Recommended content determination method, recommendation method, device, equipment and storage medium

Abstract

The present disclosure relates to a method for determining recommended content, a recommendation method, an apparatus, a device and a storage medium. The method obtains a recommended content set including multiple categories and recommended content that meets preset conditions; determines a target category and a first recommended content belonging to the target category in multiple industry categories according to evaluation parameter information corresponding to various categories determined by historical record information of historical recommended content of multiple categories by users; obtains a second recommended content recalled for a target user, and determines the target recommended content based on the first recommended content and the second recommended content. In this way, by combining the recommended content that meets the preset conditions corresponding to the target category to determine the target recommended content recommended to the target user, the exposure of the target recommended content can be increased while minimizing the loss of user experience, thereby improving the recommendation effect and user satisfaction.

Description

Technical Field

[0001] The present disclosure relates to the field of computer technology, and in particular to a method for determining recommended content, a recommendation method, an apparatus, a device, and a storage medium. Background Art

[0002] Recommendation systems usually recommend content that users are interested in based on their historical behavior data. However, always making recommendations based on users' known interests will make users feel that repetitive content is always recommended, greatly reducing the sense of freshness and surprise.

[0003] The above problems can be classified as the exploration and exploitation (EE) problem in recommendation systems, which is also one of the two most common problems in computational advertising and recommendation systems. However, existing solutions to the exploration and exploitation problem still have problems such as poor exploration recommendation effect and low user satisfaction, which need to be further improved. Summary of the invention

[0004] The present disclosure provides a method for determining recommended content, a recommendation method, an apparatus, a device, and a storage medium to at least solve at least one problem in the related art, such as poor recommendation effect and low user satisfaction. The technical solution of the present disclosure is as follows:

[0005] According to a first aspect of an embodiment of the present disclosure, a method for determining recommended content is provided, including:

[0006] Acquire a recommended content set, wherein the recommended content set includes recommended content of multiple categories and meeting preset conditions;

[0007] Determining evaluation parameter information corresponding to each category according to the historical record information of the user for the multiple categories of historical recommended content; the historical record information represents the content information of each category and the corresponding operation information;

[0008] Determining a target category among the multiple categories according to the evaluation parameter information corresponding to the various categories;

[0009] Determining a first recommended content belonging to the target category from the recommended content set;

[0010] Acquire second recommended content recalled for a target user, and determine target recommended content based on the first recommended content and the second recommended content.

[0011] As an optional implementation manner, obtaining the recommended content set includes:

[0012] Obtain feedback information on recommended content corresponding to each category within a historical preset time period;

[0013] Parsing the feedback information to obtain content data of each recommended content in at least three dimensions;

[0014] Determining recommended content that meets preset conditions based on the content data and a preset threshold;

[0015] The recommended content set is established according to the determined recommended content.

[0016] As an optional implementation manner, determining the evaluation parameter information corresponding to various categories according to the historical record information of the user for the multiple categories of historical recommended contents includes:

[0017] Acquire historical record information of the user's historical recommended content for the multiple categories; the historical record information includes content feature information and the number of historical operations corresponding to each category;

[0018] Based on the content feature information, obtaining historical average evaluation parameter information corresponding to different categories;

[0019] Based on the number of historical operations corresponding to each category and the total number of historical operations of each category, the evaluation uncertainty parameters corresponding to different categories are determined;

[0020] According to the historical average evaluation parameter information and the evaluation uncertainty parameter, the evaluation parameter information corresponding to various categories is obtained.

[0021] As an optional implementation manner, determining the target category from the multiple categories according to the evaluation parameter information corresponding to the various categories includes:

[0022] Determine the evaluation parameter information that meets the preset evaluation conditions among the evaluation parameter information corresponding to various categories as the target evaluation parameter information;

[0023] The category corresponding to the target evaluation parameter information is determined as the target category.

[0024] As an optional implementation manner, determining the target recommended content based on the first recommended content and the second recommended content includes:

[0025] Determining candidate recommended content that meets a preset targeting condition based on the first recommended content and the second recommended content;

[0026] Based on a preset recommendation strategy, the candidate recommended content is processed to obtain target recommended content for recommendation to the target user.

[0027] As an optional implementation manner, determining the candidate recommended content that meets the preset targeting condition based on the first recommended content and the second recommended content includes:

[0028] configuring the first recommended content and the second recommended content to a first channel and at least one second channel respectively, each channel including a content tag for representing an attribute of the recommended content;

[0029] Merging the recommended contents in the first channel and the at least one second channel to obtain a merged recommended content set;

[0030] Determining a desired recommended content set according to user characteristic information corresponding to the target user;

[0031] According to the preset orientation condition of the combined recommended content, performing intersection processing on the combined recommended content set and the expected recommended content set to obtain a candidate recommended content set;

[0032] The candidate recommended content set is used as candidate recommended content that meets a preset targeting condition.

[0033] As an optional implementation manner, the preset recommendation strategy includes a first recommendation strategy and a second recommendation strategy, and the content tag includes a first content tag representing the first recommended content; and the processing of the candidate recommended content based on the preset recommendation strategy to obtain the target recommended content for recommendation to the target user includes:

[0034] Based on the first recommendation strategy, processing the candidate recommended contents except those carrying the first content tag among the candidate recommended contents to obtain a first subset to be recommended;

[0035] Based on the second recommendation strategy, the first subset to be recommended and the candidate recommended content carrying the first content tag are processed to obtain a second subset to be recommended;

[0036] In the case that the second subset to be recommended is a non-empty set, using the second subset to be recommended as target recommended content for recommending to the target user;

[0037] In a case where the second subset to be recommended is an empty set, the candidate recommended content carrying the first content tag is used as the target recommended content for recommendation to the target user.

[0038] As an optional implementation manner, when the second subset to be recommended is a non-empty set, using the second subset to be recommended as target recommended content for recommending to the target user includes:

[0039] In the case that the second subset to be recommended is a non-empty set, performing diversity processing on the second subset to be recommended to obtain a processed second subset to be recommended;

[0040] using the processed second subset to be recommended as target recommended content for recommending to the target user; or

[0041] In the case where the second subset to be recommended is an empty set, taking the candidate recommended content carrying the first content tag as the target recommended content for recommendation to the target user comprises:

[0042] In the case that the second subset to be recommended is an empty set, performing diversity processing on the candidate recommended content carrying the first content tag to obtain a processed third subset to be recommended;

[0043] The processed third subset to be recommended is used as target recommended content for recommending to the target user.

[0044] As an optional implementation manner, after the step of filtering and sorting the candidate recommended contents based on a preset recommendation strategy to obtain target recommended contents for recommendation to the target user, the method further includes:

[0045] Acquiring the target user's posterior behavior data on the target recommended content;

[0046] parsing the extended interest tags of the target user according to the a posteriori behavior data;

[0047] The target recommended content is filtered based on the extended interest tag to obtain updated target recommended content.

[0048] According to a second aspect of an embodiment of the present disclosure, a recommendation method is provided, including:

[0049] Acquire a target user and target recommended content for recommending to the target user, wherein the target recommended content is obtained according to any of the above-mentioned recommended content determination methods;

[0050] The target recommended content is recommended to the target user.

[0051] As an optional implementation, the target user satisfies at least one of the following conditions: users whose user activity is greater than or equal to a preset activity threshold, users whose application usage time is greater than or equal to a preset time threshold, users whose application cold start times are greater than or equal to a preset times threshold, and users whose media content fill rate is less than or equal to a preset fill rate threshold.

[0052] According to a third aspect of an embodiment of the present disclosure, there is provided a recommended content determination device, including:

[0053] A first acquisition module is configured to execute acquisition of a recommended content set, wherein the recommended content set includes recommended content of multiple categories and meeting preset conditions;

[0054] A first determination module is configured to determine evaluation parameter information corresponding to various categories based on historical record information of historically recommended contents by the user for the multiple categories; the historical record information represents content information of various categories and corresponding operation information;

[0055] A second determination module is configured to determine a target category among the multiple categories according to the evaluation parameter information corresponding to the various categories;

[0056] A third determination module is configured to determine, from the recommended content set, a first recommended content belonging to the target category;

[0057] The target content determination module is configured to obtain second recommended content recalled for a target user, and determine target recommended content based on the first recommended content and the second recommended content.

[0058] As an optional implementation manner, the first acquisition module includes:

[0059] A first acquisition submodule is configured to acquire feedback information of recommended content corresponding to each category within a historical preset time period;

[0060] A second acquisition submodule is configured to parse the feedback information and acquire content data of each recommended content in at least three dimensions;

[0061] A determination submodule, configured to determine recommended content that meets preset conditions based on the content data and a preset threshold;

[0062] The establishing submodule is configured to establish the recommended content set according to the determined recommended content.

[0063] As an optional implementation manner, the first determining module includes:

[0064] A second acquisition module is configured to acquire historical record information of the user's historical recommended content of the multiple categories; the historical record information includes content feature information and the number of historical operations corresponding to each category;

[0065] A first determination module is configured to obtain historical average evaluation parameter information corresponding to different categories based on the content feature information;

[0066] A second determination module is configured to determine evaluation uncertainty parameters corresponding to different categories based on the number of historical operations corresponding to each category and the total number of historical operations of each category;

[0067] The score determination module is configured to obtain the evaluation parameter information corresponding to various categories according to the historical average evaluation parameter information and the evaluation uncertainty parameter.

[0068] As an optional implementation manner, the second determining module includes:

[0069] The target score determination submodule is configured to determine the evaluation parameter information that satisfies the preset evaluation conditions among the evaluation parameter information corresponding to various categories as the target evaluation parameter information;

[0070] The category determination submodule is configured to determine the category corresponding to the target evaluation parameter information as the target category.

[0071] As an optional implementation, the third determining module includes:

[0072] a candidate determination submodule, configured to determine candidate recommended content that meets a preset targeting condition based on the first recommended content and the second recommended content;

[0073] The target determination submodule is configured to process the candidate recommended content based on a preset recommendation strategy to obtain target recommended content for recommendation to the target user.

[0074] As an optional implementation manner, the candidate content determination submodule includes:

[0075] a channel configuration unit, configured to configure the first recommended content and the second recommended content to a first channel and at least one second channel respectively, each channel including a content tag for characterizing an attribute of the recommended content;

[0076] a merging unit, configured to merge the recommended contents in the first channel and the at least one second channel to obtain a merged recommended content set;

[0077] A desired content determination unit, configured to determine a desired recommended content set according to user feature information corresponding to the target user;

[0078] an intersection unit, configured to perform intersection processing on the combined recommended content set and the expected recommended content set according to a preset orientation condition of the combined recommended content, to obtain a candidate recommended content set;

[0079] The candidate content determination unit is configured to execute the candidate recommended content set as the candidate recommended content that meets the preset targeting condition.

[0080] As an optional implementation, the preset recommendation strategy includes a first recommendation strategy and a second recommendation strategy, the content tag includes a first content tag representing the first recommended content; and the target determination submodule includes:

[0081] A first processing unit is configured to filter and sort the candidate recommended contents except those carrying the first content tag among the candidate recommended contents based on the first recommendation strategy to obtain a first subset to be recommended;

[0082] A second processing unit is configured to filter and sort the first subset to be recommended and the candidate recommended content carrying the first content tag based on the second recommendation strategy to obtain a second subset to be recommended;

[0083] A first target content determination unit is configured to, when the second subset to be recommended is a non-empty set, use the second subset to be recommended as target recommended content for recommending to the target user;

[0084] The second target content determination unit is configured to, when the second to-be-recommended subset is an empty set, use the candidate recommended content carrying the first content tag as the target recommended content for recommendation to the target user.

[0085] As an optional implementation manner, the first target content determination unit includes:

[0086] A first processing sub-unit is configured to perform diversity processing on the second subset to be recommended, to obtain a processed second subset to be recommended, when the second subset to be recommended is a non-empty set;

[0087] The first target content determination subunit is configured to use the processed second subset to be recommended as target recommended content for recommendation to the target user; or

[0088] The second target content determination unit includes:

[0089] The second processing sub-unit is configured to perform diversity processing on the candidate recommended content carrying the first content tag to obtain a processed third subset to be recommended when the second subset to be recommended is an empty set;

[0090] The second target content determination subunit is configured to use the processed third to-be-recommended subset as target recommended content for recommendation to the target user.

[0091] As an optional implementation, the device further includes:

[0092] A second acquisition module is configured to execute acquisition of the target user's posterior behavior data on the target recommended content;

[0093] A parsing module, configured to parse the extended interest tags of the target user according to the a posteriori behavior data;

[0094] The filtering module is configured to filter the target recommended content based on the extended interest tag to obtain updated target recommended content.

[0095] According to a fourth aspect of an embodiment of the present disclosure, a recommendation device is provided, including:

[0096] An acquisition module, configured to acquire a target user and target recommended content for recommending to the target user, wherein the target recommended content is obtained according to any of the above-mentioned recommended content determination devices;

[0097] The recommendation module is configured to recommend the target recommended content to the target user.

[0098] As an optional implementation, the target user satisfies at least one of the following conditions: users whose user activity is greater than or equal to a preset activity threshold, users whose application usage time is greater than or equal to a preset time threshold, users whose application cold start times are greater than or equal to a preset times threshold, and users whose media content fill rate is less than or equal to a preset fill rate threshold.

[0099] According to a fifth aspect of an embodiment of the present disclosure, a computer-readable storage medium is provided. When instructions in the computer-readable storage medium are executed by a processor of an electronic device, the electronic device is enabled to execute the recommended content determination method or recommendation method as described in any of the above embodiments.

[0100] According to a sixth aspect of an embodiment of the present disclosure, there is provided an electronic device, including:

[0101] processor;

[0102] a memory for storing instructions executable by the processor;

[0103] The processor is configured to execute the instructions to implement the recommended content determination method or recommendation method as described in any of the above embodiments.

[0104] According to a seventh aspect of an embodiment of the present disclosure, a computer program product is provided, the computer program product comprising a computer program, and when the computer program is executed by a processor, the recommended content determination method or recommendation method provided in any one of the above-mentioned embodiments is implemented.

[0105] The technical solution provided by the embodiments of the present disclosure brings at least the following beneficial effects:

[0106] The disclosed embodiment obtains a recommended content set, which includes recommended content of multiple categories and meets preset conditions; determines evaluation parameter information corresponding to various categories based on historical record information of historical recommended content of the multiple categories by the user; the historical record information represents content information of various categories and corresponding operation information; determines a target category among the multiple categories based on the evaluation parameter information corresponding to the various categories; determines a first recommended content belonging to the target category from the recommended content set; obtains a second recommended content recalled for the target user, and determines the target recommended content based on the first recommended content and the second recommended content. In this way, by combining the recommended content that meets the preset conditions corresponding to the target category to determine the target recommended content recommended to the target user, it is possible to increase the exposure of the target recommended content without reducing the loss of user experience, fully tap more potential interest tags of the target user, improve the recommendation effect and user satisfaction, and thus improve user stickiness.

[0107] It is to be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the present disclosure. BRIEF DESCRIPTION OF THE DRAWINGS

[0108] The drawings herein are incorporated into the specification and constitute a part of the present disclosure, illustrate embodiments consistent with the present disclosure, and together with the description are used to explain the principles of the present disclosure, and do not constitute improper limitations on the present disclosure.

[0109] Figure 1 The invention is an architecture diagram of a system for applying a method for determining application recommended content or a method for recommending application recommended content according to an exemplary embodiment.

[0110] Figure 2 The figure is a flowchart of a method for determining recommended content according to an exemplary embodiment.

[0111] Figure 3 It is a partial flow chart of another method for determining recommended content according to an exemplary embodiment.

[0112] Figure 4 The present invention is a flowchart showing a step of determining candidate recommended content that meets a preset targeting condition according to an exemplary embodiment.

[0113] Figure 5 The present invention is a flowchart showing a step of obtaining target recommended content for recommending to the target user according to an exemplary embodiment.

[0114] Figure 6 is a partial flow chart of yet another method for determining recommended content according to an exemplary embodiment.

[0115] Figure 7 The figure is a flowchart of a recommendation method according to an exemplary embodiment.

[0116] Figure 8 The figure is a block diagram of a device for determining recommended content according to an exemplary embodiment.

[0117] Fig. 9 is a block diagram of another recommendation device according to an exemplary embodiment.

[0118] Fig.10 It is a block diagram of an electronic device according to an exemplary embodiment. DETAILED DESCRIPTION

[0119] In order to enable ordinary persons in the art to better understand the technical solutions of the present disclosure, the technical solutions in the embodiments of the present disclosure will be clearly and completely described below in conjunction with the accompanying drawings.

[0120] It should be noted that the terms "first", "second", etc. in the specification and claims of the present disclosure and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments of the present disclosure described herein can be implemented in an order other than those illustrated or described herein. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present disclosure. Instead, they are merely examples of devices and methods consistent with some aspects of the present disclosure as detailed in the appended claims.

[0121] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data for display, data for analysis, etc.) involved in this disclosure are all information and data authorized by the user or fully authorized by all parties.

[0122] Figure 1 is an architecture diagram of a system for determining a method for application recommended content or a method for recommending content according to an exemplary embodiment, see Figure 1 , the architecture diagram may include a first terminal 10, a second terminal 20 and a server 30.

[0123] Among them, the first terminal 10 and the second terminal 20 can be but are not limited to physical devices such as smart phones, tablet computers, laptops, desktop computers, smart wearable devices, digital assistants, augmented reality devices, virtual reality devices, or one or more applications or applets running in physical devices.

[0124] The server 30 may be a server corresponding to an application on the first terminal 10 and the second terminal 20. As an example only, the server 30 may be, but is not limited to, an independent server, or a server cluster or distributed system composed of multiple physical servers, or may be one or more cloud servers that provide basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, CDN, and big data and artificial intelligence platforms.

[0125] As an example only, the application programs installed on the first terminal 10 and the second terminal 20 may be provided with services by the server 30, the first terminal 10 may be a terminal corresponding to the target user, the second terminal 20 may be a terminal corresponding to the non-target user, the server 30 determines the recommended content according to the first terminal 10 and the second terminal 20, and recommends the determined recommended content to the target user corresponding to the first terminal 10. The first terminal 10 and the server 30, as well as the second terminal 20 and the server 30, may be directly or indirectly connected via wired or wireless communication, which is not limited in the embodiments of the present disclosure.

[0126] The recommended content determination method or recommendation method provided in the embodiments of the present disclosure may be executed by a recommended content determination device or recommendation device, which may be integrated in an electronic device such as a client or a terminal in the form of hardware or software.

[0127] Figure 2 is a flowchart of a method for determining recommended content according to an exemplary embodiment. Figure 2 As shown, the recommended content determination method can be applied to an electronic device, and is described by taking the electronic device as the server in the above implementation environment diagram as an example, and includes the following steps.

[0128] In step S201, a recommended content set is obtained, where the recommended content set includes recommended content of multiple categories that meets preset conditions.

[0129] Among them, the category may include but is not limited to content category, industry category, etc. Content category refers to the category to which the recommended content set belongs, such as primary content categories such as food, film and television, life, etc., or secondary content categories such as pastry, martial arts drama, square dance, etc. Industry category refers to the category of the industry to which the recommended content belongs. Just as an example, the industry category can include primary industry or secondary industry, and the secondary industry is a fine-grained division of the primary industry. For example, the primary industry may include games, and its corresponding secondary industries may include leisure and entertainment, role-playing, etc. For another example, the primary industry may include shopping, and its corresponding secondary industries may include daily necessities, clothing, cosmetics, etc.

[0130] It should be understood that the multiple categories here may refer to categories at the same level, such as multiple primary content categories, multiple secondary content categories, multiple primary industry categories, multiple secondary industry categories, etc. Of course, the level of categories is not limited to this, and may also include more fine-grained category divisions.

[0131] The preset condition can be determined by a preset threshold value, which can be measured by content data used to measure the quality of the recommended content. Exemplarily, the content data may include at least one dimension of 3-second play rate, content score, negative review rate, completion rate, negative feedback rate, conversion rate, 5-second play rate, etc. Taking the recommended content as video material as an example, the 3-second play rate refers to the ratio of the number of users who watch the video material for more than 3 seconds to the total number of users exposed to the video material. The content score refers to the content quality of the video material. The negative review rate refers to the proportion of negative reviews in the evaluation of the video material by the viewing users.

[0132] The preset thresholds correspond to the dimensions of the content data one by one. For example, the preset thresholds may include a preset 3-second play rate threshold k1, a preset content score threshold k2, a preset negative review rate threshold k3, etc. According to the size of the content data and the corresponding preset threshold, it is determined whether the recommended content meets the preset conditions.

[0133] In an optional embodiment, obtaining the recommended content set may include:

[0134] In step S2011, feedback information of recommended content corresponding to each category within a preset historical time period is obtained;

[0135] In step S2012, the feedback information is parsed to obtain content data of each recommended content in at least three dimensions;

[0136] In step S2013, based on the content data and a preset threshold, recommended content that meets a preset condition is determined;

[0137] In step S2014, the recommended content set is established according to the determined recommended content.

[0138] Among them, feedback information is used to characterize the user's behavior data on recommended content in a historical preset time period, for example, video data of user browsing, on-demand, liking, and commenting. The historical preset time period can be set within the past month, the past week, etc., and the present disclosure does not specifically limit this. By parsing the feedback information of recommended content in various industry categories within the historical preset time period, the content data corresponding to each recommended content is obtained, such as at least one dimension of 3-second play rate, content score, negative review rate, completion rate, negative feedback rate, conversion rate, 5-second play rate, etc.

[0139] In this embodiment, the feedback information can be parsed to obtain content data in at least three dimensions, and then, based on the size relationship between the content data and the preset threshold, the recommended content whose content data in at least three dimensions simultaneously meets the preset threshold is determined as the recommended content that meets the preset conditions; then, a recommended content set is constructed based on all the determined recommended content.

[0140] Preferably, taking the content data of three dimensions as an example, the three dimensions here may include 3-second play rate, content score and negative review rate. At this time, the corresponding content data obtained includes 3-second play rate, content score and negative review rate. When the 3-second play rate of the recommended content M is greater than the preset 3-second play rate threshold k1, and the content score is less than the preset content score threshold k2, and the negative review rate is less than the preset negative review rate threshold k3, it is determined that the content quality data of the recommended content M meets the preset conditions, that is, the recommended content M is a high-quality recommended content. If any dimension of the 3-second play rate, content score and negative review rate does not meet the corresponding preset conditions, it is determined that the content data of the delivered content M does not meet the preset conditions, that is, the recommended content M is a non-high-quality recommended content. As an example only, the preset 3-second play rate threshold k1 can be set to 0.5, the preset content score threshold k2 can be set to 0.35, and the preset negative review rate threshold k3 can be set to 0.3.

[0141] In the above embodiment, by analyzing the feedback information of the recommended content corresponding to each category within the historical preset time period, the content data of each recommended content in at least three dimensions is obtained; then, based on the content data and the recommended content that meets the preset conditions determined by the preset threshold, a recommended content set is established. In this way, the constructed recommended content set is high-quality content, thereby reducing the loss of user experience in the process of determining the recommended content, which is conducive to improving the recommendation effect and user satisfaction.

[0142] In step S202, evaluation parameter information corresponding to various categories is determined according to the historical record information of the user's historical recommendation content for the multiple categories.

[0143] The historical record information represents various types of content information and corresponding operation information, and the evaluation parameter information represents the estimated recommendation dedication of recommended content under different categories.

[0144] Optionally, the evaluation parameter information corresponding to various categories may be determined by using a Bandit algorithm according to the historical record information of the user's historical recommendation content for various categories.

[0145] In an exemplary embodiment, the Bandit algorithm includes at least one of the following algorithms: a naive Bandit algorithm, an UCB (Upper Confidence Bound) algorithm, and variations thereof.

[0146] In an alternative embodiment, if Figure 3 As shown, the determining of evaluation parameter information corresponding to various categories based on the historical record information of the user's historical recommendation content for the multiple categories includes:

[0147] In step S301, the historical record information of the user's historical recommended content of multiple categories is obtained; the historical record information includes content feature information and the number of historical operations corresponding to each category;

[0148] In step S302, based on the content feature information, historical average evaluation parameter information corresponding to different categories is obtained;

[0149] In step S303, based on the number of historical operations corresponding to each category and the total number of historical operations of each category, the evaluation uncertainty parameters corresponding to different categories are determined;

[0150] In step S304, the evaluation parameter information corresponding to various categories is obtained according to the historical average evaluation parameter information and the evaluation uncertainty parameter.

[0151] The content feature information may include but is not limited to the exposure, clicks, and conversions of the user's historical viewing of the content delivered in various industry categories. The degree of recommendation dedication brought by these exposures, clicks, and conversions can be used to measure the historical average evaluation parameter information under different categories. Then, the evaluation parameter information corresponding to the current user's visit volume in various categories is determined by combining the historical average evaluation parameter information and the evaluation uncertainty.

[0152] Optionally, the evaluation parameter information corresponding to various industries here may include evaluation parameter values, and correspondingly, the historical average evaluation parameter information may include historical average evaluation parameter values. As an example only, the evaluation parameter value may be determined using the UCB (Upper Confidence Bound) algorithm, which is a selection strategy algorithm based on the current level of dedication and number of explorations, and uses the confidence upper bound of the current point of interest (e.g., the secondary industry of advertising in the information flow scenario) as the estimated return value. Specifically, by trying each category once, the evaluation parameter value A brought about by the operation of each category is calculated separately according to the historical record information of each category. The iterative formula for the evaluation parameter value A can be:

[0153]

[0154] The first term Q(a) represents the historical average evaluation parameter value of the current category, and the second term Indicates the evaluation uncertainty corresponding to the current category. a is the category of the operation, N is the total number of historical operations of each category, n is the number of historical operations of the current category, and A is the estimated evaluation parameter value when the operation category is a. The operations here include but are not limited to interest exploration.

[0155] From the above iterative formula, we can see that for the selected category, getting one more exploration opportunity will make the second term in the formula becomes smaller, and eventually becomes smaller than other categories that are not selected. In addition, for the unselected categories, the confidence upper bound is It will increase with the number of rounds, and finally will be greater than the scores of other selected categories. The UCB algorithm can be seen as an optimization process, which requires multiple iterations to find a better solution. In the decision-making process, all uncertainties are gradually recognized. Finally, from the time dimension, under the interference of uncertainty, it is still possible to optimize the objective function corresponding to the evaluation parameter value A.

[0156] In the above embodiment, the historical average evaluation parameter information corresponding to different categories is determined according to the content feature information in the historical record information, and then the evaluation uncertainty parameters corresponding to different categories are determined according to the historical operation times in the historical record information corresponding to each category and the total historical operation times of each category; the evaluation parameter information corresponding to various categories is obtained according to the historical average evaluation parameter information and the evaluation uncertainty parameters. Since the evaluation parameter information corresponding to various categories is related to the historical average evaluation parameter information and the evaluation uncertainty parameters, and the historical average evaluation parameter information is proportional to the content feature information, and the evaluation uncertainty parameters are inversely proportional to the historical operation times, the recommendation dedication and the operation times of the historical recommended content of different categories can be balanced, so that the determined evaluation parameter information can better reflect the recommendation dedication brought by each category, which is conducive to improving the accuracy of the subsequent determination of recommended content.

[0157] In step S203, a target category is determined from among the multiple categories according to the evaluation parameter information corresponding to the various categories.

[0158] Optionally, after determining the evaluation parameter information corresponding to various categories, the target category can be determined from the multiple categories according to the evaluation parameter information corresponding to the various categories (for example, the size of the evaluation parameter value A). The number of target categories determined here can be one or more, and the present disclosure does not specifically limit this.

[0159] In an optional embodiment, determining the target category among the multiple categories according to the evaluation parameter information corresponding to the various categories may include:

[0160] In step S2021, the evaluation parameter information corresponding to various categories that meets the preset evaluation conditions is determined as the target evaluation parameter information;

[0161] In step S2022, the category corresponding to the target evaluation parameter information is determined as the target category.

[0162] Among them, taking the evaluation parameter information including the evaluation parameter value as an example, the preset evaluation condition may include but is not limited to at least one of the following: the evaluation parameter value is greater than or equal to the preset evaluation parameter threshold, TOP N evaluation parameter values ​​(N is a positive integer), etc. The target category may be a category corresponding to the evaluation parameter value in the evaluation parameter information reaching the preset evaluation parameter value, or may be at least one of the categories with the evaluation parameter values ​​ranked top in order, which is not specifically limited in the present disclosure.

[0163] As an example only, when determining the evaluation parameter information corresponding to various categories, the maximum evaluation parameter value can be filtered out from the evaluation parameter information as the target evaluation parameter information, and then the category corresponding to the target evaluation parameter information can be used as the target category to facilitate further interest exploration of the recommended content in the category with the largest evaluation parameter value A.

[0164] In the above embodiment, the category corresponding to the target evaluation parameter information using the evaluation parameter information that meets the preset evaluation conditions is determined as the target category. The determined target category is a category with a higher degree of recommendation dedication, thereby ensuring the quality of the recommended content and further improving the recommendation effect and user satisfaction.

[0165] In step S204, a first recommended content belonging to the target category is determined from the recommended content set.

[0166] Optionally, after the target category is determined, the recommended content corresponding to the target category can be screened out from the recommended content set as the first recommended content. Subsequent recommendations and interest explorations are only conducted on the first recommended content, without processing other recommended content in the recommended content set, which can greatly reduce the workload and help improve the quality and effect of recommendations.

[0167] In step S205, second recommended content recalled for the target user is obtained, and target recommended content is determined based on the first recommended content and the second recommended content.

[0168] Optionally, before determining the second recommended content, the target user may be determined first. The target user here refers to the user who needs to be recommended or explored for interest. As an example only, the target user may meet at least one of the following conditions: a user whose user activity is greater than or equal to a preset activity threshold, a user whose application usage time is greater than or equal to a preset time threshold, a user whose application cold start times are greater than or equal to a preset times threshold, and a user whose media content fill rate is less than or equal to a preset fill rate threshold. Among them, the media content may include but is not limited to advertisements. It should be understood that the conditions satisfied by the target user are not limited to this.

[0169] After determining the target user, the corresponding recommended content can be recalled for the target user according to the user characteristic data of the target user as the second recommended content. The user characteristic data may include the user's behavior data, user interests, hobbies, gender, etc. Then, according to the first recommended content and the second recommended content obtained from the recommended content set, the target recommended content for recommendation to the target user is determined.

[0170] As an example only, the target recommended content may include but is not limited to the union or intersection of the first recommended content and the second recommended content, or the recommended content included in the union of the first recommended content and the second recommended content and meeting the preset recommendation conditions. The preset recommendation conditions may include but are not limited to user profile conditions, recommendation demand conditions, recommended content quality conditions, etc.

[0171] In an optional embodiment, determining the target recommended content based on the first recommended content and the second recommended content may include:

[0172] In step S2051, candidate recommended content that meets a preset targeting condition is determined based on the first recommended content and the second recommended content.

[0173] Optionally, before determining the target recommended content, candidate recommended content that meets the preset targeting condition may be determined based on the first recommended content and the second recommended content obtained from the recommended content set, and then the target recommended content may be determined based on the candidate recommended content. Taking the recommended content as an advertising material as an example, the preset targeting condition may refer to the targeting user condition for the advertiser to place an advertisement, such as user gender, age, etc.

[0174] In an alternative embodiment, if Figure 4 As shown, the determining, based on the first recommended content and the second recommended content, candidate recommended content that meets the preset targeting condition includes:

[0175] In step S401, the first recommended content and the second recommended content are respectively configured to a first channel and at least one second channel, and each channel includes a content tag for representing an attribute of the recommended content;

[0176] In step S402, the recommended contents in the first channel and the at least one second channel are merged to obtain a merged recommended content set;

[0177] In step S403, a desired recommended content set is determined according to the user characteristic information corresponding to the target user;

[0178] In step S404, according to the preset orientation condition of the combined recommended content, the combined recommended content set and the expected recommended content set are subjected to intersection processing to obtain a candidate recommended content set;

[0179] In step S405, the candidate recommended content set is used as candidate recommended content that meets a preset targeting condition.

[0180] Optionally, the first recommended content is configured to a first channel in the online multiple channels through Redis (a cross-platform non-relational database), and the second recommended content is configured to at least one second channel in the online multiple channels. Each channel includes a content tag for characterizing the attributes of the recommended content to mark which channel the recommended content belongs to.

[0181] For example, the first channel includes a first content tag representing the first recommended content, that is, a high-quality content tag. The second channel includes at least one second content tag representing the second recommended content, such as a hot content tag, a personalized content tag, etc. Then, each recommended content in the first channel and at least one second channel is merged to obtain a merged recommended content set. The merged recommended content set includes both the first recommended content with the high-quality content tag and the second recommended content with other content tags. Then, according to the user feature information corresponding to the target user, the expected recommended content set is determined, for example, the expected recommended content set may include {female, 24 years old, like sports}. Taking the recommended content as an advertising material as an example, each merged recommended content has its preset targeting conditions. For example, for the recommended content c1 in the merged recommended content, its preset targeting condition is {male, over 30 years old, likes food}; for the recommended content c2 in the merged recommended content, its preset targeting condition is {female, under 25 years old, likes travel}, etc. According to the preset targeting conditions of the merged recommended content, the merged recommended content set and the expected recommended content set are subjected to intersection processing, that is, the merged recommended content is screened to obtain merged recommended content matching the target user to form a candidate recommended content set, and then the candidate recommended content set is used as the candidate recommended content that meets the preset targeting conditions.

[0182] In the above embodiment, since the first recommended content and the second recommended content are respectively configured on different channels, and different channels carry different content tags, it is convenient to quickly determine the source of each recommended content, and to perform differentiated processing on different recommended content, thereby reducing the amount of calculation. At the same time, since the candidate recommended content is determined by intersection processing based on the combined recommended content of multiple channels and the expected recommended content set determined by user feature information, more potential interest tags of the target user are fully explored, the recommendation quality of the candidate recommended content is ensured, the quality of the target recommended content is improved, and the recommendation effect and user satisfaction are further improved.

[0183] In step S2052, the candidate recommended content is processed based on a preset recommendation strategy to obtain target recommended content for recommendation to the target user.

[0184] Optionally, after the candidate recommendation content is determined, the candidate recommendation content may be subjected to at least one of screening processing, rough sorting and fine sorting, so as to filter and sort the candidate recommendation content and obtain the target recommendation content for recommendation to the target user. The screening processing may include but is not limited to at least one of diversity filtering, frequency control, reward strategy, etc.

[0185] In an optional embodiment, the determined candidate recommended contents are first screened and then sorted, and a preset number M (M is a positive integer) of candidate recommended contents that are at the top of the sorting results are coarsely sorted and finely sorted in turn to obtain target recommended contents for recommendation to the target user.

[0186] In another alternative embodiment, if Figure 5 As shown, the preset recommendation strategy includes a first recommendation strategy and a second recommendation strategy, and the content tag includes a first content tag representing the first recommended content; and the processing of the candidate recommended content based on the preset recommendation strategy to obtain the target recommended content for recommendation to the target user includes:

[0187] In step S501, based on the first recommendation strategy, the candidate recommended contents except those carrying the first content tag in the candidate recommended contents are filtered and sorted to obtain a first subset to be recommended;

[0188] In step S502, based on the second recommendation strategy, the first subset to be recommended and the candidate recommended content carrying the first content tag are filtered and sorted to obtain a second subset to be recommended;

[0189] In step S503, when the second subset to be recommended is a non-empty set, the second subset to be recommended is used as target recommended content for recommending to the target user;

[0190] In step S504, when the second subset to be recommended is an empty set, the candidate recommended content carrying the first content tag is used as the target recommended content for recommendation to the target user.

[0191] Optionally, for the candidate recommended content set obtained by the directional intersection processing, if there is a recommended content with a first content tag, the recommended content with the first content tag will be directly sent to the refined ranking, that is, only the second recommendation strategy needs to be executed for the recommended content with the first content tag, and the first recommendation strategy does not need to be executed for the recommended content with the first content tag. The first recommendation strategy can be a rough ranking strategy, and the second recommendation strategy can be a refined ranking strategy. When executing the second recommendation strategy, filtering and sorting can be performed according to the scores of indicators such as UEQ (user rating) and CPM to obtain the second subset to be recommended.

[0192] If the second subset to be recommended is a non-empty set, the second subset to be recommended is used as the target recommended content for recommendation to the target user, so that it can be exposed according to the arrangement order of the second subset to be recommended and recommended to the target user. If the second subset to be recommended is an empty set, it means that all candidate recommended contents have been excluded through fine sorting. At this time, the candidate recommended content carrying the first content tag can be used as the target recommended content for recommendation to the target user, so that at least one candidate recommended content carrying the first content tag can be randomly selected for exposure and recommended to the target user.

[0193] In an optional embodiment, when the second subset to be recommended is a non-empty set, using the second subset to be recommended as target recommended content for recommending to the target user may include:

[0194] In the case that the second subset to be recommended is a non-empty set, performing diversity processing on the second subset to be recommended to obtain a processed second subset to be recommended;

[0195] The processed second subset to be recommended is used as target recommended content for recommending to the target user.

[0196] Optionally, after obtaining the second recommended subset, the second recommended subset can be first subjected to diversity processing to reduce repeated recommended content, and then the second recommended subset subjected to diversity processing can be used as target recommended content for recommendation to the target user. This can improve the quality of the target recommended content, thereby helping to improve the recommendation effect and user satisfaction.

[0197] In an optional embodiment, when the second subset to be recommended is an empty set, taking the candidate recommended content carrying the first content tag as the target recommended content for recommending to the target user may include:

[0198] In the case that the second subset to be recommended is an empty set, performing diversity processing on the candidate recommended content carrying the first content tag to obtain a processed third subset to be recommended;

[0199] The processed third subset to be recommended is used as target recommended content for recommending to the target user.

[0200] Optionally, when the second recommendation subset is an empty set, the candidate recommended content carrying the first content tag can be first subjected to diversity processing to reduce duplicate recommended content, and then the third subset to be recommended that has undergone diversity processing can be used as the target recommended content for recommendation to the target user. This can improve the quality of the target recommended content, thereby helping to improve the recommendation effect and user satisfaction.

[0201] In the above embodiment, the candidate recommended contents with different content tags are processed differently through the first recommendation strategy and the second recommendation strategy respectively, and different recommended contents are used as target recommended contents according to whether the second subset to be recommended is a non-empty set or an empty set, thereby improving the quality of the recommended contents, improving the quality of the target recommended contents, and further improving the recommendation effect and user satisfaction.

[0202] In an alternative embodiment, if Figure 6 As shown, after the step of filtering and sorting the candidate recommended contents based on the preset recommendation strategy to obtain the target recommended contents for recommendation to the target user, the method further includes:

[0203] In step S601, the target user's a posteriori behavior data on the target recommended content is obtained;

[0204] In step S602, the extended interest tags of the target user are parsed according to the a posteriori behavior data;

[0205] In step S603, the target recommended content is filtered based on the extended interest tag to obtain updated target recommended content.

[0206] Among them, the a posteriori behavior data is used to reflect the interest of the target user in the target recommended content. The a posteriori behavior data may include but is not limited to data such as viewing time, clicks, likes, blocking, sharing, etc. By parsing the a posteriori behavior data of the target user, the extended interest tag of the target user is obtained. The extended interest tag reflects the mined potential interest points of the user. Then, the target recommended content is filtered using the extended interest tag to obtain updated target recommended content that better meets the interests of the target user.

[0207] In the above embodiment, the target user's posterior behavior data on the target recommended content is obtained; the target user's extended interest tag is obtained based on the posterior behavior data, and then the target recommended content is filtered based on the extended interest tag to obtain updated target recommended content. Therefore, the determined target recommended content is obtained by filtering and updating based on the target user's posterior behavior data, which is more in line with the target user's needs or interests, further improving the recommendation effect and user satisfaction, and increasing user stickiness.

[0208] The disclosed embodiment obtains a recommended content set, which includes recommended content of multiple categories and meets preset conditions; determines evaluation parameter information corresponding to various categories based on historical record information of historical recommended content of the multiple categories by the user; the historical record information represents content information of various categories and corresponding operation information; determines a target category among the multiple categories based on the evaluation parameter information corresponding to the various categories; determines a first recommended content belonging to the target category from the recommended content set; obtains a second recommended content recalled for the target user, and determines the target recommended content based on the first recommended content and the second recommended content. In this way, by combining the recommended content that meets the preset conditions corresponding to the target category to determine the target recommended content recommended to the target user, it is possible to increase the exposure of the target recommended content without reducing the loss of user experience, fully tap more potential interest tags of the target user, improve the recommendation effect and user satisfaction, and thus improve user stickiness.

[0209] Figure 7 is a flowchart of a recommendation method according to an exemplary embodiment. Figure 7 As shown, the recommended method can be applied to an electronic device, and the electronic device is taken as the server in the above implementation environment diagram as an example for explanation, and includes the following steps.

[0210] In step S701, a target user and target recommended content for recommending to the target user are obtained, wherein the target recommended content is obtained according to any of the above-mentioned recommended content determination methods.

[0211] In an optional embodiment, the target user satisfies at least one of the following conditions: users whose user activity is greater than or equal to a preset activity threshold, users whose application usage time is greater than or equal to a preset time threshold, users whose application cold start times are greater than or equal to a preset times threshold, and users whose media content fill rate is less than or equal to a preset fill rate threshold.

[0212] In step S702, the target recommended content is recommended to the target user.

[0213] Optionally, after the target user and the target recommended content are determined, the target recommended content may be recommended to the corresponding target user.

[0214] The disclosed embodiment obtains a recommended content set, which includes recommended content of multiple categories and meets preset conditions; determines evaluation parameter information corresponding to various categories based on historical record information of historical recommended content of the multiple categories by the user; the historical record information represents content information of various categories and corresponding operation information; determines a target category among the multiple categories based on the evaluation parameter information corresponding to the various categories; determines a first recommended content belonging to the target category from the recommended content set; obtains a second recommended content recalled for the target user, and determines the target recommended content based on the first recommended content and the second recommended content. In this way, by combining the recommended content that meets the preset conditions corresponding to the target category to determine the target recommended content recommended to the target user, it is possible to increase the exposure of the target recommended content without reducing the loss of user experience, fully tap more potential interest tags of the target user, improve the recommendation effect and user satisfaction, and thus improve user stickiness.

[0215] Figure 8 is a block diagram of a device for determining recommended content according to an exemplary embodiment. Figure 8 , the device is applied to electronic equipment, including:

[0216] A first acquisition module 810 is configured to acquire a recommended content set, wherein the recommended content set includes recommended content of multiple categories and meeting preset conditions;

[0217] The first determination module 820 is configured to determine the evaluation parameter information corresponding to each category according to the historical record information of the user's historical recommendation content of the multiple categories; the historical record information represents the content information of each category and the corresponding operation information;

[0218] A second determination module 830 is configured to determine a target category among the multiple categories according to the evaluation parameter information corresponding to the various categories;

[0219] A third determination module 840 is configured to determine, from the recommended content set, a first recommended content belonging to the target category;

[0220] The target content determination module 850 is configured to obtain the second recommended content recalled for the target user, and determine the target recommended content based on the first recommended content and the second recommended content.

[0221] As an optional implementation manner, the first acquisition module includes:

[0222] A first acquisition submodule is configured to acquire feedback information of recommended content corresponding to each category within a historical preset time period;

[0223] A second acquisition submodule is configured to parse the feedback information and acquire content data of each recommended content in at least three dimensions;

[0224] A determination submodule, configured to determine recommended content that meets preset conditions based on the content data and a preset threshold;

[0225] The establishing submodule is configured to establish the recommended content set according to the determined recommended content.

[0226] As an optional implementation manner, the first determining module includes:

[0227] A second acquisition module is configured to acquire historical record information of the user's historical recommended content of the multiple categories; the historical record information includes content feature information and the number of historical operations corresponding to each category;

[0228] A first determination module is configured to obtain historical average evaluation parameter information corresponding to different categories based on the content feature information;

[0229] A second determination module is configured to determine evaluation uncertainty parameters corresponding to different categories based on the number of historical operations corresponding to each category and the total number of historical operations of each category;

[0230] The score determination module is configured to obtain the evaluation parameter information corresponding to various categories according to the historical average evaluation parameter information and the evaluation uncertainty parameter.

[0231] As an optional implementation manner, the second determining module includes:

[0232] The target score determination submodule is configured to determine the evaluation parameter information that satisfies the preset evaluation conditions among the evaluation parameter information corresponding to various categories as the target evaluation parameter information;

[0233] The category determination submodule is configured to determine the category corresponding to the target evaluation parameter information as the target category.

[0234] As an optional implementation, the third determining module includes:

[0235] a candidate determination submodule, configured to determine candidate recommended content that meets a preset targeting condition based on the first recommended content and the second recommended content;

[0236] The target determination submodule is configured to process the candidate recommended content based on a preset recommendation strategy to obtain target recommended content for recommendation to the target user.

[0237] As an optional implementation manner, the candidate content determination submodule includes:

[0238] a channel configuration unit, configured to configure the first recommended content and the second recommended content to a first channel and at least one second channel respectively, each channel including a content tag for characterizing an attribute of the recommended content;

[0239] a merging unit, configured to merge the recommended contents in the first channel and the at least one second channel to obtain a merged recommended content set;

[0240] A desired content determination unit, configured to determine a desired recommended content set according to user feature information corresponding to the target user;

[0241] an intersection unit, configured to perform intersection processing on the combined recommended content set and the expected recommended content set according to a preset orientation condition of the combined recommended content, to obtain a candidate recommended content set;

[0242] The candidate content determination unit is configured to execute the candidate recommended content set as the candidate recommended content that meets the preset targeting condition.

[0243] As an optional implementation, the preset recommendation strategy includes a first recommendation strategy and a second recommendation strategy, the content tag includes a first content tag representing the first recommended content; and the target content determination submodule includes:

[0244] A first processing unit is configured to filter and sort the candidate recommended contents except those carrying the first content tag among the candidate recommended contents based on the first recommendation strategy to obtain a first subset to be recommended;

[0245] A second processing unit is configured to filter and sort the first subset to be recommended and the candidate recommended content carrying the first content tag based on the second recommendation strategy to obtain a second subset to be recommended;

[0246] A first target content determination unit is configured to, when the second subset to be recommended is a non-empty set, use the second subset to be recommended as target recommended content for recommending to the target user;

[0247] The second target content determination unit is configured to, when the second to-be-recommended subset is an empty set, use the candidate recommended content carrying the first content tag as the target recommended content for recommendation to the target user.

[0248] As an optional implementation manner, the first target content determination unit includes:

[0249] A first processing sub-unit is configured to perform diversity processing on the second subset to be recommended, to obtain a processed second subset to be recommended, when the second subset to be recommended is a non-empty set;

[0250] The first target content determination subunit is configured to use the processed second subset to be recommended as target recommended content for recommendation to the target user; or

[0251] The second target content determination unit includes:

[0252] The second processing sub-unit is configured to perform diversity processing on the candidate recommended content carrying the first content tag to obtain a processed third subset to be recommended when the second subset to be recommended is an empty set;

[0253] The second target content determination subunit is configured to use the processed third to-be-recommended subset as target recommended content for recommendation to the target user.

[0254] As an optional implementation, the device further includes:

[0255] A second acquisition module is configured to execute acquisition of the target user's posterior behavior data on the target recommended content;

[0256] A parsing module, configured to parse the extended interest tags of the target user according to the a posteriori behavior data;

[0257] The filtering module is configured to filter the target recommended content based on the extended interest tag to obtain updated target recommended content.

[0258] Fig. 9 is a block diagram of a recommendation device according to an exemplary embodiment. Fig. 9 , the device is applied to electronic equipment, including:

[0259] An acquisition module 910 is configured to acquire a target user and target recommended content for recommending to the target user, wherein the target recommended content is obtained according to any of the above-mentioned recommended content determination devices;

[0260] The recommendation module 920 is configured to recommend the target recommended content to the target user.

[0261] As an optional implementation, the target user satisfies at least one of the following conditions: users whose user activity is greater than or equal to a preset activity threshold, users whose application usage time is greater than or equal to a preset time threshold, users whose application cold start times are greater than or equal to a preset times threshold, and users whose media content fill rate is less than or equal to a preset fill rate threshold.

[0262] Regarding the device in the above embodiment, the specific manner in which each module performs operations has been described in detail in the embodiment of the method, and will not be elaborated here.

[0263] In an exemplary embodiment, an electronic device is also provided, the electronic device comprising a processor; a memory for storing instructions executable by the processor; wherein the processor is configured to implement any recommended content determination method or steps of the recommendation method in the above-mentioned embodiments when executing the instructions stored in the memory.

[0264] The electronic device may be a terminal, a server or a similar computing device. For example, the electronic device is a server. Fig.10 It is a block diagram of an electronic device for determining or recommending recommended content according to an exemplary embodiment. The electronic device 1000 may have relatively large differences due to different configurations or performances, and may include one or more central processing units (CPU) 1010 (the processor 1010 may include but is not limited to a processing device such as a microprocessor MCU or a programmable logic device FPGA), a memory 1030 for storing data, and one or more storage media 1020 (such as one or more mass storage devices) for storing application programs 1023 or data 1022. Among them, the memory 1030 and the storage medium 1020 can be short-term storage or permanent storage. The program stored in the storage medium 1020 may include one or more modules, and each module may include a series of instruction operations in the electronic device. Furthermore, the central processor 1010 can be configured to communicate with the storage medium 1020 and execute a series of instruction operations in the storage medium 1020 on the electronic device 1000.

[0265] The electronic device 1000 may also include one or more power supplies 1060, one or more wired or wireless network interfaces 1050, one or more input and output interfaces 1040, and / or one or more operating systems 1021, such as Windows Server, Mac OS X, Unix, Linux, FreeBSD, etc.

[0266] The input / output interface 1040 may be used to receive or send data via a network. The specific example of the network may include a wireless network provided by a communication provider of the electronic device 1000. In one example, the input / output interface 1040 includes a network adapter (Network Interface Controller, NIC), which can be connected to other network devices through a base station so as to communicate with the Internet. In an exemplary embodiment, the input / output interface 1040 may be a radio frequency (RF) module, which is used to communicate with the Internet wirelessly.

[0267] It can be understood by those skilled in the art that Fig.10 The structure shown is only for illustration and does not limit the structure of the above electronic device. Fig.10 More or fewer components as shown, or with Fig.10 Different configurations are shown.

[0268] In an exemplary embodiment, a computer-readable storage medium including instructions is also provided, such as a memory including instructions, and the above instructions can be executed by a processor of the electronic device 1000 to complete the above method. Alternatively, the computer-readable storage medium can be a ROM, a random access memory (RAM), a CD-ROM, a magnetic tape, a floppy disk, an optical data storage device, etc.

[0269] In an exemplary embodiment, a computer program product is also provided, the computer program product including a computer program or an instruction, and the computer program or the instruction is executed by a processor to implement the method provided in any of the above embodiments. Optionally, the computer program or the instruction is stored in a computer-readable storage medium. The processor of the electronic device reads the computer program or the instruction from the computer-readable storage medium, and the processor executes the computer program or the instruction, so that the electronic device performs the method provided in any of the above embodiments.

[0270] Those skilled in the art will readily appreciate other embodiments of the present disclosure after considering the specification and practicing the invention disclosed herein. This application is intended to cover any variations, uses, or adaptations of the present disclosure that follow the general principles of the present disclosure and include common knowledge or customary techniques in the art that are not disclosed in the present disclosure. The specification and examples are intended to be exemplary only, and the true scope and spirit of the present disclosure are indicated by the following claims.

[0271] It should be understood that the present disclosure is not limited to the exact structures that have been described above and shown in the drawings, and that various modifications and changes may be made without departing from the scope thereof. The scope of the present disclosure is limited only by the appended claims.

Claims

1. A method for determining recommended content, characterized in that: include: Acquire a recommended content set, wherein the recommended content set includes recommended content of multiple categories and meeting preset conditions; Determine evaluation parameter information corresponding to various categories according to the historical record information of the user for the historical recommended contents of the multiple categories; the historical record information represents the content information of various categories and the corresponding operation information; the evaluation parameter information is determined based on the historical average evaluation parameter information corresponding to the historical recommended contents of different categories and the corresponding evaluation uncertainty parameter, and the evaluation uncertainty parameter is determined based on the historical operation times corresponding to the historical recommended contents of each category and the total historical operation times of each category; Determining a target category among the multiple categories according to the evaluation parameter information corresponding to the various categories; The target category is a category corresponding to the target evaluation parameter information, and the target evaluation parameter information is a maximum evaluation parameter value in the evaluation parameter information corresponding to the multiple categories; Determining a first recommended content belonging to the target category from the recommended content set; Acquire second recommended content recalled for a target user, and determine target recommended content based on the first recommended content and the second recommended content.

2. The method for determining recommended content according to claim 1, characterized in that: The obtaining of the recommended content set comprises: Obtain feedback information on recommended content corresponding to each category within a historical preset time period; Parsing the feedback information to obtain content data of each recommended content in at least three dimensions; Determining recommended content that meets preset conditions based on the content data and a preset threshold; The recommended content set is established according to the determined recommended content.

3. The method for determining recommended content according to claim 1, characterized in that: Determining the evaluation parameter information corresponding to various categories according to the historical record information of the user's historical recommendation content of the multiple categories includes: Acquire historical record information of the user's historical recommended content for the multiple categories; the historical record information includes content feature information and the number of historical operations corresponding to each category; Based on the content feature information, obtaining historical average evaluation parameter information corresponding to different categories; Based on the number of historical operations corresponding to each category and the total number of historical operations of each category, the evaluation uncertainty parameters corresponding to different categories are determined; According to the historical average evaluation parameter information and the evaluation uncertainty parameter, the evaluation parameter information corresponding to various categories is obtained.

4. The method for determining recommended content according to any one of claims 1 to 3, characterized in that: Determining the target category among the multiple categories according to the evaluation parameter information corresponding to the various categories includes: Determine the evaluation parameter information that meets the preset evaluation conditions among the evaluation parameter information corresponding to various categories as the target evaluation parameter information; The category corresponding to the target evaluation parameter information is determined as the target category.

5. The method for determining recommended content according to any one of claims 1 to 3, characterized in that: The determining target recommended content based on the first recommended content and the second recommended content includes: Determining candidate recommended content that meets a preset targeting condition based on the first recommended content and the second recommended content; Based on a preset recommendation strategy, the candidate recommended content is processed to obtain target recommended content for recommendation to the target user.

6. The method for determining recommended content according to claim 5, characterized in that: The determining, based on the first recommended content and the second recommended content, candidate recommended content that meets a preset targeting condition includes: configuring the first recommended content and the second recommended content to a first channel and at least one second channel respectively, each channel including a content tag for representing an attribute of the recommended content; Merging the recommended contents in the first channel and the at least one second channel to obtain a merged recommended content set; Determining a desired recommended content set according to user characteristic information corresponding to the target user; According to the preset orientation condition of the combined recommended content, performing intersection processing on the combined recommended content set and the expected recommended content set to obtain a candidate recommended content set; The candidate recommended content set is used as candidate recommended content that meets a preset targeting condition.

7. The method for determining recommended content according to claim 6, characterized in that: The preset recommendation strategy includes a first recommendation strategy and a second recommendation strategy, and the content tag includes a first content tag representing the first recommended content; The processing of the candidate recommended content based on the preset recommendation strategy to obtain the target recommended content for recommendation to the target user includes: Based on the first recommendation strategy, processing the candidate recommended contents except those carrying the first content tag among the candidate recommended contents to obtain a first subset to be recommended; Based on the second recommendation strategy, the first subset to be recommended and the candidate recommended content carrying the first content tag are processed to obtain a second subset to be recommended; In the case that the second subset to be recommended is a non-empty set, using the second subset to be recommended as target recommended content for recommending to the target user; In a case where the second subset to be recommended is an empty set, the candidate recommended content carrying the first content tag is used as the target recommended content for recommendation to the target user.

8. The method for determining recommended content according to claim 7, characterized in that: In the case that the second subset to be recommended is a non-empty set, using the second subset to be recommended as target recommended content for recommending to the target user comprises: In the case that the second subset to be recommended is a non-empty set, performing diversity processing on the second subset to be recommended to obtain a processed second subset to be recommended; using the processed second subset to be recommended as target recommended content for recommending to the target user; or In the case where the second subset to be recommended is an empty set, taking the candidate recommended content carrying the first content tag as the target recommended content for recommendation to the target user comprises: In the case that the second subset to be recommended is an empty set, performing diversity processing on the candidate recommended content carrying the first content tag to obtain a processed third subset to be recommended; The processed third subset to be recommended is used as target recommended content for recommending to the target user.

9. The method for determining recommended content according to any one of claims 1 to 3, characterized in that: After the step of determining target recommended content based on the first recommended content and the second recommended content, the method further includes: Acquiring the target user's posterior behavior data on the target recommended content; parsing the extended interest tags of the target user according to the a posteriori behavior data; The target recommended content is filtered based on the extended interest tag to obtain updated target recommended content.

10. A recommendation method, characterized in that: include: Acquire a target user and target recommended content for recommending to the target user, wherein the target recommended content is obtained according to any one of the recommended content determination methods according to claims 1-9; The target recommended content is recommended to the target user.

11. The recommendation method according to claim 10, characterized in that: The target user meets at least one of the following conditions: users whose user activity is greater than or equal to a preset activity threshold, users whose application usage time is greater than or equal to a preset time threshold, users whose application cold start times are greater than or equal to a preset times threshold, and users whose media content fill rate is less than or equal to a preset fill rate threshold.

12. A device for determining recommended content, characterized in that: include: A first acquisition module is configured to execute acquisition of a recommended content set, wherein the recommended content set includes recommended content of multiple categories and meeting preset conditions; The first determination module is configured to determine evaluation parameter information corresponding to various categories according to the historical record information of the user for the historical recommended contents of the multiple categories; the historical record information represents the content information of various categories and the corresponding operation information; the evaluation parameter information is determined based on the historical average evaluation parameter information corresponding to the historical recommended contents of different categories and the corresponding evaluation uncertainty parameter, and the evaluation uncertainty parameter is determined based on the historical operation times corresponding to the historical recommended contents of each category and the total historical operation times of each category; A second determination module is configured to determine a target category among the multiple categories according to the evaluation parameter information corresponding to the various categories; The target category is a category corresponding to the target evaluation parameter information, and the target evaluation parameter information is a maximum evaluation parameter value in the evaluation parameter information corresponding to the multiple categories; A third determination module is configured to determine, from the recommended content set, a first recommended content belonging to the target category; The target content determination module is configured to obtain second recommended content recalled for a target user, and determine target recommended content based on the first recommended content and the second recommended content.

13. The recommended content determination device according to claim 12, characterized in that: The first acquisition module includes: A first acquisition submodule is configured to acquire feedback information of recommended content corresponding to each category within a historical preset time period; A second acquisition submodule is configured to parse the feedback information and acquire content data of each recommended content in at least three dimensions; A determination submodule, configured to determine recommended content that meets preset conditions based on the content data and a preset threshold; The establishing submodule is configured to establish the recommended content set according to the determined recommended content.

14. The recommended content determination device according to claim 12, characterized in that: The first determining module comprises: A second acquisition module is configured to acquire historical record information of the user's historical recommended content of the multiple categories; the historical record information includes content feature information and the number of historical operations corresponding to each category; A first determination module is configured to obtain historical average evaluation parameter information corresponding to different categories based on the content feature information; A second determination module is configured to determine evaluation uncertainty parameters corresponding to different categories based on the number of historical operations corresponding to each category and the total number of historical operations of each category; The score determination module is configured to obtain the evaluation parameter information corresponding to various categories according to the historical average evaluation parameter information and the evaluation uncertainty parameter.

15. The recommended content determination device according to any one of claims 12 to 14, characterized in that: The second determining module comprises: The target score determination submodule is configured to determine the evaluation parameter information that satisfies the preset evaluation conditions among the evaluation parameter information corresponding to various categories as the target evaluation parameter information; The category determination submodule is configured to determine the category corresponding to the target evaluation parameter information as the target category.

16. The recommended content determination device according to any one of claims 12 to 14, characterized in that: The third determination module comprises: a candidate determination submodule, configured to determine candidate recommended content that meets a preset targeting condition based on the first recommended content and the second recommended content; The target determination submodule is configured to process the candidate recommended content based on a preset recommendation strategy to obtain target recommended content for recommendation to the target user.

17. The recommended content determination device according to claim 16, characterized in that: The candidate determination submodule includes: a channel configuration unit, configured to configure the first recommended content and the second recommended content to a first channel and at least one second channel respectively, each channel including a content tag for characterizing an attribute of the recommended content; a merging unit, configured to merge the recommended contents in the first channel and the at least one second channel to obtain a merged recommended content set; A desired content determination unit, configured to determine a desired recommended content set according to user feature information corresponding to the target user; an intersection unit, configured to perform intersection processing on the combined recommended content set and the expected recommended content set according to a preset orientation condition of the combined recommended content, to obtain a candidate recommended content set; The candidate content determination unit is configured to execute the candidate recommended content set as the candidate recommended content that meets the preset targeting condition.

18. The recommended content determination device according to claim 17, characterized in that: The preset recommendation strategy includes a first recommendation strategy and a second recommendation strategy, and the content tag includes a first content tag representing the first recommended content; The target determination submodule includes: A first processing unit is configured to execute, based on the first recommendation strategy, processing the candidate recommended contents except those carrying the first content tag in the candidate recommended contents to obtain a first subset to be recommended; A second processing unit is configured to process the first subset to be recommended and the candidate recommended content carrying the first content tag based on the second recommendation strategy to obtain a second subset to be recommended; A first target content determination unit is configured to, when the second subset to be recommended is a non-empty set, use the second subset to be recommended as target recommended content for recommending to the target user; The second target content determination unit is configured to, when the second to-be-recommended subset is an empty set, use the candidate recommended content carrying the first content tag as the target recommended content for recommendation to the target user.

19. The recommended content determination device according to claim 18, characterized in that: The first target content determination unit includes: A first processing sub-unit is configured to perform diversity processing on the second subset to be recommended, to obtain a processed second subset to be recommended, when the second subset to be recommended is a non-empty set; The first target content determination subunit is configured to use the processed second subset to be recommended as target recommended content for recommendation to the target user; or The second target content determination unit includes: The second processing sub-unit is configured to perform diversity processing on the candidate recommended content carrying the first content tag to obtain a processed third subset to be recommended when the second subset to be recommended is an empty set; The second target content determination subunit is configured to use the processed third to-be-recommended subset as target recommended content for recommendation to the target user.

20. The recommended content determination device according to any one of claims 12 to 14, characterized in that: The device also includes: A second acquisition module is configured to execute acquisition of the target user's posterior behavior data on the target recommended content; A parsing module, configured to parse the extended interest tags of the target user according to the a posteriori behavior data; The filtering module is configured to filter the target recommended content based on the extended interest tag to obtain updated target recommended content.

21. A recommendation device, characterized in that: include: an acquisition module, configured to acquire a target user and target recommended content for recommending to the target user, wherein the target recommended content is obtained according to the recommended content determination device according to claim 12; The recommendation module is configured to recommend the target recommended content to the target user.

22. The recommendation device according to claim 21, characterized in that The target users include users who meet at least one of the following conditions: users whose user activity is greater than or equal to a preset activity threshold, users whose application usage time is greater than or equal to a preset time threshold, users whose application cold start times are greater than or equal to a preset times threshold, and users whose media content fill rate is less than or equal to a preset fill rate threshold.

23. An electronic device, characterized in that: include: processor; a memory for storing instructions executable by the processor; The processor is configured to execute the instructions to implement the recommended content determination method according to any one of claims 1 to 9, or the recommendation method according to claim 10 or 11.

24. A computer-readable storage medium, when the instructions in the computer-readable storage medium are executed by a processor of an electronic device, the electronic device is enabled to execute the recommended content determination method as described in any one of claims 1 to 9, or the recommendation method as described in claim 10 or 11.

25. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the computer program implements the recommended content determination method according to any one of claims 1 to 9, or the recommendation method according to claim 10 or 11.

Images