Barrage management method and device, server and computer storage medium

Abstract

The application relates to barrage control technology, and provides a barrage control method and device, a server and a computer storage medium, the method comprises the following steps: acquiring original barrage sent by a user in an online live room; performing online risk filtering on the original barrage, and screening online barrage meeting an online barrage control strategy; performing offline risk filtering on the online barrage, and screening offline barrage meeting an offline barrage control strategy; and delivering the offline barrage to an offline live room. The application realizes double risk filtering of the barrage, and ensures effective control of the barrage content in an online and offline linkage scene.

Description

Technical Field

[0001] This invention relates to the field of bullet screen control technology, and more specifically, to a bullet screen control method, device, server, and computer storage medium. Background Technology

[0002] To enhance interactivity and engagement during live streams, users are typically allowed to send comments (bullet comments) while watching. These comments allow users to share their thoughts and opinions on the live stream, making them an important channel for communication and exchange of views among multiple users.

[0003] For important events that are simultaneously broadcast online and offline, such as sports events and award ceremonies, in order to enhance the interaction between online and offline activities, a large screen is usually set up at the offline event venue. The comments from the online live stream are forwarded to the large screen at the offline event venue. How to effectively manage the comments in the scenario of online and offline interaction is a technical problem that urgently needs to be solved by those skilled in the art. Summary of the Invention

[0004] The purpose of this invention is to provide a method, device, server, and computer storage medium for controlling bullet comments, which can effectively control bullet comments in online and offline interactive scenarios.

[0005] The embodiments of the present invention can be implemented as follows:

[0006] In a first aspect, the present invention provides a method for controlling bullet comments, the method comprising:

[0007] Get the original bullet comments sent by users in the online live broadcast room;

[0008] The original bullet comments are subjected to online risk filtering to select online bullet comments that comply with the online bullet comment control strategy;

[0009] The online bullet comments are filtered for offline risks to identify those that comply with the offline bullet comment control strategy.

[0010] The offline bullet comments are then delivered to the offline live streaming room.

[0011] In an optional implementation, the step of performing offline risk filtering on the online bullet comments to select offline bullet comments that meet the offline bullet comment control strategy includes:

[0012] Get the total number of online bullet comments;

[0013] The risk level is determined from the preset levels based on the total quantity;

[0014] Based on the risk keywords belonging to the aforementioned risk level, the online bullet comments are filtered for offline risks to obtain the offline bullet comments.

[0015] In an optional implementation, the preset risk levels include high-risk, medium-risk, and low-risk levels, with risk levels decreasing sequentially. The step of determining the risk level based on the total number includes:

[0016] If the total number is less than the preset number, then the high-risk level is taken as the risk level; otherwise, the high-risk level, the medium-risk level, and the low-risk level are taken as the risk level.

[0017] In an optional implementation, the step of performing offline risk filtering on the online bullet comments based on risk keywords belonging to the risk level to obtain the offline bullet comments includes:

[0018] The online bullet comments are parsed to extract text segments.

[0019] If a text segment matching the risk keyword exists in the online bullet comments, then the online bullet comments are filtered.

[0020] The filtered online bullet comments are used as the offline bullet comments.

[0021] In an optional implementation, the offline bullet comments include bullet comments from multiple platforms, and the step of delivering the offline bullet comments to the offline live streaming room includes:

[0022] The number of offline bullet comments originating from the same platform was counted.

[0023] Based on the number of bullet comments on each platform, the bullet comment traffic from each platform is allocated to obtain the bullet comments delivered to each platform.

[0024] All the bullet comments from all platforms will be sent to the offline live broadcast room.

[0025] In an optional implementation, the step of allocating bullet comment traffic from each platform based on the number of bullet comments on each platform to obtain the bullet comment delivery for each platform includes:

[0026] The number of bullet comments for each platform is determined based on the pre-set proportion of bullet comments for each platform and the number of bullet comments for each platform.

[0027] Select a number of bullet comments from each platform that equals the number of bullet comments sent on each platform, and use these bullet comments for each platform.

[0028] In an optional implementation, before the step of determining the number of submissions for each platform based on the pre-set percentage of bullet comments on each platform and the number of bullet comments on each platform, the following steps are included:

[0029] Get reference online comments within a preset time period;

[0030] The number of bullet comments from the same platform in the reference online bullet comments is counted;

[0031] The proportion of bullet comments for each platform is set based on the number of reference online bullet comments from each platform and the total number of reference online bullet comments.

[0032] Secondly, the present invention provides a bullet screen control device, the device comprising:

[0033] The acquisition module is used to acquire the original bullet comments sent by users in the online live broadcast room;

[0034] The filtering module is used to perform online risk filtering on the original bullet comments and select online bullet comments that comply with the online bullet comment control strategy.

[0035] The filtering module is also used to perform offline risk filtering on the online bullet comments and filter out offline bullet comments that meet the offline bullet comment control strategy.

[0036] The delivery module is used to deliver the offline bullet comments to the offline live broadcast room.

[0037] Thirdly, the present invention provides a server including a processor and a memory, the memory being used to store a program, and the processor being used to implement the bullet screen control method as described in the first aspect when executing the program.

[0038] Fourthly, the present invention provides a computer storage medium storing a computer program thereon, which, when executed by a processor, implements the bullet screen control method as described in the first aspect.

[0039] Compared with the prior art, the present invention has the following beneficial effects:

[0040] This invention first performs online risk filtering on the original bullet comments sent by users in the online live broadcast room, and selects online bullet comments that meet the online bullet comment control strategy. Then, it performs offline risk filtering on the selected online bullet comments, and selects offline bullet comments that meet the offline bullet comment control strategy. Finally, the offline bullet comments are delivered to the offline live broadcast room. By performing dual filtering of online and offline risk filtering in sequence, effective control of bullet comments in online and offline interactive scenarios is ensured. Attached Figure Description

[0041] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of the present invention and should not be regarded as a limitation on the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0042] Figure 1 This is an example diagram illustrating an application scenario provided in this embodiment.

[0043] Figure 2 This is a block diagram of the server provided in this embodiment.

[0044] Figure 3 This is a flowchart illustrating the bullet screen control method provided in this embodiment.

[0045] Figure 4 This is a block diagram of the bullet screen control device provided in this embodiment.

[0046] Icons: 10-Server; 11-Processor; 12-Memory; 13-Bus; 20-Client; 100-Bullet Screen Control Device; 110-Acquisition Module; 120-Filtering Module; 130-Delivery Module. Detailed Implementation

[0047] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. The components of the embodiments of the present invention described and shown in the accompanying drawings can generally be arranged and designed in various different configurations.

[0048] Therefore, the following detailed description of the embodiments of the invention provided in the accompanying drawings is not intended to limit the scope of the claimed invention, but merely to illustrate selected embodiments of the invention. All other embodiments obtained by those skilled in the art based on the embodiments of the invention without inventive effort are within the scope of protection of the invention.

[0049] It should be noted that similar labels and letters in the following figures indicate similar items. Therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures.

[0050] In the description of this invention, it should be noted that if terms such as "upper," "lower," "inner," or "outer" are used to indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings, or the orientation or positional relationship in which the product of this invention is usually placed, they are only for the convenience of describing this invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of this invention.

[0051] Furthermore, the terms "first" and "second" are used only to distinguish descriptions and should not be interpreted as indicating or implying relative importance.

[0052] It should be noted that, where there is no conflict, the features in the embodiments of the present invention can be combined with each other.

[0053] Please refer to Figure 1 , Figure 1 This is an example diagram illustrating an application scenario provided in this embodiment. Figure 1 In this process, server 10 and client 20 are connected. Users send bullet comments (danmaku) in the online live stream room through client 20. Client 20 sends the bullet comments to server 10, which performs online risk filtering, selecting those that comply with the online bullet comment control policy and returning them to client 20. Client 20 then displays the returned bullet comments in the online live stream room. Furthermore, server 10 also performs offline risk filtering on the online bullet comments, selecting those that comply with the offline bullet comment control policy and delivering them to the offline live stream room for display.

[0054] Server 10 can be a single server, a cluster of multiple servers, or a cloud server, while client 20 can be a mobile phone, laptop, tablet, wearable device, etc.

[0055] exist Figure 1 Based on this, this embodiment also provides a block diagram example of server 10. Please refer to... Figure 2 , Figure 2 This is a block diagram of the server 10 provided in this embodiment. The server 10 includes a processor 11, a memory 12 and a bus 13. The processor 11 and the memory 12 are connected through the bus 13.

[0056] The processor 11 can be an integrated circuit chip with signal processing capabilities. In implementation, each step of the bullet screen control method described in the above embodiments can be completed by the integrated logic circuits in the hardware of the processor 11 or by software instructions. The processor 11 can be a general-purpose processor, including a CPU (Central Processing Unit), an NP (Network Processor), etc.; it can also be a DSP (Digital Signal Processor), an ASIC (Application Specific Integrated Circuit), an FPGA (Field Programmable Logic Gate Array), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components.

[0057] The memory 12 is used to store the program that implements the bullet screen control method. The program can be a software function module stored in the memory 12 in the form of software or firmware or embedded in the OS (Operating System) of the server 10.

[0058] After receiving the execution instruction, the processor 11 executes the program to implement the bullet screen control method of the aforementioned embodiment.

[0059] based on Figure 1 and Figure 2 This embodiment provides a method for controlling bullet comments applied to server 10. Please refer to... Figure 3 , Figure 3 This is a flowchart illustrating the bullet screen control method provided in this embodiment. The method includes the following steps:

[0060] Step S101: Obtain the original bullet comments sent by users in the online live broadcast room.

[0061] In this embodiment, the online live streaming room can be a virtual interactive space within a live streaming platform, where users can send bullet comments to interact with other users or streamers in real time. The original bullet comments can be unfiltered text messages sent by users in the online live streaming room or gift-type messages. It should be noted that these original bullet comments include all content that users may input, including but not limited to normal interactive content, inappropriate words, or potentially risky content. This content can be input by the user through client 20 and transmitted over the network to server 10 for processing.

[0062] Step S102: Perform online risk filtering on the original bullet comments and select online bullet comments that comply with the online bullet comment control strategy.

[0063] In this embodiment, the online bullet screen control strategy involves a series of rules and mechanisms, such as black word database checks and semantic analysis, to determine whether the original bullet screen contains sensitive words or other content unsuitable for online display. If the original bullet screen does not trigger any risk rules, it is considered to comply with the online bullet screen control strategy. Due to the high real-time requirements of bullet screen control in online live streaming rooms, the implementation of online risk filtering must balance efficiency and accuracy. A deep learning-based text classification model can be trained to identify bullet screens that may contain obscure or inappropriate content. This model can provide prediction results within milliseconds, thereby significantly enhancing the responsiveness of online bullet screens.

[0064] Step S103: Perform offline risk filtering on online bullet comments and select offline bullet comments that meet the offline bullet comment control strategy.

[0065] In this embodiment, the offline live streaming room can refer to the space area of ​​the large screen display at the offline event venue, used to simultaneously present online interactive content to the on-site audience. While the risks of online bullet comments have been controlled to a certain extent after online risk filtering, the offline live streaming room is geared towards an on-site audience, and the audience of the offline live streaming room is more diverse than that of the online live streaming room. Therefore, offline risk filtering aims to further assess whether online bullet comments are suitable for display to the audience of the offline live streaming room in an on-site environment.

[0066] In this embodiment, the offline bullet screen control strategy and the online bullet screen control strategy are different, and their focuses are also different based on different needs. The former faces a more complex audience, so the degree of risk control is more stringent. In order to better meet the needs and atmosphere of the live audience and interact better with the live audience, it is also necessary to consider that bullet screens from various platforms can be displayed in the offline live broadcast room. The latter sends and displays bullet screens online, so the real-time requirements for bullet screen control are higher.

[0067] Offline bullet comments can be understood as the result of further filtering from online bullet comments. They not only meet the requirements of online bullet comment control strategies, but also undergo strict filtering for offline scenarios.

[0068] Step S104: Send the offline bullet comments to the offline live broadcast room.

[0069] In this embodiment, the step-by-step filtering and screening process from the original bullet comments to the offline bullet comments ensures that the bullet comment content finally delivered to the offline live broadcast room not only conforms to the online interaction norms, but also adapts to the viewing experience of offline viewers, thus ensuring effective control of bullet comments in online and offline interactive scenarios.

[0070] In this embodiment, to ensure flexibility in risk management, multiple different risk levels can be used for offline risk filtering of online bullet comments. Typically, the risk level used for filtering can be selected based on different risk management requirements. However, when the number of online bullet comments is small, to ensure the display volume of offline bullet comments while keeping the risk under control and to maximize the audience's interactive experience, in addition to considering risk management requirements, the total number of online bullet comments is also considered. Therefore, this embodiment also provides a method for performing offline risk filtering on online bullet comments to obtain offline bullet comments:

[0071] First, obtain the total number of online bullet comments;

[0072] Secondly, the risk level is determined from the preset levels based on the total quantity;

[0073] In this embodiment, the risk level is a grading system used to measure the potential risk level of bullet screen content, which can include three levels: high, medium, and low. It should be noted that the division of these three levels is based on the degree of impact the bullet screen content may have on offline viewers. For example, high-risk content clearly violates laws, regulations, or social norms and has an extremely high risk; medium-risk content may cause controversy or discomfort, but has not yet reached the level of illegality; low-risk content corresponds to slightly inappropriate content, which usually does not have a significant negative impact on the overall atmosphere. Generally, the risk level used for filtering can be selected according to different risk control requirements. For example, in scenarios with extremely high risk control requirements, even low-risk bullet screens should be filtered out; in scenarios with less stringent risk control requirements, low-risk or even medium-risk bullet screens may not be filtered.

[0074] In this embodiment, when the total number of online bullet comments is small, the control can be appropriately relaxed through risk level, allowing some low-risk or medium-risk bullet comments to pass. Conversely, in order to avoid bullet comments being too dense, the control can be tightened through risk level. One way to determine the risk level based on the total number is: if the total number is less than the preset number, then the high-risk level is used as the risk level, that is, only bullet comments that meet the high-risk level are filtered; otherwise, the high-risk level, medium-risk level, and low-risk level are used as the risk level, that is, as long as the bullet comment meets any one of the high-risk level, medium-risk level, and low-risk level, the bullet comment is filtered out.

[0075] In this embodiment, the total number can be the total number of online bullet comments within the most recent preset time period. The preset number can be set according to actual needs. When the interaction requirements are high and the risk control is not too high, the preset number can be increased, and vice versa.

[0076] Finally, based on the risk keywords belonging to the risk level, the online bullet comments are filtered for offline risks to obtain the offline bullet comments.

[0077] In this embodiment, the risk level may include at least one of high-risk, medium-risk, and low-risk levels. When the risk level includes multiple levels, if any risk keyword from any of the multiple risk levels is matched, it is considered a risk level match, and the bullet comment needs to be filtered out. One implementation method is as follows:

[0078] (1) Perform text parsing on online bullet comments and extract text segments from them;

[0079] In this embodiment, the text parsing extracts independent text segments from the online bullet screens. As an implementation method, the text parsing can be achieved through a rule engine generated by the ANTLR technology. It should be noted that ANTLR is a powerful language processing tool that can structurally parse text content according to predefined grammar rules. To achieve efficient text parsing, a set of configuration files can be pre-loaded, which contain the definition of grammar rules for bullet screen content. These rules can include, but are not limited to, punctuation splitting, lexical boundary recognition, special character filtering, etc. Through the ANTLR technology, the bullet screen text can be automatically split according to these rules and converted into a series of independent text segments. For example, for a bullet screen "The game is so wonderful!", it may be parsed into two independent text segments: "game" and "so wonderful".

[0080] Furthermore, this parsing method can not only identify complete words or phrases, but also capture potential variant forms or ambiguous expressions. For example, if the bullet screen contains variant words such as "harmonious", "hexie", or "HX", they can also be effectively parsed through predefined rules, providing accurate input data for subsequent risk keyword matching.

[0081] (2) If there is a text segment in the online bullet screen that matches the risk keyword, then filter the online bullet screen;

[0082] In this embodiment, as a specific implementation method of matching, it can be:

[0083] First, load the risk keyword set: Read the keyword set related to the current risk level from the black word library. The black word library mentioned here is a database implemented based on MySQL, which stores sensitive words of different risk levels and their corresponding risk level labels. For example, the high-risk level may include clearly illegal information, the medium-risk level may involve vulgar language, and the low-risk level covers slightly inappropriate but non-affecting overall atmosphere content.

[0084] Second, compare each text segment with the risk keywords one by one: Compare each parsed text segment with the words in the risk keyword set one by one. If it is found that a certain text segment completely matches or partially matches a risk keyword, record the matching result and mark the bullet screen as having potential risks.

[0085] Furthermore, the matching includes not only exact matching. To enhance robustness, the matching can also include fuzzy matching. For example, when the text segment is a synonym or a synonymous replacement of the risk keyword, it will also be regarded as a match between the two. Exemplarily, if the risk keyword is "fool", if the bullet screen contains its synonym "idiot", it will also be determined that the two are matched.

[0086] It should be noted that the matching rules can also be flexibly adjusted through the configuration file. Based on the progress of the event or feedback from the live audience, the configuration file can be modified at any time to update the black word database or modify the matching rules to ensure that it can always adapt to actual needs.

[0087] It should also be noted that, in addition to text-based bullet comments, for gift-type bullet comments, in order to avoid displaying gift-type bullet comments for a long time and to avoid the risk of prolonged exposure, gift-type bullet comments can be pushed to the reviewer in the form of SMS, system messages, or alarm messages. After the reviewer reviews them, the bullet comment sending interface will be called back to display the gift message.

[0088] Alternatively, gift messages can be filtered using the blacklist method mentioned above. One implementation is as follows:

[0089] First, obtain the preset black word library;

[0090] Secondly, risky keywords are obtained from a pre-set blacklist for filtering gift messages;

[0091] Finally, the gift messages are filtered based on the matching results between the risk keywords and the text segments in the gift messages.

[0092] In this embodiment, since the number of gift messages is much smaller than the number of bullet screen messages, and considering user experience, the need to intercept gift messages is generally low; that is, it is not desirable to intercept too many gift messages. Therefore, the preset blacklist can be independent of the blacklist used for bullet screen messages. The preset blacklist can be an independent database set according to the gift message interception requirements. Of course, for convenient unified management, gift messages and bullet screen messages can also share the same blacklist. In this case, the preset blacklist is the blacklist used for bullet screen messages. Since the requirements for bullet screen filtering and gift filtering are different, in order to simultaneously meet the requirements for gift message filtering, corresponding matching rules can be set specifically for gift messages. The matching rules can also include exact matching and fuzzy matching. For example, for exact matching, a match can be considered only when a text segment in a gift message completely matches a certain risk keyword, thus avoiding filtering out partially matched gift messages. For fuzzy matching, the range of synonyms or near-synonyms of the risk keyword can be narrowed to reduce the probability of gift messages being filtered out.

[0093] It should be noted that the method of obtaining risk keywords for filtering gift messages is similar to the method of obtaining risk keywords for filtering online bullet comments. When the preset black word library is an independent black word library, the preset level can be set according to the filtering requirements of gift messages. When gift messages and bullet comment messages can also share the black word library, the preset number corresponding to gift messages can be set higher than the preset number corresponding to online bullet comments. In this way, the total number of gift messages will be less than the corresponding preset number in most cases, which can avoid filtering out medium-risk and low-risk gift messages. (3) Treat the filtered online bullet comments as offline bullet comments.

[0094] In this embodiment, since offline activities are usually cross-platform, and the number of bullet comments on different platforms is uneven, in order to avoid the offline live broadcast room being occupied by bullet comments from the same platform for too long, affecting the experience of the offline audience, this embodiment is based on the statistics of the number of bullet comments from different platforms, and realizes effective management of bullet comments from each platform through bullet comment traffic allocation, and finally completes the delivery operation. A specific implementation method for delivering offline bullet comments is as follows:

[0095] First, count the number of bullet comments from the same platform in offline bullet comments;

[0096] In this embodiment, all offline bullet comments within a preset period can be acquired first. These offline bullet comments within the preset period are then categorized according to the platform they originate from, with comments from the same platform grouped into one category. The number of bullet comments within each category is then counted. This statistical operation can be implemented using a pre-configured data storage structure (e.g., a Redis array). For example, suppose there are three platforms A, B, and C. The bullet comment sending activity of each platform is recorded in real time over a period of time, and this data is stored in a Redis array.

[0097] Secondly, based on the number of bullet comments on each platform, the bullet comment traffic from each platform is allocated to obtain the bullet comments delivered on each platform;

[0098] In this embodiment, the proportion of bullet comments for each platform can be preset according to the needs of the actual scenario. If the number of bullet comments on a certain platform is greater than the proportion of bullet comments on that platform, the bullet comments on that platform need to be filtered and discarded to effectively control the bullet comment traffic from different platforms, control the number of bullet comments on that platform within the proportion of bullet comments, avoid excessive use of resources or impact on user experience by bullet comments on a certain platform, and improve the overall efficiency and fairness of bullet comment management in offline live streaming rooms.

[0099] One possible implementation method for bullet screen traffic allocation is:

[0100] (1) Determine the number of bullet comments for each platform based on the pre-set proportion of bullet comments for each platform and the number of bullet comments for each platform.

[0101] (2) Select the number of bullet comments from each platform that is equal to the number of bullet comments sent on each platform as the bullet comments sent on each platform.

[0102] In this embodiment, the bullet comment ratio can be a pre-set ratio by the system administrator or operations personnel according to business needs. For example, there are three platforms A, B, and C, and their bullet comment ratio is set to 5:3:2. When the actual number of bullet comments received exceeds the preset ratio, the excess bullet comments will be randomly discarded, thereby ensuring that the final distribution of bullet comments delivered to the offline live broadcast room meets the expected ratio. For example, the current number of bullet comments on platforms A, B, and C are 10, 6, and 4, respectively. The total number of bullet comments = 10(A) + 6(B) + 4(C) = 20.

[0103] Actual percentages for each platform: Platform A's actual percentage = 10 / 20 = 50%, Platform B's actual percentage = 6 / 20 = 30%, Platform C's actual percentage = 4 / 20 = 20%. The pre-set percentage ratio for bullet comments across platforms is 5:2:3, meaning Platform B's actual percentage is greater than the pre-set percentage. Therefore, Platform B needs to discard 10% of its bullet comments.

[0104] It should be noted that the selection of bullet comments can be random or based on the current network quality. For example, if the current network quality is low, bullet comments with small amounts of information can be selected first to reduce the pressure on network transmission.

[0105] Finally, all the comments from all platforms were sent to the offline live stream room.

[0106] In this embodiment, to ensure that the proportion of online and offline bullet comments remains basically consistent across platforms and to guarantee the fairness of bullet comment delivery across platforms, this embodiment also provides an implementation method for setting the proportion of bullet comments on each platform:

[0107] First, obtain reference online comments within a preset time period;

[0108] In this embodiment, "reference online bullet comments" can be bullet comment data obtained from the online live broadcast room within a preset time period and filtered for online risks. For example, reference online bullet comments can be all online bullet comments in the past 30 minutes, which reflects the activity level and user behavior characteristics of each platform in a specific time period.

[0109] Secondly, the number of bullet comments from the same platform in the online bullet comments were statistically analyzed;

[0110] Finally, based on the number of reference online bullet comments from each platform and the total number of reference online bullet comments, the proportion of bullet comments for each platform is set.

[0111] In this embodiment, based on statistical results, the bullet comment ratio for each platform is calculated and set according to the relationship between the number of reference online bullet comments from each platform and the total number of reference online bullet comments. Here, "bullet comment ratio" refers to a dynamically allocated proportion based on actual reference online bullet comment data, used to guide subsequent bullet comment traffic allocation operations. For example, assuming that platforms A, B, and C send 500, 300, and 200 bullet comments respectively within a certain time period, and the total number of reference online bullet comments is 1000, the bullet comment ratio for each platform can be set to 5:3:2.

[0112] The method provided in this embodiment not only ensures that the proportion of bullet comments on each platform in the online live broadcast room is roughly the same as that in the offline live broadcast room, thus guaranteeing the fairness of the delivery of bullet comments on each platform, but also optimizes resource allocation in a timely manner based on actual user behavior, thereby improving user experience and management efficiency.

[0113] To perform the corresponding steps in the above embodiments and various possible implementations, an implementation method of the bullet screen control device 100 is given below. Please refer to... Figure 4 , Figure 4 This is a block diagram of the bullet screen control device provided in this embodiment. It should be noted that the basic principle and technical effects of the bullet screen control device 100 provided by the present invention are the same as those of the corresponding embodiments described above. For the sake of brevity, some parts of this embodiment are not mentioned.

[0114] The bullet screen control device 100 includes an acquisition module 110, a filtering module 120, and a delivery module 130.

[0115] The acquisition module 110 is used to acquire the original bullet comments sent by users in the online live broadcast room;

[0116] The filtering module 120 is used to perform online risk filtering on the original bullet comments and select online bullet comments that comply with the online bullet comment control strategy.

[0117] The filtering module 120 is also used to perform offline risk filtering on online bullet comments and filter out offline bullet comments that comply with the offline bullet comment control strategy.

[0118] The delivery module 130 is used to deliver offline bullet comments to the offline live broadcast room.

[0119] In an optional implementation, the filtering module 120 is specifically used for:

[0120] Get the total number of online bullet comments;

[0121] The risk level is determined from the preset levels based on the total quantity;

[0122] Based on the risk keywords belonging to the risk level, online bullet comments are filtered for offline risks to obtain offline bullet comments.

[0123] In an optional implementation, the preset risk levels include high-risk, medium-risk, and low-risk levels, with risk levels decreasing sequentially. The filtering module 120, when used to determine the risk level based on the total quantity, is specifically used for:

[0124] If the total number is less than the preset number, the high-risk level will be used as the risk level; otherwise, the high-risk, medium-risk, and low-risk levels will be used as the risk levels.

[0125] In an optional implementation, the filtering module 120 is specifically configured to include:

[0126] Perform text parsing on online bullet comments to extract text segments from them;

[0127] If there are text segments in the online bullet comments that match risk keywords, then the online bullet comments will be filtered.

[0128] The filtered online bullet comments will be used as offline bullet comments.

[0129] In an optional implementation, the offline bullet comments include bullet comments from multiple platforms, and the delivery module 130 is specifically used for:

[0130] Count the number of bullet comments from the same platform in offline bullet comments;

[0131] Based on the number of bullet comments on each platform, the bullet comment traffic from each platform is allocated to obtain the bullet comments delivered to each platform.

[0132] All the comments from various platforms will be sent to the offline live stream room.

[0133] In an optional implementation, the delivery module 130, when used to allocate bullet comment traffic from each platform based on the number of bullet comments on each platform to obtain the delivered bullet comments for each platform, is specifically used for:

[0134] The number of bullet comments for each platform is determined based on the pre-set proportion of bullet comments for each platform and the number of bullet comments for each platform.

[0135] Select a number of bullet comments from each platform that equals the number of bullet comments sent on each platform, and use these bullet comments for each platform.

[0136] In an optional implementation, the delivery module 130 is further configured to:

[0137] Get reference online comments within a preset time period;

[0138] The statistics refer to the number of bullet comments from the same platform in the online bullet comments;

[0139] The proportion of bullet comments for each platform is set based on the number of reference online bullet comments from each platform and the total number of reference online bullet comments.

[0140] This embodiment provides a computer storage medium storing a computer program, which, when executed by a processor, implements the bullet screen control method as described in any of the foregoing embodiments.

[0141] In summary, the embodiments of the present invention provide a method, device, server, and computer storage medium for controlling bullet comments. The method includes: acquiring original bullet comments sent by users in an online live broadcast room; performing online risk filtering on the original bullet comments to select online bullet comments that meet the online bullet comment control strategy; performing offline risk filtering on the online bullet comments to select offline bullet comments that meet the offline bullet comment control strategy; and delivering the offline bullet comments to an offline live broadcast room. Compared with the prior art, this embodiment has at least the following advantages: (1) Dual risk filtering of bullet comments ensures effective control of bullet comment content in online and offline linkage scenarios; (2) Multiple risk levels are used to filter online bullet comments offline, improving the flexibility of risk control; (3) When the number of online bullet comments is small, the risk control requirements and the total number of online bullet comments are considered at the same time, ensuring the display of offline bullet comments under controllable risk, and satisfying the audience's interactive experience as much as possible; (4) Fuzzy matching can be used when matching with risk keys, enhancing the robustness of offline risk control; (5) The proportion of bullet comments corresponding to each platform is preset according to the needs of the actual scenario, avoiding excessive use of resources or impact on user experience by bullet comments of a certain platform, and improving the overall efficiency and fairness of bullet comment management in offline live broadcast rooms; (6) The proportion of bullet comments of each platform is set using bullet comment data obtained from online live broadcast rooms within a preset time and after online risk filtering, ensuring that the proportion of online and offline bullet comments of each platform is basically consistent, and ensuring that the delivery of online and offline bullet comments of each platform is fairly equal.

[0142] The above descriptions are merely various embodiments of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A method for controlling bullet comments, characterized in that, The method includes: Get the original bullet comments sent by users in the online live broadcast room; The original bullet comments are subjected to online risk filtering to select online bullet comments that comply with the online bullet comment control strategy; The online bullet comments are filtered for offline risks to identify those that meet the offline bullet comment control strategy. These offline bullet comments come from multiple platforms. The number of offline bullet comments originating from the same platform was counted. The number of bullet comments for each platform is determined based on the pre-set proportion of bullet comments for each platform and the number of bullet comments for each platform. Select a number of bullet comments from each platform that equals the number of bullet comments sent on each platform, and use these bullet comments for each platform. All the bullet comments from all platforms will be sent to the offline live broadcast room.

2. The method according to claim 1, characterized in that, The steps of performing offline risk filtering on the online bullet comments and selecting offline bullet comments that meet the offline bullet comment control strategy include: Get the total number of online bullet comments; The risk level is determined from the preset levels based on the total quantity; Based on the risk keywords belonging to the aforementioned risk level, the online bullet comments are filtered for offline risks to obtain the offline bullet comments.

3. The method according to claim 2, characterized in that, The preset risk levels include high-risk, medium-risk, and low-risk levels, with risk levels decreasing sequentially. The step of determining the risk level based on the total number includes: If the total number is less than the preset number, then the high-risk level is taken as the risk level; otherwise, the high-risk level, the medium-risk level, and the low-risk level are taken as the risk level.

4. The method according to claim 2, characterized in that, The step of performing offline risk filtering on the online bullet comments based on risk keywords belonging to the risk level to obtain the offline bullet comments includes: The online bullet comments are parsed to extract text segments. If a text segment matching the risk keyword exists in the online bullet comments, then the online bullet comments are filtered. The filtered online bullet comments are used as the offline bullet comments.

5. The method according to claim 1, characterized in that, Before the step of determining the number of submissions for each platform based on the pre-set proportion and number of bullet comments for each platform, the following steps are included: Get reference online comments within a preset time period; The number of bullet comments from the same platform in the reference online bullet comments is counted; The proportion of bullet comments for each platform is set based on the number of reference online bullet comments from each platform and the total number of reference online bullet comments.

6. A bullet screen control device, characterized in that, The device includes: The acquisition module is used to acquire the original bullet comments sent by users in the online live broadcast room; The filtering module is used to perform online risk filtering on the original bullet comments and select online bullet comments that comply with the online bullet comment control strategy. The filtering module is also used to perform offline risk filtering on the online bullet comments, and to filter out offline bullet comments that meet the offline bullet comment control strategy. The offline bullet comments come from multiple platforms. The delivery module is used to: count the number of bullet comments from the same platform in the offline bullet comments; determine the delivery quantity for each platform based on the pre-set bullet comment ratio and the number of bullet comments on each platform; select bullet comments from the bullet comments from each platform whose quantity is equal to the delivery quantity for each platform as the delivery bullet comments for each platform; and deliver the delivery bullet comments from all platforms to the offline live broadcast room.

7. A server, characterized in that, It includes a processor and a memory, the memory being used to store a program, and the processor being used to implement the bullet screen control method as described in any one of claims 1-5 when executing the program.

8. A computer storage medium, characterized in that, It stores a computer program that, when executed by a processor, implements the bullet screen control method as described in any one of claims 1-5.

Images