Data recommendation method, device, equipment and storage medium

Abstract

The embodiment of the present invention provides a data recommendation method, device, equipment and storage medium, the method comprising: obtaining a first data set corresponding to a user; filtering out data that does not meet the group characteristics in the first data set according to the group characteristics corresponding to the user to obtain a second data set; and recommending the second data set to the user. When data such as songs and videos need to be recommended for a certain user, the user group characteristics corresponding to the user are combined to make the data set recommended for the user meet the group characteristics, which can enhance the possibility that the data is preferred by the user, thereby enhancing user stickiness.

Description

Technical Field

[0001] The present invention relates to the field of Internet technology, and in particular to a data recommendation method, device, equipment and storage medium. Background Art

[0002] With the continuous development of artificial intelligence technology and Internet technology, more and more smart devices have been widely used in people's lives and work. Moreover, the way people interact with smart devices has also changed, from traditional buttons and touch methods to voice interaction and other methods.

[0003] Take smart speakers as an example. When users use smart speakers to listen to songs, they may not give a clear intention to listen to a certain song, but only give the intention to listen to songs. At this time, the smart speaker will recommend songs to users according to the set strategy to play for the users. If the songs recommended by the smart speaker are not liked by the user, then the user will have a poor user experience. Therefore, it is necessary to make targeted data recommendations such as songs for users to enhance user stickiness. Summary of the invention

[0004] Embodiments of the present invention provide a data recommendation method, apparatus, device and storage medium to recommend more suitable data to users.

[0005] In a first aspect, an embodiment of the present invention provides a data recommendation method, the method comprising:

[0006] Acquire a first data set corresponding to the user;

[0007] According to the group characteristics corresponding to the users, filtering out the data in the first data set that does not conform to the group characteristics, so as to obtain a second data set;

[0008] The second data set is recommended to the user.

[0009] In a second aspect, an embodiment of the present invention provides a data recommendation device, the device comprising:

[0010] An acquisition module, used to acquire a first data set corresponding to the user;

[0011] A filtering module, configured to filter out data in the first data set that does not conform to the group characteristics according to the group characteristics corresponding to the users, so as to obtain a second data set;

[0012] An output module is used to recommend the second data set to the user.

[0013] In a third aspect, an embodiment of the present invention provides an electronic device, comprising: a memory and a processor; wherein the memory stores executable code, and when the executable code is executed by the processor, the processor can at least implement the data recommendation method as described in the first aspect.

[0014] In a fourth aspect, an embodiment of the present invention provides a non-temporary machine-readable storage medium having executable code stored thereon. When the executable code is executed by a processor of an electronic device, the processor can at least implement the data recommendation method as described in the first aspect.

[0015] In a fifth aspect, an embodiment of the present invention provides another data recommendation method, the method comprising:

[0016] Determine the group characteristics of users;

[0017] Acquire a data set matching the group characteristics;

[0018] The data set is recommended to the user.

[0019] In a sixth aspect, an embodiment of the present invention provides another data recommendation device, the device comprising:

[0020] A determination module, used to determine group characteristics of users;

[0021] An acquisition module, used to acquire a data set matching the group characteristics;

[0022] An output module is used to recommend the second data set to the user.

[0023] In the seventh aspect, an embodiment of the present invention provides an electronic device, comprising: a memory and a processor; wherein the memory stores executable code, and when the executable code is executed by the processor, the processor can at least implement the data recommendation method as described in the fifth aspect.

[0024] In an eighth aspect, an embodiment of the present invention provides a non-temporary machine-readable storage medium having executable code stored thereon. When the executable code is executed by a processor of an electronic device, the processor can at least implement the data recommendation method as described in the fifth aspect.

[0025] In a ninth aspect, an embodiment of the present invention provides a data recommendation method, the method comprising:

[0026] Recommend data corresponding to various group characteristics to users;

[0027] filtering the data according to the user's feedback behavior on the data corresponding to each of the multiple group characteristics;

[0028] The filtered data is recommended to the user.

[0029] In a tenth aspect, an embodiment of the present invention provides a data recommendation device, the device comprising:

[0030] The first recommendation module is used to recommend data corresponding to various group characteristics to users;

[0031] A filtering module, used for filtering the data according to the feedback behavior of the user on the data corresponding to the multiple group characteristics;

[0032] The second recommendation module is used to recommend the filtered data to the user.

[0033] In the eleventh aspect, an embodiment of the present invention provides an electronic device, comprising: a memory and a processor; wherein the memory stores executable code, and when the executable code is executed by the processor, the processor can at least implement the data recommendation method as described in the ninth aspect.

[0034] In the twelfth aspect, an embodiment of the present invention provides a non-temporary machine-readable storage medium having executable code stored thereon. When the executable code is executed by a processor of an electronic device, the processor can at least implement the data recommendation method as described in the ninth aspect.

[0035] In a thirteenth aspect, an embodiment of the present invention provides a data recommendation method, the method comprising:

[0036] A first recommendation module, used for recommending first data corresponding to a plurality of group characteristics to a user;

[0037] a determination module, configured to determine at least one group characteristic matching the user among the multiple group characteristics according to the user's feedback behavior on the first data corresponding to each of the multiple group characteristics;

[0038] The second recommendation module is used to recommend second data corresponding to the at least one group characteristic to the user.

[0039] In a fourteenth aspect, an embodiment of the present invention provides a data recommendation device, the device comprising:

[0040] recommending first data corresponding to each of the plurality of group characteristics to the user;

[0041] Determining at least one group feature matching the user among the multiple group features according to the user's feedback behavior on the first data corresponding to each of the multiple group features;

[0042] Second data corresponding to the at least one group characteristic is recommended to the user.

[0043] In the fifteenth aspect, an embodiment of the present invention provides an electronic device, comprising: a memory and a processor; wherein the memory stores executable code, and when the executable code is executed by the processor, the processor can at least implement the data recommendation method as described in the thirteenth aspect.

[0044] In the sixteenth aspect, an embodiment of the present invention provides a non-temporary machine-readable storage medium having executable code stored thereon. When the executable code is executed by a processor of an electronic device, the processor can at least implement the data recommendation method as described in the thirteenth aspect.

[0045] In the data recommendation scheme provided in the embodiment of the present invention, when it is necessary to recommend data such as songs and videos to a certain user, the user group characteristics corresponding to the user are combined so that the data set recommended to the user is data that meets the group characteristics, which can enhance the possibility that these data are preferred by the user, thereby enhancing user stickiness. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0047] Figure 1 A flowchart of a data recommendation method provided by an embodiment of the present invention;

[0048] Figure 2 A schematic diagram of a data recommendation operation triggering method provided by an embodiment of the present invention;

[0049] Figure 3 A schematic diagram of a feedback behavior triggering method provided by an embodiment of the present invention;

[0050] Figure 4 A schematic diagram of a process for filtering data according to user group characteristics provided by an embodiment of the present invention;

[0051] Figure 5 A flowchart of another data recommendation method provided by an embodiment of the present invention;

[0052] Figure 6 A schematic diagram of a data recommendation scenario provided by an embodiment of the present invention;

[0053] Figure 7 A schematic diagram of another data recommendation scenario provided by an embodiment of the present invention;

[0054] Figure 8 A flowchart of another data recommendation method provided by an embodiment of the present invention;

[0055] Fig. 9 A flowchart of another data recommendation method provided by an embodiment of the present invention;

[0056] Fig.10 A flowchart of another data recommendation method provided by an embodiment of the present invention;

[0057] Fig.11 A schematic diagram of the structure of a data recommendation device provided by an embodiment of the present invention;

[0058] Fig.12 For Fig.11 A schematic diagram of the structure of an electronic device corresponding to the data recommendation device provided in the illustrated embodiment;

[0059] Fig.13 A schematic diagram of the structure of another data recommendation device provided by an embodiment of the present invention;

[0060] Fig.14 For Fig.13 A schematic diagram of the structure of an electronic device corresponding to the data recommendation device provided in the illustrated embodiment;

[0061] Fig.15 A schematic diagram of the structure of another data recommendation device provided by an embodiment of the present invention;

[0062] Fig.16 For Fig.15 A schematic diagram of the structure of an electronic device corresponding to the data recommendation device provided in the illustrated embodiment;

[0063] Fig.17 A schematic diagram of the structure of another data recommendation device provided by an embodiment of the present invention;

[0064] Fig.18 For Fig.17 A schematic structural diagram of an electronic device corresponding to the data recommendation device provided in the illustrated embodiment. DETAILED DESCRIPTION

[0065] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0066] The terms used in the embodiments of the present invention are only for the purpose of describing specific embodiments, and are not intended to limit the present invention. The singular forms "a", "said" and "the" used in the embodiments of the present invention and the appended claims are also intended to include plural forms, unless the context clearly indicates other meanings, and "multiple" generally includes at least two.

[0067] As used herein, the words "if" and "if" may be interpreted as "at the time of" or "when" or "in response to determining" or "in response to detecting", depending on the context. Similarly, the phrases "if it is determined" or "if (stated condition or event) is detected" may be interpreted as "when it is determined" or "in response to determining" or "when detecting (stated condition or event)" or "in response to detecting (stated condition or event)", depending on the context.

[0068] In addition, the step sequence in the following method embodiments is only an example and not a strict limitation.

[0069] The data recommendation method provided in the embodiment of the present invention can be executed by an electronic device, which can be a terminal device such as a PC, a laptop, a smart phone, a smart speaker, etc., or a server. The server can be a physical server including an independent host, or a virtual server, or a cloud server.

[0070] The data recommendation method provided in the embodiment of the present invention can be applied to different data recommendation scenarios. For example, the data that needs to be recommended to the user includes: songs, videos, articles and other types of data.

[0071] The data recommendation method provided by the embodiment of the present invention can be applied to situations where the user has not clearly stated the required data features. Taking the data as songs as an example, assuming that the user currently wants to use a smart speaker to listen to songs, the user only triggers a voice command such as "I want to listen to songs" to the smart speaker, and the voice command does not clearly indicate what kind of songs he wants to listen to. For another example, the user uses a music APP to listen to songs. Assuming that the homepage interface of the APP includes the function item "Listen Randomly", the user clicks on the function item to trigger a playback request that does not clearly indicate what kind of songs he wants to listen to.

[0072] Although the user has not explicitly stated what kind of data he or she specifically needs, the method provided in the embodiment of the present invention can be used to recommend data to the user. This can be combined with the user group characteristics corresponding to the user, so that the data set recommended to the user is data that meets the group characteristics, which can enhance the possibility that the data is preferred by the user, thereby enhancing user stickiness.

[0073] The following examples are used to illustrate the implementation process of the data recommendation method provided in this article.

[0074] Figure 1 A flowchart of a data recommendation method provided by an embodiment of the present invention is shown in FIG. Figure 1 As shown, the method comprises the following steps:

[0075] 101. Obtain a first data set corresponding to a user.

[0076] 102. According to group characteristics corresponding to the users, filter out data in the first data set that does not conform to the group characteristics to obtain a second data set.

[0077] 103. Recommend the second data set to the user.

[0078] The core idea of ​​the data recommendation solution provided in this embodiment is first summarized: first, when a user triggers a data recommendation operation, a first data set consisting of a large number of data that the user may be interested in can be obtained first; secondly, on the basis of the first data set, combined with the user's group characteristics, the data that does not match the user's group characteristics is filtered out to obtain a second data set; finally, the second data set is recommended to the user.

[0079] In actual applications, users can trigger data recommendation operations through some triggering methods. For example, taking the song recommendation scenario as an example, Figure 2 As shown, the user can say "XX Elf, play music!" to the smart speaker to trigger the song recommendation operation. After the smart speaker receives the voice command, it can execute the process of recommending songs for the user.

[0080] Specifically, for the acquisition process of the first data set, the first data set can be obtained by one or more acquisition methods. When multiple data acquisition methods are used, a certain amount of data can be obtained by each data acquisition method, and finally the data obtained by all data acquisition methods are aggregated to obtain the first data set.

[0081] Optionally, the above-mentioned multiple data acquisition methods include the following two acquisition methods:

[0082] First acquisition method: obtaining the first data set by combining individual user feedback behavior on recommended historical data.

[0083] Second acquisition method: when the user's feedback behavior cannot be obtained, the first data set can be obtained according to the default acquisition method.

[0084] Among them, the above feedback behavior reflects whether the user likes the data recommended to him.

[0085] In fact, the inability to obtain user feedback may include two actual situations: when a user is a newly registered user, when he triggers the data recommendation operation for the first time, it is impossible to obtain the user's feedback; when the user is not triggering the data recommendation operation for the first time, but the user has not triggered feedback on the historical data recommended to him, it is also impossible to obtain the user's feedback.

[0086] In the song recommendation scenario, the default acquisition method may be, for example, acquiring popular songs and new songs to form the first data set.

[0087] Taking the song recommendation scenario as an example, if a user collects or downloads a song, it is considered that the user likes the song. If the user blacklists a song or triggers a song change (play next song) operation when the song has just been played for a short time, it is considered that the user does not like the song.

[0088] In actual applications, a time range can be set, such as the past month or half a month. Therefore, after a user currently triggers a data recommendation operation, it is possible to query whether the user has triggered a certain feedback behavior on the historical data recommended to him within the time range, and determine which acquisition method to use to obtain the first data set based on the query result. For example, assuming that the query result is empty, that is, the user has not triggered feedback on the historical data recommended to him within the above time range, at this time, the above second acquisition method can be used to obtain the first data set. Assuming that it is queried that the user has triggered feedback on some historical data recommended to him within the above time range, at this time, the above first acquisition method can be used to obtain the first data set, or the first acquisition method and the second acquisition method can be used at the same time to obtain the first data set.

[0089] It is understandable that since the user's feedback behavior on historical data can be queried, it means that after some data is recommended to the user, the corresponding relationship between the user identifier, the recommended data and the user's feedback behavior on the recommended data can be recorded. In practical applications, the life cycle of the above-mentioned recording results can be set, such as one month or three months. After reaching the life cycle, the recording results can be deleted to reduce the occupancy of storage resources. In addition, optionally, when there are multiple data to be recommended to the user, only part of the multiple data may be actually used by the user, and the remaining part may not be used by the user. Therefore, the above-mentioned recording operation can also be performed when the user actually uses a certain recommended data. Moreover, if the user only uses the recommended data and does not trigger feedback behavior on the recommended data, the feedback behavior corresponding to the recommended data can be recorded as empty.

[0090] For different electronic devices, the way users trigger feedback behavior can be different. For example, in the song recommendation scenario, if a user uses a music app installed on a mobile phone to listen to music, when playing any recommended song, the user can use the preset touch buttons in the display interface of the music app to trigger feedback behaviors such as collecting, liking, replaying, blacklisting, switching to the next song, etc. for the currently playing song. For another example, if a user uses a smart speaker to listen to music, when playing any recommended song, the user can use voice interaction to indicate whether he is interested in the currently playing song. Figure 3 As shown, the user can say "Collect this song" to the smart speaker, which means that the user likes the currently playing song. Therefore, the smart speaker can respond to the user "OK, the currently playing song has been collected to your playlist", and can also add the currently playing song to the playlist. It can record that the user has triggered the feedback behavior of liking the song. On the contrary, if the smart speaker receives the user's voice saying "I don't like this song", the smart speaker can record that the user has triggered the feedback behavior of not liking the song.

[0091] In summary, optionally, obtaining the first data set corresponding to the user can be implemented as follows: obtaining historical data preferred by the user based on the user's feedback behavior on the recommended historical data; obtaining data that meets the similarity condition between the historical data preferred by the user, so that the first data set includes the data that meets the similarity condition.

[0092] It can be understood that obtaining the historical data of user preferences based on the user's feedback behavior on the recommended historical data means that the user's feedback behavior can be divided into two types: one is positive feedback behavior, indicating that the user likes a certain historical data; the other is negative feedback behavior, indicating that the user does not like a certain historical data. Therefore, the historical data corresponding to the positive feedback behavior can be filtered out as the historical data of the user's preferences. Taking the song recommendation scenario as an example, positive feedback behaviors include collection, download, and like, and negative feedback behaviors include blacklisting and switching songs. After obtaining the historical data of user preferences, the data features of each historical data of the user's preferences can be determined, and then, any historical data X of the user's preferences can be used to identify the historical data of the user's preferences. i For example, based on historical data X i The data characteristics of each data in the database are used to determine the historical data X i The similarity between each data in the database is calculated, and all or a set number of data with similarity higher than a preset threshold are selected in the database as data meeting the similarity condition.

[0093] Data features can be one or more attributes of data. For example, the singer, style, and era of a song can all be used as features of the song. Therefore, determining the similarity of two data based on data features can be implemented as follows: determining the similarity between the two data based on the degree of matching of at least one data feature of the two data. For example, assuming that the data features of a song are singer, style, and era, and assuming that song A and song B are consistent in these three features, the similarity between the two songs can be considered to be 100%.

[0094] In practical applications, the data in the first data set includes some data determined by expanding the correlation (similarity) data based on some feedback behaviors of users on the recommended historical data. However, the fact that a user likes a certain historical data does not mean that the user actually likes other data similar to the historical data. If these data obtained based on similarity are recommended to users, it may cause dissatisfaction among users. Therefore, it is necessary to provide a solution to further measure whether the data in the first data set matches the current user. The solution provided by the embodiment of the present invention is to further filter the data in the first data set from the perspective of user group characteristics to filter out the data in the first data set that does not meet the group characteristics of the user to obtain the second data set.

[0095] Optionally, the group characteristics of users may include one or more dimensions such as age, gender, and region.

[0096] In actual applications, a variety of group characteristics can be preset, so that, for a current user, the group characteristics corresponding to the user can be determined based on the personal information (such as registration information) provided by the user.

[0097] For example, assuming that the dimensions for defining a group feature include gender and age, multiple age groups can be pre-set, such as the post-70s (1970-1979), post-80s, and post-90s age groups, thus obtaining the following group features:

[0098] [Born in the 1970s, male]; [Born in the 1970s, female]; [Born in the 1980s, male]; [Born in the 1980s, female]; [Born in the 1990s, male]; [Born in the 1990s, female].

[0099] Assuming that the user who currently triggers the data recommendation operation is a female born in 1988, the group characteristic corresponding to the user is [born in 1980s, female]. Based on this, in summary, for each data included in the first data set, it can be determined whether each data matches the group characteristic corresponding to the user, and the data that does not match the group characteristic is filtered out from the first data set to obtain the second data set.

[0100] Combine the following Figure 4 The embodiment shown in the figure exemplifies a scheme for filtering out data in the first data set that does not conform to the group characteristics of the user. Figure 4 As shown, the filtering scheme may include the following steps:

[0101] 401. Acquire multiple user groups, each of which corresponds to a different group feature, and the multiple group features corresponding to the multiple user groups include the group feature corresponding to the current user.

[0102] Taking the song recommendation scenario as an example, a large number of users who have used the song recommendation function can be collected, such as 9,000 users. Based on the personal information provided by these users, the group characteristics corresponding to each of these users can be determined. Therefore, users corresponding to the same group characteristics can be classified into a user group.

[0103] Still taking the several group characteristics mentioned above as an example, assume that these 9,000 users correspond to the several group characteristics mentioned above, and assume that the user group corresponding to [born in the 1970s, male] is called user group 1; the user group corresponding to [born in the 1970s, female] is called user group 2; the user group corresponding to [born in the 1980s, male] is called user group 3; the user group corresponding to [born in the 1980s, female] is called user group 4; the user group corresponding to [born in the 1990s, male] is called user group 5; the user group corresponding to [born in the 1990s, female] is called user group 6. Assume that the user who currently triggers the data recommendation operation corresponds to the group characteristic: [born in the 1980s, female].

[0104] Optionally, in actual applications, for each group characteristic, a set number of users matching the group characteristic may be obtained as a user group corresponding to the group characteristic, so that the numbers of users in the above multiple user groups are equal to each other.

[0105] 402. Determine, for target data in the first data set, the preferences of multiple user groups for the target data.

[0106] In this embodiment, the target data may be any data in the first data set.

[0107] Optionally, determining the preferences of multiple user groups for target data may be implemented as follows:

[0108] For any user group among the multiple user groups, a first ratio of the number of users in the user group who have used the target data to the total number of users corresponding to the user group is determined as the preference of the user group for the target data.

[0109] For example, suppose the target data is represented by Y i , taking user group 1 as an example, assuming that user group 1 includes 1000 users, and assuming that statistics of these 1000 users on the target data Y in the past set time i If we find that 300 of them have used this target data Y i , then it can be determined that user group 1 has the target data Y i The preference is: 300 / 1000.

[0110] Taking the song recommendation scenario as an example, the target data Y i is a song, and the user has used target data Y i It means that the user has listened to this song.

[0111] In addition, as mentioned above, for any user, the historical data recommended to the user and the user's feedback behavior on the historical data can be recorded. Based on this, optionally, the first ratio can also be adjusted according to the feedback behavior of users in any user group on the target data.

[0112] For example, as mentioned above, feedback behavior can be positive feedback behavior and negative feedback behavior. For the 300 users in the above-mentioned user group 1 who have used the target data, it is possible to count whether there are positive feedback behaviors for the target data among these 300 users. When the statistical results show that a certain number of users have triggered positive feedback behaviors for the target data, the first ratio can be multiplied by a weight coefficient greater than 1 according to the set rules. On the contrary, it is also possible to count whether there are negative feedback behaviors for the target data among the remaining 700 users. When the statistical results show that a certain number of users have triggered negative feedback behaviors for the target data, the first ratio can be multiplied by a weight coefficient less than 1 according to the set rules.

[0113] 403. Determine a target user group corresponding to the target data according to the preferences of the multiple user groups for the target data.

[0114] Optionally, determining the target user group corresponding to the target data may be implemented as follows:

[0115] Determine a second ratio of the number of users in the multiple user groups who have used the target data to the total number of users corresponding to the multiple user groups; determine the target user group as a user group in the multiple user groups for which the first ratio is greater than the second ratio.

[0116] The first ratio is considered from the perspective of each user group, and the second ratio is considered from the perspective of multiple user groups. Therefore, the second ratio can be considered as a measure of the overall user impact of multiple user groups on the target data Y. i The process of determining the second ratio is similar to that of determining the first ratio, and will not be described in detail here.

[0117] Now assume that the first ratio corresponding to user group 1 is a, the first ratio corresponding to user group 2 is b, the first ratio corresponding to user group 3 is c, the first ratio corresponding to user group 4 is d, the first ratio corresponding to user group 5 is e, and the first ratio corresponding to user group 6 is f. Assume that the second ratio is g. In addition, assume that a and b are greater than g, and a is greater than b; c, d, e, and f are all less than g.

[0118] Therefore, optionally, it can be considered that the first user group and the second user group corresponding to a and b are both i The corresponding target user group, or the first user group corresponding to a is considered to be the target user group.

[0119] In practical applications, in addition to the above-mentioned method, optionally, after determining each user group's response to the target data Y i After determining the preference, the user group with the highest preference can be directly selected from multiple user groups as the target user group, or the user group with a preference greater than a preset threshold can be selected from multiple user groups as the target user group.

[0120] 404. If the group characteristics corresponding to the target user group do not match the group characteristics corresponding to the user, filter out the target data from the first data set.

[0121] Based on the above example, assuming that the target data Y i The corresponding target user groups are the first user group and the second user group. The group characteristics of these two user groups are: [born in the 1970s, male] and [born in the 1970s, female]. The group characteristics of the user are [born in the 1980s, female]. Therefore, the group characteristics of the target user group do not match the group characteristics of the user, so the target data Y is filtered out from the first data set. i .

[0122] Based on the above process, after the above filtering process is performed on each data in the first data set, part of the data in the first data set may be filtered out to obtain the second data set, and then the second data set is recommended to the user.

[0123] Among them, taking the song recommendation scenario as an example, when the user listens to music through a smart speaker, the recommendation may mean that the smart speaker plays multiple songs in the second data set one by one in sequence.

[0124] Of course, optionally, before recommending the data in the second data set to the user, the data in the second data set may be sorted.

[0125] In summary, based on the above data recommendation scheme, when it is necessary to recommend data such as songs and videos for a certain user, the user group characteristics corresponding to the user are combined so that the data set recommended to the user is data that meets the group characteristics, which can enhance the possibility that these data are preferred by the user, thereby enhancing user stickiness.

[0126] Figure 5 A flowchart of another data recommendation method provided by an embodiment of the present invention is shown in FIG. Figure 5 As shown, the method comprises the following steps:

[0127] 501. Obtain a first data set corresponding to a user.

[0128] 502. Obtain historical data recommended to the user within a set time period, and if the first data set includes the historical data, filter out the historical data from the first data set.

[0129] 503. According to the user's feedback behavior on the recommended historical data, obtain the historical data that the user does not like. If the first data set includes the historical data that the user does not like, filter out the historical data that the user does not like from the first data set.

[0130] 504. According to the group characteristics corresponding to the users, filter out the data in the first data set that does not conform to the group characteristics to obtain a second data set.

[0131] 505. Recommend the second data set to the user.

[0132] In this embodiment, in addition to filtering the data in the first data set based on the group characteristics of users, the following two filtering strategies may also be included: fatigue filtering (corresponding to step 502) and blacklist filtering (corresponding to step 503).

[0133] Among them, fatigue filtering is to avoid recommending the same data to users repeatedly, and blacklist filtering is to avoid recommending data that users have clearly expressed that they do not like.

[0134] For example, the current first data set includes song A, song B and song C. It is determined that song C was recommended to the user 3 days ago, so song C can be filtered out from the first data set. After filtering out song C, song A and song B remain in the first data set.

[0135] For example, song A, song B and song C have been recommended to the user, but the user is not interested in song C and has added song C to the blacklist. At this time, if song C is included in the first data set, song C can be directly filtered out from the first data set. After song C is filtered out, song A and song B remain in the first data set.

[0136] As described above, the above fatigue filtering and blacklist filtering can be implemented based on the record results of the historical data recommended to the user and the user's feedback behavior on the historical data.

[0137] It is worth noting that, for fatigue filtering, in actual applications, the number of times that a certain data is allowed to be repeatedly recommended within a set time range can be set, such as 0 or 1, where 0 means that repeated recommendation is not allowed, and 1 means that it can be repeatedly recommended once. Therefore, for a certain data in the first data set, if the number of repeated recommendations is set to 0, it means that if the data appears in the historical data recommended for the user, it needs to be filtered out from the first data set; and if the number of repeated recommendations is set to 1, it means that if the data appears in the historical data recommended for the user once, it can currently be retained in the first data set.

[0138] For ease of understanding, the following Figure 6 , Figure 7 The execution process of the above data recommendation method in different data recommendation scenarios is exemplified.

[0139] exist Figure 6In the example, it is assumed that the user currently wants to use a smart speaker to listen to songs, and the user triggers a voice command such as "I want to listen to songs" to the smart speaker. After receiving the voice command, the smart speaker can obtain the first song set corresponding to the user in combination with the user's feedback behavior on the songs recommended to him in the past. The first song set includes songs a, b, c and d. Assuming that the user who currently triggers the data recommendation operation is a female born in 1988, the group characteristics corresponding to the user are [born in the 1980s, female]. Based on the group characteristics corresponding to the user, the songs in the first song set that do not meet the group characteristics of [born in the 1980s, female] can be filtered out to obtain the second song set. Assuming that song d is a song that does not meet the group characteristics of [born in the 1980s, female], the second song set includes songs a, b and c. Finally, songs a, b and c can be played in sequence through the smart speaker.

[0140] exist Figure 7 In the example, suppose the user wants to browse some news. At this time, the user can open the news information APP in the mobile phone. The homepage of the news information APP can automatically recommend some news that the user is interested in. After starting the news information APP, the first news set corresponding to the user can be obtained. Suppose the user often reads some entertainment news in the past. At this time, some entertainment news can be found as news in the first news set, and the first news set includes news A, news B, news C and news D. Suppose the user who currently triggers the data recommendation operation is a female born in 1990, then the group characteristics corresponding to the user are [post-90s, female]. Based on the group characteristics corresponding to the user, the news in the first news set that does not meet the group characteristics of [post-90s, female] can be filtered out to obtain the second news set. Suppose news C is news that does not meet the group characteristics of [post-90s, female], and the second news set includes news A, news B and news D. Finally, news A, news B and news D can be displayed on the homepage of the news information APP.

[0141] Figure 6 and Figure 7 For the specific implementation process in the illustrated embodiment, reference may be made to the relevant descriptions in the aforementioned embodiments.

[0142] Figure 8 A flowchart of another data recommendation method provided by an embodiment of the present invention, such as Figure 8 As shown, the data recommendation method may include the following steps:

[0143] 801. Determine group characteristics of users.

[0144] 802. Obtain a data set that matches the group characteristics.

[0145] 803. Recommend the data set to the user.

[0146] The core idea of ​​the data recommendation solution provided in this embodiment is first summarized: First, when a user triggers a data recommendation operation, a data set matching the user's group characteristics can be directly obtained based on the user's group characteristics, and the data set can be recommended to the user.

[0147] Optionally, the group characteristics of users may include one or more dimensions such as age, gender, and region.

[0148] In actual applications, a variety of group characteristics can be preset, so that, for a current user, the group characteristics corresponding to the user can be determined based on the personal information (such as registration information) provided by the user.

[0149] For example, assuming that the dimensions for defining a group feature include gender and age, multiple age groups can be pre-set, such as the post-70s (1970-1979), post-80s, and post-90s age groups, thus obtaining the following group features:

[0150] [Born in the 1970s, male]; [Born in the 1970s, female]; [Born in the 1980s, male]; [Born in the 1980s, female]; [Born in the 1990s, male]; [Born in the 1990s, female].

[0151] Assuming that the user who currently triggers the data recommendation operation is a female born in 1988, the group characteristics corresponding to this user are [born in the 1980s, female].

[0152] Based on this, a data set matching the group characteristics of the user can be obtained from the database based on the group characteristics of the user.

[0153] The above-mentioned acquisition of a data set matching the group characteristics can be implemented as follows: acquiring multiple user groups, each of the multiple user groups corresponding to a different group characteristic, and these group characteristics include the group characteristic corresponding to the current user; for target data in multiple candidate data, determining the preference of the multiple user groups for the target data; determining the target user group corresponding to the target data based on the preference of the multiple user groups for the target data; if the group characteristic corresponding to the target user group matches the group characteristic corresponding to the user, determining that the data set includes the target data.

[0154] The candidate data may be all the data contained in the database. For example, in the song recommendation scenario, the candidate data may be all the songs contained in the song database.

[0155] Optionally, after acquiring the above data set, the following steps may be further included: acquiring historical data recommended to the user within a set time; if the data set includes the historical data, filtering out the historical data from the data set.

[0156] Optionally, after obtaining the above data set, the following steps may also be included: obtaining historical data that the user does not like based on the user's feedback behavior on the recommended historical data; if the data set includes the historical data, filtering the historical data from the data set.

[0157] The acquisition process and filtering process of the above data set can refer to the relevant descriptions in the above other embodiments, which will not be repeated here.

[0158] Based on the data recommendation solution provided in this embodiment, when it is necessary to recommend data such as songs and videos to a certain user, the user group characteristics corresponding to the user are combined so that the data set recommended to the user is data that meets the group characteristics, which can enhance the possibility that these data are preferred by the user, thereby enhancing user stickiness.

[0159] Fig. 9 A flowchart of another data recommendation method provided by an embodiment of the present invention is shown in FIG. Fig. 9 As shown, the data recommendation method may include the following steps:

[0160] 901. Recommend data corresponding to various group characteristics to users.

[0161] 902. Filter the data corresponding to various group characteristics according to the user's feedback behavior.

[0162] 903. Recommend the filtered data to the user.

[0163] In the data recommendation method provided in this embodiment, firstly, for multiple group characteristics, multiple pieces of data corresponding to each group characteristic are obtained, so that a data set consisting of multiple pieces of data corresponding to each of the multiple group characteristics is recommended to the user. Secondly, when the user triggers a certain feedback behavior on some data in the data set, the data set can be filtered according to the feedback behavior to filter out the data that the user does not like. Finally, the remaining data set after filtering is recommended to the user.

[0164] Among them, the above filtering process, for example, is: assuming that the user triggers reverse feedback behavior (such as blacklisting, song switching, etc.) on some of the multiple data corresponding to group feature A, it means that the user does not like these data. At this time, all or part of the data corresponding to group feature A can be filtered out from the above data set.

[0165] For ease of understanding, take an actual application scenario as an example, assuming that the group feature is specifically implemented as an age feature, and assuming that the above multiple group features include: post-70s, post-80s and post-90s, that is, the 70s, 80s and 90s. Assume that initially, 10 songs matching post-70s, 10 songs matching post-80s, and 10 songs matching post-90s are recommended to the user. If the user triggers reverse feedback behaviors such as song switching and blacklisting for some of the 10 songs matching post-70s, it means that the user is likely to dislike such old songs. Therefore, these 10 songs matching post-70s can be deleted.

[0166] In addition, in this embodiment, optionally, the multiple group features in step 901 may be all the group features that have been generated, or some of the group features selected therefrom.

[0167] Among them, when some group characteristics are selected from all group characteristics as the multiple group characteristics, optionally, some can be randomly selected, or a certain number of high-level group characteristics can be selected according to the levels of these group characteristics. Among them. The level of a certain group characteristic can be determined by counting the preference of the data corresponding to the group characteristic by the majority of users. For example, assuming that within a certain statistical period, it is counted that there are 1,000 positive feedback behaviors for the data corresponding to group characteristic X, and there are 700 positive feedback behaviors for the data corresponding to group characteristic Y, then it can be considered that the level of group characteristic X is higher than the level of group characteristic Y.

[0168] In addition, in order to improve the relevance of the above-mentioned multiple group characteristics to the user who currently needs data recommendation, optionally, the above-mentioned multiple group characteristics corresponding to the user can also be determined based on the user's feedback behavior on the historical recommended data. Specifically, assuming that before the current data recommendation is made to the user, a lot of data has been recommended to the user, and these data are the historical recommended data. Assuming that the user has triggered positive feedback behavior on some of the historical recommended data, it means that the user prefers this part of the data, and thus, the group characteristics corresponding to each of these parts of the data are determined, and the group characteristics corresponding to these parts of the data are summarized to obtain multiple group characteristics corresponding to the user.

[0169] It is worth noting that, in this embodiment, the determination of the correspondence between data and group characteristics can refer to the relevant descriptions in the aforementioned other embodiments, which will not be repeated here.

[0170] Fig.10 A flowchart of another data recommendation method provided by an embodiment of the present invention, such as Fig.10 As shown, the data recommendation method may include the following steps:

[0171] 1001. Recommend first data corresponding to a plurality of group characteristics to a user.

[0172] 1002. Determine at least one group feature that matches the user among the multiple group features according to the user's feedback behavior on the first data corresponding to each of the multiple group features.

[0173] 1003. Recommend second data corresponding to the at least one group characteristic to the user.

[0174] The core idea of ​​the data recommendation method provided in this embodiment is: first recommend a small amount of data (the amount of data can be preset) corresponding to each of the multiple group characteristics to the user, and then, based on the user's feedback behavior on these data, guess which one or several group characteristics of the multiple group characteristics the user may match, and then obtain a large amount of data corresponding to the one or several group characteristics and continue to recommend them to the user.

[0175] Among them, in step 1002, based on the user's feedback behavior on the first data corresponding to each of the multiple group characteristics, at least one group characteristic among the multiple group characteristics that matches the user is determined. Simply put, it can be implemented as follows: if the user triggers a positive feedback behavior on any first data corresponding to a certain group characteristic, then this group characteristic is considered to match the user, wherein the positive feedback behavior includes, for example, likes, favorites, etc.

[0176] In this embodiment, the methods for obtaining the above-mentioned various group characteristics can refer to Fig. 9 The description in the illustrated embodiment will not be repeated here. In addition, the determination of the correspondence between data and group characteristics can refer to the relevant description in the aforementioned other embodiments, which will not be repeated here.

[0177] In addition, in another optional embodiment, the data recommendation method can also be implemented as the following steps:

[0178] recommending data corresponding to the first group characteristic to the user;

[0179] If the user triggers a reverse feedback behavior for the data corresponding to the first group feature, the data corresponding to the second group feature is recommended to the user. At this time, the data corresponding to the first group feature can be stopped from being recommended to the user.

[0180] The data recommendation device of one or more embodiments of the present invention will be described in detail below. Those skilled in the art will appreciate that these data recommendation devices can be configured using commercially available hardware components through the steps taught in this solution.

[0181] Fig.11 A structural diagram of a data recommendation device provided by an embodiment of the present invention is shown in FIG. Fig.11As shown, the device includes: an acquisition module 11, a filtering module 12, and an output module 13.

[0182] The acquisition module 11 is used to acquire a first data set corresponding to the user.

[0183] The filtering module 12 is used to filter out data in the first data set that does not conform to the group characteristics according to the group characteristics corresponding to the users, so as to obtain a second data set.

[0184] The output module 13 is configured to recommend the second data set to the user.

[0185] Optionally, the acquisition module 11 can be specifically used to: acquire the historical data preferred by the user based on the user's feedback behavior on the recommended historical data; acquire data that meets the similarity condition with the historical data preferred by the user, and the first data set includes the data that meets the similarity condition.

[0186] Optionally, the filtering module 12 may also be used to: obtain historical data recommended to the user within a set time; and if the first data set includes the historical data, filter out the historical data from the first data set.

[0187] Optionally, the filtering module 12 may also be used to: obtain historical data that the user does not like based on the user's feedback behavior on the recommended historical data; and filter out the historical data from the first data set if the first data set includes the historical data.

[0188] Optionally, the filtering module 12 can also be used to: obtain multiple user groups, each of the multiple user groups corresponds to a different group characteristic, and the group characteristics include the group characteristics corresponding to the user; for the target data in the first data set, determine the preferences of the multiple user groups for the target data; based on the preferences of the multiple user groups for the target data, determine the target user group corresponding to the target data; if the group characteristics corresponding to the target user group do not match the group characteristics corresponding to the user, filter out the target data from the first data set.

[0189] Optionally, the filtering module 12 can also be used to: for any user group among the multiple user groups, determine a first ratio of the number of users in the any user group who have used the target data to the total number of users corresponding to the any user group as the preference of the any user group for the target data.

[0190] Optionally, the filtering module 12 may also be configured to adjust the first ratio according to feedback behavior of users in any user group on the target data.

[0191] Optionally, the filtering module 12 can also be used to: determine a second ratio of the number of users in the multiple user groups who have used the target data to the total number of users corresponding to the multiple user groups; determine the target user group as the user group among the multiple user groups for which the first ratio is greater than the second ratio.

[0192] Optionally, the filtering module 12 can also be used to: if the first ratio of at least two user groups among the multiple user groups is greater than the second ratio, determine that the user group with the largest first ratio among the at least two user groups is the target user group, or determine that both of the at least two user groups are the target user groups.

[0193] Optionally, the second data set includes a plurality of songs, and the output module 13 can be specifically used to control the audio playback device to play the plurality of songs in sequence.

[0194] Fig.11 The device shown can perform the aforementioned Figures 1 to 7 The data recommendation method provided in the illustrated embodiment, the detailed execution process and technical effects can be found in the description of the aforementioned embodiment, which will not be repeated here.

[0195] In one possible design, the above Fig.11 The structure of the data recommendation device shown can be implemented as an electronic device, such as Fig.12 As shown, the electronic device may include: a first processor 21 and a first memory 22. The first memory 22 stores executable code. When the executable code is executed by the first processor 21, the first processor 21 can at least implement the above-mentioned Figures 1 to 7 The data recommendation method provided in the illustrated embodiment.

[0196] Optionally, the electronic device may further include a first communication interface 23 for communicating with other devices.

[0197] In addition, an embodiment of the present invention provides a non-transitory machine-readable storage medium, wherein an executable code is stored on the non-transitory machine-readable storage medium. When the executable code is executed by a processor of an electronic device, the processor can at least implement the above-mentioned Figures 1 to 7 The data recommendation method provided in the illustrated embodiment.

[0198] Fig.13 A structural diagram of another data recommendation device provided by an embodiment of the present invention, such as Fig.13 As shown, the device includes: a determination module 31, an acquisition module 32, and an output module 33.

[0199] The determination module 31 is used to determine the group characteristics of users.

[0200] The acquisition module 32 is used to acquire a data set matching the group characteristics.

[0201] The output module 33 is used to recommend the second data set to the user.

[0202] Optionally, the acquisition module 32 can be specifically used to: acquire multiple user groups, each of the multiple user groups corresponds to a different group characteristic, and the group characteristics include the group characteristics corresponding to the user; for target data in multiple candidate data, determine the preferences of the multiple user groups for the target data; based on the preferences of the multiple user groups for the target data, determine the target user group corresponding to the target data; if the group characteristics corresponding to the target user group match the group characteristics corresponding to the user, determine that the data set includes the target data.

[0203] Optionally, in the process of determining the preferences of the multiple user groups for the target data, the acquisition module 32 can be specifically used to: for any user group among the multiple user groups, determine a first ratio of the number of users in the any user group who have used the data to the total number of users corresponding to the any user group as the preference of the any user group for the target data.

[0204] Optionally, the acquisition module 32 may also be configured to adjust the first ratio according to feedback behavior of users in any user group on the target data.

[0205] Optionally, the acquisition module 32 can also be used to: determine a second ratio of the number of users in the multiple user groups who have used the data to the total number of users corresponding to the multiple user groups; determine the target user group as the user group in the multiple user groups for which the first ratio is greater than the second ratio.

[0206] Optionally, the acquisition module 32 may also be used to: if the first ratio of at least two user groups among the multiple user groups is greater than the second ratio, determine that the user group with the largest first ratio among the at least two user groups is the target user group, or determine that both of the at least two user groups are the target user groups.

[0207] Optionally, the plurality of data included in the data set are a plurality of songs. The output module 33 may be specifically used to control the audio playback device to play the plurality of songs in sequence.

[0208] Optionally, the acquisition module 32 may also be used to: acquire historical data recommended to the user within a set time; and if the data set includes the historical data, filter out the historical data from the data set.

[0209] Optionally, the acquisition module 32 may also be used to: acquire historical data that the user does not like based on the user's feedback behavior on the recommended historical data; and filter out the historical data from the data set if the data set includes the historical data.

[0210] Fig.13 The device shown can perform the aforementioned Figure 8 The data recommendation method provided in the illustrated embodiment, the detailed execution process and technical effects can be found in the description of the aforementioned embodiment, which will not be repeated here.

[0211] In one possible design, the above Fig.13 The structure of the data recommendation device shown can be implemented as an electronic device, such as Fig.14 As shown, the electronic device may include: a second processor 41 and a second memory 42. The second memory 42 stores executable code. When the executable code is executed by the second processor 41, the second processor 41 can at least implement the above-mentioned Figure 8 The data recommendation method provided in the illustrated embodiment.

[0212] Optionally, the electronic device may further include a second communication interface 43 for communicating with other devices.

[0213] Fig.15 A structural diagram of another data recommendation device provided by an embodiment of the present invention, such as Fig.15 As shown, the device includes: a first recommendation module 51, a filtering module 52, and a second recommendation module 53.

[0214] The first recommendation module 51 is used to recommend data corresponding to various group characteristics to users.

[0215] The filtering module 52 is used to filter the data according to the feedback behavior of the user on the data corresponding to the multiple group characteristics.

[0216] The second recommendation module 53 is used to recommend the filtered data to the user.

[0217] Fig.15 The device shown can perform the aforementioned Fig. 9 The data recommendation method provided in the illustrated embodiment, the detailed execution process and technical effects can be found in the description of the aforementioned embodiment, which will not be repeated here.

[0218] In one possible design, the above Fig.15 The structure of the data recommendation device shown can be implemented as an electronic device, such as Fig.16 As shown, the electronic device may include: a third processor 61 and a third memory 62. The third memory 62 stores executable code. When the executable code is executed by the third processor 61, the third processor 61 can at least implement the above-mentioned Fig. 9 The data recommendation method provided in the illustrated embodiment.

[0219] Optionally, the electronic device may further include a third communication interface 63 for communicating with other devices.

[0220] Fig.17 A structural diagram of another data recommendation device provided by an embodiment of the present invention, such as Fig.17 As shown, the device includes: a first recommendation module 71, a determination module 72, and a second recommendation module 73.

[0221] The first recommendation module 71 is used to recommend first data corresponding to various group characteristics to users.

[0222] The determination module 72 is configured to determine at least one group characteristic matching the user among the multiple group characteristics according to the user's feedback behavior on the first data corresponding to each of the multiple group characteristics.

[0223] The second recommendation module 73 is configured to recommend second data corresponding to the at least one group characteristic to the user.

[0224] Fig.17 The device shown can perform the aforementioned Fig.10 The data recommendation method provided in the illustrated embodiment, the detailed execution process and technical effects can be found in the description of the aforementioned embodiment, which will not be repeated here.

[0225] In one possible design, the above Fig.17 The structure of the data recommendation device shown can be implemented as an electronic device, such as Fig.18 As shown, the electronic device may include: a fourth processor 81 and a fourth memory 82. The fourth memory 82 stores executable code. When the executable code is executed by the fourth processor 81, the fourth processor 81 can at least implement the above-mentioned Fig.10 The data recommendation method provided in the illustrated embodiment.

[0226] Optionally, the electronic device may further include a fourth communication interface 83 for communicating with other devices.

[0227] The device embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separated. Some or all of the modules may be selected according to actual needs to achieve the purpose of the present embodiment. Those of ordinary skill in the art may understand and implement the present invention without creative effort.

[0228] Through the description of the above embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by adding a necessary general hardware platform, and of course can also be implemented by combining hardware and software. Based on such an understanding, the above technical solution is essentially or the part that contributes to the prior art can be embodied in the form of a computer product, and the present invention can be in the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program codes.

[0229] The data recommendation method provided in the embodiment of the present invention can be executed by a certain program / software, and the program / software can be provided by the network side. The electronic device mentioned in the above embodiment can download the program / software to a local non-volatile storage medium, and when it needs to execute the above data recommendation method, the program / software is read into the memory by the CPU, and then the CPU executes the program / software to implement the data recommendation method provided in the above embodiment. The execution process can refer to the above Figures 1 to 10 Instructions in .

[0230] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A data recommendation method, It is characterized in that include: Acquire a first data set corresponding to a user and a plurality of user groups, wherein each of the plurality of user groups corresponds to a different group characteristic, and the different group characteristics include a group characteristic corresponding to the user; For the target data in the first data set, determining the preference of the multiple user groups for the target data respectively; determining a target user group corresponding to the target data according to the preferences of the multiple user groups for the target data, the target user group having a higher preference for the target data; If the group characteristics corresponding to the target user group do not match the group characteristics corresponding to the user, filtering out the target data from the first data set to obtain a second data set; The second data set is recommended to the user.

2. The method according to claim 1, It is characterized in that The obtaining of a first data set corresponding to the user includes: Acquiring the historical data preferred by the user according to the user's feedback behavior on the recommended historical data; Data meeting a similarity condition with the historical data preferred by the user is obtained, and the first data set includes the data meeting the similarity condition.

3. The method according to claim 1 or 2, It is characterized in that The method further comprises: Obtain historical data recommended for the user within a set period of time; If the first data set includes the historical data, the historical data is filtered out from the first data set.

4. The method according to claim 1 or 2, It is characterized in that The method further comprises: Acquire historical data that the user dislikes according to the user's feedback behavior on the recommended historical data; If the first data set includes the historical data, the historical data is filtered out from the first data set.

5. The method according to claim 1, It is characterized in that The determining the preferences of the plurality of user groups for the target data respectively includes: For any user group among the multiple user groups, a first ratio of the number of users in the any user group who have used the target data to the total number of users corresponding to the any user group is determined as the preference of the any user group for the target data.

6. The method according to claim 5, It is characterized in that The method further comprises: The first ratio is adjusted according to feedback behavior of users in any one user group on the target data.

7. The method according to claim 5, It is characterized in that The determining the target user group corresponding to the target data according to the preferences of the multiple user groups for the target data respectively includes: determining a second ratio of the number of users in the multiple user groups who have used the target data to the total number of users corresponding to the multiple user groups; The target user group is determined to be a user group among the multiple user groups for which the first ratio is greater than the second ratio.

8. The method according to claim 7, It is characterized in that The determining the target user group as a user group among the multiple user groups for which the first ratio is greater than the second ratio includes: If the first ratio of at least two user groups among the multiple user groups is greater than the second ratio, the user group with the largest first ratio among the at least two user groups is determined as the target user group, or both of the at least two user groups are determined as the target user groups.

9. The method according to claim 1, It is characterized in that The second data set includes a plurality of songs; and recommending the second data set to the user includes: The audio playback device is controlled to play the multiple songs in sequence.

10. A data recommendation device, It is characterized in that include: An acquisition module, configured to acquire a first data set corresponding to a user and a plurality of user groups, wherein each of the plurality of user groups corresponds to a different group characteristic, and the different group characteristics include a group characteristic corresponding to the user; a filtering module, for determining, for the target data in the first data set, the preferences of the multiple user groups for the target data, and determining, based on the preferences of the multiple user groups for the target data, a target user group corresponding to the target data, the target user group having a higher preference for the target data, and if the group characteristics corresponding to the target user group do not match the group characteristics corresponding to the user, filtering the target data from the first data set to obtain a second data set; An output module is used to recommend the second data set to the user.

11. An electronic device, It is characterized in that include: A memory and a processor; wherein the memory stores executable code, and when the executable code is executed by the processor, the processor executes the data recommendation method as described in any one of claims 1 to 9.

12. A non-transitory machine-readable storage medium, It is characterized in that The non-transitory machine-readable storage medium stores executable code, and when the executable code is executed by a processor of an electronic device, the processor is caused to execute the data recommendation method as described in any one of claims 1 to 9.

13. A data recommendation method, It is characterized in that include: Determine the group characteristics of users; The data set matching the group characteristics is obtained by: obtaining multiple user groups, each of which corresponds to a different group characteristic, wherein the different group characteristics include the group characteristic corresponding to the user; for target data in multiple candidate data, determining the preference of each of the multiple user groups for the target data; determining a target user group corresponding to the target data according to the preference of each of the multiple user groups for the target data, wherein the target user group has a higher preference for the target data; If the group characteristics corresponding to the target user group match the group characteristics corresponding to the user, it is determined that the data set includes the target data; and the data set is recommended to the user.

14. The method according to claim 13, It is characterized in that The determining the preferences of the plurality of user groups for the target data respectively includes: For any user group among the multiple user groups, a first ratio of the number of users in the any user group who have used the data to the total number of users corresponding to the any user group is determined as the preference of the any user group for the target data.

15. The method according to claim 14, It is characterized in that The method further comprises: The first ratio is adjusted according to feedback behavior of users in any one user group on the target data.

16. The method according to claim 14, It is characterized in that The determining the target user group corresponding to the target data according to the preferences of the multiple user groups for the target data respectively includes: determining a second ratio of the number of users in the multiple user groups who have used the data to the total number of users corresponding to the multiple user groups; The target user group is determined to be a user group among the multiple user groups for which the first ratio is greater than the second ratio.

17. The method according to claim 16, It is characterized in that The determining the target user group as a user group among the multiple user groups for which the first ratio is greater than the second ratio includes: If the first ratio of at least two user groups among the multiple user groups is greater than the second ratio, the user group with the largest first ratio among the at least two user groups is determined as the target user group, or both of the at least two user groups are determined as the target user groups.

18. The method according to claim 13, It is characterized in that The plurality of data included in the data set are a plurality of songs; and recommending the data set to the user comprises: The audio playback device is controlled to play the multiple songs in sequence.

19. The method according to claim 13, It is characterized in that The method further comprises: Obtain historical data recommended for the user within a set period of time; If the data set includes the historical data, the historical data is filtered out from the data set.

20. The method according to claim 13, It is characterized in that The method further comprises: Acquire historical data that the user dislikes according to the user's feedback behavior on the recommended historical data; If the data set includes the historical data, the historical data is filtered out from the data set.

21. A data recommendation device, It is characterized in that include: A determination module, used to determine group characteristics of users; An acquisition module is used to acquire a data set matching the group characteristics in the following manner: acquiring multiple user groups, each of the multiple user groups corresponds to a different group characteristic, and the different group characteristics include the group characteristic corresponding to the user; for target data in multiple candidate data, determining the preference of the multiple user groups for the target data; determining a target user group corresponding to the target data according to the preference of the multiple user groups for the target data, the target user group having a higher preference for the target data; If the group characteristics corresponding to the target user group match the group characteristics corresponding to the user, determining that the data set includes the target data; An output module is used to recommend the data set to the user.

22. An electronic device, It is characterized in that include: A memory and a processor; wherein the memory stores executable code, and when the executable code is executed by the processor, the processor executes the data recommendation method as described in any one of claims 13 to 20.

23. A non-transitory machine-readable storage medium, It is characterized in that The non-transitory machine-readable storage medium stores executable code, and when the executable code is executed by a processor of an electronic device, the processor executes the data recommendation method as described in any one of claims 13 to 20.

Images