to be audited;
determining a real-time scheduling sequence of the information to be audited by using a scheduling index corresponding to the information to be audited;
and performing auditing scheduling on the information to be audited according to the real-time scheduling sequence.
An embodiment of the present application provides an information scheduling apparatus, including:
the data extraction module is used for extracting information to be audited from the push information when the push information pushed to a user is received, and extracting real-time feedback data corresponding to the information to be audited from the real-time behavior data corresponding to the push information;
the index determining module is used for determining a scheduling index corresponding to the information to be audited based on the real-time feedback data; the scheduling index represents the abnormal influence degree of the information to be audited;
the order determining module is used for determining a real-time scheduling order of the information to be audited by using the scheduling index corresponding to the information to be audited;
and the auditing scheduling module is used for auditing and scheduling the information to be audited according to the real-time scheduling sequence.
In some embodiments of the present application, the index determining module is further configured to construct an index weight corresponding to the to-be-audited information according to the real-time feedback data and the scheduling index calculation rule; the index weight represents the abnormal degree of the information to be audited, and the index calculation rule is a temporary index calculation rule with the best scheduling effect selected from a plurality of temporary index calculation rules; and calculating the scheduling index corresponding to the information to be audited by adopting the index weight corresponding to the information to be audited.
In some embodiments of the present application, the real-time feedback data includes an information rejection amount, an information distribution amount, and a click rate;
the index determining module is further configured to compare the information rejection amount with the information release amount to obtain an information source rejection rate corresponding to the information to be audited; the information rejection quantity represents the quantity of information of which the information source corresponding to the information to be audited is off-shelf, and the information release quantity represents the total quantity of the information released by the information source; according to the index calculation rule, carrying out logarithm operation on the rejection rate of the information source and the click rate to obtain the index weight corresponding to the information to be audited; the click rate characterizes the number of times the information to be audited is displayed.
In some embodiments of the present application, the real-time feedback data includes an information rejection amount, an information distribution amount, a click rate, and a click rate acceleration;
the index determining module is further configured to compare the information rejection amount with the information release amount to obtain an information source rejection rate corresponding to the information to be audited; carrying out weight adjustment on the information source rejection rate to obtain the adjusted information source rejection rate; according to the scheduling index calculation rule, carrying out logarithm operation on the adjusted information source rejection rate, the click rate and the click rate increase to obtain the index weight corresponding to the information to be audited; the click rate acceleration represents the increase speed of the click rate of the information to be audited.
In some embodiments of the present application, the index determining module is further configured to calculate a first index component by using the click rate and a preset weight construction parameter; calculating a second index component by using the click rate acceleration rate and the preset weight construction parameter; multiplying the first index component, the second index component and the adjusted information source rejection rate to obtain an intermediate multiplication result corresponding to the information to be audited; and carrying out logarithmic operation on the intermediate multiplication result to obtain the index weight corresponding to the information to be audited.
In some embodiments of the present application, the index determining module is further configured to reduce the weight of the index corresponding to the information to be checked, so as to obtain the reduced weight of the index; and calculating the scheduling index corresponding to the information to be audited by using the reduced index weight.
In some embodiments of the present application, the data extraction module is further configured to select the real-time feedback data from the real-time behavior data according to a comparison result between the real-time behavior data and a real-time data threshold; the real-time behavior data is returned by the user aiming at the push information, and the real-time behavior data comprises at least one of real-time click quantity and real-time comment quantity; and screening corresponding information to be audited from the push information by utilizing the real-time feedback data.
In some embodiments of the application, the data extraction module is further configured to compare the acquired historical behavior data corresponding to the push information with a historical data threshold value to obtain a historical data comparison result; the historical behavior data comprises at least one of historical click rate, historical negative feedback rate, historical click acceleration rate and historical browsing duration; screening information to be audited from the push information according to the historical data comparison result; and screening out the real-time feedback data corresponding to the information to be audited from the real-time behavior data.
In some embodiments of the present application, the order determining module is further configured to determine an index determining order of the information to be checked according to the obtained issuing time of the information to be checked; the index determination sequence indicates the sequence of the scheduling indexes corresponding to the information to be audited when determining;
correspondingly, the index determining module is further configured to determine the scheduling index corresponding to the information to be audited according to the index determining sequence and by using the real-time feedback data.
In some embodiments of the present application, the information scheduling apparatus further includes: a rule selection module;
the data extraction module is further configured to obtain at least one piece of test audit information and at least one piece of test feedback data corresponding to the at least one piece of test audit information, and divide the at least one piece of test audit information into at least two test information sets;
the index determining module is further configured to calculate a test index weight corresponding to the test audit information in each test information set according to the test feedback data corresponding to the test audit information in each test information set and the temporary index calculation rule corresponding to each test information set; constructing a test scheduling index corresponding to the test auditing information in each test information set by using the test index weight;
the auditing scheduling module is further used for auditing and scheduling the test auditing information in each test information set according to the sequence of the test scheduling indexes;
the rule selection module is also used for acquiring scheduling effect data of the test auditing information in each test information set; the scheduling effect data describes the auditing and scheduling conditions of the test auditing information in each test information set; and selecting the scheduling index calculation rule with the best effect from the temporary index calculation rules corresponding to each test information set according to the scheduling effect data.
In some embodiments of the present application, the rule selecting module is further configured to determine, according to the scheduling effect data, a scheduling effect corresponding to each test information set; and taking the temporary index calculation rule corresponding to the test information set with the best scheduling effect as the scheduling index calculation rule.
In some embodiments of the present application, the order determining module is further configured to update a real-time scheduling order of the information to be audited based on latest feedback data corresponding to the acquired information to be audited, so as to obtain an updated scheduling order;
correspondingly, the audit scheduling module is further configured to perform audit scheduling on the information to be audited according to the updated scheduling sequence.
An embodiment of the present application provides an information scheduling apparatus, including:
a memory for storing executable information scheduling instructions;
and the processor is used for realizing the information scheduling method provided by the embodiment of the application when the executable information scheduling instruction stored in the memory is executed.
The embodiment of the present application provides a computer-readable storage medium, which stores executable information scheduling instructions for causing a processor to execute, so as to implement the information scheduling method provided in the embodiment of the present application.
The embodiment of the application has the following beneficial effects:
the information scheduling equipment can construct the scheduling index corresponding to the information to be audited by utilizing real-time feedback data corresponding to the information to be audited, so that the abnormal influence degree of the information to be audited is determined according to the user feedback, and then the real-time scheduling sequence of the information to be audited is determined according to the abnormal influence degree of the information to be audited, namely the sequence of the information to be audited during auditing is determined.
Drawings
Fig. 1 is a schematic diagram of an alternative architecture of an information scheduling system 100 provided in an embodiment of the present application;
fig. 2 is a schematic structural diagram of an information scheduling apparatus in fig. 1 according to an embodiment of the present application;
fig. 3 is a first alternative flow chart of an information scheduling method provided in an embodiment of the present application;
fig. 4 is a schematic diagram illustrating an optional flow of an information scheduling method according to an embodiment of the present application;
fig. 5 is a schematic flow chart diagram three of an alternative information scheduling method provided in the embodiment of the present application;
fig. 6 is a first exemplary process diagram for extracting real-time feedback data and information to be reviewed according to an embodiment of the present application;
fig. 7 is a second exemplary process diagram for extracting real-time feedback data and information to be reviewed according to the embodiment of the present application;
fig. 8 is a fourth alternative flowchart of an information scheduling method according to an embodiment of the present application;
FIG. 9 is a diagram of an example process for determining a scheduling index according to an embodiment of the present application;
fig. 10(a) is a diagram of an example of a process for auditing subscription content according to an embodiment of the present application;
fig. 10(b) is a diagram of an example of a process for selecting a priority algorithm provided in an embodiment of the present application.
Detailed Description
In order to make the objectives, technical solutions and advantages of the present application clearer, the present application will be described in further detail with reference to the attached drawings, the described embodiments should not be considered as limiting the present application, and all other embodiments obtained by a person of ordinary skill in the art without creative efforts shall fall within the protection scope of the present application.
In the following description, reference is made to "some embodiments" which describe a subset of all possible embodiments, but it is understood that "some embodiments" may be the same subset or different subsets of all possible embodiments, and may be combined with each other without conflict.
In the following description, references to the terms "first", "second", and the like are only used for distinguishing similar objects and do not denote a particular order or importance, but rather the terms "first", "second", and the like may be used interchangeably with the order of priority or the order in which they are expressed, where permissible, to enable embodiments of the present application described herein to be practiced otherwise than as specifically illustrated and described herein.
Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. The terminology used herein is for the purpose of describing embodiments of the present application only and is not intended to be limiting of the application.
Before further detailed description of the embodiments of the present application, terms and expressions referred to in the embodiments of the present application will be described, and the terms and expressions referred to in the embodiments of the present application will be used for the following explanation.
1) Cloud Technology refers to a hosting Technology for unifying resources of hardware, software, network and the like in a wide area network or a local area network to realize calculation, storage, processing and sharing of data.
The cloud technology is based on the general names of network technology, information technology, integration technology, management platform technology, application technology and the like applied in the cloud computing business model, can form a resource pool, is used as required, and is flexible and convenient. Cloud computing technology will become an important support. Background services of the technical network system require a large amount of computing and storage resources, such as video websites, picture-like websites and more web portals. With the high development and application of the internet industry, each article may have its own identification mark and needs to be transmitted to a background system for logic processing, data in different levels are processed separately, and various industrial data need strong system background support and can only be realized through cloud computing.
2) The push information aggregation refers to aggregating push information of different information sources subscribed by a user, that is, different information sources. Push message aggregation can help users to continuously obtain the latest feed, namely the latest information published by the feed.
3) And pushing the information flow, and representing the information flow presented to the user after the pushed information is aggregated. The push information stream is generally displayed on a display interface of the user terminal, and in the push information stream, a title and a cover map of each push information are generally displayed, and in the title and the cover map, a jump link is attached, so that when the user clicks the title or the cover map, the user can jump to a detail page of the push information.
4) The review mainly refers to pushing the push information to a terminal of a user, and then determining whether the push information is abnormal content which brings discomfort to the user according to the response of the user to different dimensions of the push information, for example, determining whether the push information is content such as a security sensitive topic, a rumor, and the like. When the push information is judged to be the bad content by the review, the bad content needs to be deleted in time, so that the bad content is prevented from being continuously spread on line to cause bad influence.
5) The information scheduling is to determine an audit priority for each piece of push information to be audited under the condition that a plurality of pieces of push information to be audited exist, so that the more urgent push information can be audited preferentially.
6) The Page View volume (PV), also known as the click volume, reflects the number of pages clicked on by the user. PV is the primary indicator of the measure of a source of information, even a piece of pushed information. In other words, PV measures how well the user receives the teletext content page.
7) The Video View (VV) reflects the number of times that the Video is played, and when the number of times that the Video is played is larger, the Video is more easily accepted by the general public. Therefore, VV also reflects the degree of reception of video content by the user.
The social application provides great convenience for communication and exchange of users, and various information sources are registered in the social application, so that various types of pushed information can be spread in the social application, and the browsing amount of the pushed information is increased. When a user uses a social application, the user often actively subscribes some interested information sources in various information sources registered in the social application, so that the user can conveniently acquire the interested content in time. After the information source issues the latest pushed information, the social application starts a pushed information aggregation process, that is, the latest pushed information is aggregated in a pushed information stream corresponding to a user who subscribes to the information source, so that the latest pushed information is pushed to the user who subscribes to the information source in time.
The user can also generate feedback of different dimensions in the process of browsing the push information. However, not all push messages can have a positive impact on the user, for example, some information channels may maliciously distribute rumors as the latest information messages, or push messages related to sensitive topics. If these abnormal pushed information is allowed to propagate on the network, it will have a bad influence on information security, public opinion security, and even user experience. In order to solve the problem, the abnormal push information needs to be found out by means of review so as to delete the abnormal push information in time.
In the related art, each piece of push information is reviewed according to the time sequence of each piece of push information, that is, the review sequence is determined according to the release time of each piece of push information, so that the first review of the first release and the second review of the second release are realized. However, considering the online environment, the push information with problems needs to be checked and off-shelf in time, if the review is performed based on the time sequence, the review sequence for issuing the late abnormal push information is backward, so that the limited review manpower cannot be arranged for the most urgent content, and finally the efficiency of checking the abnormal push information is low, and the online environment is unstable.
Meanwhile, due to the fact that the abnormal pushing information is difficult to be checked in time, the abnormal pushing information is difficult to be off shelf in time, the abnormal pushing information still has an opportunity to be transmitted on the network, adverse effects are continuously brought to users, and experience of the whole product is affected.
In addition, when the auditing sequence of various types of push information is determined according to the time sequence, when the auditing speed is lower than the auditing speed of the push information, the push information to be audited is accumulated, so that some push information to be audited cannot be audited in real time, and the auditing real-time degree of the push information is lower.
The embodiment of the application provides an information scheduling method, an information scheduling device, information scheduling equipment and a computer readable storage medium, which can improve the stability of an online environment. An exemplary application of the information scheduling apparatus provided in the embodiment of the present application is described below, and the information scheduling apparatus provided in the embodiment of the present application may be implemented as a terminal, or may be implemented as a server. The server may be an independent physical server, a server cluster or a distributed system formed by a plurality of physical servers, or a cloud server providing basic cloud computing services such as cloud service, a cloud database, cloud computing, a cloud function, cloud storage, network service, cloud communication, middleware service, domain name service, security service, CDN, big data and artificial intelligence platform. The terminal may be, but is not limited to, a smart phone, a tablet computer, a laptop computer, a desktop computer, a smart speaker, a smart watch, and the like. The terminal and the server may be directly or indirectly connected through wired or wireless communication, and the application is not limited herein. Next, an exemplary application of the information scheduling apparatus will be explained.
Referring to fig. 1, fig. 1 is an optional architecture diagram of an information scheduling system 100 provided in this embodiment of the present application, in order to implement supporting an information scheduling application, terminals (terminal 400-1 and terminal 400-2 are exemplarily shown) are connected to an information scheduling apparatus 200 through a network 300, and the network 300 may be a wide area network or a local area network, or a combination of both. The information scheduling apparatus 200 is also provided with a database 500, and the database 500 is used to provide data services for the information scheduling apparatus 200.
The terminal 400-1 (information source side) is configured to edit the push information and send the push information to the information scheduling apparatus 200 for distribution. After the manager of each information source finishes editing the push information, the manager clicks a distribution button provided on the graphical interface 400-11 of the terminal 400-1 to distribute the push information through the network 300, and the terminal 400-1 further sends the push information to the information scheduling apparatus 200.
The terminal 400-2 (information display side) is configured to display the released push information on the graphical interface 400-21, obtain corresponding real-time behavior data of the user on the push information, and then send the real-time behavior data to the information scheduling apparatus 200.
When the information scheduling apparatus 200 receives the push information pushed to the user, the real-time feedback data corresponding to the information to be audited is extracted from the real-time behavior data corresponding to the push information. Then, the information scheduling device 200 determines a scheduling index corresponding to the information to be audited based on the real-time feedback data, where the scheduling index represents an abnormal influence degree of the information to be audited. Then, the information scheduling device 200 determines a real-time scheduling sequence of the information to be audited by using the scheduling index corresponding to the information to be audited. Finally, the information scheduling device 200 performs audit scheduling on the information to be audited according to the real-time scheduling data, for example, the information to be audited is sent to a work terminal of an auditor to audit the information to be audited.
Referring to fig. 2, fig. 2 is a schematic structural diagram of the information scheduling apparatus in fig. 1 according to an embodiment of the present application, and the information scheduling apparatus 200 shown in fig. 2 includes: at least one processor 210, memory 250, at least one network interface 220, and a user interface 230. The various components in the information scheduling apparatus 200 are coupled together by a bus system 240. It is understood that the bus system 240 is used to enable communications among the components. The bus system 240 includes a power bus, a control bus, and a status signal bus in addition to a data bus. For clarity of illustration, however, the various buses are labeled as bus system 240 in fig. 2.
The Processor 210 may be an integrated circuit chip having Signal processing capabilities, such as a general purpose Processor, a Digital Signal Processor (DSP), or other programmable logic device, discrete gate or transistor logic device, discrete hardware components, or the like, wherein the general purpose Processor may be a microprocessor or any conventional Processor, or the like.
The user interface 230 includes one or more output devices 231, including one or more speakers and/or one or more visual display screens, that enable the presentation of media content. The user interface 230 also includes one or more input devices 232, including user interface components that facilitate user input, such as a keyboard, mouse, microphone, touch screen display, camera, other input buttons and controls.
The memory 250 may be removable, non-removable, or a combination thereof. Exemplary hardware devices include solid state memory, hard disk drives, optical disk drives, and the like. Memory 250 optionally includes one or more storage devices physically located remotely from processor 210.
The memory 250 includes volatile memory or nonvolatile memory, and may include both volatile and nonvolatile memory. The nonvolatile memory may be a Read Only Memory (ROM), and the volatile memory may be a Random Access Memory (RAM). The memory 250 described in embodiments herein is intended to comprise any suitable type of memory.
In some embodiments, memory 250 is capable of storing data, examples of which include programs, modules, and data structures, or a subset or superset thereof, to support various operations, as exemplified below.
An operating system 251 including system programs for processing various basic system services and performing hardware-related tasks, such as a framework layer, a core library layer, a driver layer, etc., for implementing various basic services and processing hardware-based tasks;
a network communication module 252 for communicating to other computing devices via one or more (wired or wireless) network interfaces 220, exemplary network interfaces 220 including: bluetooth, wireless-compatibility authentication (Wi-Fi), and Universal Serial Bus (USB), etc.;
a presentation module 253 to enable presentation of information (e.g., a user interface for operating peripherals and displaying content and information) via one or more output devices 231 (e.g., a display screen, speakers, etc.) associated with the user interface 230;
an input processing module 254 for detecting one or more user inputs or interactions from one of the one or more input devices 232 and translating the detected inputs or interactions.
In some embodiments, the information scheduling apparatus provided in the embodiments of the present application may be implemented in software, and fig. 2 illustrates the information scheduling apparatus 255 stored in the memory 250, which may be software in the form of programs and plug-ins, and includes the following software modules: the data extraction module 2551, the index determination module 2552, the order determination module 2553, the audit scheduling module 2554 and the rule selection module 2555 are logical and thus may be arbitrarily combined or further split depending on the functions implemented.
The functions of the respective modules will be explained below.
In other embodiments, the information scheduling apparatus provided in this embodiment may be implemented in hardware, and for example, the information scheduling apparatus provided in this embodiment may be a processor in the form of a hardware decoding processor, which is programmed to execute the information scheduling method provided in this embodiment, for example, the processor in the form of the hardware decoding processor may employ one or more Application Specific Integrated Circuits (ASICs), DSPs, Programmable Logic Devices (PLDs), Complex Programmable Logic Devices (CPLDs), Field Programmable Gate Arrays (FPGAs), or other electronic components.
Illustratively, an embodiment of the present application provides an information scheduling apparatus, including:
a memory for storing executable information scheduling instructions;
and the processor is used for realizing the information scheduling method provided by the embodiment of the application when the executable information scheduling instruction stored in the memory is executed.
In the following, the information scheduling method provided by the embodiment of the present application will be described in conjunction with exemplary applications and implementations of the information scheduling apparatus provided by the embodiment of the present application. It should be noted that the embodiments of the present application may be implemented by means of cloud technology.
Referring to fig. 3, fig. 3 is a first alternative flowchart of an information scheduling method provided in the embodiment of the present application, and will be described with reference to the steps shown in fig. 3.
S101, when push information pushed to a user is received, extracting information to be audited from the push information, and extracting real-time feedback data corresponding to the information to be audited from real-time behavior data corresponding to the push information.
The embodiment of the application is realized in a scene of auditing the published push information, for example, auditing the published hot articles or auditing the published short videos. In this scenario, the information scheduling apparatus receives the push information pushed by the information source to the user, that is, when receiving the push information, it will determine that information auditing needs to be started, so as to start the information scheduling process, at this time, the information scheduling apparatus will extract the push information that needs to be audited again from the push information, and obtain the information to be audited. Meanwhile, the information scheduling equipment can also acquire real-time feedback data corresponding to the information to be audited from the real-time behavior data corresponding to the push information.
It should be noted that the push information in the embodiment of the present application may be issued after being audited, or may be directly issued without being audited. However, for a certain piece of pushed information that has already been published, whether the pushed information is audited or not during publishing, the pushed information is highly likely to cause a certain degree of influence when being clicked and browsed by a user or in the process of network propagation, and whether the specifically caused influence endangers an online environment or not needs to be further audited to be determined, so that some pieces of pushed information need to be audited as information to be audited after being published.
It is understood that the information to be audited may be image and text information, or, for example, popular articles, popular pictures, etc., may also be video information, such as short videos recorded by the user, news release videos, etc., and may also be audio information, such as popular songs, recorded contents, etc.
In some embodiments of the present application, the information to be audited may refer to all pieces of push information that have been issued, and in this case, the information scheduling is implemented for all pieces of push information that have been issued. In other embodiments, the information to be audited may also be some pieces of pushed information selected from all pieces of pushed information already published, that is, the information scheduling device selects, from all pieces of pushed information already published, pushed information with a wide propagation range, for example, pushed information in which the number of people browsing reaches a certain threshold, or pushed information in which the number of times of review forwarding reaches a certain threshold (such pushed information needs to be focused during auditing due to its large influence), or pushed information with potential adverse effects, for example, pushed information in which the number of people in negative feedback reaches a threshold, or pushed information in which public opinion is guided in a negative direction, and then the selected pushed information is used as the information to be audited. Of course, the information scheduling device may also obtain the information to be checked in other ways, and the embodiment of the present application is not limited herein.
It should be noted that the real-time behavior data is data obtained by counting the feedback of the user for the pushed information at the current time, for example, the number of comments, the number of negative feedbacks, the total time length of the user staying on the page of the information to be audited, the number of positive feedbacks, the click amount, the play amount, and the like. Similarly, the real-time feedback data is data obtained by counting the feedback of the user for the information to be audited at the current moment, and the real-time feedback data can reflect the propagation degree of the information to be audited at the current moment, the emotional tendency of the user for the information to be audited at the current moment, and the like.
It is understood that in some embodiments of the present application, the information scheduling apparatus may be an immediate information scheduling process when receiving the push information. In other embodiments, the information scheduling apparatus may also start the information scheduling process when the push information is received and the preset time is reached.
And S102, determining a scheduling index corresponding to the information to be audited based on the real-time feedback data.
After the information scheduling device obtains the real-time feedback data, the information scheduling device determines a scheduling index corresponding to the information to be audited for the information to be audited by using the user feedback corresponding to the information to be audited. The real-time feedback data can reflect the propagation degree of the information to be audited at the current moment and the emotion tendency of the user to the information to be audited at the current moment, the propagation degree reflects the influence of the information to be audited, and the emotion tendency of the information to be audited of the user reflects the positive influence or the negative influence of the information to be audited, so that the scheduling index constructed by the real-time feedback data represents the abnormal influence of the information to be audited.
It should be noted that, in the embodiment of the present application, the scheduling index may be in the form of an abnormal influence level, for example, a high degree of abnormality, a medium degree of abnormality, a normal, and the like. The scheduling index may also be in the form of an anomaly impact score, e.g., 10000 points, 0 points, 500 points, etc.
In some embodiments of the present application, when the form of the scheduling indicator is an abnormal influence level, the information scheduling apparatus may directly determine whether the information to be audited is positively or negatively influenced (for example, negatively influenced when the negative feedback quantity is greater than the positive feedback quantity) according to a magnitude relationship between the positive feedback quantity and the negative feedback quantity in the real-time feedback data corresponding to the information to be audited, and determine the magnitude of the influence of the information to be audited (for example, the magnitude of the influence is greater when any one of the quantity of comments, the total time that the user stays on the page of the information to be audited, the click quantity, and the play quantity exceeds a preset threshold value) according to a magnitude relationship between the quantity of comments, the total time that the user stays on the page of the information to be audited, the click quantity, and the play quantity of the information to be audited, and then according to whether the negative influence is a positive influence or a negative influence, and (4) dividing the grade of the information to be audited according to whether the influence is large or small, wherein the obtained grade is the scheduling index. For example, when the information to be audited is negative influence and the influence is large, the scheduling index is determined to be highly abnormal, and when the information to be audited is positive influence and the influence is small, the scheduling index is determined to be normal.
When the form of the scheduling index is the abnormal influence score, the information scheduling device can calculate the real-time feedback data according to the scheduling index calculation rule, so that the scheduling index of the information to be audited is determined. Further, the scheduling index calculation rule may be a temporary index calculation rule with the best effect selected from a plurality of different temporary index calculation rules in the test stage according to the scheduling effect of the audit scheduling.
S103, determining a real-time scheduling sequence of the information to be audited by using the scheduling index corresponding to the information to be audited.
After the information scheduling index obtains the scheduling index corresponding to the information to be audited, the order of the information to be audited during auditing can be determined for the information to be audited according to the scheduling index, namely the abnormal influence degree of the information to be audited, so that the information scheduling equipment obtains the real-time scheduling order of the information to be audited, and the auditing scheduling can be performed according to the real-time scheduling order.
It should be noted that, in practice, in order to maintain the online environment as much as possible, it is necessary to check the push information having a large influence and a negative influence as soon as possible, that is, check the push information having a large abnormal influence as soon as possible. Therefore, the information scheduling equipment determines the auditing priority degree of the information to be audited by using the abnormal influence degree, and can realize that the information to be audited with larger abnormal influence, namely larger adverse influence, is audited first.
It can be understood that, when the scheduling index is the abnormal influence score of the audit, the information scheduling device may directly sequence the scheduling indexes corresponding to the information to be audited in a descending order, and the obtained sequencing result is the real-time scheduling sequence of each information to be audited.
When the scheduling index is the abnormal influence level, the information scheduling device may first sort the information to be audited corresponding to the highest abnormal influence according to the sequence of the release time to obtain the sub-scheduling sequence corresponding to the highest abnormal influence level, then sort the information to be audited corresponding to the next-highest abnormal influence level according to the sequence of the release time to obtain the sub-scheduling sequence corresponding to the next-highest abnormal influence level, until all the abnormal influence levels are sorted, obtain the sub-scheduling sequence corresponding to the abnormal influence level, and finally sort the sub-scheduling sequences according to the abnormal influence levels from high to low to finally obtain the real-time scheduling sequence of the information to be audited.
Illustratively, when the abnormal influence level is high abnormal and normal, the information scheduling device firstly sequences the information to be audited 1 and the information to be audited 2 corresponding to the high abnormal according to a time sequence, and the obtained sub-scheduling sequence is < the information to be audited 2 and the information to be audited 1 >; the information scheduling device sequences the normally corresponding information to be audited 3 and information to be audited 4 according to the time sequence, and the obtained sub-scheduling sequence is < information to be audited 3 and information to be audited 4 >. And finally, the information scheduling equipment arranges the two sub-scheduling sequences according to the highly abnormal and normal sequence, and the obtained real-time scheduling sequence is < information to be audited 2, information to be audited 1, information to be audited 3 and information to be audited 4 >.
And S104, performing auditing scheduling on the information to be audited according to the real-time scheduling sequence.
After the information scheduling device obtains the real-time scheduling sequence, the information to be audited is sequentially distributed to the terminals of the auditors or the machine devices capable of automatically auditing the information according to the real-time scheduling sequence, so that the auditing scheduling of the information to be audited is completed. Therefore, the to-be-audited information with high auditing urgency can be audited first, the auditing efficiency of the abnormal consultation information is improved, the to-be-audited information with problems can be positioned in time and processed, and the stability of the online environment is guaranteed.
It should be noted that, in some embodiments of the present application, the information scheduling device may perform automatic audit of information, and at this time, when the information scheduling device performs audit scheduling on information to be audited according to a real-time scheduling sequence, the method may be changed to: and auditing the information to be audited one by one according to the real-time scheduling sequence. At this time, the information scheduling device may determine whether the information to be audited has an illegal image by means of a computer vision technology, for example, an object recognition technology and a face recognition technology, may analyze whether the information to be audited has illegal text contents by means of a natural language processing technology, for example, entity word extraction and semantic understanding, and may determine whether the information to be audited has illegal voice by means of a voice recognition technology, thereby extracting push information to be off-shelved from the information to be audited, and off-shelving the consultation information to be off-shelved, thereby ensuring stability of an online environment and ensuring user experience.
In the embodiment of the application, the information scheduling device can utilize real-time feedback data corresponding to the information to be audited to construct a scheduling index corresponding to the information to be audited, so that the abnormal influence degree of the information to be audited is determined according to the feedback of a user, and then, the real-time scheduling sequence of the information to be audited is determined according to the abnormal influence degree of the information to be audited, namely, the sequence of the information to be audited during auditing is determined.
In some embodiments of the application, the information scheduling device performs audit scheduling on the information to be audited according to the real-time scheduling sequence, and after the information to be audited is audited, whether the information to be audited needs to be off-shelf or not can be judged. When the information to be audited needs to be off-shelved, the information scheduling equipment deletes the information to be audited from the line, and at this time, when a user triggers a display link for displaying the information to be audited in a push information stream, the information scheduling equipment feeds back that the information to be audited is off-shelved to the terminal, namely, off-shelve prompt information aiming at the information to be audited is sent to the terminal, and the terminal displays off-shelve prompt information corresponding to the information to be audited; when the information to be audited does not need to be off-shelf, the information scheduling equipment cannot delete the information to be audited from the line, and at the moment, when a user triggers a display link for displaying the information to be audited at the terminal, the terminal can normally display the information to be audited. It is understood that the display link of the information to be checked may be a title, a cover page, a video cover page, etc. of the information to be checked, and the application is not limited herein.
In some embodiments of the present application, the pushed information stream is displayed in a first preset area of a display interface of the terminal. When the terminal displays the off-shelf prompt information, the terminal can enter an off-shelf prompt page to display, or pop up a floating window on the pushed information stream to display. Similarly, when the terminal displays the information to be checked, the terminal may enter an information display interface to display the information, or pop up a floating window on the pushed information stream to display the information, which is not limited herein.
Referring to fig. 4, fig. 4 is a schematic view illustrating an optional flow chart of an information scheduling method according to an embodiment of the present application. In some embodiments of the present application, determining, based on the real-time feedback data, a scheduling indicator corresponding to the information to be audited, that is, a specific implementation process of S102, may include: S1021-S1022, as follows:
and S1021, determining the index weight corresponding to the information to be audited according to the real-time feedback data and the scheduling index calculation rule.
When the information scheduling equipment determines the scheduling index, the index calculation parameters are firstly analyzed from the real-time feedback data, and then the index calculation parameters are calculated according to the scheduling index calculation rule, so that the index weight corresponding to the information to be audited is determined. It should be noted that the index weight represents the abnormal degree of the information to be audited. That is to say, the information scheduling device determines the index weight of the information to be audited, and actually determines the abnormal degree of the information to be audited according to each index calculation parameter of the information to be audited. The abnormal degree may mean that the information to be audited has a wide spread range, i.e. causes a great influence in the public, or means that the information to be audited has a potentially adverse effect.
In some embodiments of the present application, the index calculation parameter may include a click amount and a browsing amount of the information to be audited, or may include related parameters of an information source of the information to be audited (i.e., an information source of the information to be audited), for example, the information source is rejected, the number of information to be off-shelf, the total amount of information issued by the information source, or may include an increase in the click amount, an increase in the browsing amount, and the like, which is not limited herein.
It should be noted that, in the embodiment of the present application, the index calculation rule is a temporary index calculation rule with the best scheduling effect selected from the multiple temporary index calculation rules, so that it can be ensured that a relatively accurate scheduling index is obtained in the information scheduling process.
And S1022, calculating the scheduling index corresponding to the information to be audited by adopting the index weight corresponding to the information to be audited.
After the information scheduling device constructs the index weight corresponding to the information to be audited, the information scheduling device firstly obtains a preset calculation parameter for calculating the scheduling index, and then calculates the scheduling index by using the preset parameter and the index weight. The index weight represents the abnormal degree of the information to be audited, and the abnormal degree of the information to be audited can reflect the influence caused by the abnormal degree of the information to be audited, so that the information scheduling equipment can construct the scheduling index representing the abnormal influence degree of the information to be audited according to the index weight.
In some embodiments of the present application, the preset calculation parameters may include a preset ratio configuration parameter, a preset difference parameter, and a preset index score. At this time, the information scheduling apparatus may first use the inverse number of the index weight as an index, use the natural number e as an index, construct an index result, then sum the index result and the preset duty ratio construction parameter, then use the sum result as a denominator, use the preset duty ratio construction parameter as a numerator, and calculate the duty ratio result. And then, the information scheduling equipment performs difference on the occupation ratio result and a preset difference value parameter to obtain a difference value result, and then multiplies the difference value result by a preset index fraction to obtain a product result, namely the scheduling index.
It can be understood that the specific numerical values of the preset ratio construction parameter, the preset difference parameter and the preset index score can be set according to actual conditions. For example, the preset proportion configuration parameter may be 1 or 2; the preset difference parameter can be 0.5 or 0.3; the preset index score may be 20000 or 5000.
Illustratively, the present application provides a formula for calculating a scheduling index, see formula (1):
Figure BDA0002760487520000181
wherein x isweightFor the index weight, 1 is an empirical value of a preset ratio construction parameter, 0.5 is an empirical value of a preset difference parameter, 20000 is an empirical value of a preset index score, and score is a scheduling index. After obtaining the specific value of the index weight, the information scheduling apparatus can substitute the specific value of the index weight into equation (1) to obtain the specific value of the scheduling index.
In the embodiment of the application, the information scheduling equipment can determine the index weight capable of representing the abnormal degree of the information to be audited according to the real-time feedback data and the scheduling index calculation rule, so that the abnormal influence degree can be accurately analyzed based on the abnormal degree of the information to be audited, and the accuracy of the scheduling index is improved.
In some embodiments of the present application, the real-time feedback data includes an information rejection amount, an information issue amount, and a click amount, and at this time, according to the real-time feedback data and a scheduling index calculation rule, an index weight corresponding to the information to be audited is determined, that is, a specific implementation process of S1021 may include: S201-S202, as follows:
s201, comparing the utilization information rejection quantity with the information release quantity to obtain an information source rejection rate corresponding to the information to be audited.
The information rejection quantity represents the quantity of information sources off shelves corresponding to the information to be audited, and the information release quantity represents the total quantity of information released by the information sources;
when the real-time feedback data comprises the information rejection amount, the information release amount and the click amount, the information scheduling equipment can simultaneously refer to the information source factors and the propagation factors according to the parameters to determine the abnormal degree of the information to be audited. At this time, when the information scheduling apparatus constructs the index weight, the information rejection amount is used as a numerator, the information distribution amount is used as a denominator, and ratio calculation is performed, and an obtained ratio result is an information source rejection rate.
It can be understood that the source reject rate is a proportion of the information distributed by the information on the shelves, and the information on the shelves generally has adverse effects on the public, so that the source reject rate can illustrate the possibility of potential adverse effects on the content distributed by the information source from the side, and whether important attention needs to be paid in auditing. For example, when the rejection rate of a certain information source is high, it indicates that the information source often issues some abnormal push information, and therefore, the possibility of the potential adverse effect of the information to be audited belonging to the information source is high.
In other embodiments of the present application, the information scheduling device may further construct an information source rejection rate based on a bayesian averaging algorithm, that is, perform bayesian averaging on the information rejection amount and the information distribution amount to obtain the information source rejection rate. At this time, the information scheduling device may first obtain a preset numerator parameter and a preset denominator parameter of the bayesian average, then sum the information rejection amount and the preset numerator parameter to obtain a numerator and a result, sum the information distribution amount and the preset denominator parameter to obtain a denominator and a result, and finally compare the numerator and the result with the denominator and the result to obtain a ratio, which is the information source rejection rate constructed based on the bayesian average algorithm.
It is understood that the specific values of the preset numerator parameter and the preset denominator parameter can be set according to practical situations, for example, the preset numerator parameter is set to 1, or set to 2.5, and the preset denominator parameter is set to 3, or set to 5, which is not limited herein.
S202, according to the index calculation rule, carrying out logarithm operation on the rejection rate and the click quantity of the information source to obtain the index weight corresponding to the information to be audited.
After the information scheduling device constructs the information source rejection rate, the information source rejection rate corresponding to the information to be audited can be multiplied by the click rate of the information to be audited, then a first product result obtained by the multiplication is used as a logarithm, the logarithm operation is carried out on the logarithm result and a preset base number, and the obtained logarithm operation result is the index weight of the information to be audited. In other embodiments, the information scheduling device may further obtain a preset weight construction parameter, add the click rate to the preset weight construction parameter to obtain a summation result, multiply the summation result with the information source rejection rate to obtain a second product result, and then perform a logarithm operation on the second product result and a preset base number by using the second product result as a logarithm, where the operation result is the index weight.
It is understood that the predetermined base number may be a natural number e, or may be other designated values, such as 10, or 2.
It should be noted that, in the embodiment of the present application, no matter whether the information to be checked is video information, image-text information, or audio information, the information to be checked can be displayed and played through a click operation of a user, so that the click amount reflects the number of times that the user plays or browses the information to be checked, that is, the click amount represents the number of times that the information to be checked is displayed.
It can be understood that there are positive correlations between the index weight and the click rate, and between the index weight and the information source rejection rate. The reason is that when the click quantity is large, it indicates that more users watch the information to be audited, so that the information to be audited is spread widely, and when the rejection rate of the information source is large, it indicates that the information source corresponding to the information to be audited frequently issues bad content, so that the possibility of potential bad influence of the information to be audited is also large, and the possibility of wide spread and large bad influence indicates that the abnormal degree of the information to be audited is large.
For example, the embodiment of the present application provides a formula for constructing an index weight, see formula (2):
Figure BDA0002760487520000201
wherein x ispv/vvThe click rate, x, of the information to be auditedrejectAs a quantity of information rejection, xtotalFor the information distribution amount, a is a preset weight construction parameter, 1 is an empirical value of a preset numerator parameter, 5 is an empirical value of a preset denominator parameter,
Figure BDA0002760487520000202
rejection rate of information source, x, constructed based on Bayesian average algorithmweightIs the index weight. After the information scheduling device obtains the specific numerical values of the click rate, the information rejection rate and the information distribution rate, the specific numerical value of the index weight can be obtained through calculation.
In the embodiment of the application, when the real-time feedback data comprises the information rejection amount, the information release amount and the click amount, the information scheduling device can calculate the information source rejection rate firstly, so that whether the information source frequently releases the content with adverse effect is determined, and then the abnormal degree of the information to be audited is judged according to the information source rejection rate and the click amount.
Referring to fig. 5, fig. 5 is a third optional flowchart of the information scheduling method according to the embodiment of the present application. In some embodiments of the present application, the real-time feedback data includes an information rejection amount, an information distribution amount, a click amount, and a click rate increase rate, and in this case, an index weight corresponding to the information to be audited is determined according to the real-time feedback data and a scheduling index calculation rule, that is, a specific implementation process of S1021 may include: S203-S205, as follows:
and S203, comparing the information rejection amount with the information release amount to obtain the information source rejection rate corresponding to the information to be audited.
In the embodiment of the present application, the scheduling index calculation rule determines the index weight by giving an emphasis to the propagation factor, and at this time, the index calculation parameters corresponding to the scheduling index calculation rule may include an information amount increase rate in addition to an information rejection amount, an information distribution amount, and a click amount. In this step, the information scheduling device may directly compare the information rejection amount with the information release amount to obtain an information source rejection rate, and may further construct the information source rejection rate based on a bayesian average algorithm, which is not limited in this application.
S204, carrying out weight adjustment on the rejection rate of the information source to obtain the adjusted rejection rate of the information source.
After the information scheduling device obtains the information source rejection rate, the weight of the information source rejection rate is adjusted, so that the influence of the information source rejection rate in the construction process of the index weight is changed, and the adjusted information source rejection rate is obtained. Further, the information scheduling device may reduce the weight of the information source rejection rate to obtain the adjusted information source rejection rate.
In some embodiments of the present application, the information scheduling device may perform an exponential operation on the information source rejection rate to achieve a weight adjustment on the information source rejection rate, for example, perform an exponential operation on the information source rejection rate and a preset exponent by using the information source rejection rate as a base number, so as to obtain an adjusted information source rejection rate. The preset index may be set according to actual conditions, for example, set to 0.1, 0.3, or even 0.01, and the application is not limited herein.
In other embodiments of the present application, the information scheduling apparatus may further perform a multiplication operation on the information source rejection rate to implement a weight adjustment on the information source rejection rate. For example, the information source rejection rate is multiplied by a preset weighting parameter, and the multiplication result is used as the adjusted information source rejection rate. The preset weighting parameter may be set according to actual conditions, for example, set to 0.2, set to 0.05, and the like, and the application is not limited herein.
Of course, in some embodiments, the information scheduling apparatus may also directly set the preset weighting parameter to 0, in this case, the index weight is constructed by directly using the click amount and the click amount acceleration rate.
S205, carrying out logarithm operation on the adjusted rejection rate, click rate and click rate increase of the information source according to a scheduling index calculation rule to obtain index weight corresponding to the information to be audited.
After obtaining the adjusted rejection rate of the information source, the information scheduling device may first increase the click rate and the click rate, and multiply the adjusted rejection rate of the information source to obtain a third multiplication result, and then perform a logarithmic operation on the third multiplication result as a logarithm and a preset base number, and use the result obtained by the logarithmic operation as an index weight.
In other embodiments, when the information scheduling device performs logarithmic operation on the adjusted rejection rate, click rate and click rate increase rate of the information source to obtain the index weight corresponding to the information to be audited, a preset weight construction parameter may be further obtained, a first index component is calculated by using the click rate and the preset weight construction parameter, a second index component is calculated by using the click rate increase rate and the preset weight construction parameter, the first index component, the second index component and the adjusted rejection rate of the information source are multiplied to obtain an intermediate multiplication result corresponding to the information to be audited, and then the intermediate multiplication result is subjected to logarithmic operation to obtain the index weight corresponding to the information to be audited.
Further, the information scheduling device may add the click rate and a preset weight construction parameter to obtain a first index component, or subtract the click rate and the preset weight construction parameter to obtain the first index component. Similarly, the information scheduling device may add the click rate increase rate and the preset weight structure parameter to obtain a second index component, and subtract the click rate increase rate and the preset weight structure parameter to obtain the second index component.
It should be noted that the click rate increase represents a rate of increase of the click rate of the information to be checked, and the rate of increase can represent how fast the information to be checked propagates, and both the click rate and the click rate can represent a propagation condition of the information to be checked.
It can be understood that there is a positive correlation between the index weight and the click rate increase rate. This is because, when the click rate of the information to be audited increases greatly, it indicates that the information is spread very quickly, however, there is a high possibility that the pushed information which is spread very quickly is abnormal, and therefore, when the click rate increases greatly, the index weight also increases accordingly.
For example, the embodiment of the present application provides another formula for constructing the index weight, as shown in formula (3):
Figure BDA0002760487520000221
wherein x ispv/vvxpv/vvThe click rate, x, of the information to be auditedrejectAs a quantity of information rejection, xtotalFor the information distribution amount, a is a preset weight construction parameter, 1 is an empirical value of a preset numerator parameter, 5 is an empirical value of a preset denominator parameter, Δ xpv/vvFor increasing the click rate of the information to be audited, p is a preset weighting parameter,
Figure BDA0002760487520000231
for adjusted information source rejection rate, xweightIs the index weight. After the information scheduling device knows the specific values of the parameters, the information scheduling device can substitute the specific values of the parameters into equation (3), so that the specific values of the index weights can be obtained.
In the embodiment of the application, the information scheduling device can adjust the weight of the rejection rate of the information source after calculating the rejection rate of the information source, and then adjust the weight of the rejection rate of the information source, so that when the index weight is constructed, the click rate and the click rate acceleration can be emphasized, namely the propagation factor of the information to be audited is emphasized to determine the abnormal degree of the information to be audited, and therefore in the subsequent process, whether the information to be audited needs to be audited first or not can be determined based on the influence of the information to be audited in the public, therefore, bad contents with high influence can be discovered as early as possible, the online environment is further maintained, and the use experience of subscription push information of a user is guaranteed.
In some embodiments of the present application, calculating a scheduling index corresponding to information to be audited by using the index weight corresponding to the information to be audited, that is, a specific implementation process of S1022 may include: S301-S302, as follows:
s301, performing weight reduction on the index weight corresponding to the information to be audited to obtain the weight of the index after the weight reduction.
In practice, due to reasons such as the index weight construction mode, the index weights corresponding to the information to be audited are relatively close, that is, the index weights corresponding to the information to be audited are concentrated in a smaller interval, so that the subsequently determined scheduling indexes cannot be layered, for example, the scheduling indexes are concentrated near 2000, and therefore, when auditing scheduling is performed based on the scheduling indexes, a good effect is difficult to achieve. In order to avoid such problems, the information scheduling device may perform weight reduction on the index weight corresponding to the information to be audited by using a preset weight reduction parameter when calculating the scheduling index corresponding to the information to be audited, so as to obtain the weight of the index after weight reduction.
In the application, the information scheduling device may multiply a preset weight reduction parameter and an index weight, and an obtained product result is the weight of the index after weight reduction; the information scheduling device may further perform an exponential operation with the preset weight reduction parameter as an index and the index weight as a base number, where an obtained exponential operation result is the weight of the index after the weight reduction.
It is understood that the preset weight-reducing parameter may be set to 0.2, may also be set to-0.1, and may also be set to other values according to practical situations, and the present application is not limited herein.
And S302, calculating a scheduling index corresponding to the information to be audited by using the reduced index weight.
After the information scheduling equipment obtains the reduced index weight, the preset proportion construction parameter, the preset difference parameter and the preset index fraction are obtained, then the opposite number of the reduced index weight is used as an index, a natural number e is used as an index to construct an index result, the index result and the preset proportion construction parameter are summed, then the summed result is used as a denominator, the preset proportion construction parameter is used as a numerator, and the proportion result is calculated. And then, the information scheduling equipment performs difference on the occupation ratio result and a preset difference parameter to obtain a difference result, and then multiplies the difference result by a preset index fraction to obtain a multiplication result, namely the scheduling index.
For example, another formula for calculating the scheduling index is provided in the embodiments of the present application, as shown in formula (4):
Figure BDA0002760487520000241
wherein x isweightFor the index weight, 1 is an empirical value of a preset ratio construction parameter, 0.5 is an empirical value of a preset difference parameter, 20000 is an empirical value of a preset index score, b is a preset weight reduction parameter, and score is a scheduling index. After obtaining the specific value of the parameter, the information scheduling apparatus can substitute the specific value of the index weight into equation (4) to obtain the specific value of the scheduling index.
In the embodiment of the application, the information scheduling device can also perform weight reduction on the index weight corresponding to the information to be audited, and then calculate the scheduling index by using the reduced index weight, so that the scheduling index can be layered in the aspect of value, and the subsequent auditing and scheduling based on the scheduling index are facilitated.
In some embodiments of the present application, extracting information to be audited from the push information, and extracting real-time feedback data corresponding to the information to be audited from the real-time behavior data corresponding to the push information, that is, a specific implementation process of S101 may include: S1011-S1012, as follows:
s1011, selecting real-time feedback data from the real-time behavior data according to the comparison result of the real-time behavior data and the real-time data threshold.
The information scheduling equipment acquires a preset real-time data threshold, compares the real-time behavior data with the real-time data threshold, and judges the size relation between the real-time behavior data and the real-time data threshold. And then, the information scheduling equipment takes the judged size relationship as a comparison result of the real-time behavior data and the real-time data threshold. And finally, the information scheduling equipment selects the real-time behavior data larger than the real-time data threshold value according to the comparison result, and uses the selected real-time behavior data as real-time feedback data, or selects the real-time behavior data equal to the real-time data threshold value as the real-time feedback data.
It can be understood that the push information is all the push information that has been published, that is, all the push information that exists on the line, and the user can generate feedback of various dimensions for the push information, and the feedback is real-time data, and thus, the real-time behavior data is real-time data returned by the user for the push information.
In the embodiment of the application, the real-time behavior data includes at least one of a real-time click rate and a real-time comment rate. The real-time click table shows the propagation condition of the initial information, and the real-time comment table shows the influence of the initial information. The real-time behavior data may only include the real-time click rate of the pushed information, or only include the real-time comment amount of the pushed initial information, and of course, the real-time behavior data may also include both the real-time click rate and the real-time comment amount of the pushed information. In other embodiments of the present application, other information, such as real-time approval, real-time bad assessment, etc., may also be included in the real-time behavior data.
Further, when the information scheduling apparatus compares the real-time behavior data with the real-time data threshold, the content included in the real-time behavior data is actually compared with the real-time data threshold. The real-time data threshold value should correspond to the content included in the real-time behavior data, for example, when the real-time behavior data only includes the real-time click amount, the real-time data threshold value is the real-time click amount threshold value, when the real-time behavior data only includes the real-time comment amount, the real-time data threshold value is the real-time comment amount threshold value, and when the real-time behavior data includes both the real-time click amount and the real-time comment amount, the real-time data threshold value includes the real-time click amount threshold value and the real-time comment amount threshold value.
It is understood that the real-time data threshold may be set according to actual situations, for example, the real-time comment amount threshold may be set to 100, or the real-time click amount threshold may be set to 1000, and the application is not limited herein.
In some embodiments of the application, the information scheduling device may select real-time data with a real-time click rate greater than or equal to a real-time number threshold to obtain real-time feedback data, or select real-time data with a real-time comment rate greater than or equal to a real-time number threshold to obtain real-time feedback data, or select real-time data with a real-time click rate and a real-time comment rate both greater than or equal to a real-time number threshold to obtain real-time feedback data.
And S1012, screening corresponding information to be checked from the pushed information by utilizing the real-time feedback data.
The real-time feedback data is the selected real-time behavior data, the real-time behavior data and the push information have a corresponding relation, and after the information scheduling device obtains the real-time feedback data, the information scheduling device can extract the push information corresponding to the real-time feedback data from the push information based on the corresponding relation between the real-time behavior data and the push information, and then the extracted prompt information is used as the information to be audited.
For example, an example of a process for extracting real-time feedback data and information to be audited is provided in the embodiment of the present application, referring to fig. 6, an information scheduling device first acquires push information 6-1 and real-time behavior data 6-2, and a corresponding relationship exists between the push information 6-1 and the real-time behavior data 6-2. The information scheduling equipment compares each data in the real-time behavior data 6-2 with the real-time data threshold 6-3, so that the real-time feedback data 6-21 which is greater than or equal to the real-time data threshold is selected from the real-time behavior data 6-2. And then, the information scheduling equipment selects the information to be audited 6-11 corresponding to the real-time feedback data 6-21 from the push information 6-1.
In the embodiment of the application, the information scheduling device can directly select the push information with large influence and fast propagation at the current moment according to the size relationship between the real-time behavior data corresponding to the push information and the real-time data threshold value, so that the push information is used as the information to be audited, and the follow-up audit can be performed on the push information with large influence and fast propagation at the current moment.
In some embodiments of the present application, extracting information to be audited from the push information, and extracting real-time feedback data corresponding to the information to be audited from the real-time behavior data corresponding to the push information, that is, a specific implementation process of S101 may include: S1013-S1015, as follows:
and S1013, comparing the acquired historical behavior data corresponding to the pushing information with a historical data threshold value to obtain a historical data comparison result.
The information scheduling device can select the information to be audited based on the real-time feedback of the user to the issued push information, and can also select the information to be audited based on the historical feedback of the user to the issued push information. At this time, the information scheduling device needs to obtain historical behavior data corresponding to the push information, that is, obtain data reflecting historical feedback of the user on the push information. The information scheduling equipment acquires the historical data threshold value, and compares the historical behavior data corresponding to the push information with the historical data threshold value, so that the size relation between the historical behavior data corresponding to the push information and the historical data threshold value can be determined, and the historical data comparison result corresponding to the push information is obtained.
It should be noted that the historical behavior data includes at least one of a historical click amount, a historical negative feedback amount (number of steps, reports, and the like), a historical click speed increase rate, and a historical browsing time length, that is, the information scheduling apparatus may use any one of the historical click amount, the historical negative feedback amount, the historical click speed increase rate, and the historical browsing time length as the historical behavior data, or use a plurality of the historical click amount, the historical negative feedback amount, the historical click speed increase rate, and the historical browsing time length as the historical data.
When the historical behavior data corresponding to the push information is compared with the historical data threshold, the information scheduling apparatus actually compares the content included in the historical behavior data with the historical data threshold. Further, the historical data threshold may be corresponding to the content included in the historical behavior data, that is, the historical data threshold may be a historical click amount threshold, a historical negative feedback amount threshold, a historical click speed increasing threshold, a historical browsing duration threshold, and the like.
It is understood that the specific value of the historical data threshold may be set according to actual situations, for example, the historical negative feedback amount threshold is set to 500, or the historical click acceleration rate is set to 1000/h, etc., and the present application is not limited herein.
And S1014, screening the information to be audited from the push information according to the historical data comparison result.
After the information scheduling device obtains the historical data comparison result, the information scheduling device selects the information of which the historical behavior data is greater than the historical data threshold value from the push information, and at this time, the selected information is the information to be audited.
And S1015, screening out real-time feedback data corresponding to the information to be audited from the real-time behavior data.
Because the corresponding relationship exists between the push information and the real-time behavior data, and the information to be audited is the push information of which the historical behavior data is greater than the historical data threshold, the information scheduling device can select the real-time behavior data corresponding to the information to be audited according to the corresponding relationship between the push information and the real-time behavior data, and the real-time behavior data is used as the real-time feedback data corresponding to the information to be audited.
For example, an example of a process for extracting real-time feedback data and information to be audited is provided in the embodiment of the present application, as shown in fig. 7, the information scheduling device first extracts push information 7-1, real-time behavior data 7-2 corresponding to the push information, and historical behavior data 7-3 of the push information. Then, the information scheduling device compares historical behavior data corresponding to the push information with a historical data threshold value 7-4, so that historical data 7-31 larger than the historical data threshold value 7-4 are selected from the total historical behavior data 7-3, then the information scheduling device selects information to be audited 7-11 corresponding to the historical data 7-31 from the push information 7-1, and finally the information scheduling device selects behavior data corresponding to the information to be audited 7-11 from the real-time behavior data according to the corresponding relation between the push information and the real-time behavior data, wherein the behavior data are real-time feedback data 7-21.
In the embodiment of the application, the information scheduling device can select the push information which has a large influence and is fast to propagate in the historical time according to the size relationship between the historical behavior data corresponding to the push information and the historical data threshold value, so that the push information is used as the information to be audited, and then the real-time feedback data corresponding to the information to be audited is selected from the real-time behavior data, so that the follow-up push information which has a large influence and is fast to propagate in the historical time can be audited.
Referring to fig. 8, fig. 8 is a fourth alternative flowchart of the information scheduling method according to the embodiment of the present application. In some embodiments of the present application, after extracting information to be audited from the push information, and extracting real-time feedback data corresponding to the information to be audited from the real-time behavior data corresponding to the push information, before determining a scheduling index corresponding to the information to be audited based on the real-time feedback data, that is, after S101 and before S102, the method further includes: s105, the following steps are carried out:
and S105, determining the index determination sequence of the information to be audited according to the acquired issuing time of the information to be audited.
The index determination sequence indicates the sequence of determining the scheduling indexes corresponding to the information to be audited.
In practice, since there is a high possibility of more information in the information to be audited, the information scheduling device may determine the order of constructing the scheduling index for the information to be audited according to the issue time of the information to be audited when constructing the scheduling index. At this time, the information scheduling device obtains the release time of the information to be checked, and then sorts each information in the information to be checked according to the sequence of the release time, where the obtained sorting result is the index determination sequence of the information to be checked. Correspondingly, the process of configuring the scheduling index corresponding to the information to be audited by the information scheduling device and implementing the feedback data, that is, the process of S102, will correspondingly become: and according to the index determination sequence, determining the scheduling index corresponding to the information to be audited by using the real-time feedback data. Therefore, under the condition that the information to be audited is scheduled based on the influence degree of the information to be audited to be abnormal, the scheduling index is constructed for the information to be audited which is issued first, and then auditing scheduling is performed by using the scheduling index, so that the issuing time of the information to be audited can be considered during scheduling, and the scheduling time fairness is improved.
In some embodiments of the application, the information scheduling device may sequence the information to be audited according to a principle that the release time is from front to back, so that the information to be audited can be guaranteed to be audited in time as much as possible under a condition that the audit scheduling is performed according to the abnormal influence degree, the information to be audited is not accumulated, and the instantaneity of the audit on the abnormal information is improved.
In other embodiments, the information scheduling device may also sequence the information to be checked according to a principle that the release time is from back to front, so that the newly released problem push information can be checked as quickly as possible.
For example, an example diagram of a process of determining a scheduling index is provided in the embodiment of the present application, as shown in fig. 9, an information scheduling device obtains 3 pieces of information to be audited, which are information to be audited 9-1 issued at 12:00, information to be audited 9-2 issued at 13:06, and information to be audited 9-3 issued at 10: 58. The information scheduling device sorts the information to be audited 9-1, the information to be audited 9-2 and the information to be audited 9-3 according to the sequence of the release time, so as to obtain the sequence to be audited (index determination sequence): information to be audited 9-3, information to be audited 9-1, and information to be audited 9-2. And finally, the information scheduling equipment determines the scheduling index of the information to be audited according to the sequence.
In the embodiment of the application, the information scheduling device can obtain the index determination sequence of the information to be audited by sequencing according to the release time of the information to be audited, and then determine the scheduling index of the information to be audited according to the index scheduling sequence, so that the release time of the information to be audited can be considered during scheduling, and the scheduling time fairness is improved.
In some embodiments of the present application, before determining the index weight corresponding to the information to be audited according to the real-time feedback data and the scheduling index calculation rule, that is, before S1021, the method may further include: S1023-S1027, as follows:
s1023, obtaining at least one piece of test audit information and at least one piece of test feedback data corresponding to the at least one piece of test audit information, and dividing the at least one piece of test audit information into at least two test information sets.
The embodiment of the application is applied to a test stage, and the scene of the index calculation rule with the best effect is selected from a plurality of temporary index calculation rules. At this time, the information scheduling device may obtain one or more pieces of pushed information that have already been issued, use the pushed information as test audit information, and obtain feedback data corresponding to the test audit information. Then, the information scheduling device takes the obtained push information as test audit information, and takes feedback data corresponding to the test audit information as test feedback data, so that the information scheduling device can obtain at least one piece of test audit information and at least one piece of test feedback data. Then, the information scheduling device divides at least one piece of test audit information into different test information sets, that is, performs bucket division on at least one piece of test audit information. It should be noted that the temporary index calculation rules corresponding to each test information set are different, that is, each test information set has its own corresponding audit scheduling manner, so that the index calculation rules are selected according to the scheduling effect of each test information set in the following.
It will be appreciated that the test audit information may be specified good published push information, such as specifying good news, articles, or videos, etc. The test audit information may also be any push information that has been issued, and the application is not limited herein.
It should be noted that the information scheduling device may randomly divide the test audit information into at least two test information sets, or may divide the information in the white list into the test information sets corresponding to the white list and divide the information in the black list into the test information sets corresponding to the black list.
In some embodiments of the application, the at least two test information sets may further include a test information set for performing audit scheduling according to a time sequence, so that a scheduling effect corresponding to each temporary index calculation rule can be compared with a scheduling effect corresponding to the time sequence, and a result of the time sequence corresponding to each temporary index calculation rule is determined.
And S1024, calculating the test index weight corresponding to the test audit information in each test information set according to the test feedback data corresponding to the test audit information in each test information set and the temporary index calculation rule corresponding to each test information set.
And the information scheduling equipment carries out scheduling index construction on each test auditing information in each test information set according to the temporary index calculation rule corresponding to each test information set, so as to obtain the test weight index corresponding to each test auditing information in each test information set.
It can be understood that the temporary index calculation rule corresponding to each test information set may be to construct a test index weight when constructing the index weight based on the information rejection rate and the click rate, or may determine the test index weight by emphasizing the propagation factor, which is not limited herein.
S1025, constructing a test scheduling index corresponding to the test auditing information in each test information set by using the test index weight, and auditing and scheduling the test auditing information in each test information set according to the sequence of the test scheduling index.
After obtaining the test index weight, the audit scheduling device constructs a corresponding test scheduling index for each test audit information in each test information set by using the test index weight corresponding to each test audit information in each test information set, and then performs audit scheduling on each test audit information in each test information set according to the obtained test scheduling index.
And S1026, acquiring scheduling effect data of the test auditing information in each test information set.
The scheduling effect data describes the auditing scheduling effect of the test auditing information in each test information set.
After the information scheduling device completes the audit scheduling of each test audit information in each test information set, scheduling effect data corresponding to each test audit information in each test information set is obtained respectively, so that the scheduling effect of each test information set is judged by using the scheduling effect data in the subsequent process, and the effect of the temporary index calculation rule corresponding to each test information set is further judged.
It can be understood that the scheduling effect data may include the click rate of the test audit information, and the time from the release of the test audit information to the audit, so that the information scheduling device may determine whether the test audit information having the abnormal influence is audited in time. The scheduling effect data can also comprise effect data which indicates whether the test audit information is adversely affected, the click rate of the test audit information is increased, and the user pictures with the most clicks on the test audit information.
S1027, according to the scheduling effect data, selecting a scheduling index calculation rule with the best effect from the temporary index calculation rules corresponding to each test information set.
And finally, the information scheduling equipment reads the scheduling effect data and determines the scheduling condition of each test audit information in each test information set, so that the temporary index calculation rule with the best effect is selected from the temporary index calculation rules corresponding to each test information set according to the scheduling condition of each test audit information, namely whether each test audit information is audited in time or not, and is used as the final scheduling index calculation rule.
In the embodiment of the application, the information scheduling device can select a temporary index calculation rule with the best auditing scheduling effect as a final index calculation rule according to the auditing scheduling effects corresponding to the temporary index calculation rules, so that the scheduling index is constructed based on the index calculation rule with the best scheduling effect, and the auditing scheduling effect is ensured.
In some embodiments of the application, the selecting, according to the scheduling effect data, a scheduling index calculation rule with a best effect from the temporary index calculation rules corresponding to each test information set, that is, a specific implementation process of S1027 may include: s1027a-S1027b, as follows:
s1027a, according to the scheduling effect data, determining the scheduling effect corresponding to each test information set.
S1027b, using the temporary index calculation rule corresponding to the test information set with the best scheduling effect as the scheduling index calculation rule.
The information scheduling equipment judges the scheduling effect of each test information set according to the effect scheduling data corresponding to each test auditing information in each test information set, and then compares the scheduling effects of each test information set with each other, so as to judge which test information set corresponds to the best scheduling effect. And then, the information scheduling equipment acquires a temporary index calculation rule corresponding to the test information set with the best scheduling effect, and the temporary index calculation rule is used as a scheduling index calculation rule finally used for scheduling index determination of the actual information to be checked.
In the embodiment of the application, the information scheduling device can firstly confirm the scheduling effect corresponding to each test information set according to the scheduling effect data, and then screen out the scheduling index calculation rule according to the scheduling effect, so that the scheduling index with the best scheduling effect can be understood to calculate the crime and the malice, and the information to be audited is determined according to the scheduling index.
In some embodiments of the present application, after determining a real-time scheduling order of the information to be audited by using the scheduling index corresponding to the information to be audited, and before performing audit scheduling on the information to be audited according to the real-time scheduling order, that is, after S103 and before S104, the method may further include: s106, the following steps are carried out:
and S106, updating the real-time scheduling sequence of the information to be audited based on the latest feedback data corresponding to the acquired information to be audited to obtain the updated scheduling sequence.
After the information scheduling device determines the real-time scheduling sequence, the information scheduling device can also continue to obtain the latest user feedback corresponding to the information to be audited, so as to obtain the latest feedback data corresponding to the information to be audited. Then, the information scheduling device determines the latest scheduling index corresponding to the information to be audited by using the latest feedback data according to the scheduling index calculation rule, and updates the real-time scheduling data based on the latest scheduling index, thereby obtaining the updated scheduling sequence. At this time, according to the real-time scheduling sequence, the audit scheduling is performed on the information to be audited, that is, the specific implementation process of S104 is changed to: and according to the updated scheduling sequence, performing audit scheduling on the information to be audited. Therefore, the scheduling sequence corresponding to the information to be audited can be updated in real time, and the auditing scheduling can be performed by using the latest scheduling sequence.
In the embodiment of the application, the information scheduling device can also update the real-time scheduling sequence corresponding to the information to be audited according to the latest feedback data, so that the auditing scheduling can be performed according to the latest real-time scheduling sequence in the following, and the auditing efficiency of the abnormal push information is further improved.
Next, an exemplary application of the embodiment of the present application in a practical application scenario will be described.
The embodiment of the application is realized under the scene of auditing and scheduling the subscription content according to the behavior data of the user. Referring to fig. 10(a), fig. 10(a) is a diagram illustrating a process of auditing subscription content according to an embodiment of the present application. The background 10-a (information scheduling device) will continuously obtain the subscription content (push information) recommended to the terminal 10-C of the user from different information sources 10-B, the user will also continuously generate different-dimension feedback during browsing, and the feedback is the user behavior data, and the background 10-a will collect the user behavior data (real-time behavior data). The background 10-a finds that the subscription content hits a predetermined review (audit) policy, for example, hot text review, feedback review, suspected low quality, random access check, by means of real-time triggering or timing triggering, and then reviews the subscription content (to-be-reviewed information) that will hit the review policy, and at this time, the subscription content that will hit the review policy is the review content. The background 10-a calculates priority data (scheduling index) of the contents to be checked by using a determined priority algorithm (scheduling index calculation rule), and distributes the contents to be checked to the auditors according to a priority sequence (real-time scheduling sequence) for scheduling and checking. In addition, the background 10-a continuously collects user behavior data (obtains the latest feedback data), for example, collects negative feedback, comments, PV/VV, etc., so as to continuously update the priority (obtain the updated scheduling order) when the review content is pending. After the auditor finishes auditing the audited contents and determines that the audited contents need to be off-shelf, the background 10-A will off-shelf the audited contents from the line, and when the terminal 10-C of the user clicks the title or the cover page of the audited contents, the background 10-A will send an off-shelf prompt character 'the contents have been off-shelf' to the terminal 10-C.
Before the auditing and scheduling are started, the background can also utilize the online content to test the scheduling effect of each priority algorithm (temporary index calculation rule), so that the priority algorithm with the best effect is selected to obtain the determined priority algorithm. Fig. 10(b) is a diagram of an example of a process for selecting a priority algorithm provided in an embodiment of the present application, and referring to fig. 10(b), the process may be divided into:
s401, recommending information; each recommending side (information source) pushes subscription information (push information) to the user side.
S402, recommending information to be stored; and each recommending side sends the subscribed information to the background, and the background stores the subscribed information into the database.
S403, behavior data collection; the background collects real-time data (real-time behavior data) of the user's reaction to the subscription information, such as comments, negative feedback, dwell time, PV/VV, etc., and stores the data in the database.
S404, extracting the audition content; and the background pushes some contents on the line to an audit sending pool according to the post audit sending strategy. Further, the posterior submission strategy mainly includes the following two types:
1) timing strategy (offline real-time order fetching user's real-time data) (screening information to be checked based on historical behavior data): the historical negative feedback quantity of the content is larger than a certain threshold value, and the historical report quantity is larger than a certain threshold value (the historical negative feedback quantity and the historical report quantity are collectively called as the historical negative feedback quantity); the historical PV/VV speed-up (historical click speed-up) of the content is greater than a certain threshold (historical data threshold); the historical browsing duration of the content is less than a certain threshold.
2) Real-time policy (screening information to be audited based on real-time data): the real-time PV/VV (real-time click volume) of the content is greater than a certain threshold (real-time data threshold); the real-time comment amount of the content is larger than a certain threshold value.
S405, dividing the barrel.
The background distributes the content to be checked to different content buckets (at least two test information sets) according to a set strategy (white list, random classification and the like), wherein the content buckets comprise an experiment bucket 1-an experiment bucket i and a comparison bucket (the comparison bucket is scheduled according to a time sequence), and each experiment bucket is firstly configured with a priority algorithm (a temporary index calculation rule corresponding to each test information set).
S406, in-bucket sorting (not shown); the background uses the priority algorithm of the experiment barrel to calculate the priority data (test scheduling index) of the contents to be checked in each barrel according to the content pool time sequence (the release time of each information to be checked), and uses the priority data to discharge the priority sequence in the barrel.
S407, updating to be checked (not shown); and when the contents to be checked are to be checked in the experiment barrel, the real-time priority can be continuously acquired, and the priority sequence of the contents in the barrel is updated.
S408, a collar slip; and obtaining the contents to be reviewed from different content buckets by the auditors according to the priority order and reviewing.
S409, collecting data; and (4) acquiring review data reported by the auditors by the background, performing report analysis, and selecting the priority algorithm with the best effect.
It should be noted that these characteristics should be satisfied during review: idempotent, namely the same content only enters the same experiment barrel, and the dispatching of the experiment barrel with the consistent content is always unchanged; real-time performance, namely, real-time user behavior data should be always acquired, real-time priority data is given, and the priority scheduling sequence in the bucket is updated; fairness, which calculates priority data according to the order of content delivery (index determination order).
By the mode, the background can effectively identify problem contents (abnormal pushing information) which have a large influence on the online in the submission contents, and the contents which have a large influence on the online are processed in a limited manner, so that the contents can be put off shelf in time, the influence of the contents on the user experience on the online is reduced, and the user experience is optimized.
Continuing with the exemplary structure of the information scheduling apparatus 255 provided in the embodiments of the present application as software modules, in some embodiments, as shown in fig. 2, the software modules stored in the information scheduling apparatus 255 of the memory 250 may include:
the data extraction module 2551 is configured to, when push information pushed to a user is received, extract information to be checked from the push information, and extract real-time feedback data corresponding to the information to be checked from real-time behavior data corresponding to the push information;
an index determining module 2552, configured to determine, based on the real-time feedback data, a scheduling index corresponding to the to-be-audited information; the scheduling index represents the abnormal influence degree of the information to be audited;
a sequence determining module 2553, configured to determine a real-time scheduling sequence of the information to be audited by using a scheduling index corresponding to the information to be audited;
and the audit scheduling module 2554 is configured to perform audit scheduling on the information to be audited according to the real-time scheduling sequence.
In some embodiments of the present application, the index determining module 2552 is further configured to determine an index weight corresponding to the to-be-audited information according to the real-time feedback data and the scheduling index calculation rule; the index weight represents the abnormal degree of the information to be audited, and the index calculation rule is a temporary index calculation rule with the best scheduling effect selected from a plurality of temporary index calculation rules; and calculating the scheduling index corresponding to the information to be audited by adopting the index weight corresponding to the information to be audited.
In some embodiments of the present application, the real-time feedback data includes an information rejection amount, an information distribution amount, and a click rate;
the index determining module 2552 is further configured to compare the information rejection amount with the information release amount, and obtain an information source rejection rate corresponding to the information to be audited; the information rejection quantity represents the quantity of information of which the information source corresponding to the information to be audited is off-shelf, and the information release quantity represents the total quantity of the information released by the information source; according to the index calculation rule, carrying out logarithm operation on the rejection rate of the information source and the click rate to obtain the index weight corresponding to the information to be audited; the click rate characterizes the number of times the information to be audited is displayed.
In some embodiments of the present application, the real-time feedback data includes an information rejection amount, an information distribution amount, a click rate, and a click rate acceleration;
the index determining module 2552 is further configured to compare the information rejection amount with the information release amount, and obtain an information source rejection rate corresponding to the information to be audited; carrying out weight adjustment on the information source rejection rate to obtain the adjusted information source rejection rate; according to the scheduling index calculation rule, carrying out logarithm operation on the adjusted information source rejection rate, the click rate and the click rate increase to obtain the index weight corresponding to the information to be audited; the click rate acceleration represents the increase speed of the click rate of the information to be audited.
In some embodiments of the present application, the index determining module 2552 is further configured to calculate a first index component by using the click rate and a preset weight construction parameter; calculating a second index component by using the click rate acceleration and the preset weight construction parameter; multiplying the first index component, the second index component and the adjusted information source rejection rate to obtain an intermediate multiplication result corresponding to the information to be audited; and carrying out logarithmic operation on the intermediate multiplication result to obtain the index weight corresponding to the information to be audited.
In some embodiments of the present application, the index determining module 2552 is further configured to perform weight reduction on the index weight corresponding to the information to be audited, so as to obtain a reduced index weight; and calculating the scheduling index corresponding to the information to be audited by using the reduced index weight.
In some embodiments of the present application, the data extraction module 2551 is further configured to select the real-time feedback data from the real-time behavior data according to a comparison result between the real-time behavior data and a real-time data threshold; the real-time behavior data is returned by the user aiming at the push information, and the real-time behavior data comprises at least one of real-time click quantity and real-time comment quantity; and screening out corresponding information to be audited from the push information by utilizing the real-time feedback data.
In some embodiments of the application, the data extraction module 2551 is further configured to compare the acquired historical behavior data corresponding to the push information with a historical data threshold, so as to obtain a historical data comparison result; the historical behavior data comprises at least one of historical click rate, historical negative feedback rate, historical click acceleration rate and historical browsing duration; screening information to be audited from the push information according to the historical data comparison result; and screening out the real-time feedback data corresponding to the information to be audited from the real-time behavior data.
In some embodiments of the present application, the order determining module 2553 is further configured to determine an index determining order of the information to be checked according to the obtained issuing time of the information to be checked; the index determination sequence indicates the sequence of the scheduling indexes corresponding to the information to be audited when determining;
correspondingly, the index determining module 2552 is further configured to determine the scheduling index corresponding to the information to be audited by using the real-time feedback data according to the index determining sequence.
In some embodiments of the present application, the information scheduling apparatus 255 further includes: a rule selection module 2555;
in some embodiments of the present application, the data extraction module 2551 is further configured to obtain at least one test audit information and at least one test feedback data corresponding to the at least one test audit information, and divide the at least one test audit information into at least two test information sets;
the index determining module 2552 is further configured to calculate a test index weight corresponding to the test audit information in each test information set according to the test feedback data corresponding to the test audit information in each test information set and the temporary index calculation rule corresponding to each test information set; constructing a test scheduling index corresponding to the test auditing information in each test information set by using the test index weight;
the audit scheduling module 2554 is further configured to perform audit scheduling on the test audit information in each test information set according to the sequence of the test scheduling indexes;
the rule selecting module 2555 is further configured to collect scheduling effect data of the test audit information in each test information set; the scheduling effect data describes the auditing and scheduling conditions of the test auditing information in each test information set; and selecting the scheduling index calculation rule with the best effect from the temporary index calculation rules corresponding to each test information set according to the scheduling effect data.
In some embodiments of the present application, the rule selecting module 2555 is further configured to determine a scheduling effect corresponding to each test information set according to the scheduling effect data; and taking the temporary index calculation rule corresponding to the test information set with the best scheduling effect as the scheduling index calculation rule.
In some embodiments of the present application, the order determining module 2553 is further configured to update a real-time scheduling order of the information to be audited based on the latest feedback data corresponding to the acquired information to be audited, so as to obtain an updated scheduling order;
correspondingly, the audit scheduling module 2554 is further configured to perform audit scheduling on the information to be audited according to the updated scheduling sequence.
Embodiments of the present application provide a computer program product or computer program comprising computer instructions stored in a computer readable storage medium. The processor of the computer device reads the computer instructions from the computer-readable storage medium, and the processor executes the computer instructions, so that the computer device executes the information scheduling method described in the embodiment of the present application.
Embodiments of the present application provide a computer-readable storage medium storing executable instructions, where the executable instructions are stored, and when executed by a processor, the executable instructions will cause the processor to execute an information scheduling method provided by embodiments of the present application, for example, the method shown in fig. 3, fig. 4, fig. 5, and fig. 8.
In some embodiments, the computer-readable storage medium may be memory such as FRAM, ROM, PROM, EP ROM, EEPROM, flash memory, magnetic surface memory, optical disk, or CD-ROM; or may be various devices including one or any combination of the above memories.
In some embodiments, the executable information scheduling instructions may be in the form of a program, software module, script, or code, written in any form of programming language, including compiled or interpreted languages, or declarative or procedural languages, and may be deployed in any form, including as a stand-alone program or as a module, component, subroutine, or other unit suitable for use in a computing environment.
By way of example, the executable information scheduling instructions may, but need not, correspond to files in a file system, may be stored in a portion of a file that holds other programs or data, such as in one or more scripts in a hypertext Markup Language (HTML) document, in a single file dedicated to the program in question, or in multiple coordinated files (e.g., files that store one or more modules, sub-programs, or portions of code).
By way of example, the executable information scheduling instructions may be deployed to be executed on one computing device or on multiple computing devices located at one site or distributed across multiple sites and interconnected by a communication network.
The above description is only an example of the present application, and is not intended to limit the scope of the present application. Any modification, equivalent replacement, and improvement made within the spirit and scope of the present application are included in the protection scope of the present application.

Claims (15)

1. An information scheduling method, comprising:
when push information pushed to a user is received, extracting information to be audited from the push information, and extracting real-time feedback data corresponding to the information to be audited from real-time behavior data corresponding to the push information;
determining a scheduling index corresponding to the information to be audited based on the real-time feedback data; the scheduling index represents the abnormal influence degree of the information to be audited;
determining a real-time scheduling sequence of the information to be audited by using a scheduling index corresponding to the information to be audited;
and according to the real-time scheduling sequence, performing audit scheduling on the information to be audited.
2. The method according to claim 1, wherein the determining a scheduling index corresponding to the to-be-audited information based on the real-time feedback data includes:
determining index weight corresponding to the information to be audited according to the real-time feedback data and a scheduling index calculation rule; the index weight represents the abnormal degree of the information to be audited, and the index calculation rule is a temporary index calculation rule with the best scheduling effect selected from a plurality of temporary index calculation rules;
and calculating the scheduling index corresponding to the information to be audited by adopting the index weight corresponding to the information to be audited.
3. The method of claim 2, wherein the real-time feedback data includes an amount of information rejection, an amount of information release, and an amount of clicks;
the determining the index weight corresponding to the information to be audited according to the real-time feedback data and the scheduling index calculation rule comprises the following steps:
comparing the information rejection quantity with the information release quantity to obtain an information source rejection rate corresponding to the information to be audited;
the information rejection quantity represents the quantity of information of which the information source corresponding to the information to be audited is off-shelf, and the information release quantity represents the total quantity of the information released by the information source;
according to the index calculation rule, carrying out logarithm operation on the rejection rate of the information source and the click rate to obtain the index weight corresponding to the information to be audited; the click rate indicates the number of times the information to be audited is displayed.
4. The method of claim 2, wherein the real-time feedback data includes an amount of information rejection, an amount of information distribution, an amount of clicks, and a click rate increase;
the determining the index weight corresponding to the information to be audited according to the real-time feedback data and the scheduling index calculation rule comprises the following steps:
comparing the information rejection quantity with the information release quantity to obtain an information source rejection rate corresponding to the information to be audited;
carrying out weight adjustment on the information source rejection rate to obtain the adjusted information source rejection rate;
according to the scheduling index calculation rule, carrying out logarithm operation on the adjusted information source rejection rate, the click rate and the click rate increase to obtain the index weight corresponding to the information to be audited; the click rate acceleration represents the increase speed of the click rate of the information to be audited.
5. The method according to claim 4, wherein the obtaining the index weight corresponding to the information to be audited by performing a logarithmic operation on the adjusted information source rejection rate, the click rate, and the click rate increase includes:
constructing parameters by using the click rate and preset weight, and calculating a first index component;
calculating a second index component by using the click rate acceleration and the preset weight construction parameter;
multiplying the first index component, the second index component and the adjusted information source rejection rate to obtain an intermediate multiplication result corresponding to the information to be audited;
and carrying out logarithmic operation on the intermediate multiplication result to obtain the index weight corresponding to the information to be audited.
6. The method according to claim 2, wherein the calculating the scheduling index corresponding to the information to be checked by using the index weight corresponding to the information to be checked includes:
performing weight reduction on the index weight corresponding to the information to be audited to obtain the weight of the index after the weight reduction;
and calculating the scheduling index corresponding to the information to be audited by using the reduced index weight.
7. The method according to any one of claims 1 to 6, wherein the extracting information to be checked from the push information and extracting real-time feedback data corresponding to the information to be checked from real-time behavior data corresponding to the push information includes:
selecting the real-time feedback data from the real-time behavior data according to the comparison result of the real-time behavior data and the real-time data threshold;
the real-time behavior data is returned by the user aiming at the push information, and the real-time behavior data comprises at least one of real-time click quantity and real-time comment quantity;
and screening corresponding information to be audited from the push information by utilizing the real-time feedback data.
8. The method according to any one of claims 1 to 6, wherein the extracting information to be checked from the push information and extracting real-time feedback data corresponding to the information to be checked from real-time behavior data corresponding to the push information includes:
comparing historical behavior data corresponding to the acquired push information with a historical data threshold value to obtain a historical data comparison result; the historical behavior data comprises at least one of historical click rate, historical negative feedback rate, historical click acceleration rate and historical browsing duration;
screening information to be audited from the push information according to the historical data comparison result;
and screening out the real-time feedback data corresponding to the information to be audited from the real-time behavior data.
9. The method according to any one of claims 1 to 6, wherein after the information to be audited is extracted from the push information and the real-time feedback data corresponding to the information to be audited is extracted from the real-time behavior data corresponding to the push information, and before the scheduling index corresponding to the information to be audited is determined based on the real-time feedback data, the method further comprises:
determining an index determination sequence of the information to be audited according to the obtained issuing time of the information to be audited; the index determination sequence indicates the sequence of constructing the scheduling indexes corresponding to the information to be audited;
correspondingly, the determining the scheduling index corresponding to the to-be-audited information based on the real-time feedback data includes:
and according to the index determination sequence, determining the scheduling index corresponding to the information to be audited by using the real-time feedback data.
10. The method according to claim 2, wherein before determining the index weight corresponding to the information to be audited according to the real-time feedback data and the scheduling index calculation rule, the method further comprises:
acquiring at least one piece of test auditing information and at least one piece of test feedback data corresponding to the at least one piece of test auditing information, and dividing the at least one piece of test auditing information into at least two test information sets;
calculating the test index weight corresponding to the test audit information in each test information set according to the test feedback data corresponding to the test audit information in each test information set and the temporary index calculation rule corresponding to each test information set;
constructing a test scheduling index corresponding to the test auditing information in each test information set by using the test index weight, and performing auditing scheduling on the test auditing information in each test information set according to the sequence of the test scheduling index;
acquiring scheduling effect data of the test auditing information in each test information set; the scheduling effect data describes the auditing and scheduling conditions of the test auditing information in each test information set;
and selecting the scheduling index calculation rule with the best scheduling effect from the temporary index calculation rules corresponding to each test information set according to the scheduling effect data.
11. The method according to claim 10, wherein the selecting the scheduling index calculation rule with the best scheduling effect from the temporary index calculation rules corresponding to each test information set according to the scheduling effect data includes:
judging the scheduling effect corresponding to each test information set according to the scheduling effect data;
and taking the temporary index calculation rule corresponding to the test information set with the best scheduling effect as the scheduling index calculation rule.
12. The method according to any one of claims 1 to 6, 10, or 11, wherein after the determining of the real-time scheduling order of the information to be audited by using the scheduling index corresponding to the information to be audited, and before performing audit scheduling on the information to be audited according to the real-time scheduling order, the method further comprises:
updating the real-time scheduling sequence of the information to be audited based on the latest feedback data corresponding to the acquired information to be audited to obtain an updated scheduling sequence;
correspondingly, the performing audit scheduling on the information to be audited according to the real-time scheduling sequence includes:
and according to the updated scheduling sequence, performing audit scheduling on the information to be audited.
13. An information scheduling apparatus, comprising:
the data extraction module is used for extracting information to be audited from the push information when the push information pushed to a user is received, and extracting real-time feedback data corresponding to the information to be audited from the real-time behavior data corresponding to the push information;
the index determining module is used for determining a scheduling index corresponding to the information to be audited based on the real-time feedback data; the scheduling index represents the abnormal influence degree of the information to be audited;
the order determining module is used for determining a real-time scheduling order of the information to be audited by using the scheduling index corresponding to the information to be audited;
and the auditing scheduling module is used for auditing and scheduling the information to be audited according to the real-time scheduling sequence.
14. An information scheduling apparatus, comprising:
a memory for storing executable information scheduling instructions;
a processor configured to implement the method of any one of claims 1 to 12 when executing the executable information scheduling instructions stored in the memory.
15. A computer-readable storage medium having stored thereon executable information scheduling instructions for, when executed by a processor, implementing the method of any one of claims 1 to 12.
CN202011216286.9A 2020-11-04 2020-11-04 Information scheduling method, device and equipment and computer readable storage medium Pending CN114443943A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202011216286.9A CN114443943A (en) 2020-11-04 2020-11-04 Information scheduling method, device and equipment and computer readable storage medium

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202011216286.9A CN114443943A (en) 2020-11-04 2020-11-04 Information scheduling method, device and equipment and computer readable storage medium

Publications (1)

Publication Number Publication Date
CN114443943A true CN114443943A (en) 2022-05-06

Family

ID=81360733

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202011216286.9A Pending CN114443943A (en) 2020-11-04 2020-11-04 Information scheduling method, device and equipment and computer readable storage medium

Country Status (1)

Country Link
CN (1) CN114443943A (en)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114996165A (en) * 2022-08-01 2022-09-02 飞狐信息技术（天津）有限公司 Business data auditing method and device, storage medium and electronic equipment
CN117271100A (en) * 2023-11-21 2023-12-22 北京国科天迅科技股份有限公司 Algorithm chip cluster scheduling method, device, computer equipment and storage medium

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114996165A (en) * 2022-08-01 2022-09-02 飞狐信息技术（天津）有限公司 Business data auditing method and device, storage medium and electronic equipment
CN114996165B (en) * 2022-08-01 2022-12-13 飞狐信息技术（天津）有限公司 Business data auditing method and device, storage medium and electronic equipment
CN117271100A (en) * 2023-11-21 2023-12-22 北京国科天迅科技股份有限公司 Algorithm chip cluster scheduling method, device, computer equipment and storage medium
CN117271100B (en) * 2023-11-21 2024-02-06 北京国科天迅科技股份有限公司 Algorithm chip cluster scheduling method, device, computer equipment and storage medium

Similar Documents

Publication Publication Date Title
US11128582B2 (en) Emoji recommendation method and apparatus
TWI491217B (en) Method to increase content relevance using insights obtained from user activity updates
US8793312B2 (en) Bridging social silos for knowledge discovery and sharing
US9331971B2 (en) Message subscription based on message aggregate characteristics
AU2016261088A1 (en) Social media events detection and verification
CN112153426B (en) Content account management method and device, computer equipment and storage medium
CN108288179A (en) A kind of user preference source of houses computational methods and system
US20200334697A1 (en) Generating survey responses from unsolicited messages
CN108470057B (en) Generating and pushing method, device, terminal, server and medium of integrated information
CN114443943A (en) Information scheduling method, device and equipment and computer readable storage medium
CN111435369B (en) Music recommendation method, device, terminal and storage medium
CN111078931A (en) Singing sheet pushing method and device, computer equipment and storage medium
CN105991722B (en) Downloader recommendation method, application server, terminal and system
US9582586B2 (en) Massive rule-based classification engine
CN110245684B (en) Data processing method, electronic device, and medium
CN103906066A (en) Disturbance shielding method and apparatus for user generation content referring
CN106293650A (en) A kind of folder attribute method to set up and device
CN116932906A (en) Search term pushing method, device, equipment and storage medium
KR101568800B1 (en) Real-time issue search word sorting method and system
CN116089490A (en) Data analysis method, device, terminal and storage medium
CN113343133B (en) Display page generation method, related device and computer program product
CN114143571B (en) User processing method, device, equipment and storage medium
CN116028723A (en) Data recommendation method, device, equipment and computer storage medium
KR101862178B1 (en) Method for customized posting and server implementing the same
CN117240819B (en) Mail configuration method, device, equipment and computer readable storage medium

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