Method and device for training fatigue prediction model and predicting fatigue

Abstract

The present disclosure provides a method and apparatus for training a fatigue prediction model and predicting fatigue, which relate to artificial intelligence fields such as deep learning, large models, natural language processing, and intelligent recommendation. The method for training the fatigue prediction model may include: obtaining historical consumption records corresponding to a predetermined resource recommendation application, where the historical consumption records include: operation information of a recommended object who opens the resource recommendation application for the recommended resource object; generating a training sample set according to the historical consumption records; training the fatigue prediction model according to the training sample set, and the fatigue prediction model is used to determine a fatigue prediction result of the target recommended object for the candidate resource to be recommended when recommending a resource to the target recommended object to be processed.

Description

Technical Field

[0001] The present disclosure relates to the field of artificial intelligence technology, and particularly to the training of a fatigue prediction model and the methods and devices for fatigue prediction in the fields of deep learning, large models, natural language processing, and intelligent recommendation, etc. Background Art

[0002] Currently, short video platforms are favored by more and more users. When recommending videos to users, it may involve multiple stages such as recall, rough ranking, fine ranking, and re-ranking. Among them, for the multi-objective fusion operation in the fine ranking stage, the modeled objectives can be divided into two categories: positive feedback objectives and negative feedback objectives. Among them, positive feedback objectives include satisfaction, completion rate, playback completion, etc., while the research on modeling negative feedback objectives is relatively less, thus affecting the accuracy of the fusion result, etc. Summary of the Invention

[0003] The present disclosure provides a method and device for training a fatigue prediction model and a method and device for fatigue prediction.

[0004] A method for training a fatigue prediction model includes:

[0005] Obtaining the historical consumption records corresponding to a predetermined resource recommendation application, where the historical consumption records include: the operation information of the recommended object who opens the resource recommendation application for the recommended resource object;

[0006] Generating a training sample set according to the historical consumption records;

[0007] Training the fatigue prediction model according to the training sample set, where the fatigue prediction model is used to determine the fatigue prediction result of the target recommendation object for the candidate resources to be recommended when recommending resources to the target recommendation object to be processed.

[0008] A method for fatigue prediction includes:

[0009] For a target recommendation object who opens a resource recommendation application, determining the candidate resources to be recommended;

[0010] Using the fatigue prediction model to respectively determine the fatigue prediction results of the target recommendation object for each candidate resource, where the fatigue prediction model is trained according to a training sample set, and the training sample set is generated according to the historical consumption records corresponding to the resource recommendation application, and the historical consumption records include: the operation information of the recommended object who opens the resource recommendation application for the recommended resource object.

[0011] A device for training a fatigue prediction model includes: an information acquisition module, a sample construction module, and a model training module;

[0012] The information acquisition module is configured to acquire the historical consumption records corresponding to a predetermined resource recommendation application, where the historical consumption records include: the operation information of the recommended object who opens the resource recommendation application for the recommended resource object;

[0013] The sample construction module is configured to generate a training sample set according to the historical consumption records;

[0014] The model training module is configured to train the fatigue feeling prediction model according to the training sample set, and the fatigue feeling prediction model is used to determine the fatigue feeling prediction result of the target recommendation object for the candidate resources to be recommended when recommending resources to the target recommendation object to be processed.

[0015] A fatigue feeling prediction device includes: a resource determination module and a result determination module;

[0016] The resource determination module is configured to determine the candidate resources to be recommended for the target recommendation object who opens the resource recommendation application;

[0017] The result determination module is configured to use the fatigue feeling prediction model to respectively determine the fatigue feeling prediction results of the target recommendation object for each candidate resource. The fatigue feeling prediction model is obtained by training according to the training sample set, and the training sample set is generated according to the historical consumption records corresponding to the resource recommendation application. The historical consumption records include: the operation information of the recommended object who opens the resource recommendation application for the recommended resource object.

[0018] An electronic device includes:

[0019] At least one processor; and

[0020] A memory communicatively connected to the at least one processor; wherein,

[0021] The memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the method as described above.

[0022] A non-transitory computer-readable storage medium storing computer instructions, where the computer instructions are used to cause a computer to execute the method as described above.

[0023] A computer program product includes computer programs / instructions, and when the computer programs / instructions are executed by a processor, the method as described above is implemented.

[0024] It should be understood that the content described in this section is not intended to identify the key or important features of the embodiments of the present disclosure, nor is it used to limit the scope of the present disclosure. Other features of the present disclosure will become readily understood through the following description. Description of the Drawings

[0025] The drawings are used to better understand the solution and do not constitute a limitation to the present disclosure. Among them:

[0026] Figure 1 is a flowchart of an embodiment of the fatigue feeling prediction model training method described in the present disclosure;

[0027] Figure 2 is a flowchart of an embodiment of the fatigue feeling prediction method described in the present disclosure;

[0028] Figure 3 is a schematic diagram of the overall implementation process of the fatigue feeling prediction model training and fatigue feeling prediction methods described in the present disclosure;

[0029] Figure 4 is a schematic diagram of the composition structure of Embodiment 400 of the fatigue feeling prediction model training device described in the present disclosure;

[0030] Figure 5 is a schematic diagram of the composition structure of Embodiment 500 of the fatigue feeling prediction device described in the present disclosure;

[0031] Figure 6 Shows a schematic block diagram of an electronic device 600 that can be used to implement the embodiments of the present disclosure. Detailed Embodiments

[0032] The following describes exemplary embodiments of the present disclosure in conjunction with the drawings. Various details of the embodiments of the present disclosure are included to assist in understanding, and they should be considered merely exemplary. Therefore, those of ordinary skill in the art should recognize that various changes and modifications can be made to the embodiments described herein without departing from the scope and spirit of the present disclosure. Similarly, for the sake of clarity and conciseness, descriptions of well-known functions and structures are omitted in the following description.

[0033] In addition, it should be understood that the term "and / or" herein is merely a description of the association relationship of associated objects, indicating that three relationships can exist. For example, A and / or B can represent: A exists alone, A and B exist simultaneously, and B exists alone. In addition, the character " / " herein generally represents an "or" relationship between the associated objects before and after.

[0034] Figure 1 is a flowchart of an embodiment of the fatigue feeling prediction model training method described in the present disclosure. As Figure 1 shown, it includes the following specific implementation manners.

[0035] In step 101, obtain the historical consumption records corresponding to a predetermined resource recommendation application. The historical consumption records include: operation information of the recommended object who opens the resource recommendation application for the recommended resource object.

[0036] In step 102, generate a training sample set according to the historical consumption records.

[0037] In step 103, train a fatigue prediction model according to the training sample set. The fatigue prediction model is used to determine the fatigue prediction result of the target recommended object for the candidate resource to be recommended when recommending resources to the target recommended object to be processed.

[0038] Adopting the solution described in the above method embodiment, the fatigue prediction model can be trained using the training sample set. Furthermore, when recommending resources to the target recommended object, the trained fatigue prediction model can be used to determine the fatigue prediction result of the target recommended object for the candidate resource to be recommended. The fatigue prediction result can be used as a negative feedback target, thus enriching the content of the existing negative feedback target and further improving the accuracy of the fusion result, etc.

[0039] The recommended object usually refers to a user. Correspondingly, the operation information refers to user behavior information.

[0040] For a predetermined resource recommendation application, historical consumption records can be obtained. Generally speaking, after a recommended object opens the resource recommendation application, resource objects will be recommended to the recommended object. Among them, for each recommended resource object, a corresponding consumption record will be generated respectively. The consumption record can include the operation information of the recommended object for the resource object. The specific content included in the operation information can be determined according to actual needs.

[0041] The historical consumption record refers to the consumption record generated before the current time. In addition, obtaining the historical consumption record can refer to obtaining the historical consumption record generated within a recently predetermined time period. The specific value of the recently predetermined time period can be determined according to actual needs.

[0042] According to the obtained historical consumption records, a training sample set corresponding to the fatigue prediction model can be generated. The training sample set includes positive samples and negative samples.

[0043] In some embodiments of the present disclosure, the resource recommendation application may include: a short video platform. Correspondingly, the resource object type may be a video type. Each historical consumption record may respectively correspond to a recommended object and a first video recommended. The method for generating a training sample set based on historical consumption records may include: in response to determining that there is a second video in the first video that meets any one of the first negative sample screening condition, the second negative sample screening condition, and the third negative sample screening condition, determining that the negative sample video includes the second video; in response to determining that there is a third video in the first video that meets the positive sample screening condition, determining that the positive sample video includes the third video; generating negative samples in the training sample set based on the negative sample videos, and generating positive samples in the training sample set based on the positive sample videos.

[0044] That is, for the first video corresponding to each historical consumption record, negative sample videos and positive sample videos can be respectively screened out from it, and then negative samples and positive samples in the training sample set can be respectively generated based on the negative sample videos and positive sample videos. The whole process is simple and convenient to implement, thus laying a good foundation for subsequent processing.

[0045] Among them, the screening methods for negative sample videos may include Method 1, Method 2, and Method 3, which are introduced respectively below.

[0046] 1) Method 1

[0047] In some embodiments of the present disclosure, any fourth video (the video corresponding to any historical consumption record) may be determined in the first video, and the first playback duration of the fourth video may be determined according to the corresponding historical consumption record. In response to determining that the first playback duration is less than the first threshold, it may be determined that the fourth video meets the first negative sample screening condition, and then the fourth video may be determined as a negative sample video.

[0048] The historical consumption record will include the playback duration information of the video, that is, the viewing duration information of the recommended object for the video. If it is determined that the playback duration of the fourth video is less than the first threshold, it can be determined that the fourth video meets the first negative sample screening condition, and the fourth video can be determined as a negative sample video.

[0049] The specific value of the first threshold may be determined according to actual needs. For example, if the playback duration of the fourth video is less than 3 seconds, the fourth video can be considered a fast-scrolling video. Fast scrolling is usually caused by the recommended object not being interested in the fourth video. Therefore, the fourth video in this case can be determined as a negative sample video.

[0050] 2) Method 2

[0051] In some embodiments of the present disclosure, any fourth video may be determined in a first video, and a sixth video whose video duration belongs to the same duration interval as that of the fourth video may be obtained from a fifth video. The fifth video includes other videos in the first video except the fourth video. The target playback duration of each video in the video set is determined according to the corresponding historical consumption record. The video set includes the fourth video and the sixth video. That is, the sixth video whose video duration belongs to the same duration interval as that of the fourth video may be selected from the videos corresponding to other historical consumption records. The selected sixth video and the fourth video are used to form a video set. Then, the target playback duration of each video in the video set may be determined according to the corresponding historical consumption record, and the videos in the video set may be sorted in descending order of the target playback duration. Correspondingly, in response to determining that there is a skip operation for the fourth video and determining that the sorting position serial number of the fourth video is greater than the product of M and x, it may be determined that the fourth video meets the second negative sample screening condition, and the fourth video may be determined as a negative sample video. M represents the number of videos in the video set, and x represents a first percentage greater than 0 and less than 1. The existence of a skip operation means that the recommended object corresponding to the historical consumption record exits the short video platform when playing the fourth video.

[0052] According to the video durations of the videos, multiple different duration intervals may be preset in advance. There is no overlap between any two duration intervals, and moreover, the video duration of each video belongs to one duration interval respectively. The video duration is usually in seconds as the unit.

[0053] Suppose 99 sixth videos are selected in total. Then, the 99 sixth videos and the fourth video may be used to form a video set. Then, the 100 videos in the video set may be sorted in descending order of the playback duration recorded in the corresponding historical consumption records of the 100 videos. Suppose the value of the first percentage is 20%. Then, if it is determined that the recommended object corresponding to the fourth video exits the short video platform when playing the fourth video, and it is determined that the sorting position of the fourth video is after the 20th (100 * 20%) position, then it may be determined that the fourth video meets the second negative sample screening condition, and the fourth video may be determined as a negative sample video.

[0054] The negative sample video that meets the second negative sample screening condition may be called a confident skip video. In other words, a confident skip video means a video with a skip operation and non-satisfactory distribution. Satisfactory distribution may be defined as: ranking in the TOP 20% (i.e., the top 20%) in terms of playback duration among all videos with video durations belonging to the same duration interval. Correspondingly, non-satisfactory distribution means being in the last 80%.

[0055] 3) Method 3

[0056] In some embodiments of the present disclosure, any fourth video can be determined in the first video, and the video set can be obtained and the videos in the video set can be sorted in the manner described in 2). Additionally, the number of videos belonging to the same category as the fourth video that have been recommended before the fourth video in the same session can be counted. The session refers to the consumption process of the recommended object corresponding to the historical consumption record from opening the short video platform to exiting the short video platform. Correspondingly, in response to determining that the following conditions are simultaneously met: the number of videos is greater than the second threshold, the sorting position serial number of the fourth video is greater than the product of M and y, and the video duration of the fourth video is less than the third threshold, it can be determined that the fourth video meets the third negative sample screening condition, and the fourth video can be determined as a negative sample video. y represents a second percentage greater than 0 and less than 1.

[0057] The specific values of the second threshold, the third threshold, and the second percentage can all be determined according to actual needs.

[0058] A session refers to the consumption process of the recommended object from opening the short video platform to exiting the short video platform. The consumption process refers to the complete session process. During this process, the recommended object may have watched multiple recommended videos and generated corresponding operation information for each of the recommended videos.

[0059] Additionally, how to divide each video into different categories and what specific categories are included can also be determined according to actual needs. For example, the categories can include sports, entertainment, military, etc.

[0060] Assume that the recommended object corresponding to the fourth video is recommended object a, and assume that the value of the second threshold is 3, the value of the third threshold is 100 seconds, and the second percentage is 50%. Then, the number of videos belonging to the same category as the fourth video that have been recommended to recommended object a before the fourth video in the same session can be counted. Assume it is 4, and assume that the sorting position serial number of the fourth video is greater than M * 50%. At the same time, assume that the video duration of the fourth video is less than 100 seconds. Then, since the following conditions are simultaneously met: the number of videos 4 is greater than 3, the sorting position serial number of the fourth video is greater than M * 50%, and the video duration of the fourth video is less than 100 seconds, it can be determined that the fourth video meets the third negative sample screening condition, and the fourth video can be determined as a negative sample video.

[0061] The negative sample video that meets the third negative sample screening condition can be called a fatigue video. In other words, a fatigue video refers to a video in the same session that has consumed more than the second threshold of videos in the same category and is not satisfactory. Not satisfactory can be defined as: the playback duration ranks after the TOP 50% among all videos with video durations in the same duration interval and the video duration is less than 100 seconds.

[0062] By using the processing methods in the above-mentioned Method 1, Method 2, and Method 3, fast-scrolling videos, confidence-jumping videos, and fatigue videos can all be determined as negative sample videos. That is, significant negative operations such as fast-scrolling operations, jumping operations, and fatigue operations after multiple consumptions of videos of the same category within the same session can be combined to determine negative sample videos, thereby realizing the joint modeling of multiple negative operations, enriching the expression ability of fatigue, enhancing the learning ability of the model, and correspondingly improving the accuracy of the fatigue prediction results generated by the subsequent model utilization, etc.

[0063] In addition, in some embodiments of the present disclosure, any fourth video can be determined in the first video, and the video set can be obtained and the videos in the video set can be sorted in the manner described in 2). Correspondingly, in response to determining that no predetermined negative operation has occurred for the fourth video and determining that the sorting position serial number of the fourth video is less than the product of M and z, it can be determined that the fourth video meets the positive sample screening conditions, and the fourth video can be determined as a positive sample video, where z represents a third percentage greater than 0 and less than 1.

[0064] The specific value of the third percentage can be determined according to actual needs, such as 50%. If it is determined that no predetermined negative operation has occurred for the fourth video and it is determined that the sorting position serial number of the fourth video is less than the product of M * 50%, it can be determined that the fourth video meets the positive sample screening conditions, and the fourth video can be determined as a positive sample video. The predetermined negative operation can include the aforementioned fast-scrolling operation, jumping operation, and fatigue operation, etc.

[0065] Based on the determined negative sample videos, negative samples can be generated. In addition, through the above processing, positive sample videos can be determined from the videos corresponding to each historical consumption record, and positive samples can be generated based on the positive sample videos. Furthermore, the model can be trained by combining the positive samples and negative samples, thereby improving the training accuracy of the model, and further improving the accuracy of the fatigue prediction results generated by the subsequent model utilization, etc.

[0066] In addition, as can be seen from the above introduction, the fourth video can be any video in the first video. When the fourth video meets the first negative sample screening condition, the second negative sample screening condition, or the third negative sample screening condition, the fourth video is the second video. When the fourth video meets the positive sample screening conditions, the fourth video is the third video.

[0067] Correspondingly, in some embodiments of the present disclosure, for each negative sample video, a corresponding negative sample may be generated respectively. Each negative sample may respectively include: the first feature information corresponding to the corresponding negative sample video and the first label, where the first label is used to indicate that the corresponding sample is a negative sample. In addition, for each positive sample video, a corresponding positive sample may be generated respectively. Each positive sample may respectively include: the second feature information corresponding to the corresponding positive sample video and the second label, where the second label is used to indicate that the corresponding sample is a positive sample.

[0068] In practical applications, before generating positive and negative samples, optimization processing may also be performed on the determined negative sample videos and positive sample videos. For example, according to the ratio requirements of positive and negative samples, some negative sample videos or some positive sample videos may be discarded.

[0069] Furthermore, for each negative sample video, a corresponding negative sample may be generated respectively, and for each positive sample video, a corresponding positive sample may be generated respectively. Each negative sample respectively includes the first feature information and the first label, and each positive sample respectively includes the second feature information and the second label. For example, the value of the first label may be 0, and the value of the second label may be 1, or vice versa, as long as the model can understand the meaning of the corresponding label.

[0070] In some embodiments of the present disclosure, the acquisition methods of the first feature information and the second feature information may be the same. For example, it may include: for a target video, where the target video is any negative sample video or any positive sample video, acquiring the object feature of the recommended object corresponding to the target video, the recommended object corresponding to the target video being the recommended object corresponding to the same historical consumption record as the target video, acquiring the video feature of the target video, acquiring the cross feature between the target video and the recommended object corresponding to the target video, acquiring the request feature, where the request is the request that triggers the recommendation of the target video, and acquiring the historical session feature of the recommended object corresponding to the target video. The session refers to the consumption process of the recommended object corresponding to the target video from opening the short video platform to exiting the short video platform. Determining the acquired object feature, video feature, cross feature, request feature, and historical session feature as the feature information corresponding to the target video.

[0071] The above-mentioned target video may refer to a negative sample video or a positive sample video, and the corresponding feature information acquisition methods are the same.

[0072] The specific content included in each of the various feature information may be determined according to actual needs. In addition, the specific acquisition methods of the various feature information may also be determined according to actual needs.

[0073] It should be noted that various feature information and historical consumption records involved in the solutions described in this disclosure can be obtained through various public, legal and compliant means. For example, they can be obtained from the user after the user's authorization, that is, they can be obtained with the user's knowledge and consent. In the technical solutions of this disclosure, the collection, storage, use, processing, transmission, provision, and disclosure of the user's personal information comply with the provisions of relevant laws and regulations and do not violate public order and good customs.

[0074] Among them, the object features may include: the age, gender, education level, life stage (such as going to school, working, etc.), activity level on the short video platform (such as high, medium, low), and whether it is a new user of the recommended object, etc.

[0075] The video features may include: the video length of the target video, the category to which the target video belongs, the semantic features of the category to which it belongs, and the overall market posterior probability of the category to which it belongs (such as average satisfaction, average completion rate, average play duration, etc.).

[0076] The cross features may include: whether the category to which the target video belongs is a category that the recommended object is interested in, etc. For example, the categories of interest can be determined based on the user profile of the recommended object.

[0077] The request may refer to the user performing a swiping up operation on the screen, etc. Correspondingly, the request features may include: the number of the request in the same session that triggers the recommendation of the target video, the device information used by the recommended object (such as mobile phone or tablet, etc.), the time stage when the request to trigger the recommendation of the target video is sent (such as morning, evening, etc.), and the date type when the request to trigger the recommendation of the target video is sent (such as working day, holiday, etc.).

[0078] The historical session features may include: some features determined based on each historical session of the recommended object, such as the user's historical interaction situation, user engagement, etc.

[0079] It can be seen that the above feature information includes features in different dimensions such as the recommended object, video, request, and historical session, thus enriching the content of the feature information. Moreover, in practical applications, in addition to the feature information such as object features, video features, cross features, request features, and historical session features, the positive and negative samples may further include some other features as needed, which is very flexible and convenient. For example, it may further include the features used in the existing methods for predicting feedback targets such as satisfaction, completion rate, and fast swiping.

[0080] The generated positive samples and negative samples can be used to form a training sample set, and the training sample set can be used to train the fatigue prediction model. Among them, the fatigue prediction model can adopt the structure of Extreme Gradient Boosting (XGBoost).

[0081] After the training of the fatigue prediction model is completed, it can be applied to actual fatigue prediction. This will be illustrated by the following embodiments.

[0082] Figure 2 This is a flowchart of an embodiment of the fatigue prediction method described in the present disclosure. As Figure 2 shown, it includes the following specific implementation manners.

[0083] In step 201, for the target recommendation object who opens the resource recommendation application, the candidate resources to be recommended are determined.

[0084] In step 202, the fatigue prediction results of the target recommendation object for each candidate resource are respectively determined by using the fatigue prediction model. The fatigue prediction model is obtained by training according to the training sample set, and the training sample set is generated according to the historical consumption records corresponding to the resource recommendation application. The historical consumption records include: the operation information of the recommendation object who opens the resource recommendation application for the recommended resource object.

[0085] Adopting the solution described in the above method embodiment, when recommending resources to the target recommendation object, the fatigue prediction results of the candidate resources to be recommended can be determined by using the fatigue prediction model. The fatigue prediction results can be used as negative feedback targets, thus enriching the content of the existing negative feedback targets, and further improving the accuracy of the fusion results, etc.

[0086] In some embodiments of the present disclosure, for any candidate resource, the third feature information corresponding to the candidate resource can be obtained, and the third feature information can be input into the fatigue prediction model to obtain the output fatigue prediction result.

[0087] Among them, the fatigue prediction result can be 1 or 0. The fatigue prediction model can be the fatigue prediction model trained by the method in the embodiment shown as Figure 1 shown.

[0088] In addition, in some embodiments of the present disclosure, the resource recommendation application may include: a short video platform. Correspondingly, the resource object type may be a video type. For any candidate video, the method for obtaining the third feature information corresponding to the candidate resource may include: obtaining the object features of the target recommendation object, obtaining the video features of the candidate video, obtaining the cross features between the candidate video and the target recommendation object, obtaining the request features, where the request is a request that triggers the recommendation of the candidate video, obtaining the historical session features of the target recommendation object, and a session refers to the consumption process of the target recommendation object from opening the short video platform to exiting the short video platform. The object features, video features, cross features, request features, and historical session features are determined as the third feature information corresponding to the candidate video.

[0089] Combined with the above introduction, Figure 3 is a schematic diagram of the overall implementation process of the fatigue feeling prediction model training and fatigue feeling prediction method described in the present disclosure. As Figure 3 shown, for a certain short video platform, historical consumption records can be obtained. Among them, for any historical consumption record, in response to determining that the video corresponding to the historical consumption record meets any one of the first negative sample screening condition, the second negative sample screening condition, and the third negative sample screening condition, the video corresponding to the historical consumption record can be determined as a negative sample video. For any historical consumption record, in response to determining that the video corresponding to the historical consumption record meets the positive sample screening condition, the video corresponding to the historical consumption record can be determined as a positive sample video. Then, the first feature information corresponding to each negative sample video and the second feature information corresponding to each positive sample video can be obtained respectively through feature extraction and the like, and negative samples and positive samples can be generated according to the extracted feature information and the like to form a training sample set. Furthermore, the fatigue feeling prediction model can be trained using the training sample set. After the training is completed, for any target recommendation object that opens the short video platform, when it is necessary to recommend videos to it, for each determined candidate resource, feature extraction can be performed respectively, and the fatigue feeling prediction results of each candidate resource can be determined using the extracted third feature information and the fatigue feeling prediction model. Furthermore, subsequent related processing, such as multi-object fusion, can be performed using the fatigue feeling prediction results, and finally the recommended video can be determined and recommended to the target recommendation object. Further, new historical consumption records can also be generated according to the operation information of the target recommendation object for the recommended video and the like for subsequent optimization of the fatigue feeling prediction model and the like.

[0090] It should be noted that for the foregoing method embodiments, for the sake of simple description, they are all expressed as a series of action combinations. However, those skilled in the art should be aware that the present disclosure is not limited by the described action sequence, because according to the present disclosure, certain steps can be performed in other sequences or simultaneously. Secondly, those skilled in the art should also be aware that the embodiments described in the specification are all preferred embodiments, and the actions and modules involved are not necessarily essential to the present disclosure. In addition, for the parts not detailed in a certain embodiment, reference can be made to the relevant descriptions in other embodiments.

[0091] The above is the introduction to the method embodiments. The following further illustrates the solution of the present disclosure through device embodiments.

[0092] Figure 4 It is a schematic structural diagram of the composition of the fatigue feeling prediction model training device embodiment 400 according to the present disclosure. As Figure 4 shown, it includes: an information acquisition module 401, a sample construction module 402, and a model training module 403.

[0093] The information acquisition module 401 is configured to acquire the historical consumption records corresponding to a predetermined resource recommendation application, and the historical consumption records include: the operation information of the recommended object for opening the resource recommendation application with respect to the recommended resource object.

[0094] The sample construction module 402 is configured to generate a training sample set according to the historical consumption records.

[0095] The model training module 403 is configured to train the fatigue feeling prediction model according to the training sample set. The fatigue feeling prediction model is used to determine the fatigue feeling prediction result of the target recommended object for the candidate resource to be recommended when recommending resources to the target recommended object to be processed.

[0096] By adopting the solution described in the above device embodiment, the fatigue feeling prediction model can be trained by using the training sample set. Furthermore, when recommending resources to the target recommended object, the fatigue feeling prediction result of the candidate resource to be recommended can be determined by using the trained fatigue feeling prediction model. The fatigue feeling prediction result can be used as a negative feedback target, thereby enriching the content of the existing negative feedback target and further improving the accuracy of the fusion result, etc.

[0097] In some embodiments of the present disclosure, the resource recommendation application may include: a short video platform. Correspondingly, the resource object type may be a video type. Each historical consumption record respectively corresponds to a recommended object and a first video recommended thereby. The manner in which the sample construction module 402 generates a training sample set according to the historical consumption records may include: in response to determining that there is a second video in the first video that meets any one of the first negative sample screening condition, the second negative sample screening condition, and the third negative sample screening condition, determining that the negative sample video includes the second video; in response to determining that there is a third video in the first video that meets the positive sample screening condition, determining that the positive sample video includes the third video; generating negative samples in the training sample set according to the negative sample videos; and generating positive samples in the training sample set according to the positive sample videos.

[0098] In some embodiments of the present disclosure, the sample construction module 402 may determine any fourth video in the first video, and may determine the first playback duration of the fourth video according to the corresponding historical consumption record. In response to determining that the first playback duration is less than the first threshold, it may be determined that the fourth video meets the first negative sample screening condition, and further the fourth video may be determined as a negative sample video.

[0099] In some embodiments of the present disclosure, for any historical consumption record, the sample construction module 402 may determine any fourth video in the first video, and may obtain a sixth video from the fifth videos whose video duration belongs to the same duration interval as that of the fourth video. The fifth videos include other videos in the first video except the fourth video. The target playback durations of the videos in the video set are respectively determined according to the corresponding historical consumption records. The video set includes the fourth video and the sixth video. The videos in the video set are sorted in descending order according to the target playback durations. Correspondingly, in response to determining that there is a jump-out operation for the fourth video, and determining that the sorting position serial number of the fourth video is greater than the product of M and x, it may be determined that the fourth video meets the second negative sample screening condition, and the fourth video may be determined as a negative sample video. M represents the number of videos in the video set, and x represents a first percentage greater than 0 and less than 1. The existence of a jump-out operation means that the recommended object corresponding to the historical consumption record exits the short video platform when playing the fourth video.

[0100] In some embodiments of the present disclosure, the sample construction module 402 may further count the number of videos belonging to the same category as the fourth video that have been recommended before the fourth video in the same session. The session refers to the consumption process of the recommended object corresponding to the historical consumption record from opening the short video platform to exiting the short video platform. Accordingly, in response to determining that the following conditions are met simultaneously: the number of videos is greater than a second threshold, the sorting position serial number of the fourth video is greater than the product of M and y, and the video duration of the fourth video is less than a third threshold, it can be determined that the fourth video meets the third negative sample screening condition, and the fourth video can be determined as a negative sample video, where y represents a second percentage greater than 0 and less than 1.

[0101] In addition, in some embodiments of the present disclosure, the sample construction module 402, in response to determining that no predetermined negative operation has occurred for the fourth video and determining that the sorting position serial number of the fourth video is less than the product of M and z, determines that the fourth video meets the positive sample screening condition, and the fourth video can be determined as a positive sample video, where z represents a third percentage greater than 0 and less than 1.

[0102] Further, in some embodiments of the present disclosure, for each negative sample video, the sample construction module 402 may respectively generate a corresponding negative sample. Each negative sample may respectively include: the first feature information corresponding to the corresponding negative sample video and a first label, where the first label is used to indicate that the corresponding sample is a negative sample. In addition, for each positive sample video, a corresponding positive sample may be respectively generated. Each positive sample may respectively include: the second feature information corresponding to the corresponding positive sample video and a second label, where the second label is used to indicate that the corresponding sample is a positive sample.

[0103] In some embodiments of the present disclosure, the acquisition methods of the first feature information and the second feature information may be the same. For example, the sample construction module 402 may perform the following processing: for a target video, where the target video is any negative sample video or any positive sample video, obtain the object feature of the recommended object corresponding to the target video. The recommended object corresponding to the target video is the recommended object corresponding to the same historical consumption record as the target video. Obtain the video feature of the target video, obtain the cross feature between the target video and the recommended object corresponding to the target video, obtain the request feature, where the request is the request that triggers the recommendation of the target video, obtain the historical session feature of the recommended object corresponding to the target video. The session refers to the consumption process of the recommended object corresponding to the target video from opening the short video platform to exiting the short video platform. Determine the obtained object feature, video feature, cross feature, request feature, and historical session feature as the feature information corresponding to the target video.

[0104] Figure 5 This is a schematic structural diagram of Embodiment 500 of the fatigue prediction device according to the present disclosure. As Figure 5 shown, it includes: a resource determination module 501 and a result determination module 502.

[0105] A resource determination module 501 is configured to determine candidate resources to be recommended for a target recommendation object for which resources are recommended for opening an application.

[0106] A result determination module 502 is configured to use a fatigue prediction model to respectively determine fatigue prediction results of the target recommendation object for each candidate resource. The fatigue prediction model is obtained by training according to a training sample set, and the training sample set is generated according to historical consumption records corresponding to the resource recommendation application. The historical consumption records include: operation information of a recommendation object for which resources are recommended for opening the resource recommendation application for the recommended resource object.

[0107] By adopting the solution described in the above device embodiment, when recommending resources to a target recommendation object, the fatigue prediction model can be used to determine the fatigue prediction result of the candidate resources to be recommended, and the fatigue prediction result can be used as a negative feedback target, thereby enriching the content of the existing negative feedback target, and further improving the accuracy of the fusion result, etc.

[0108] In some embodiments of the present disclosure, for any candidate resource, the result determination module 502 can respectively obtain third feature information corresponding to the candidate resource, and can input the third feature information into the fatigue prediction model to obtain the output fatigue prediction result.

[0109] In addition, in some embodiments of the present disclosure, the resource recommendation application may include: a short video platform. Correspondingly, the resource object type may be a video type. For any candidate video, the manner in which the result determination module 502 obtains the third feature information corresponding to the candidate resource may include: obtaining the object feature of the target recommendation object, obtaining the video feature of the candidate video, obtaining the cross feature between the candidate video and the target recommendation object, obtaining a request feature, where the request is a request for triggering the recommendation of the candidate video, obtaining the historical session feature of the target recommendation object, and the session refers to the consumption process of the target recommendation object from opening the short video platform to exiting the short video platform. Determining the object feature, video feature, cross feature, request feature, and historical session feature as the third feature information corresponding to the candidate video.

[0110] Figure 4 and Figure 5 The specific working process of the device embodiment shown can refer to the relevant description in the foregoing method embodiment and will not be elaborated herein.

[0111] In summary, by adopting the solution of the present disclosure, the content of the existing negative feedback target can be enriched, thereby improving the accuracy of the fusion result and the recommendation result, etc. Moreover, the solution of the present disclosure is applicable to various resource recommendation applications, that is, it has wide applicability.

[0112] The solution described in this disclosure can be applied to the field of artificial intelligence, particularly in the fields of deep learning, large models, natural language processing, and intelligent recommendation. Artificial intelligence is a discipline that studies how to make computers simulate certain thinking processes and intelligent behaviors of humans (such as learning, reasoning, thinking, planning, etc.). It has both hardware-level technologies and software-level technologies. Artificial intelligence hardware technologies generally include technologies such as sensors, dedicated artificial intelligence chips, cloud computing, distributed storage, and big data processing. Artificial intelligence software technologies mainly include several major directions such as computer vision technology, speech recognition technology, natural language processing technology, and machine learning / deep learning, big data processing technology, and knowledge graph technology.

[0113] According to an embodiment of the present disclosure, the present disclosure also provides an electronic device, a readable storage medium, and a computer program product.

[0114] Figure 6 FIG. shows a schematic block diagram of an electronic device 600 that can be used to implement the embodiments of the present disclosure. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital assistants, cellular phones, smart phones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present disclosure described and / or claimed herein.

[0115] As Figure 6 shown, the electronic device 600 includes a computing unit 601, which can execute various appropriate actions and processes according to a computer program stored in a read-only memory (ROM, Read-Only Memory) 602 or a computer program loaded from a storage unit 608 into a random access memory (RAM, Random Access Memory) 603. In the RAM 603, various programs and data required for the operation of the electronic device 600 can also be stored. The computing unit 601, the ROM 602, and the RAM 603 are connected to each other via a bus 604. An input / output (I / O, Input / Output) interface 605 is also connected to the bus 604.

[0116] Multiple components in the electronic device 600 are connected to the I / O interface 605, including: an input unit 606, such as a keyboard, a mouse, etc.; an output unit 607, such as various types of displays, speakers, etc.; a storage unit 608, such as a disk, an optical disc, etc.; and a communication unit 609, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 609 allows the electronic device 600 to exchange information / data with other devices via a computer network such as the Internet and / or various telecommunication networks.

[0117] The computing unit 601 can be various general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of the computing unit 601 include but are not limited to a central processing unit (CPU), a graphic processing unit (GPU), various dedicated artificial intelligence (AI) computing chips, various computing units running machine learning model algorithms, a digital signal processor (DSP), and any suitable processor, controller, microcontroller, etc. The computing unit 601 executes the various methods and processes described above, such as the methods described in the present disclosure. For example, in some embodiments, the methods described in the present disclosure can be implemented as a computer software program, which is tangibly contained in a machine-readable medium, such as the storage unit 608. In some embodiments, part or all of the computer program can be loaded and / or installed onto the electronic device 600 via the ROM 602 and / or the communication unit 609. When the computer program is loaded into the RAM 603 and executed by the computing unit 601, one or more steps of the methods described in the present disclosure can be executed. Alternatively, in other embodiments, the computing unit 601 can be configured to execute the methods described in the present disclosure by any other suitable means (e.g., by means of firmware).

[0118] The various embodiments of the systems and techniques described above in this document may be implemented in digital electronic circuitry, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard parts (ASSPs), system on chip systems (SOCs), complex programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments may include: being implemented in one or more computer programs that are executable and / or interpretable on a programmable system including at least one programmable processor, which may be a special-purpose or general-purpose programmable processor that receives data and instructions from, and transmits data and instructions to, a storage system, at least one input device, and at least one output device.

[0119] The program code for implementing the methods of the present disclosure may be written in any combination of one or more programming languages. These program codes may be provided to a processor or controller of a general purpose computer, special purpose computer, or other programmable data processing apparatus, such that the program codes, when executed by the processor or controller, cause the functions / operations specified in the flowchart and / or block diagram to be implemented. The program code may execute entirely on the machine, partly on the machine, as a stand-alone software package partly on the machine and partly on a remote machine or entirely on the remote machine or server.

[0120] In the context of this disclosure, a machine-readable medium can be a tangible medium that can contain or store a program for use by or in connection with an instruction execution system, apparatus, or device. A machine-readable medium can be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium can include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination of the foregoing. More specific examples of a machine-readable storage medium would include an electrical connection based on one or more wires, a portable computer diskette, a hard disk, a random access memory, a read-only memory, an erasable programmable read-only memory (EPROM), a flash memory, an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0121] To provide for interaction with a user, the systems and techniques described herein can be implemented on a computer having: a display device (e.g., a cathode ray tube (CRT) or a liquid crystal display (LCD) monitor) for displaying information to the user; and a keyboard and a pointing device (e.g., a mouse or a trackball) by which the user can provide input to the computer. Other kinds of devices can also be used to provide for interaction with the user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including acoustic input, voice input, or tactile input).

[0122] The systems and techniques described herein can be implemented in a computing system that includes back-end components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes front-end components (e.g., a user computer having a graphical user interface or a web browser through which the user can interact with an implementation of the systems and techniques described herein), or a computing system that includes any combination of such back-end components, middleware components, or front-end components. The components of the system can be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of a communication network include: a local area network (LAN), a wide area network (WAN), and the Internet.

[0123] A computer system may include a client and a server. The client and the server are generally far from each other and usually interact through a communication network. The client-server relationship is generated by computer programs running on the respective computers and having a client-server relationship with each other. The server may be a cloud server, a server of a distributed system, or a server incorporating a blockchain.

[0124] It should be understood that various forms of the processes shown above can be used, steps can be reordered, added or deleted. For example, the steps described in this disclosure can be executed in parallel, sequentially, or in a different order, as long as the desired results of the technical solutions disclosed in this disclosure can be achieved, and no limitations are imposed herein.

[0125] The above specific embodiments do not constitute a limitation on the protection scope of this disclosure. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this disclosure shall be included within the protection scope of this disclosure.

Claims

1. A method for training a fatigue prediction model, comprising: Obtaining historical consumption records corresponding to a predetermined resource recommendation application, where the historical consumption records include: operation information of a recommended object who opens the resource recommendation application for a recommended resource object; wherein, the resource recommendation application includes: a short video platform, the resource object type is a video type, and each historical consumption record corresponds to a recommended object and a first video recommended to the recommended object; Generating a training sample set according to the historical consumption records, where the training sample set includes negative samples and positive samples, wherein, in response to determining that there is a second video in the first video that meets any one of the first negative sample screening conditions, the second negative sample screening conditions, and the third negative sample screening conditions, determining that the negative sample video includes the second video, and generating the negative sample according to the negative sample video; Training the fatigue prediction model according to the training sample set, where the fatigue prediction model is used to determine a fatigue prediction result of a target recommended object for a candidate resource to be recommended when recommending a resource to the target recommended object to be processed; Wherein, determining to meet the second negative sample screening condition includes: determining any fourth video in the first video; obtaining a sixth video from a fifth video, where the video duration of the sixth video belongs to the same duration interval as the video duration of the fourth video, and the fifth video includes other videos in the first video except the fourth video; respectively determining the target playback duration of each video in the video set according to the corresponding historical consumption records, where the video set includes the fourth video and the sixth video; sorting the videos in the video set in descending order according to the target playback duration; in response to determining that there is a jump-out operation for the fourth video and determining that the sorting position serial number of the fourth video is greater than the product of M and x, determining that the fourth video meets the second negative sample screening condition, M represents the number of videos in the video set, and x represents a first percentage greater than 0 and less than 1, and the jump-out operation means that the recommended object corresponding to the historical consumption record exits the short video platform when playing the fourth video.

2. The method according to claim 1, wherein, The generating a training sample set according to the historical consumption records includes: In response to determining that there is a third video in the first video that meets the positive sample screening condition, determining that the positive sample video includes the third video; Generating the positive sample according to the positive sample video.

3. The method according to claim 1, wherein Determining to meet the first negative sample screening condition includes: Determining a first playback duration of the fourth video according to the corresponding historical consumption record; In response to determining that the first playback duration is less than a first threshold, determining that the fourth video meets the first negative sample screening condition.

4. The method according to claim 1, wherein, In response to determining that the following conditions are simultaneously met, it is determined that the fourth video meets the third negative sample screening condition: the number of videos is greater than a second threshold, the sorting position serial number of the fourth video is greater than the product of M and y, and the video duration of the fourth video is less than a third threshold, where y represents a second percentage greater than 0 and less than 1; Wherein, the number of videos is the number of videos in the same session that have been recommended before the fourth video and belong to the same category as the fourth video, and the session refers to the consumption process of the recommended object corresponding to the historical consumption record from opening the short video platform to exiting the short video platform.

5. The method according to claim 2, wherein, In response to determining that no predetermined negative operation has occurred for the fourth video and determining that the sorting position serial number of the fourth video is less than the product of M and z, it is determined that the fourth video meets the positive sample screening condition, where z represents a third percentage greater than 0 and less than 1.

6. The method according to claim 2, wherein, Generating the negative sample according to the negative sample video includes: For each negative sample video, generating a corresponding negative sample, and the negative sample includes: the first feature information corresponding to the corresponding negative sample video and a first label, where the first label is used to indicate that the corresponding sample is a negative sample; Generating the positive sample according to the positive sample video includes: For each positive sample video, generating a corresponding positive sample, and the positive sample includes: the second feature information corresponding to the corresponding positive sample video and a second label, where the second label is used to indicate that the corresponding sample is a positive sample.

7. The method according to claim 6, wherein, Both the first feature information and the second feature information are obtained by the following methods, including: For a target video in any of the negative sample videos or any of the positive sample videos, obtaining the object feature of the recommended object corresponding to the target video, where the recommended object corresponding to the target video is the recommended object corresponding to the same historical consumption record as the target video; Obtaining the video feature of the target video; Obtaining the cross feature between the target video and the recommended object corresponding to the target video; Obtaining a request feature, where the request is the request that triggers the recommendation of the target video; Obtaining the historical session feature of the recommended object corresponding to the target video, where the session refers to the consumption process of the recommended object corresponding to the target video from opening the short video platform to exiting the short video platform; Determining the object feature, the video feature, the cross feature, the request feature, and the historical session feature as the feature information corresponding to the target video.

8. A method for estimating fatigue, including: For a target recommended object that opens a resource recommendation application, determining candidate resources to be recommended; Using a fatigue estimation model to respectively determine the fatigue estimation results of the target recommended object for each candidate resource, where the fatigue estimation model is trained according to the method described in any one of claims 1 to 7.

9. The method according to claim 8, wherein, The fatigue prediction results of the target recommended object for each candidate resource determined by using the fatigue prediction model respectively include: For any candidate resource, obtain the third feature information corresponding to the candidate resource respectively; Input the third feature information into the fatigue prediction model to obtain the output fatigue prediction result.

10. The method according to claim 9, wherein The resource recommendation application includes: a short video platform; the resource object type is a video type; The obtaining of the third feature information corresponding to the candidate video includes: Obtain the object features of the target recommended object; Obtain the video features of the candidate video; Obtain the cross features between the candidate video and the target recommended object; Obtain the request feature, where the request is a request for triggering the recommendation of the candidate video; Obtain the historical session features of the target recommended object, where the session refers to the consumption process of the target recommended object from opening the short video platform to exiting the short video platform; Determine the object features, the video features, the cross features, the request features, and the historical session features as the third feature information.

11. A device for training a fatigue prediction model, comprising: An information acquisition module, a sample construction module, and a model training module; The information acquisition module is used to obtain the historical consumption records corresponding to a predetermined resource recommendation application, and the historical consumption records include: operation information of the recommended object who opens the resource recommendation application for the recommended resource object; wherein, the resource recommendation application includes: a short video platform, the resource object type is a video type, and each historical consumption record corresponds to a recommended object and a first video recommended; The sample construction module is used to generate a training sample set according to the historical consumption records, and the training sample set includes negative samples and positive samples. Among them, in response to determining that there is a second video in the first video that meets any one of the first negative sample screening condition, the second negative sample screening condition, and the third negative sample screening condition, determine that the negative sample video includes the second video, and generate the negative sample according to the negative sample video; The model training module is used to train the fatigue prediction model according to the training sample set. The fatigue prediction model is used to determine the fatigue prediction result of the target recommended object for the candidate resource to be recommended when recommending resources to the target recommended object to be processed; Among them, the sample construction module determines any fourth video in the first video, obtains a sixth video from a fifth video, where the video duration of the sixth video belongs to the same duration interval as that of the fourth video. The fifth video includes other videos in the first video except the fourth video. The target playback duration of each video in the video set is determined according to the corresponding historical consumption records. The video set includes the fourth video and the sixth video. The videos in the video set are sorted in descending order according to the target playback duration. In response to determining that there is a jump-out operation for the fourth video and determining that the sorting position serial number of the fourth video is greater than the product of M and x, it is determined that the fourth video meets the second negative sample screening condition. M represents the number of videos in the video set, and x represents a first percentage greater than 0 and less than 1. The jump-out operation means that the recommended object corresponding to the historical consumption record exits the short video platform when playing the fourth video.

12. The device according to claim 11, wherein, The sample construction module, in response to determining that there is a third video in the first video that meets the positive sample screening condition, determines that the positive sample video includes the third video, and generates the positive sample according to the positive sample video.

13. The device according to claim 11, wherein, The sample construction module determines the first playback duration of the fourth video according to the corresponding historical consumption record. In response to determining that the first playback duration is less than the first threshold, it is determined that the fourth video meets the first negative sample screening condition.

14. The device according to claim 11, wherein, The sample construction module, in response to determining that the following conditions are met simultaneously, determines that the fourth video meets the third negative sample screening condition: the number of videos is greater than the second threshold, the sorting position serial number of the fourth video is greater than the product of M and y, and the video duration of the fourth video is less than the third threshold. y represents a second percentage greater than 0 and less than 1; Among them, the number of videos is the number of videos of the same category that have been recommended before the fourth video in the same session. The session refers to the consumption process of the recommended object corresponding to the historical consumption record from opening the short video platform to exiting the short video platform.

15. The device according to claim 12, wherein, The sample construction module, in response to determining that there is no predetermined negative operation for the fourth video and determining that the sorting position serial number of the fourth video is less than the product of M and z, determines that the fourth video meets the positive sample screening condition. z represents a third percentage greater than 0 and less than 1.

16. The device according to claim 12, wherein, The sample construction module generates corresponding negative samples for each negative sample video. The negative samples include: the first feature information corresponding to the corresponding negative sample video and the first label, where the first label is used to indicate that the corresponding sample is a negative sample. And for each positive sample video, corresponding positive samples are generated respectively. The positive samples include: the second feature information corresponding to the corresponding positive sample video and the second label, where the second label is used to indicate that the corresponding sample is a positive sample.

17. The apparatus according to claim 16, wherein The sample construction module obtains the first feature information and the second feature information in the same way, including: for the target video in any of the negative sample videos or any of the positive sample videos, obtaining the object feature of the recommended object corresponding to the target video, where the recommended object corresponding to the target video is the recommended object corresponding to the same historical consumption record as the target video; obtaining the video feature of the target video; obtaining the cross feature between the target video and the recommended object corresponding to the target video; obtaining the request feature, where the request is the request that triggers the recommendation of the target video; obtaining the historical session feature of the recommended object corresponding to the target video, where the session refers to the consumption process of the recommended object corresponding to the target video from opening the short video platform to exiting the short video platform; and determining the object feature, the video feature, the cross feature, the request feature, and the historical session feature as the feature information corresponding to the target video.

18. A fatigue prediction device, comprising: A resource determination module and a result determination module; The resource determination module is used to determine the candidate resources to be recommended for the target recommended object that opens the resource recommendation application. The result determination module is used to use the fatigue prediction model to respectively determine the fatigue prediction results of the target recommended object for each candidate resource, where the fatigue prediction model is trained according to the method described in any one of claims 1 to 7.

19. The apparatus according to claim 18, wherein For any candidate resource, the result determination module respectively obtains the third feature information corresponding to the candidate resource, and inputs the third feature information into the fatigue prediction model to obtain the output fatigue prediction result.

20. The apparatus according to claim 19, wherein The resource recommendation application includes: a short video platform; the resource object type is a video type; The result determination module obtains the object feature of the target recommended object, obtains the video feature of the candidate video, obtains the cross feature between the candidate video and the target recommended object, obtains the request feature, where the request is the request that triggers the recommendation of the candidate video, obtains the historical session feature of the target recommended object, where the session refers to the consumption process of the target recommended object from opening the short video platform to exiting the short video platform, and determines the object feature, the video feature, the cross feature, the request feature, and the historical session feature as the third feature information.

21. An electronic device, including: At least one processor; And A memory communicatively connected to the at least one processor; wherein, the memory stores instructions executable by the at least one processor, and when the instructions are executed by the at least one processor, the at least one processor is enabled to execute the method according to any one of claims 1-10.

22. A non-transitory computer-readable storage medium storing computer instructions, wherein, The computer instructions are for causing a computer to execute the method according to any one of claims 1-10.

23. A computer program product comprising a computer program / instructions which, when executed by a processor, implement the method according to any one of claims 1-10.

Images