Information processing method and device, electronic equipment, medium and program product

Abstract

The present disclosure relates to an information processing method and device, electronic equipment, medium and program product, applied to the live broadcast technical field, the method comprises the steps of: receiving the live broadcast information of the live broadcast room input through the information input interface; determining the type to which the live broadcast room belongs according to the live broadcast information of the live broadcast room; determining the resource transfer information estimation model corresponding to the live broadcast room according to the type to which the live broadcast room belongs; processing the live broadcast information through the resource transfer information estimation model corresponding to the live broadcast room to obtain initial resource transfer estimation information. The present disclosure can improve the accuracy of resource transfer information estimation.

Description

Technical Field

[0001] This disclosure relates to the field of live streaming technology, and in particular to an information processing method, apparatus, electronic device, medium, and program product. Background Technology

[0002] Currently, the e-commerce sector is booming with various business models, among which live-streaming e-commerce has seen the fastest growth in the past two years. Live-streaming, with its interest-driven product sales, differs significantly from traditional e-commerce. The resource transfer information prediction methods used in traditional e-commerce are not applicable to live-streaming e-commerce; therefore, how to predict resource transfer information in live-streaming e-commerce has become a pressing technical problem to be solved. Summary of the Invention

[0003] In order to solve the above-mentioned technical problems, or at least partially solve the above-mentioned technical problems, this disclosure provides an information processing method, apparatus, electronic device, medium and program product.

[0004] According to a first aspect of this disclosure, an information processing method is provided, comprising:

[0005] Receive live streaming information from the live streaming room entered through the information input interface;

[0006] Based on the live streaming information of the live streaming room, determine the type of the live streaming room;

[0007] Based on the type of the live streaming room, determine the resource transfer information prediction model corresponding to the live streaming room;

[0008] The live stream information is processed by the resource transfer information prediction model corresponding to the live stream room to obtain initial resource transfer prediction information.

[0009] Optionally, after obtaining the initial resource transfer prediction information, the method further includes:

[0010] Obtain information about the resources and the broadcaster from the live stream information;

[0011] Based on the information of the resources and the information of the broadcaster, the relevance of the broadcaster resources is determined;

[0012] Based on the initial resource transfer prediction information and the relevance of the streamer resources, the revised resource transfer prediction information is determined.

[0013] Optionally, the type of the live stream room is determined based on its live stream information, including:

[0014] The live streaming information of the live streaming room is input into a pre-trained live streaming room classification model;

[0015] The probability that the target resource transfer value of the live room is higher than a preset resource transfer threshold is predicted using the live room classification model.

[0016] Based on the prediction results, determine whether the live streaming room belongs to the first category of live streaming rooms.

[0017] Optionally, the method further includes:

[0018] If it is determined that the live room does not belong to the first type of live room, the number of objects associated with the streamer is obtained from the live information, and the time difference between the start time of the live broadcast and the current time is determined according to the live broadcast time information in the live information.

[0019] The type of the live stream room is determined based on the number of objects associated with the streamer and the time difference.

[0020] Optionally, determining the type of the live stream room based on the number of objects associated with the streamer and the time difference includes:

[0021] If the number of objects associated with the streamer is less than the preset number of objects, or the time difference is less than the time difference threshold, the live stream room is determined to belong to the second type of live stream room.

[0022] If the number of objects associated with the streamer is greater than or equal to the preset number of objects, and the time difference is greater than or equal to the time difference threshold, the live stream room is determined to belong to the third type of live stream room.

[0023] Optionally, determining the resource transfer information prediction model corresponding to the live stream room based on its type includes:

[0024] If the live stream room belongs to the first type of live stream room, the resource transfer information prediction model corresponding to the live stream room is determined to be the pre-trained first tree model.

[0025] If the live stream room belongs to the second type of live stream room, the resource transfer information prediction model corresponding to the live stream room is determined to be the pre-trained first time series model.

[0026] If the live stream room belongs to the third type of live stream room, the resource transfer information prediction model corresponding to the live stream room is determined to be the pre-trained second tree model and second time series model.

[0027] Optionally, if the resource transfer information prediction model corresponding to the live broadcast room is a pre-trained second tree model and second time series model, the live broadcast information also includes: multiple historical resource transfer information of the broadcaster;

[0028] The live broadcast information is processed using the second tree model and the second time series model to obtain initial resource transfer prediction information, including:

[0029] The first estimated information is obtained by processing multiple historical resource transfer information of the anchor through the second time series model.

[0030] The second tree model is used to process the information in the live broadcast information other than the multiple historical resource transfer information to obtain the second estimated information;

[0031] Initial resource transfer forecast information is determined based on the first forecast information and the second forecast information.

[0032] Optionally, if the resource transfer information prediction model corresponding to the live broadcast room is a pre-trained first tree model, the first tree model includes multiple sub-models;

[0033] The live stream information is processed using the first tree model to obtain initial resource transfer prediction information, including:

[0034] The live broadcast information is processed by the multiple sub-models respectively to obtain the indicator data corresponding to each sub-model;

[0035] Based on multiple indicator data, the initial resource transfer forecast information is determined.

[0036] Optionally, the method further includes:

[0037] Receive promotional attribute information input through the information input interface;

[0038] Obtain the promotion conversion information corresponding to the streamer from the live broadcast information.

[0039] Based on the promotion attribute information and the promotion conversion information, determine the estimated information for the transfer of additional resources;

[0040] The final resource transfer forecast information is determined based on the revised resource transfer forecast information and the additional resource transfer forecast information.

[0041] Optionally, the method further includes:

[0042] The final resource transfer estimate information, the revised resource transfer estimate information, and the additional resource transfer estimate information are displayed; the final resource transfer estimate information includes the final resource transfer estimate value.

[0043] If the difference between the reported resource transfer value and the estimated final resource transfer value is greater than a preset difference, a prompt message is determined and displayed based on the estimated final resource transfer value and the reported resource transfer value. The prompt message is used to indicate adjustments to the live broadcast information and / or the promotion attribute information.

[0044] Optionally, the method further includes:

[0045] Display the live stream information and the promotion attribute information;

[0046] In response to a modification operation on the live stream information, the live stream information is modified to obtain modified live stream information; and / or

[0047] In response to the modification operation on the promotion attribute information, the promotion attribute information is modified to obtain the modified promotion attribute information;

[0048] Based on the modified live streaming information and / or the modified promotion attribute information, the final resource transfer estimate is redefined so that the difference between the reported resource transfer value and the redefined final resource transfer estimate is less than or equal to the preset difference.

[0049] According to a second aspect of this disclosure, an information processing apparatus is provided, comprising:

[0050] The live streaming information receiving module is used to receive live streaming information from the live streaming room entered through the information input interface;

[0051] The first live room type determination module is used to determine the type of the live room based on the live room's live information;

[0052] The resource transfer information prediction model determination module is used to determine the resource transfer information prediction model corresponding to the live broadcast room based on the type of the live broadcast room.

[0053] The initial resource transfer prediction information determination module is used to process the live broadcast information through the resource transfer information prediction model corresponding to the live broadcast room to obtain the initial resource transfer prediction information.

[0054] Optionally, the information processing device further includes:

[0055] The information acquisition module is used to acquire resource information and streamer information from the live broadcast information;

[0056] The anchor resource relevance determination module is used to determine the anchor resource relevance based on the information of the resource and the information of the anchor.

[0057] The resource transfer prediction information correction module is used to determine the corrected resource transfer prediction information based on the initial resource transfer prediction information and the relevance of the streamer's resources.

[0058] Optionally, the first live room type determination module is specifically used to input the live room information into a pre-trained live room classification model; use the live room classification model to predict the probability that the target resource transfer value of the live room is higher than a preset resource transfer threshold; and determine whether the live room belongs to the first type of live room based on the prediction result.

[0059] Optionally, the information processing device further includes:

[0060] The time difference determination module is used to, if it is determined that the live room does not belong to the first type of live room, obtain the number of objects that are associated with the anchor from the live information, and determine the time difference between the start time of the live broadcast and the current time based on the live broadcast time information in the live information;

[0061] The second live stream type determination module is used to determine the type of the live stream based on the number of objects associated with the streamer and the time difference.

[0062] Optionally, the second live room type determination module is specifically used to determine that the live room belongs to the second type of live room if the number of objects associated with the streamer is less than the preset number of objects, or the time difference is less than the time difference threshold; and to determine that the live room belongs to the third type of live room if the number of objects associated with the streamer is greater than or equal to the preset number of objects, and the time difference is greater than or equal to the time difference threshold.

[0063] Optionally, the resource transfer information prediction model determination module is specifically used to determine the resource transfer information prediction model corresponding to the live room as a pre-trained first tree model if the live room belongs to the first type of live room; determine the resource transfer information prediction model corresponding to the live room as a pre-trained first time series model if the live room belongs to the second type of live room; and determine the resource transfer information prediction model corresponding to the live room as a pre-trained second tree model and second time series model if the live room belongs to the third type of live room.

[0064] Optionally, if the resource transfer information prediction model corresponding to the live broadcast room is a pre-trained second tree model and second time series model, the live broadcast information also includes: multiple historical resource transfer information of the broadcaster;

[0065] The initial resource transfer prediction information determination module processes the live broadcast information using the second tree model and the second time series model through the following steps to obtain the initial resource transfer prediction information:

[0066] The first estimated information is obtained by processing multiple historical resource transfer information of the anchor through the second time series model.

[0067] The second tree model is used to process the information in the live broadcast information other than the multiple historical resource transfer information to obtain the second estimated information;

[0068] Initial resource transfer forecast information is determined based on the first forecast information and the second forecast information.

[0069] Optionally, if the resource transfer information prediction model corresponding to the live broadcast room is a pre-trained first tree model, the first tree model includes multiple sub-models;

[0070] The initial resource transfer prediction information determination module processes the live broadcast information using the first tree model to obtain the initial resource transfer prediction information through the following steps:

[0071] The live broadcast information is processed by the multiple sub-models respectively to obtain the indicator data corresponding to each sub-model;

[0072] Based on multiple indicator data, the initial resource transfer forecast information is determined.

[0073] Optionally, the information processing device further includes:

[0074] The promotion attribute information receiving module is used to receive promotion attribute information input through the information input interface;

[0075] The promotion conversion information acquisition module is used to acquire the promotion conversion information corresponding to the anchor in the live broadcast information;

[0076] The module for determining additional resource transfer prediction information is used to determine additional resource transfer prediction information based on the promotion attribute information and the promotion conversion information.

[0077] The final resource transfer forecast information determination module is used to determine the final resource transfer forecast information based on the revised resource transfer forecast information and the additional resource transfer forecast information.

[0078] Optionally, the information processing device further includes:

[0079] The forecast information display module is used to display the final resource transfer forecast information, the revised resource transfer forecast information, and the additional resource transfer forecast information; the final resource transfer forecast information includes the final resource transfer estimate value.

[0080] The prompt information display module is used to determine and display prompt information based on the final resource transfer estimate and the reported resource transfer value when the difference between the reported resource transfer value and the final resource transfer estimate is greater than a preset difference. The prompt information is used to indicate adjustments to the live broadcast information and / or the promotion attribute information.

[0081] Optionally, the information processing device further includes:

[0082] The live streaming and promotion conversion information display module is used to display the live streaming information and the promotion attribute information;

[0083] The live stream information modification module is used to respond to a modification operation on the live stream information, modify the live stream information, and obtain the modified live stream information; and / or

[0084] The promotion attribute information modification module is used to respond to the modification operation on the promotion attribute information, modify the promotion attribute information, and obtain the modified promotion attribute information;

[0085] The final resource transfer estimate modification module is used to redetermine the final resource transfer estimate based on the modified live broadcast information and / or the modified promotion attribute information, so that the difference between the reported resource transfer value and the redetermined final resource transfer estimate is less than or equal to the preset difference.

[0086] According to a third aspect of this disclosure, an electronic device is provided, comprising: a processor configured to execute a computer program stored in a memory, wherein the computer program, when executed by the processor, implements the method described in the first aspect.

[0087] According to a fourth aspect of this disclosure, a computer-readable storage medium is provided having a computer program stored thereon, which, when executed by a processor, implements the method described in the first aspect.

[0088] According to a fifth aspect of this disclosure, a computer program product is provided that, when run on a computer, causes the computer to perform the method described in the first aspect.

[0089] The technical solution provided in this disclosure has the following advantages compared with the prior art:

[0090] By providing an information input interface, broadcasters, merchants, or operators can input live stream information, such as broadcaster information and information about resources to be traded. Based on this information, live streams can be categorized. For different types of live streams, resource transfer information prediction models applicable to that type can be pre-trained. Thus, different resource transfer information prediction models are used for different live stream types to process the live stream information and obtain initial resource transfer prediction information, improving the accuracy of the initial resource transfer prediction. Furthermore, compared to manual resource transfer value prediction, this reduces a significant amount of manual estimation work and improves operational efficiency. Attached Figure Description

[0091] The accompanying drawings, which are incorporated in and form a part of this specification, illustrate embodiments consistent with this disclosure and, together with the description, serve to explain the principles of this disclosure.

[0092] To more clearly illustrate the technical solutions in the embodiments of this disclosure or the prior art, the accompanying drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, for those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0093] Figure 1 A schematic diagram of a system architecture for an exemplary application environment in which the information processing method of the present disclosure embodiments can be applied is shown;

[0094] Figure 2 This is a flowchart of an information processing method in an embodiment of this disclosure;

[0095] Figure 3 This is another flowchart of the information processing method in the embodiments of this disclosure;

[0096] Figure 4 This is a schematic diagram of a twin-tower model;

[0097] Figure 5 This is a schematic diagram of an information processing method in an embodiment of this disclosure;

[0098] Figure 6 This is a schematic diagram showing the final resource transfer forecast information, the revised resource transfer forecast information, and the additional resource transfer forecast information in an embodiment of this disclosure.

[0099] Figure 7 This is yet another schematic diagram showing the final resource transfer forecast information, the revised resource transfer forecast information, and the additional resource transfer forecast information in an embodiment of this disclosure;

[0100] Figure 8This is a schematic diagram of the structure of an information processing device in an embodiment of this disclosure;

[0101] Figure 9 This is a schematic diagram of the structure of an electronic device according to an embodiment of this disclosure. Detailed Implementation

[0102] To better understand the above-mentioned objectives, features, and advantages of this disclosure, the solutions disclosed herein will be further described below. It should be noted that, unless otherwise specified, the embodiments and features described herein can be combined with each other.

[0103] Numerous specific details are set forth in the following description in order to provide a full understanding of this disclosure, but this disclosure may also be implemented in other ways different from those described herein; obviously, the embodiments in the specification are only some, and not all, of the embodiments of this disclosure.

[0104] This disclosure provides an information processing method, apparatus, electronic device, medium, and program product that can estimate the transfer value of goods resources in live-streaming e-commerce scenarios, improving the accuracy of the estimation. Furthermore, it can reduce the workload of manual estimation and improve operational efficiency.

[0105] Figure 1 A schematic diagram of a system architecture for an exemplary application environment in which the information processing methods of embodiments of the present disclosure can be applied is shown.

[0106] like Figure 1 As shown, system architecture 100 may include one or more of terminal devices 101 and 102, a network 103, and a server 104. Network 103 serves as the medium for providing a communication link between terminal devices 101, 102, and server 104. Network 103 may include various connection types, such as wired or wireless communication links or fiber optic cables. Terminal devices 101 and 102 may be various electronic devices with displays, including but not limited to desktop computers, laptops, smartphones, and tablets. It should be understood that... Figure 1 The number of terminal devices, networks, and servers shown is merely illustrative. Depending on implementation needs, there can be any number of terminal devices, networks, and servers. For example, server 104 could be a server cluster consisting of multiple servers.

[0107] The information processing method provided in this embodiment is generally executed by server 104, and correspondingly, the information processing device can be disposed in server 104. However, it is readily understood by those skilled in the art that the information processing method provided in this embodiment can also be executed by terminal device 101 and terminal device 102. For example, server 104 can pre-train and generate a resource transfer information prediction model using the information processing method provided in this embodiment. Terminal device 101 and terminal device 102 can input live broadcast information of the live broadcast room in the information input interface, and the live broadcast information can be uploaded to server 104 via network 103. Server 104 processes the live broadcast information based on the information processing method, determines the initial resource transfer prediction information, and sends the initial resource transfer prediction information to terminal device 101 and terminal device 102.

[0108] See Figure 2 , Figure 2 This is a flowchart of an information processing method according to an embodiment of the present disclosure, which may include the following steps:

[0109] Step S210: Receive the live streaming information of the live streaming room entered through the information input interface.

[0110] In this embodiment, an information input interface can be provided, allowing broadcasters, merchants, and operators to input live streaming information. A live streaming room refers to a room where resources can be transferred, such as goods transactions; for example, a broadcaster can conduct live-streaming sales in their live streaming room. The live streaming information may include: resource information, broadcaster information, and live streaming time information. Resources can be items / goods, rights, etc., that are traded; the quantity of resources can be one or more, and resource information may include: the price, category, and name of the item / goods or rights. Broadcaster information may include: broadcaster type, broadcaster's follower count level, historical product category distribution information, etc. Live streaming time information refers to the time information after the current moment, including the start time, end time, and duration of the live stream.

[0111] Step S220: Based on the live streaming information of the live streaming room, determine the type of the live streaming room, and based on the type of the live streaming room, determine the resource transfer information prediction model corresponding to the live streaming room.

[0112] The resource transfer information includes, for example, transaction value and volume. Because the size and type of the target audience associated with different livestreaming rooms and streamers vary significantly, the sales volume of resources in a streamer's livestreaming room also differs considerably, including the target audience, such as users. For example, based on sales performance, livestreaming rooms can be divided into three categories: top-tier livestreaming rooms, mid-tier livestreaming rooms, and bottom-tier livestreaming rooms. Top-tier livestreaming rooms are those with strong sales performance (e.g., the resource transfer value of a top-tier livestreaming room is greater than a first preset threshold), bottom-tier livestreaming rooms are those with weak sales performance (e.g., the resource transfer value of a bottom-tier livestreaming room is less than a second preset threshold, and the second preset threshold is less than the first preset threshold), and mid-tier livestreaming rooms have sales performance that is lower than top-tier livestreaming rooms but higher than bottom-tier livestreaming rooms. If resource transfer information is estimated using the same method, it is difficult to achieve high accuracy across different types of livestreaming rooms.

[0113] Based on this, different methods can be used to predict resource transfer information for different types of live streaming rooms. For example, all three types of live streaming rooms (top-tier, mid-tier, and bottom-tier) use neural network models to predict resource transfer information, and three different resource transfer information prediction models can be used to predict resource transfer information.

[0114] The resource transfer information prediction model is a pre-trained model used to process live stream information and obtain corresponding initial resource transfer prediction information. There are multiple resource transfer information prediction models available; different models can be used for different types of live streams, thus better adapting to different live streams and improving the accuracy of the predictions.

[0115] Step S230: Process the live broadcast information using the resource transfer information prediction model corresponding to the live broadcast room to obtain the initial resource transfer prediction information.

[0116] After determining the resource transfer information prediction model corresponding to the live broadcast room, the live broadcast information can be directly input into the resource transfer information prediction model. The model processes the live broadcast information to obtain the corresponding initial resource transfer prediction information. The initial resource transfer prediction information can be the resource transfer value (e.g., the transaction volume or total transaction amount of goods), other resource transfer information, etc.

[0117] The information processing method of this disclosure provides an information input interface, allowing broadcasters, merchants, or operators to input live broadcast information, such as broadcaster information and information about resources to be traded. Based on the live broadcast information, the live broadcast rooms are categorized. For different types of live broadcast rooms, a resource transfer information prediction model suitable for that live broadcast room can be pre-trained. This disclosure considers the impact of live broadcast room type on resource transfer information prediction, using different resource transfer information prediction models for different live broadcast room types to process the live broadcast information and obtain initial resource transfer prediction information, thereby improving the accuracy of the initial resource transfer value prediction. Furthermore, compared to manual resource transfer value prediction, this method reduces a significant amount of manual estimation work and improves operational efficiency.

[0118] See Figure 3 , Figure 3 This is another flowchart of the information processing method in this disclosure, which may include the following steps:

[0119] Step S310: Receive the live streaming information of the live streaming room entered through the information input interface.

[0120] In this disclosed embodiment, the live streaming information includes not only resource information, streamer information, and live streaming time information, but may also include multiple historical resource transfer values ​​of the streamer. Multiple historical resource transfer values ​​of the streamer refer to the resource transfer values ​​of the streamer during multiple historical time periods of live streaming.

[0121] Step S320: Using a pre-trained live room classification model, process the live room information, determine the type of the live room, and determine the corresponding resource transfer information prediction model based on the type of the live room.

[0122] This disclosure allows for the classification of live streaming rooms using a pre-trained live streaming room classification model. Based on different types of live streaming rooms, different methods are used to predict resource transfer values, thereby improving the accuracy of resource transfer value prediction. The live streaming room classification model can be a tree model (e.g., XGBoost model) or other network models. By processing live streaming information, the type of live streaming room can be determined.

[0123] When training a live stream classification model, live stream data from multiple live streams can be used as sample data. This sample data can be categorized into various live stream types, optionally including "top-tier" and "non-top-tier" live streams. However, the classification method is not limited to these. For each sample data point, the live stream's classification as a "top-tier" live stream can be determined based on its product resource transfer value. For example, if the product resource transfer value is greater than a preset value, the live stream can be classified as a "top-tier" live stream; otherwise, it can be classified as a "non-top-tier" live stream. After classifying the sample data, the live stream data can be used as input data, and the live stream type as label data to train and generate the live stream classification model.

[0124] During training, the Focal Loss function can be used to address the imbalance between positive and negative samples in top-tier live streams, thereby improving the accuracy of the trained live stream classification model. The live stream information is input into the pre-trained live stream classification model. The model predicts the probability that the target resource transfer value of the live stream is higher than a preset resource transfer threshold (e.g., 3 million, 5 million, etc.). Based on the prediction results, it is determined whether the live stream belongs to the first category, such as the aforementioned top-tier live streams. If the probability that the target resource transfer value of the live stream is higher than the preset resource transfer threshold (e.g., the probability that the resource transfer value generated by the next live stream is higher than the preset resource transfer threshold) is greater than a preset probability threshold (e.g., 0.7, 0.8, etc.), the live stream can be considered a first-category live stream; otherwise, it is not a first-category live stream. The target resource transfer value can be the predicted resource transfer value generated by a live stream occurring after the current time.

[0125] If a live stream does not belong to the first category, the number of people associated with the streamer (e.g., fans, users interacting with the streamer) can be obtained from the streamer's information. The time difference between the start time of the live stream and the current time can be determined. Based on the number of people associated with the streamer and the time difference, the live stream can be further subdivided. If the number of people associated with the streamer is less than a preset number (e.g., 200,000, 300,000), or the time difference is less than a time difference threshold (e.g., 5 days, 6 days), the live stream is determined to belong to the second category. If the number of people associated with the streamer is not less than the preset number, and the time difference is not less than the time difference threshold, the live stream is determined to belong to the third category.

[0126] In one optional implementation, if the live stream belongs to the first type of live stream, it indicates that the live stream has a high sales performance. The initial resource transfer prediction information for the first type of live stream can be accurately determined using a tree model. Therefore, the resource transfer information prediction model corresponding to the live stream can be a pre-trained first tree model (e.g., the XGBoost model). If the live stream belongs to the second type of live stream, it indicates that the live stream has a low sales performance, or that the current time is close to the start time of the live stream (i.e., it belongs to a short-term prediction problem). The prediction information for the second type of live stream can be accurately determined using a time series model. Therefore, the resource transfer information prediction model corresponding to the live stream can be a pre-trained first time series model (e.g., the ARIMA model). If the live stream belongs to the third type of live stream, neither a tree model nor a time series model can accurately determine the prediction information. Therefore, a fusion network model can be used, i.e., the resource transfer information prediction model corresponding to the live stream can be a pre-trained second tree model and a second time series model to improve the accuracy of the prediction information. The second tree model and the first tree model can be the same or different. The second time series model and the first time series model can also be the same or different.

[0127] Step S330: Process the live broadcast information using the resource transfer information prediction model corresponding to the live broadcast room to obtain the initial resource transfer prediction information.

[0128] As mentioned earlier, different resource transfer information prediction models can be used to process live streaming information for different types of live streaming rooms. In one optional implementation, if the resource transfer information prediction model corresponding to the live streaming room is a pre-trained first tree model, the live streaming information can be input into the first tree model, and the initial resource transfer prediction information can be directly output. That is to say, the first tree model can be an independent model.

[0129] In another optional implementation, the first tree model may include multiple independent sub-models. The live broadcast information is processed by each sub-model to obtain corresponding indicator data. Based on this multi-indicator data, initial resource transfer prediction information is determined. This improves the accuracy of determining the initial resource transfer prediction information.

[0130] For example, if there are four sub-models, processing the live stream data through each sub-model yields four corresponding metrics: the first click count of users associated with the streamer, the first resource transfer value generated by users associated with the streamer playing the resource, the second click count of users not associated with the streamer (e.g., clicks through search, recommendations, etc.), and the second resource transfer value generated by users not associated with the streamer playing the resource. Based on the first click count, the first resource transfer value, the second click count, and the second resource transfer value, initial resource transfer prediction information is obtained. Of course, this is not limited to these metrics; besides click counts for products, it could also include the number of items added to the shopping cart, the number of items purchased, etc. The resource transfer value could be the transaction amount or the transaction volume, etc.

[0131] In one optional implementation, the first number of clicks and the first transaction amount can be multiplied to obtain a third estimated value; the second number of clicks and the second transaction amount can be multiplied to obtain a fourth estimated value; the third and fourth estimated values ​​can be summed to obtain an initial resource transfer estimated value, which is used as the initial resource transfer estimated information. The processing of the number of items added to the cart and the quantity purchased is similar to that of the number of product clicks, and the processing of the number of product transactions and the transaction amount is also similar, so it will not be described in detail here.

[0132] If the resource transfer information prediction model corresponding to the live broadcast room is a pre-trained first time series model, multiple historical resource transfer values ​​of the anchor in the live broadcast information can be input into the first time series model. The first time series model can analyze multiple historical resource transfer values ​​and output initial resource transfer prediction information.

[0133] If the resource transfer information prediction model corresponding to the live broadcast room is a pre-trained second tree model and a second time series model, the second time series model can be used to process multiple historical resource transfer values ​​of the anchor to obtain a first prediction value; the second tree model can be used to process information in the live broadcast information other than multiple historical resource transfer values ​​(such as the aforementioned resource information, anchor information, and live broadcast time information, etc.) to obtain a second prediction value; the first prediction value and the second prediction value are weighted to obtain an initial resource transfer prediction value, which is used as the initial resource transfer prediction information.

[0134] Understandably, the greater the time difference between the start of the live stream and the current time, the lower the accuracy of the second estimated value output by the second time series model; conversely, the smaller the time difference, the higher the accuracy of the second estimated value output by the second time series model. Therefore, the first weighting coefficient corresponding to the first estimated value and the second weighting coefficient corresponding to the second estimated value can be determined based on the time difference, where the sum of the first weighting coefficient and the second weighting coefficient equals 1. The initial resource transfer estimated value is obtained by summing the product of the first estimated value and the first weighting coefficient with the product of the second estimated value and the second weighting coefficient.

[0135] In one alternative implementation, the time difference can be square-rooted and then inversely calculated to obtain a first weighting coefficient, and the difference between 1 and the first weighting coefficient can be used as a second weighting coefficient. Alternatively, the time difference can be cube-rooted and then inversely calculated to obtain a first weighting coefficient, and the difference between 1 and the first weighting coefficient can be used as the second weighting coefficient, and so on. The determination of the first and second weighting coefficients is not limited to these methods.

[0136] Step S340: Obtain resource information and streamer information from the live broadcast information; determine the streamer resource relevance based on the resource information and streamer information; and determine the corrected resource transfer prediction information based on the initial resource transfer prediction information and the streamer resource relevance.

[0137] Understandably, during live-streaming e-commerce, the higher the relevance between the streamer and the resources, the greater the sales volume of those resources. Therefore, based on resource information and streamer information, the relevance between the streamer and the resources can be determined—that is, the streamer-resource relevance—to further improve the accuracy of the predicted information. Relevance can include multiple levels, such as high relevance, medium relevance, and low relevance. Specifically, the relevance between the streamer and the product can be determined based on the type of product and the type of streamer. For example, if the product is clothing and the streamer is a clothing-selling streamer, the relevance can be determined as high; if the product is clothing and the streamer is a footwear-selling streamer, the relevance can be determined as medium; and if the product is clothing and the streamer is a toy-selling streamer, the relevance can be determined as low. Furthermore, relevance can also be a specific numerical value. When numerical, it is usually a number between 0 and 1, where 1 represents perfect relevance and 0 represents no relevance.

[0138] Based on the initial resource transfer forecast information, the initial resource transfer forecast information is adjusted according to the relevance of the streamer's resources to obtain the revised resource transfer forecast information. Optionally, the initial resource transfer forecast information and the streamer's resource relevance can be directly multiplied to obtain the revised resource transfer forecast information.

[0139] This disclosure improves accuracy by modifying the initial resource transfer information, which helps to determine the validity of user-submitted information.

[0140] In this embodiment of the disclosure, since resource information and broadcaster information belong to two different types of information, a dual-tower model can be used to determine the correlation between them. See also Figure 4 , Figure 4 This is a schematic diagram of a dual-tower model. During training, the resources of a live stream can be divided into several sub-lists. In each sub-list, the resource transfer value of the left resource is higher than that of the right resource. The relevance between each pair of left and right resources and the streamer is calculated, denoted as d1 and d2. Specifically, features are extracted from the resource information and the streamer information, and corresponding feature vectors are output. The feature vectors of both have the same dimension. The relevance is obtained by performing a dot product operation on the two feature vectors.

[0141] Based on the relevance d1 and d2 between the left and right resources and the broadcaster, respectively, a loss function (e.g., the pairwise loss function) can be constructed. The pairwise loss function can be expressed as follows:

[0142] Loss = max(0, margin + d2 - d1), where margin = 0.1.

[0143] Based on this loss function, a dual-tower model can be trained. Then, the resource model within the dual-tower model can be used to extract features from the resources in the live stream, obtaining item feature vectors; the anchor model within the dual-tower model can be used to extract features from the anchor in the live stream, obtaining anchor feature vectors; finally, the item feature vectors and anchor feature vectors are multiplied to obtain the anchor-resource relevance.

[0144] As mentioned earlier, the initial resource transfer prediction information and the streamer resource relevance can be directly multiplied to obtain the corrected resource transfer prediction information. Alternatively, the streamer resource relevance can be reasonably transformed before being multiplied with the initial resource transfer prediction information to obtain the corrected resource transfer prediction information, and so on.

[0145] Step S350: Receive promotion attribute information entered through the information input interface and obtain the promotion conversion information corresponding to the streamer in the live broadcast information.

[0146] Since promotional information (such as advertisements) can be considered another type of transaction, additional resource transfer estimates can be determined based on the promotional information. These additional estimates can then be added to the revised resource transfer estimates to obtain the final resource transfer estimates. In the information input interface, broadcasters, merchants, or operators can also input promotional attribute information (such as advertising budget, event budget, advertisement name, event name, etc.).

[0147] The conversion information for a livestreamer can include conversion rates (such as ROI or return on investment for advertising and promotions). For ROI, conversion information can include the livestreamer's first historical ROI and the resource provider's (e.g., merchant's) second historical ROI. For return on investment, conversion information can include the livestreamer's first historical return on investment and the resource provider's second historical return on investment.

[0148] Step S360: Determine the estimated information for the transfer of additional resources based on the promotion attribute information and promotion conversion information.

[0149] In this embodiment, assuming that the estimated information for additional resource transfer is determined by combining the first historical ROI and the second historical ROI, a weighted average (e.g., a weighted average) can be applied to the first historical ROI corresponding to the streamer and the second historical ROI corresponding to the resource provider to obtain the total ROI. The weight of the first historical ROI could be 0.6, and the weight of the second historical ROI could be 0.4, etc. Then, the promotion attribute information and the total ROI are multiplied to obtain the estimated value of additional resource transfer, which is used as the estimated information for additional resource transfer. Of course, the method for determining the estimated information for additional resource transfer is not limited to this.

[0150] Step S370: Determine the final resource transfer forecast information based on the revised resource transfer forecast information and the additional resource transfer forecast information.

[0151] In this embodiment, the revised resource transfer estimate and the additional resource transfer estimate can be summed or weighted to obtain the final resource transfer estimate, which is then used as the final resource transfer estimate information. After obtaining the final resource transfer estimate information, the final resource transfer estimate information, the revised resource transfer estimate information, and the additional resource transfer estimate information can be displayed. On the one hand, this allows broadcasters, merchants, and operators to intuitively view the final resource transfer estimate information and its composition (i.e., the revised resource transfer estimate information and the additional resource transfer estimate information). On the other hand, as needed, the revised resource transfer estimate information or the additional resource transfer estimate information can be clicked to modify the relevant information.

[0152] When entering live stream information and promotional attribute information through the information input interface, users can also input a proposed resource transfer value, i.e., the expected target to be achieved. The final estimated resource transfer value is compared with the proposed resource transfer value to confirm whether the target is met. For example, if the difference between the proposed resource transfer value and the final estimated resource transfer value is greater than a preset difference, it indicates that the final estimated resource transfer value is significantly different from the proposed resource transfer value, and the target is difficult to achieve. A prompt message can be displayed to indicate adjustments to the live stream information and / or promotional attribute information. This allows merchants, streamers, or operators to adjust the live stream information and / or promotional attribute information to make the final estimated resource transfer value closer to the proposed resource transfer value.

[0153] See Figure 5 , Figure 5 This is a schematic diagram of an information processing method in this embodiment. The broadcaster, merchant, or operator can input live broadcast information and promotional attribute information through an information input interface. The live broadcast information is processed by a resource transfer information prediction model, outputting revised resource transfer prediction information. The promotional attribute information is processed by a preset algorithm, such as the algorithm in step S360, outputting an additional resource transfer prediction value. The revised resource transfer prediction value and the additional resource transfer prediction value are summed to obtain the final resource transfer prediction value. The final resource transfer prediction value is compared with the reported resource transfer value to confirm whether it meets the target. If it meets the target, the input live broadcast information and promotional attribute information are directly used as the final information, and live broadcasts can be conducted according to this information in the future.

[0154] If the target is not met, a prompt message will be displayed. To modify live stream information, click "Correct Resource Transfer Estimate Information"; to modify promotion attribute information, click "Attach Resource Transfer Estimate Information". Responding to the click action of "Correct Resource Transfer Estimate Information", the live stream information is displayed; responding to the modification action of the live stream information, the live stream information is modified, resulting in the modified live stream information. For example, the resource structure, the host, the live stream time, and the live stream duration can be adjusted. Responding to the click action of "Attach Resource Transfer Estimate Information", the promotion attribute information is displayed; responding to the modification action of the promotion attribute information, the promotion attribute information is modified, resulting in the modified promotion attribute information. Based on the modified live stream information and the modified promotion attribute information, the final resource transfer estimate information is re-determined. It should be noted that both the live stream information and the promotion attribute information can be adjusted multiple times to ensure that the difference between the reported resource transfer value and the re-determined final resource transfer estimate information does not exceed a preset difference.

[0155] This disclosure determines the final resource transfer estimate based on promotion conversion information and promotion attribute information, which improves accuracy. Furthermore, it can determine the final resource transfer information corresponding to the adjusted promotion conversion information by adjusting the promotion conversion information, thereby providing effective suggestions to the reporting users and ultimately determining reasonable reporting targets.

[0156] See Figure 6 , Figure 6 This is a schematic diagram illustrating the display of final resource transfer estimate information, revised resource transfer estimate information, and additional resource transfer estimate information in an embodiment of this disclosure. The diagram shows the final resource transfer estimate information, revised resource transfer estimate information, and additional resource transfer estimate information. When "Revise Resource Transfer Estimate Information" is clicked, a live stream information modification interface is displayed. For example, the live stream start time, live stream end time, and resources can be modified, and the changes can be saved.

[0157] See Figure 7 , Figure 7 This is another schematic diagram illustrating the final resource transfer estimate, the revised resource transfer estimate, and the additional resource transfer estimate in this embodiment of the present disclosure. The diagram shows the final resource transfer estimate, the revised resource transfer estimate, and the additional resource transfer estimate. When "Additional Resource Transfer Estimate" is clicked, a modification interface for promotional attribute information is displayed. Users can slide the slider on the interface to modify the promotional attribute information and save the changes.

[0158] After the modifications are completed, the values ​​of the final resource transfer estimate, the revised resource transfer estimate, and the additional resource transfer estimate in the interface will also change. If the difference between the reported resource transfer value and the changed final resource transfer estimate is not greater than the preset difference, no further adjustments are needed, and live streaming can be conducted in the future according to the adjusted live streaming information and promotion attribute information.

[0159] The information processing method of this disclosure uses a first tree model for the first type of live streaming room, a first time series model for the second type of live streaming room, and a fusion model of a second tree model and a second time series model for the third type of live streaming room. This method can solve the problem of prediction error of live streaming room resource transfer value caused by various situations such as short-term prediction, long-term prediction, uneven distribution of anchor fans, and unstable traffic.

[0160] Furthermore, the first-tree model can be decomposed into multiple independent sub-models, each of which outputs different indicator data. By using multiple indicator data to determine the initial resource transfer prediction information, the accuracy of resource transfer value prediction can be further improved compared with determining the initial resource transfer prediction information through a single first-tree model.

[0161] Corresponding to the above method embodiments, this disclosure also provides an information processing apparatus, see below. Figure 8 The information processing device 800 includes:

[0162] The live streaming information receiving module 810 is used to receive live streaming information from the live streaming room entered through the information input interface;

[0163] The first live room type determination module 820 is used to determine the type of the live room based on the live room's information.

[0164] The resource transfer information prediction model determination module 830 is used to determine the resource transfer information prediction model corresponding to the live broadcast room based on the type of the live broadcast room.

[0165] The initial resource transfer prediction information determination module 840 is used to process the live broadcast information through the resource transfer information prediction model corresponding to the live broadcast room to obtain the initial resource transfer prediction information.

[0166] Optionally, the information processing device 800 described above further includes:

[0167] The information acquisition module is used to obtain information about resources and the streamer from the live broadcast information;

[0168] The anchor resource relevance determination module is used to determine the relevance of anchor resources based on resource information and anchor information;

[0169] The resource transfer prediction information correction module is used to determine the corrected resource transfer prediction information based on the initial resource transfer prediction information and the relevance of the streamer's resources.

[0170] Optionally, the first live room type determination module is specifically used to input the live room information into a pre-trained live room classification model; use the live room classification model to predict the probability that the target resource transfer value of the live room is higher than the preset resource transfer threshold; and determine whether the live room belongs to the first type of live room based on the prediction result.

[0171] Optionally, the information processing device 800 described above further includes:

[0172] The time difference determination module is used to determine the number of objects related to the streamer from the live broadcast information if it is determined that the live broadcast room does not belong to the first type of live broadcast room, and to determine the time difference between the start time of the live broadcast and the current time based on the live broadcast time information in the live broadcast information.

[0173] The second live stream type determination module is used to determine the type of live stream based on the number of objects that are associated with the streamer and the time difference.

[0174] Optionally, the second live room type determination module is specifically used to determine that the live room belongs to the second type of live room if the number of objects associated with the streamer is less than the preset number of objects, or the time difference is less than the time difference threshold; and to determine that the live room belongs to the third type of live room if the number of objects associated with the streamer is greater than or equal to the preset number of objects, and the time difference is greater than or equal to the time difference threshold.

[0175] Optionally, the resource transfer information prediction model determination module is specifically used to determine the resource transfer information prediction model corresponding to the live room as a pre-trained first tree model if the live room belongs to the first type of live room; to determine the resource transfer information prediction model corresponding to the live room as a pre-trained first time series model if the live room belongs to the second type of live room; and to determine the resource transfer information prediction model corresponding to the live room as a pre-trained second tree model and second time series model if the live room belongs to the third type of live room.

[0176] Optionally, if the resource transfer information prediction model corresponding to the live broadcast room is a pre-trained second tree model and second time series model, the live broadcast information also includes: multiple historical resource transfer information of the broadcaster;

[0177] The initial resource transfer prediction information determination module obtains the initial resource transfer prediction information by processing the live broadcast information through the second tree model and the second time series model through the following steps:

[0178] The first estimated information is obtained by processing multiple historical resource transfer information of the anchor through the second time series model;

[0179] The second tree model is used to process the information in the live broadcast information, excluding multiple historical resource transfer information, to obtain the second estimated information;

[0180] The initial resource transfer forecast information is determined based on the first and second forecast information.

[0181] Optionally, if the resource transfer information prediction model corresponding to the live broadcast room is a pre-trained first tree model, the first tree model includes multiple sub-models;

[0182] The initial resource transfer prediction information determination module processes the live broadcast information using the first tree model to obtain the initial resource transfer prediction information through the following steps:

[0183] The live broadcast information is processed by multiple sub-models to obtain the corresponding indicator data for each sub-model.

[0184] Based on multiple indicator data, the initial resource transfer forecast information is determined.

[0185] Optionally, the information processing device 800 described above further includes:

[0186] The promotion attribute information receiving module is used to receive promotion attribute information entered through the information input interface;

[0187] The promotion conversion information acquisition module is used to acquire the promotion conversion information corresponding to the streamer in the live broadcast information;

[0188] The module for determining additional resource transfer forecast information is used to determine additional resource transfer forecast information based on promotion attribute information and promotion conversion information.

[0189] The final resource transfer forecast information determination module is used to determine the final resource transfer forecast information based on the revised resource transfer forecast information and the additional resource transfer forecast information.

[0190] Optionally, the information processing device 800 described above further includes:

[0191] The forecast information display module is used to display the final resource transfer forecast information, the revised resource transfer forecast information, and the additional resource transfer forecast information; the final resource transfer forecast information includes the final resource transfer estimate value.

[0192] The prompt information display module is used to determine and display prompt information based on the final resource transfer estimate and the proposed resource transfer value when the difference between the reported resource transfer value and the final estimated resource transfer value is greater than a preset difference. The prompt information is used to indicate adjustments to the live broadcast information and / or promotion attribute information.

[0193] Optionally, the information processing device 800 described above further includes:

[0194] The live streaming and promotion conversion information display module is used to display live streaming information and promotion attribute information;

[0195] The live stream information modification module is used to respond to modification operations on live stream information, modify the live stream information, and obtain the modified live stream information; and / or

[0196] The promotion attribute information modification module is used to respond to modification operations on promotion attribute information, modify the promotion attribute information, and obtain the modified promotion attribute information;

[0197] The final resource transfer estimate modification module is used to redetermine the final resource transfer estimate based on the modified live broadcast information and / or modified promotion attribute information, so that the difference between the reported resource transfer value and the redetermined final resource transfer estimate is less than or equal to a preset difference.

[0198] The specific details of each module or unit in the above-mentioned device have been described in detail in the corresponding methods, so they will not be repeated here.

[0199] It should be noted that although several modules or units for the device used to perform actions have been mentioned in the detailed description above, this division is not mandatory. In fact, according to embodiments of this disclosure, the features and functions of two or more modules or units described above can be embodied in one module or unit. Conversely, the features and functions of one module or unit described above can be further divided and embodied by multiple modules or units.

[0200] This disclosure also provides an electronic device, including: a processor; and a memory for storing processor-executable instructions; wherein the processor is configured to perform the steps of the information processing method described in this example embodiment.

[0201] Figure 9 This is a schematic diagram of the structure of an electronic device according to an embodiment of this disclosure. It should be noted that... Figure 9 The electronic device 900 shown is merely an example and should not be construed as limiting the functionality and scope of use of the embodiments disclosed herein.

[0202] like Figure 9 As shown, the electronic device 900 includes a central processing unit (CPU) 901, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 902 or a program loaded from a storage section 908 into a random access memory (RAM) 903. The RAM 903 also stores various programs and data required for system operation. The CPU 901, ROM 902, and RAM 903 are interconnected via a bus 904. An input / output (I / O) interface 905 is also connected to the bus 904.

[0203] The following components are connected to I / O interface 905: an input section 906 including a keyboard, mouse, etc.; an output section 907 including a cathode ray tube (CRT), liquid crystal display (LCD), etc., and speakers, etc.; a storage section 908 including a hard disk, etc.; and a communication section 909 including a network interface card such as a local area network (LAN) card, modem, etc. The communication section 909 performs communication processing via a network such as the Internet. A drive 910 is also connected to I / O interface 905 as needed. A removable medium 911, such as a disk, optical disk, magneto-optical disk, semiconductor memory, etc., is installed on drive 910 as needed so that computer programs read from it can be installed into storage section 908 as needed.

[0204] In particular, according to embodiments of this disclosure, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments of this disclosure include a computer program product comprising a computer program carried on a computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via communication section 909, and / or installed from removable medium 911. When the computer program is executed by central processing unit 901, it performs various functions defined in the apparatus of this disclosure.

[0205] In this embodiment of the disclosure, a computer-readable storage medium is also provided, on which a computer program is stored, which, when executed by a processor, implements the above-described information processing method.

[0206] It should be noted that the computer-readable storage medium disclosed herein can be, for example,—but not limited to—an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples of computer-readable storage media may include, but are not limited to: electrical connections having one or more wires, portable computer disks, hard disks, random access memory, read-only memory, erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof. In this disclosure, a computer-readable storage medium can be any tangible medium containing or storing a program that can be used by or in connection with an instruction execution system, apparatus, or device. Program code contained on a computer-readable storage medium can be transmitted using any suitable medium, including but not limited to: wireless, wire, optical fiber, radio frequency, etc., or any suitable combination thereof.

[0207] In this embodiment of the disclosure, a computer program product is also provided, which, when run on a computer, causes the computer to execute the above-described information processing method.

[0208] It should be noted that, in this document, relational terms such as "first" and "second" are used merely to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0209] The above description is merely a specific embodiment of this disclosure, enabling those skilled in the art to understand or implement it. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of this disclosure. Therefore, this disclosure is not to be limited to the embodiments described herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. An information processing method, characterized in that, The method includes: Receive live streaming information from a live streaming room input through an information input interface; the live streaming information from the live streaming room includes at least one of resource information, streamer information, and live streaming time information; Using a pre-trained live streaming room classification model, the live streaming information of the live streaming room is processed to determine the type of the live streaming room; the types of live streaming rooms include first-class live streaming rooms, second-class live streaming rooms, and third-class live streaming rooms; Based on the type of the live streaming room, a resource transfer information prediction model corresponding to the live streaming room is determined. The resource transfer information prediction model includes a first tree model, a first time series model, a second tree model, and a second time series model. The first type of live streaming room corresponds to the first tree model, the second type of live streaming room corresponds to the first time series model, and the third type of live streaming room corresponds to the second tree model and the second time series model. The live stream information is processed by the resource transfer information prediction model corresponding to the live stream room to obtain initial resource transfer prediction information.

2. The method according to claim 1, characterized in that, After obtaining the initial resource transfer prediction information, the method further includes: Obtain information about the resources and the broadcaster from the live stream information; Based on the information of the resources and the information of the broadcaster, the relevance of the broadcaster resources is determined; Based on the initial resource transfer prediction information and the relevance of the streamer resources, the revised resource transfer prediction information is determined.

3. The method according to claim 1 or 2, characterized in that, Based on the live streaming information of the live streaming room, the type of the live streaming room is determined, including: The live streaming information of the live streaming room is input into a pre-trained live streaming room classification model; The probability that the target resource transfer value of the live room is higher than a preset resource transfer threshold is predicted using the live room classification model. Based on the prediction results, determine whether the live streaming room belongs to the first type of live streaming room.

4. The method according to claim 3, characterized in that, The method further includes: If it is determined that the live room does not belong to the first type of live room, the number of objects associated with the streamer is obtained from the live information, and the time difference between the start time of the live broadcast and the current time is determined according to the live broadcast time information in the live information. The type of the live stream room is determined based on the number of objects associated with the streamer and the time difference.

5. The method according to claim 4, characterized in that, The step of determining the type of the live stream room based on the number of objects associated with the streamer and the time difference includes: If the number of objects associated with the streamer is less than the preset number of objects, or the time difference is less than the time difference threshold, the live stream room is determined to belong to the second type of live stream room. If the number of objects associated with the streamer is greater than or equal to the preset number of objects, and the time difference is greater than or equal to the time difference threshold, the live stream room is determined to belong to the third type of live stream room.

6. The method according to claim 5, characterized in that, The step of determining the resource transfer information prediction model corresponding to the live stream room based on its type includes: If the live stream room belongs to the first type of live stream room, the resource transfer information prediction model corresponding to the live stream room is determined to be the pre-trained first tree model. If the live stream room belongs to the second type of live stream room, the resource transfer information prediction model corresponding to the live stream room is determined to be the pre-trained first time series model. If the live stream room belongs to the third type of live stream room, the resource transfer information prediction model corresponding to the live stream room is determined to be the pre-trained second tree model and second time series model.

7. The method according to claim 6, characterized in that, If the resource transfer information prediction model corresponding to the live broadcast room is a pre-trained second tree model and second time series model, the live broadcast information also includes: multiple historical resource transfer information of the broadcaster; The live broadcast information is processed using the second tree model and the second time series model to obtain initial resource transfer prediction information, including: The first estimated information is obtained by processing multiple historical resource transfer information of the anchor through the second time series model. The second tree model is used to process the information in the live broadcast information other than the multiple historical resource transfer information to obtain the second estimated information; Initial resource transfer forecast information is determined based on the first forecast information and the second forecast information.

8. The method according to claim 6, characterized in that, If the resource transfer information prediction model corresponding to the live broadcast room is a pre-trained first tree model, the first tree model includes multiple sub-models; The live stream information is processed using the first tree model to obtain initial resource transfer prediction information, including: The live broadcast information is processed by the multiple sub-models respectively to obtain the indicator data corresponding to each sub-model; Based on multiple indicator data, the initial resource transfer forecast information is determined.

9. The method according to claim 2, characterized in that, The method further includes: Receive promotional attribute information input through the information input interface; Obtain the promotion conversion information corresponding to the streamer from the live broadcast information; Based on the promotion attribute information and the promotion conversion information, determine the estimated information for the transfer of additional resources; The final resource transfer forecast information is determined based on the revised resource transfer forecast information and the additional resource transfer forecast information.

10. The method according to claim 9, characterized in that, The method further includes: The final resource transfer estimate information, the revised resource transfer estimate information, and the additional resource transfer estimate information are displayed; the final resource transfer estimate information includes the final resource transfer estimate value. If the difference between the reported resource transfer value and the estimated final resource transfer value is greater than a preset difference, a prompt message is determined and displayed based on the estimated final resource transfer value and the reported resource transfer value. The prompt message is used to indicate adjustments to the live broadcast information and / or the promotion attribute information.

11. The method according to claim 10, characterized in that, The method further includes: Display the live stream information and the promotion attribute information; In response to a modification operation on the live stream information, the live stream information is modified to obtain modified live stream information; and / or In response to the modification operation on the promotion attribute information, the promotion attribute information is modified to obtain the modified promotion attribute information; Based on the modified live streaming information and / or the modified promotion attribute information, the final resource transfer estimate is redefined so that the difference between the reported resource transfer value and the redefined final resource transfer estimate is less than or equal to the preset difference.

12. An information processing device, characterized in that, The device includes: The live streaming information receiving module is used to receive live streaming information of the live streaming room input through the information input interface; the live streaming information of the live streaming room includes at least one of the following: resource information, anchor information, and live streaming time information; The first live room type determination module is used to process the live room information using a pre-trained live room classification model to determine the type of the live room; the types of live rooms include first type live room, second type live room and third type live room; The resource transfer information prediction model determination module is used to determine the resource transfer information prediction model corresponding to the live room according to the type of the live room. The resource transfer information prediction model includes a first tree model, a first time series model, a second tree model, and a second time series model. The first type of live room corresponds to the first tree model, the second type of live room corresponds to the first time series model, and the third type of live room corresponds to the second tree model and the second time series model. The initial resource transfer prediction information determination module is used to process the live broadcast information through the resource transfer information prediction model corresponding to the live broadcast room to obtain the initial resource transfer prediction information.

13. An electronic device, characterized in that, include: A processor for executing a computer program stored in a memory, wherein the computer program, when executed by the processor, implements the steps of the method according to any one of claims 1-11.

14. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1-11.

15. A computer program product, characterized in that, When the computer program product is run on a computer, it causes the computer to perform the steps of the method according to any one of claims 1-11.

Images