Rumor recognition method and system

Abstract

The embodiment of the invention discloses a rumor recognition method and system, and relates to the technical field of computers. According to the rumor recognition method, when a user browses a target webpage, a browser plug-in captures content information of the target webpage, the content information is input into a rumor recognition engine, the rumor recognition engine carries out rumor recognition on the content information based on a rumor recognition model, and then the rumor recognition engine carries out rumor recognition on the content information based on the rumor recognition model. And the browser plug-in performs rumor prompting on the target webpage based on a rumor recognition result returned by the rumor recognition engine. According to the rumor recognition method and system provided by the embodiment of the invention, when the user browses the target webpage, the rumor can be recognized and prompted in real time, so that the user can know the rumor information in time, and the user is prevented from being misguided by the rumor.

Description

Technical Field

[0001] The present application relates to the field of computer technology, and in particular to a rumor identification method and system. Background Art

[0002] With the rapid development of Internet information, the speed and scope of rumor propagation have greatly increased, bringing many negative impacts to society. Traditional rumor identification mainly relies on manual review and expert judgment, which is inefficient and difficult to cope with the massive flow of information on the Internet. Users cannot know whether online information is a rumor in a timely manner and are easily misled by rumors. Summary of the invention

[0003] The embodiments of the present application provide a rumor identification method and system for enabling users to promptly understand whether network information is a rumor, thereby preventing users from being misled by the rumor.

[0004] In a first aspect, an embodiment of the present application provides a rumor identification method, which is applied to a browser plug-in, and the method includes: When a user browses a target webpage, crawling content information of the target webpage; Inputting the content information into a rumor identification engine, so that the rumor identification engine performs rumor identification on the content information based on a rumor identification model and returns a rumor identification result; Based on the rumor identification result, a rumor prompt is provided to the target webpage.

[0005] In a second aspect, an embodiment of the present application provides a rumor identification method, which is applied to a rumor identification engine, and the method includes: Receiving content information of a target webpage input by a browser plug-in; the content information of the target webpage is captured by the browser plug-in when a user browses the target webpage; Performing rumor identification on the content information based on a rumor identification model; The rumor identification result is returned to the browser plug-in so that the browser plug-in can provide rumor prompts to the target web page.

[0006] In a third aspect, an embodiment of the present application provides a rumor identification method, which is applied to a rumor identification system, wherein the rumor identification system includes: a browser plug-in and a rumor identification engine, and the method includes: When a user browses a target webpage, the browser plug-in captures content information of the target webpage and inputs the content information into a rumor identification engine; The rumor identification engine performs rumor identification on the content information based on a rumor identification model, and returns the rumor identification result to the browser plug-in; The browser plug-in issues a rumor prompt to the target webpage based on the rumor identification result.

[0007] In a fourth aspect, an embodiment of the present application provides a rumor identification system, the rumor identification system comprising: a browser plug-in, a rumor identification engine; The browser plug-in is used to capture content information of the target webpage when the user browses the target webpage, and input the content information into the rumor identification engine; The rumor identification engine is used to identify rumors on the content information based on a rumor identification model, and return the rumor identification result to the browser plug-in; The browser plug-in is also used to provide rumor prompts to the target web page based on the rumor identification result.

[0008] The rumor identification method and system provided in the embodiments of the present application, when a user browses the target webpage, a browser plug-in captures the content information of the target webpage, inputs the content information into a rumor identification engine, and the rumor identification engine performs rumor identification on the content information. Then, the browser plug-in issues a rumor prompt to the target webpage based on the rumor identification result returned by the rumor identification engine. When a user browses the target webpage, the browser plug-in can perform real-time identification and prompt of rumors, so that the user can learn about the rumor information in a timely manner and avoid being misled by the rumors. BRIEF DESCRIPTION OF THE DRAWINGS

[0009] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0010] Figure 1 is a structural diagram of a rumor identification system according to an embodiment of the present application; Figure 2 yes Figure 1 Schematic diagram of the interaction process in the rumor identification system shown; Figure 3 yes Figure 1 A schematic diagram of the rumor identification process in the rumor identification engine shown; Figure 4 is a structural diagram of a rumor identification system according to another embodiment of the present application; Figure 5 yes Figure 4 Schematic diagram of the interaction process in the rumor identification system shown; Figure 6 yes Figure 4A schematic diagram of the model training process in the model training engine shown; Figure 7 is a structural diagram of a rumor identification system according to another embodiment of the present application; Figure 8 yes Figure 7 Schematic diagram of the interaction process in the rumor identification system shown; Fig. 9 It is a flowchart of a rumor identification method applied to a browser plug-in according to an embodiment of the present application; Fig.10 It is a flowchart of a rumor identification method applied to a rumor identification engine according to an embodiment of the present application. DETAILED DESCRIPTION

[0011] Embodiments of the present application provide a rumor identification system and a rumor identification method.

[0012] In order to enable those skilled in the art to better understand the technical solutions in this application, the technical solutions in the embodiments of this application will be clearly and completely described below in conjunction with the drawings in the embodiments of this application. Obviously, the described embodiments are only part of the embodiments of this application, not all of them. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of this application.

[0013] Figure 1 It is a structural diagram of a rumor identification system according to an embodiment of the present application. Figure 2 yes Figure 1 The interactive process diagram of the rumor identification system shown in FIG. Figure 1 , Figure 2 , the rumor identification system and rumor identification method provided in the embodiments of the present application are described in detail.

[0014] like Figure 1 , Figure 2 As shown, the rumor identification system 100 according to an embodiment of the present application includes: a browser plug-in 110 and a rumor identification engine 120, and the rumor identification method applied to the rumor identification system 100 includes the following steps: S110, when a user browses a target web page, the browser plug-in 110 captures content information of the target web page and inputs the content information into the rumor identification engine 120.

[0015] The browser plug-in may be a plug-in of a Web browser, which may be installed in the Web browser. When a user browses a target web page through the Web browser, the browser plug-in may capture content information of the target web page in real time.

[0016] The target webpage is a webpage to be subjected to rumor identification, including but not limited to: a news webpage, an article webpage, etc. The content information of the target webpage may be text content information in the target webpage, including but not limited to: news content text and news comment text in a news webpage, article content text and article comment text in an article webpage, etc.

[0017] S120 , the rumor identification engine 120 performs rumor identification on the content information based on a rumor identification model, and returns the rumor identification result to the browser plug-in 110 .

[0018] In one implementation, the rumor identification result may include: whether there is rumor information in the content information.

[0019] In another implementation, when rumor information exists in the content information, the rumor identification result may further include: rumor refuting information corresponding to the rumor information. The rumor refuting information may include: rumor refuting source information, rumor refuting clue information, etc.

[0020] S130, the browser plug-in 110 issues a rumor prompt to the target web page based on the rumor identification result.

[0021] In one implementation, when rumor information exists in the content information, the rumor information can be marked and displayed on the target webpage with a rumor identifier, where the rumor identifier is used to indicate that the content information in the target webpage is a rumor.

[0022] In another implementation, on the basis of marking and displaying the rumor identifier, the rumor information can be further marked and displayed with rumor-refuting source information, wherein the rumor-refuting source information is used to indicate which rumor-refuting platform determined the rumor information. By further marking and displaying the rumor-refuting source information, it is easier for users to understand the confirmed subject of the rumor, and the credibility of rumor identification is improved.

[0023] In this embodiment, when a user browses a target webpage, the browser plug-in captures the content information of the target webpage, inputs the content information into a rumor recognition engine, and the rumor recognition engine performs rumor recognition on the content information based on a rumor recognition model. Then, the browser plug-in performs a rumor prompt on the target webpage based on the rumor recognition result returned by the rumor recognition engine. When the user browses the target webpage, the rumor can be recognized and prompted in real time, so that the user can learn about the rumor information in time and avoid being misled by the rumor. This method recognizes and prompts rumors from the source of information dissemination, which is convenient for coping with the massive network data flow of the Internet.

[0024] In one implementation of the above embodiment, in the above step S120, the rumor identification engine 120 may input the content information into a rumor identification model so that the rumor identification model performs rumor identification on the content information.

[0025] In another implementation of the above embodiment, in order to better identify rumors on content information and obtain more accurate rumor identification results, in the above step S120, the rumor identification engine 120 can perform semantic analysis on the content information, and input the semantic analysis results into the rumor identification model, so that the rumor identification model can identify rumors on the content information based on the semantic analysis results of the content information. Figure 3 yes Figure 1 The rumor identification process diagram in the rumor identification engine is shown in FIG. Figure 3 As shown, in this implementation, in the above step S120, rumor identification is performed on the content information based on the rumor identification model, specifically including the following steps: S122: Perform semantic analysis on the content information to obtain semantic summary information of the content information.

[0026] The semantic summary information of the content information is, for example, the semantic key summary information of the content information, which is obtained by interpreting and summarizing the content information by taking into full account the context information of the content information through semantic analysis technology. This summary information has more depth and breadth, and can help the rumor recognition model better understand the content information and output more accurate rumor recognition results.

[0027] In a specific implementation, the content information may be input into a NLP (Natural Language Processing) engine, so that the NLP engine performs semantic analysis on the content information and outputs semantic summary information of the content information.

[0028] S124, inputting the semantic summary information into a rumor recognition model, causing the rumor recognition model to perform rumor recognition based on the semantic summary information, and outputting a rumor recognition result.

[0029] By performing semantic analysis on the content information to obtain semantic summary information of the content information, and inputting the semantic summary information into a rumor recognition model, so that the rumor recognition model performs rumor recognition based on the semantic summary information, the accuracy of rumor recognition can be improved and a more accurate rumor recognition result can be obtained.

[0030] In the above embodiment, the rumor recognition model is a deep learning model, which can be trained based on a rumor recognition data set. The rumor recognition data set may include multiple rumor recognition data, each of which may include: sample content information and rumor refuting information corresponding to the sample content information, and the rumor refuting information may include: whether there is rumor information in the sample content information, and if there is rumor information in the sample content information, the rumor refuting information corresponding to the rumor information, such as rumor refuting source channels, clues, rumor situation information, etc. The sample content information and the corresponding rumor refuting information can be captured from public news information platforms, social platforms, knowledge sharing platforms, entertainment and leisure platforms, and rumor refuting platforms and other Internet platforms.

[0031] It is understandable that rumor information will change over time. Further, to improve the accuracy of rumor identification, the rumor identification model in the above embodiment can be dynamically updated, and the rumor identification engine 120 can always perform rumor identification based on the latest rumor identification model. Figure 4 It is a structural diagram of a rumor identification system according to another embodiment of the present application. Figure 5 yes Figure 4 The interactive process diagram of the rumor identification system shown in FIG. Figure 4 , Figure 5 , the rumor identification system and rumor identification method provided in the embodiments of the present application are further explained.

[0032] like Figure 4 , Figure 5 As shown, a rumor identification system 100 according to another embodiment of the present application is substantially the same as the rumor identification system 100 according to the above embodiment, except that the rumor identification system 100 according to this embodiment further includes: a model training engine 130, and a rumor identification method applied to the rumor identification system 100 further includes the following steps: S140, the model training engine 130 captures new rumor identification data in real time, and dynamically trains the rumor identification model based on the new rumor identification data, and dynamically updates the rumor identification model.

[0033] The step S140 and the steps S110 to S130 are independent of each other and can be executed in parallel.

[0034] In this embodiment, by capturing newly added rumor identification data in real time, dynamically training the rumor identification model based on the newly added rumor identification data, and dynamically updating the rumor identification model, the rumor identification engine 120 can always identify rumors on content information based on the latest rumor identification model that has learned the newly added rumor knowledge, thereby improving the accuracy of rumor identification.

[0035] Figure 6 yes Figure 4 The model training process diagram in the model training engine is shown in Figure 1. Figure 6 As shown, specifically, in one implementation of the above embodiment, in the above step S140, the real-time capture of updated rumor identification data, dynamic training of the rumor identification model based on the updated rumor identification data, and dynamic updating of the rumor identification model include the following steps: S142, capturing new rumor-refuting information and new content information corresponding to the new rumor-refuting information from the rumor-refuting platform in real time to obtain new rumor identification data.

[0036] S144, saving the newly added rumor identification data to the full rumor identification data set.

[0037] S146, at first intervals, an incremental rumor identification data set is obtained based on the top N rumor identification data with the highest fermentation trend scores in the full rumor identification data set, and the rumor identification model is incrementally updated and trained based on the incremental rumor identification data set to update the rumor identification model.

[0038] Among them, the fermentation trend score is used to comprehensively evaluate the fermentation status of the new content information in the new rumor identification data. The higher the fermentation trend score, the more serious the fermentation status.

[0039] In one implementation, the fermentation trend score corresponding to the rumor identification data may be calculated based on the tag information of the target tag corresponding to the content information in the rumor identification data and the fermentation trend scoring rule.

[0040] The target tags include but are not limited to: content category, release time, number of visits, etc.

[0041] For example, the fermentation trend scoring rule may be: fermentation trend score = content category score + release time score + number of visits score, the content category score = the score corresponding to the content category, the release time score = the time difference between the current time and the release time * the first weight, the number of visits score = the number of visits * the second weight. Based on this rule and the content category, release time, and number of visits corresponding to the content information in the rumor identification data, the fermentation trend score corresponding to the rumor identification data can be calculated.

[0042] In a specific implementation, before step S144, feature extraction can be performed on the newly added content information to obtain feature information such as content category, release time, and number of visits corresponding to the newly added content information, and label information of target tags such as content category, release time, and number of visits can be annotated for the newly added content information based on these feature information. In step S144, the newly added rumor identification data and the label information of the target tag corresponding to the newly added content information in the newly added rumor identification data can be saved together in the full rumor identification data set. In step S146, the fermentation trend score corresponding to the newly added information can be calculated based on the label information of the target tag such as content category, release time, and number of visits of the newly added content information, and the rumor identification data in the full rumor identification data set can be sorted based on the fermentation trend score corresponding to the newly added content information and the fermentation trend score corresponding to the original content information of the original rumor identification data in the full rumor identification data set, and the set of rumor identification data ranked in the top N is determined as the incremental rumor identification data set.

[0043] Wherein, N is a positive integer, which may be an empirical value. For example, N may be 1000, and a set of 1000 rumor identification data with the highest fermentation trend score in the full rumor identification data set may be determined as the incremental rumor identification data set.

[0044] S148, performing full update training on the rumor recognition model based on the full rumor recognition data set at second intervals to update the rumor recognition model.

[0045] The first time is shorter than the second time. For example, the first time is 4 hours, and the second time is 3 months.

[0046] By capturing new rumor identification data in real time, the rumor identification model is incrementally updated and trained based on the incremental rumor identification data set at a first interval, and the rumor identification model is fully updated and trained based on the full rumor identification data set at a second interval. This can take into account both the update efficiency and recognition accuracy of the rumor identification model, thereby improving the update efficiency of the rumor identification model while improving the recognition accuracy of the rumor identification model.

[0047] In the above embodiment, the browser plug-in performs real-time rumor identification and marking on the target web page when the user browses the target web page, and can provide the user with a real-time rumor refuting function when the user browses the web page.

[0048] Furthermore, based on the above embodiments, the rumor identification system and rumor identification method provided in the embodiments of the present application can also provide users with a rumor identification question and answer function. Figure 7It is a structural diagram of a rumor identification system according to another embodiment of the present application. Figure 8 yes Figure 7 The interactive process diagram in the rumor identification system is shown in FIG. Figure 7 , Figure 8 As shown, the rumor identification system and rumor identification method according to another embodiment of the present application are basically the same as the above-mentioned embodiment, except that the rumor identification system 100 according to another embodiment of the present application further includes: a rumor identification dialogue component 140, and the rumor identification method applied to the rumor identification system 100 further includes the following steps: S150 , the rumor identification dialogue component 140 receives a rumor identification question input by a user, and inputs content information of the rumor identification question into the rumor identification engine 120 .

[0049] S160 , the rumor identification engine 120 performs rumor identification on the content information based on a rumor identification model, and returns the rumor identification result to the rumor identification dialogue component 140 .

[0050] Among them, the rumor identification engine 120 can perform semantic analysis on the content information, determine the content information to be rumor identified, input the content information to be rumor identified into a rumor identification model, and the rumor identification model performs rumor identification on the content information to be rumor identified, and outputs the rumor identification result.

[0051] S170, the rumor identification dialogue component 140 outputs answer information of the rumor identification question based on the rumor identification result.

[0052] In this embodiment, by further providing a rumor identification dialogue component, the rumor identification dialogue component receives rumor identification questions input by users, inputs the content information of the rumor identification questions into the rumor identification engine for rumor identification, and outputs the answer information of the rumor identification questions based on the rumor identification results output by the rumor identification engine. This can further expand the rumor identification scenarios and prevent users from being misled by rumors.

[0053] Furthermore, the rumor identification dialogue component 140 can receive the user's evaluation information on the answer information, and feed the evaluation information back to the model training engine 130, so that the model training engine 130 processes the rumor identification data set based on the evaluation information to further update the rumor identification model.

[0054] The above is the rumor identification system provided in the embodiments of the present application and the rumor identification method applied to the rumor identification system.

[0055] Based on the same inventive concept, an embodiment of the present application also provides a rumor identification method, which applies the browser plug-in in the above embodiment. Fig. 9 FIG. 1 is a flow chart of a rumor identification method applied to a browser plug-in according to an embodiment of the present application. Fig. 9 As shown, the rumor identification method applied to the browser plug-in includes the following steps: S210, when the user browses the target webpage, crawling the content information of the target webpage; S220: inputting the content information into a rumor recognition engine, so that the rumor recognition engine performs rumor recognition on the content information based on a rumor recognition model and returns a rumor recognition result; S230: Based on the rumor identification result, a rumor prompt is provided to the target webpage.

[0056] In one implementation, the target web page includes: a news web page and an article web page.

[0057] In one implementation, the rumor identification result includes: whether there is rumor information in the content information; The step of providing a rumor prompt to the target webpage based on the rumor identification result includes: When rumor information exists in the content information, the rumor information is marked and a rumor logo is displayed on the target webpage.

[0058] The rumor identification method applied to the browser plug-in provided in this embodiment is the same as the method performed by the browser plug-in 110 in the aforementioned embodiment. The specific processing can be referred to the aforementioned embodiment, which will not be described in detail.

[0059] Based on the same inventive concept, an embodiment of the present application also provides a rumor identification method, which is applied to the rumor identification engine in the above embodiment. Fig.10 It is a flowchart of a rumor identification method applied to a rumor identification engine according to an embodiment of the present application. Fig.10 As shown, the rumor identification method applied to the rumor identification engine includes the following steps: S310, receiving content information of a target webpage input by a browser plug-in; the content information of the target webpage is captured by the browser plug-in when a user browses the target webpage; S320, performing rumor identification on the content information based on a rumor identification model; S330, returning the rumor identification result to the browser plug-in, so that the browser plug-in can provide a rumor prompt to the target web page.

[0060] In one implementation, the step of performing rumor identification on the content information based on a rumor identification model includes: Performing semantic analysis on the content information to obtain semantic summary information of the content information; The semantic summary information is input into a rumor recognition model, so that the rumor recognition model performs rumor recognition based on the semantic summary information and outputs a rumor recognition result.

[0061] In one implementation, the rumor identification model is captured by a model training engine in real time with newly added rumor identification data, and the rumor identification model is dynamically trained based on the newly added rumor identification data to be dynamically updated.

[0062] In one implementation, the real-time capture of newly added rumor identification data and the dynamic training of the rumor identification model based on the newly added rumor identification data to dynamically update the rumor identification model include: Capture new rumor-refuting information and new content information corresponding to the new rumor-refuting information from the rumor-refuting platform in real time to obtain new rumor identification data; Saving the newly added rumor identification data to the full rumor identification data set; At first intervals, an incremental rumor identification data set is obtained based on the top N rumor identification data with the highest fermentation trend score in the full rumor identification data set, and the rumor identification model is incrementally updated and trained based on the incremental rumor identification data set to update the rumor identification model; wherein N is a positive integer; At second intervals, based on the full rumor identification data set, the rumor identification model is fully updated and trained to update the rumor identification model; The first time is shorter than the second time.

[0063] In one implementation, the method further includes: Receiving content information of a rumor identification question sent by a rumor identification dialogue component, wherein the rumor identification question is input by a user into the rumor identification dialogue component; Performing rumor identification on the content information based on a rumor identification model; The rumor identification result is returned to the rumor identification dialogue component, so that the rumor identification dialogue component outputs answer information of the rumor identification question based on the rumor identification result.

[0064] The rumor identification method applied to the rumor identification engine provided in this embodiment is the same as the method performed by the rumor identification engine 120 in the aforementioned embodiment. The specific processing can be referred to the aforementioned embodiment, which will not be described in detail.

[0065] Based on the same inventive concept, the present application embodiment also provides a rumor identification system, such as Figure 1 As shown, Figure 1 As shown, the rumor identification system includes: a browser plug-in 110 and a rumor identification engine 120.

[0066] The browser plug-in 110 is used to capture the content information of the target webpage when the user browses the target webpage, and input the content information into the rumor identification engine; The rumor identification engine 120 is used to identify rumors on the content information based on a rumor identification model, and return the rumor identification result to the browser plug-in; The browser plug-in 120 is also used to provide rumor prompts to the target web page based on the rumor identification result.

[0067] In one implementation, Figure 4 As shown, the rumor identification system 100 also includes: a model training engine 130; The model training engine 130 is used to capture new rumor identification data in real time, dynamically train the rumor identification model based on the new rumor identification data, and dynamically update the rumor identification model.

[0068] In one implementation, Figure 7 As shown, the rumor identification system 100 further includes: a rumor identification dialogue component 140; The rumor identification dialogue component 140 is used to receive a rumor identification question input by a user, and input content information of the rumor identification question into the rumor identification engine 120 .

[0069] The rumor identification engine 120 is further used to perform rumor identification on the content information based on a rumor identification model, and return the rumor identification result to the rumor identification dialogue component 140 .

[0070] The rumor identification dialogue component 140 is also used to output answer information of the rumor identification question based on the rumor identification result.

[0071] The specific processing in the rumor identification system 100 can be found in the aforementioned embodiment and will not be described in detail here.

[0072] The embodiment of the present application also provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, each process of the rumor identification method embodiment applied to a browser plug-in or the rumor identification method embodiment applied to a rumor identification engine is implemented, and the same technical effect can be achieved. To avoid repetition, it is not repeated here. The computer-readable storage medium is, for example, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk.

[0073] An embodiment of the present application also provides a computer program product. When the computer program in the computer program product is executed by a processor, the various processes of the above-mentioned rumor identification method embodiment applied to a browser plug-in or the rumor identification method embodiment applied to a rumor identification engine are implemented, and the same technical effect can be achieved. To avoid repetition, it will not be repeated here.

[0074] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present application may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program codes.

[0075] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0076] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0077] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process in the computer or other programmable device. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0078] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0079] Memory may include non-permanent storage in a computer-readable medium, in the form of random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM. Memory is an example of a computer-readable medium.

[0080] Computer-readable media include permanent and non-permanent, removable and non-removable media that can be used to store information by any method or technology. Information can be computer-readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disk read-only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, magnetic cassettes, magnetic tape magnetic disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. According to the definition in this article, computer-readable media does not include temporary computer-readable media (transitory media), such as modulated data signals and carrier waves.

[0081] It should also be noted that the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, commodity or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, commodity or device. In the absence of more restrictions, the elements defined by the sentence "comprises a ..." do not exclude the existence of other identical elements in the process, method, commodity or device including the elements.

[0082] The above is only an embodiment of the present application and is not intended to limit the present application. For those skilled in the art, the present application may have various changes and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application should be included in the scope of the claims of the present application.

Claims

1. A rumor identification method, characterized in that: Applied to a browser plug-in, the method comprises: When a user browses a target webpage, crawling content information of the target webpage; Inputting the content information into a rumor identification engine, so that the rumor identification engine performs rumor identification on the content information based on a rumor identification model and returns a rumor identification result; Based on the rumor identification result, a rumor prompt is provided to the target webpage.

2. The method according to claim 1, characterized in that The target web pages include: news web pages and article web pages.

3. The method according to claim 1, characterized in that The rumor identification result includes: whether there is rumor information in the content information; The step of providing a rumor prompt to the target webpage based on the rumor identification result includes: When rumor information exists in the content information, the rumor information is marked and displayed with a rumor identifier on the target webpage.

4. A rumor identification method, characterized in that: Applied to a rumor identification engine, the method includes: Receiving content information of a target webpage input by a browser plug-in; the content information of the target webpage is captured by the browser plug-in when a user browses the target webpage; Performing rumor identification on the content information based on a rumor identification model; The rumor identification result is returned to the browser plug-in so that the browser plug-in can provide rumor prompts to the target web page.

5. The rumor identification method according to claim 4, characterized in that: The performing rumor identification on the content information based on the rumor identification model includes: Performing semantic analysis on the content information to obtain semantic summary information of the content information; The semantic summary information is input into a rumor recognition model, so that the rumor recognition model performs rumor recognition based on the semantic summary information and outputs a rumor recognition result.

6. The rumor identification method according to claim 4, characterized in that: The rumor identification model is captured by a model training engine in real time with newly added rumor identification data, and the rumor identification model is dynamically trained based on the newly added rumor identification data to be dynamically updated.

7. The rumor identification method according to claim 6, characterized in that: The real-time capture of newly added rumor identification data and the dynamic training of the rumor identification model based on the newly added rumor identification data to dynamically update the rumor identification model include: Capture new rumor-refuting information and new content information corresponding to the new rumor-refuting information from the rumor-refuting platform in real time to obtain new rumor identification data; Saving the newly added rumor identification data to the full rumor identification data set; At first intervals, an incremental rumor identification data set is obtained based on the top N rumor identification data with the highest fermentation trend score in the full rumor identification data set, and the rumor identification model is incrementally updated and trained based on the incremental rumor identification data set to update the rumor identification model; wherein N is a positive integer; At second intervals, based on the full rumor identification data set, the rumor identification model is fully updated and trained to update the rumor identification model; The first time is shorter than the second time.

8. The rumor identification method according to claim 7, characterized in that: The method further comprises: Receiving content information of a rumor identification question sent by a rumor identification dialogue component, wherein the rumor identification question is input by a user into the rumor identification dialogue component; Performing rumor identification on the content information based on a rumor identification model; The rumor identification result is returned to the rumor identification dialogue component, so that the rumor identification dialogue component outputs answer information of the rumor identification question based on the rumor identification result.

9. A rumor identification method, characterized in that: Applied to a rumor identification system, the rumor identification system includes: a browser plug-in and a rumor identification engine, the method includes: When a user browses a target webpage, the browser plug-in captures content information of the target webpage and inputs the content information into the rumor identification engine; The rumor identification engine performs rumor identification on the content information based on a rumor identification model, and returns the rumor identification result to the browser plug-in; The browser plug-in issues a rumor prompt to the target webpage based on the rumor identification result.

10. A rumor identification system, characterized in that: The rumor identification system includes: a browser plug-in and a rumor identification engine; The browser plug-in is used to capture content information of the target webpage when the user browses the target webpage, and input the content information into the rumor identification engine; The rumor identification engine is used to identify rumors on the content information based on a rumor identification model, and return the rumor identification result to the browser plug-in; The browser plug-in is also used to provide rumor prompts to the target web page based on the rumor identification result.

Images