CN113688325A

CN113688325A - Content recommendation method and device, electronic equipment and computer readable medium

Info

Publication number: CN113688325A
Application number: CN202111056474.4A
Authority: CN
Inventors: 张秀军
Original assignee: Beijing Jingdong Century Trading Co Ltd; Beijing Wodong Tianjun Information Technology Co Ltd
Current assignee: Beijing Jingdong Century Trading Co Ltd; Beijing Wodong Tianjun Information Technology Co Ltd
Priority date: 2021-09-09
Filing date: 2021-09-09
Publication date: 2021-11-23

Abstract

The embodiment of the disclosure discloses a content recommendation method, a content recommendation device, an electronic device and a medium. One embodiment of the method comprises: determining a first form content sequence corresponding to the content acquisition request, wherein the first form content in the first form content sequence has associated item information; selecting a first target number of second form contents from a second form content pool to obtain a second form content sequence; matching a second target number of first form contents which are ranked at the top in the first form content sequence with the second form content sequence to determine the first form contents of the matched second form contents; replacing the first form content of at least one matched second form content in the first form content sequence with the corresponding second form content to obtain a recommended content sequence; and pushing the recommended content sequence to the terminal. The embodiment can push the recommended content information which is richer in form and more preferable for the user to the terminal used by the user.

Description

Content recommendation method and device, electronic equipment and computer readable medium

Technical Field

Embodiments of the present disclosure relate to the field of computer technologies, and in particular, to a content recommendation method and apparatus, an electronic device, and a computer-readable medium.

Background

With the continuous development of streaming media, methods for content recommendation are also diversified. When a scene is recommended for an article, the recommendation method generally adopted is as follows: item content recommendations are made in a single form, for example, often in a textual form.

However, when item content recommendation is performed in the above manner, there are often the following technical problems:

the item content form is too single, so that the user has little interest in the recommended item content. The sides may allow fewer and fewer users to use the target application, which may correspondingly reduce network traffic consumption using the target application.

Disclosure of Invention

This summary is provided to introduce a selection of concepts in a simplified form that are further described below in the detailed description. This summary is not intended to identify key features or essential features of the claimed subject matter, nor is it intended to be used to limit the scope of the claimed subject matter.

Some embodiments of the present disclosure propose content recommendation methods, apparatuses, electronic devices and computer readable media to solve one or more of the technical problems mentioned in the background section above.

In a first aspect, some embodiments of the present disclosure provide a content recommendation method, including: in response to a content acquisition request which is sent by a terminal and aims at a first channel, determining a first form content sequence corresponding to the content acquisition request, wherein article information associated with first form content exists in the first form content sequence; selecting a first target quantity of second form contents from a second form content pool based on the content quality scores of the second form contents to obtain a second form content sequence, wherein the second form contents have associated item information; according to the associated item information, matching a second target number of first form contents ranked at the top in the first form content sequence with the second form content sequence to determine the first form contents of the matched second form contents; replacing the first form content of at least one matched second form content in the first form content sequence with the corresponding second form content to obtain a recommended content sequence; and pushing the recommended content sequence to the terminal.

Optionally, the content quality score of the second form content is obtained by determining a weighted sum of the author quality indicator, the user feedback indicator, and the timeliness indicator of the second form content.

Optionally, before the selecting a first target amount of the second-form content from the second-form content pool based on the content quality score of the second-form content to obtain the second-form content sequence, the method further includes: deleting the second-form content which is effectively played on the second channel from the second-form content pool; and/or deleting the second form content of which the associated item information does not meet the preset condition from the second form content pool.

Optionally, before the pushing the recommended content sequence to the terminal, the method further includes: reordering the recommended content sequence so that the obtained reordered recommended content sequence satisfies at least one of the following constraints: the contents of any two second forms are not adjacent and the interval between the contents of any two second forms meets the preset condition; the article information corresponding to any two adjacent contents is different; and the pushing of the recommended content sequence to the terminal includes: and pushing the rearranged recommended content sequence to the terminal.

Optionally, the method further includes: in response to receiving the content acquisition request for the second channel sent by the terminal, performing the following content screening operation to obtain a screened content sequence: dividing a content pool into a concerned content pool and an unconcerned content pool according to an concerned list of a user corresponding to the terminal; dividing the attention content pool according to the content form to obtain a first attention content group, a second attention content group and a third attention content group; dividing the non-attention content pool according to content forms to obtain a first non-attention content group, a second non-attention content group and a third non-attention content group; sequencing each content group in each content group according to the content quality score to obtain each content sequence; filtering the content with the correlation score smaller than the target value from each content sequence to obtain each filtered content sequence; and inserting the content among the content sequences according to a first preset proportion to obtain the screened content sequences.

Optionally, the relevance score of each content in the content sequences is generated by: determining a target label set corresponding to the content; determining score information corresponding to each target label in the target label set; and determining the score information corresponding to each target label as the relevance score of the content.

Optionally, the determining score information corresponding to each target tag in the target tag set includes: acquiring a historical content set in a target time period; performing word segmentation on the historical content set to obtain a word set; and determining score information corresponding to each target label in the target label set according to the word set.

Optionally, the determining the target tag set corresponding to the content includes: acquiring a label set of the content; at least one label of the article information related to the content is fused into the label set of the content to obtain a fused label set; and screening out at least one label of the target level from the merged label set to serve as the target label set.

In a second aspect, some embodiments of the present disclosure provide a content recommendation apparatus including: the content acquisition device comprises a determining unit, a judging unit and a display unit, wherein the determining unit is configured to respond to a content acquisition request which is sent by a terminal and aims at a first channel, and determine a first form content sequence corresponding to the content acquisition request, and article information associated with first form content exists in the first form content sequence; the selecting unit is configured to select a first target amount of second form content from a second form content pool based on the content quality score of the second form content to obtain a second form content sequence, wherein the second form content has associated item information; a matching unit configured to match a second target number of first form contents ranked in the first form content sequence with the second form content sequence according to the associated item information to determine first form contents of the matched second form contents; a replacing unit configured to replace a first form content of at least one matching second form content in the first form content sequence with a corresponding second form content, resulting in a recommended content sequence; a pushing unit configured to push the recommended content sequence to the terminal.

Optionally, the apparatus further comprises: deleting the second-form content which is effectively played on the second channel from the second-form content pool; and/or deleting the second form content of which the associated item information does not meet the preset condition from the second form content pool.

Optionally, the apparatus further comprises: reordering the recommended content sequence so that the obtained reordered recommended content sequence satisfies at least one of the following constraints: the contents of any two second forms are not adjacent and the interval between the contents of any two second forms meets the preset condition; the article information corresponding to any two adjacent contents is different; and the pushing unit is further configured to: and pushing the rearranged recommended content sequence to the terminal.

Optionally, the apparatus further comprises: in response to receiving the content acquisition request for the second channel sent by the terminal, performing the following content screening operation to obtain a screened content sequence: dividing a content pool into a concerned content pool and an unconcerned content pool according to an concerned list of a user corresponding to the terminal; dividing the attention content pool according to the content form to obtain a first attention content group, a second attention content group and a third attention content group; dividing the non-attention content pool according to content forms to obtain a first non-attention content group, a second non-attention content group and a third non-attention content group; sequencing each content group in each content group according to the content quality score to obtain each content sequence; filtering the content with the correlation score smaller than the target value from each content sequence to obtain each filtered content sequence; and inserting the content among the content sequences according to a first preset proportion to obtain the screened content sequences.

In a third aspect, some embodiments of the present disclosure provide an electronic device, comprising: one or more processors; a storage device having one or more programs stored thereon, which when executed by one or more processors, cause the one or more processors to implement the method as described in any of the implementations of the first aspect.

In a fourth aspect, some embodiments of the disclosure provide a computer readable medium having a computer program stored thereon, where the program when executed by a processor implements a method as described in any of the implementations of the first aspect.

The above embodiments of the present disclosure have the following beneficial effects: the content recommendation method of some embodiments of the disclosure can push the recommended content information with richer forms and more favorable for the user to the terminal used by the user. In particular, the item content format is too single, resulting in less interest by the user in the recommended item content. The side surface reduces the viscosity of the target application used by the user, thereby greatly reducing the user traffic for the target application. Based on this, the content recommendation method of some embodiments of the present disclosure may first determine, in response to receiving a content acquisition request for a first channel sent by a terminal, a first-form content sequence corresponding to the content acquisition request, where there is article information associated with first-form content in the first-form content sequence. Here, the first-format content sequence may be a sequence of item contents that the terminal user may prefer. By determining the first form content sequence, it can be ensured that the subsequently pushed content is the preferred content of the target user with a high probability. The side surface increases user viscosity and increases user traffic for the target application. Then, based on the content quality scores of the second form contents, a first target quantity of the second form contents are selected from the second form content pool, and a second form content sequence is obtained. Wherein, the second form content has associated item information. Here, the second-form content in the selected second-form content sequence is made more favorite by the user by the content quality score of the second-form content. In addition, the content sequence of the second form is selected for the generation of the recommended content sequence including the multi-content form. And matching a second target number of first form contents ranked in the first form content sequence with the second form content sequence according to the associated article information to determine the first form contents of the matched second form contents. Here, by matching the first-format content sequence with the second-format content sequence, it can be ensured that the recommended content in the subsequent recommended content sequence in the multi-content format is the content that the terminal user may personally like, and may also be the content with higher quality. And then, replacing the first form content of at least one matched second form content in the first form content sequence with the corresponding second form content to obtain a recommended content sequence. Here, by the substitution between the first form content sequence and the second form content sequence, a recommended content sequence having a higher content quality that the terminal user person may like can be generated. In addition, the profile increases user viscosity and user traffic for the target application. And finally, pushing the recommended content information which is richer in form and more preferable for the user to the terminal used by the user.

Drawings

The above and other features, advantages and aspects of various embodiments of the present disclosure will become more apparent by referring to the following detailed description when taken in conjunction with the accompanying drawings. Throughout the drawings, the same or similar reference numbers refer to the same or similar elements. It should be understood that the drawings are schematic and that elements and features are not necessarily drawn to scale.

FIG. 1 is a schematic diagram of one application scenario of a content recommendation method according to some embodiments of the present disclosure;

FIG. 2 is a flow diagram of some embodiments of a content recommendation method according to the present disclosure;

FIG. 3 is a schematic diagram of a second form of content substitution in some embodiments of a content recommendation method according to the present disclosure;

FIG. 4 is a flow diagram of further embodiments of a content recommendation method according to the present disclosure;

FIG. 5 is a schematic block diagram of some embodiments of a content recommendation device according to the present disclosure;

FIG. 6 is a schematic structural diagram of an electronic device suitable for use in implementing some embodiments of the present disclosure.

Detailed Description

Embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. While certain embodiments of the present disclosure are shown in the drawings, it is to be understood that the disclosure may be embodied in various forms and should not be construed as limited to the embodiments set forth herein. Rather, these embodiments are provided for a more thorough and complete understanding of the present disclosure. It should be understood that the drawings and embodiments of the disclosure are for illustration purposes only and are not intended to limit the scope of the disclosure.

It should be noted that, for convenience of description, only the portions related to the related invention are shown in the drawings. The embodiments and features of the embodiments in the present disclosure may be combined with each other without conflict.

It should be noted that the terms "first", "second", and the like in the present disclosure are only used for distinguishing different devices, modules or units, and are not used for limiting the order or interdependence relationship of the functions performed by the devices, modules or units.

It is noted that references to "a", "an", and "the" modifications in this disclosure are intended to be illustrative rather than limiting, and that those skilled in the art will recognize that "one or more" may be used unless the context clearly dictates otherwise.

The names of messages or information exchanged between devices in the embodiments of the present disclosure are for illustrative purposes only, and are not intended to limit the scope of the messages or information.

The present disclosure will be described in detail below with reference to the accompanying drawings in conjunction with embodiments.

Fig. 1 is a schematic diagram of an application scenario of a content recommendation method according to some embodiments of the present disclosure.

In the application scenario of fig. 1, in response to receiving a content acquisition request for a first channel sent by the terminal 102, the electronic device 101 may determine a first-form content sequence 103 corresponding to the content acquisition request, where there is article information associated with first-form content in the first-form content sequence 103. In the application scenario, the first-form content sequence 103 may include: first form content 1031, first form content 1032, first form content 1033. The item information corresponding to the first format content 1031 may be item information 104. The item information corresponding to the first format content 1032 may be the item information 105. The item information corresponding to the first form content 1033 may be the item information 107. The item information 104 may be a "hat". Item information 105 may be "shoes". The item information 107 may be a "key fob". The electronic device 101 may then select a first target amount of the second-form content from the second-form content pool 108 based on the content quality score of the second-form content, resulting in a second-form content sequence 109. Wherein, the second form content has associated item information. In the present application scenario, the content sequence 109 in the second form may include: second form content 1091, second form content 1092, third form content 1093. The content quality score corresponding to the second form content 1091 may be: 0.9. the content quality score corresponding to the second form content 1092 may be: 0.78. the content quality score corresponding to the second form content 1093 may be: 0.6. the item information corresponding to the second form content 1091 may be the item information 104. The item information corresponding to the second form content 1092 may be the item information 106. The item information corresponding to the second form content 1093 may be the item information 107. The item information 106 may be: a jacket. Furthermore, the electronic device 101 may match a second target number of first format contents ranked in the first format content sequence 103 with the second format content sequence 109 according to the associated item information, so as to determine the first format contents of the matched second format contents. In the present application scenario, the first form content for which there is a matching second form content may include: first form content 1031 and second form content 1033. The second form content matching the first form content 1031 may be the second form content 1091. The first form content 1031 and the second form content 1091 have the same associated item information 104. Next, the electronic device 101 replaces the first-form content of at least one matching second-form content in the first-form content sequence 103 with the corresponding second-form content, so as to obtain a recommended content sequence 113. In the first form content sequence 103, the first form content 1031 of the matching second form content is replaced by the corresponding second form content 1091, so as to obtain the recommended content 1131. The recommended content 1133 is obtained by replacing the first form content 1032 of the matching second form content in the first form content sequence 103 with the corresponding second form content 1093. The recommended content sequence 113 includes: recommended content 1131, recommended content 1132, and recommended content 1133. The recommended content 1132 may be the second form content 1092. Finally, the electronic device 101 may push the recommended content sequence 113 to the terminal 102.

The electronic device 101 may be hardware or software. When the electronic device is hardware, the electronic device may be implemented as a distributed cluster formed by a plurality of servers or terminal devices, or may be implemented as a single server or a single terminal device. When the electronic device is embodied as software, it may be installed in the above-listed hardware devices. It may be implemented, for example, as multiple software or software modules to provide distributed services, or as a single software or software module. And is not particularly limited herein.

It should be understood that the number of electronic devices in fig. 1 is merely illustrative. There may be any number of electronic devices, as desired for implementation.

With continued reference to fig. 2, a flow 200 of some embodiments of a content recommendation method according to the present disclosure is shown. The content recommendation method comprises the following steps:

step 201, in response to receiving a content acquisition request for a first channel sent by a terminal, determining a first-form content sequence corresponding to the content acquisition request.

In some embodiments, in response to receiving a content acquisition request for a first channel sent by a terminal, an executing entity (e.g., the electronic device shown in fig. 1) of the content recommendation method may determine a first-form content sequence corresponding to the content acquisition request. The terminal may be a terminal used by a target user. Such as a mobile terminal used by the target user. The first channel may be a channel for presenting the second form of content. The first form content may be a first target form content. As an example, the first form content may be a teletext form content. The content obtaining request of the first channel may be a request triggered by a target user logging in the first channel on the terminal. The first-format content sequence may be a content sequence that is pre-filtered for the target user and is preferred by the target user.

It should be noted that, for the first channel, the automatic playing is performed according to the sliding position of the user, and only the second-form content displayed on the current screen corresponding to the first channel is played. When the user sliding ends, it is necessary to judge whether to start/stop the playback. And if the user does not leave the current screen corresponding to the first channel, performing circular playing. If the user leaves the current screen, the previous video being played is paused. And if the user slides back to the current screen, continuing playing.

In addition, the information (feeds) stream page of the first channel may be set to mute by default. The user may jump to the corresponding video details page by clicking on some second modality content corresponding card in the first channel's information flow page.

It should be noted that the user may operate the same for the second channel and the first channel.

As an example, if the target user is interested in sports related content, the first formal sequence may be: "basketball-related graphics context, football-related graphics context, ping-pong-related graphics context, and football-related graphics context".

As another example, in response to receiving a content acquisition request for a first channel transmitted by a terminal, the determining, by the execution main body, a first-form content sequence corresponding to the content acquisition request may include:

first, identification information corresponding to the terminal is determined.

As an example, the execution subject may determine the identification information corresponding to the terminal by querying a database.

And secondly, determining a user tag set of a user corresponding to the terminal according to the identification information. Here, the execution body may acquire the user tab set by browsing the content in the history of the user.

And thirdly, screening a predetermined number of first form contents from the first form content pool according to the user tag set to serve as a first form content sequence. Here, the execution subject may filter out a predetermined number of first-form contents from the first-form content pool by a coincident tag between a tag of the first-form contents and a tag set of the user, and sort the predetermined number of first-form contents.

Step 202, selecting a first target amount of second form content from a second form content pool based on the content quality score of the second form content, and obtaining a second form content sequence.

In some embodiments, the execution subject may select a first target amount of the second-form content from the second-form content pool based on the content quality score of the second-form content, resulting in the second-form content sequence. Wherein the second form content may be a second target form content. As an example, the second target-form content described above may be content in a video form. For example, the second target-form content may be content in the form of short video. The content quality score of the second form content may laterally represent the popularity of the second form content and may also laterally represent the feedback information of the user. Wherein, the first target number may be preset. For example, the first target number may be 3. The second form content pool stores each second form content for each type of article.

As an example, the content quality score of each second-form content in the second-form content pool described above may be determined by the number of praise for the second-form content. For example, the second form content score with the praise number between 0 and 100 is 30 points. The content score of the second form between 100 and 1000 praise numbers is 60 points. The second form content score with the praise number between 1000 and 10000 is 80 points. The content score of the second form with the praise number of 10000-100000 is 90 points. The content score of the second form with the number of praise being 100000+ is 100.

As yet another example, the executing entity may select a first target number of content quality scores from the second content pool to obtain a second content set. Then, the execution subject may sort the second form content sets according to the order of the content quality scores from high to low, resulting in the second form content sequence.

In some optional implementations of some embodiments, the content quality score of the second form of content is obtained by determining a weighted sum of an author quality indicator, a user feedback indicator, and a timeliness indicator of the second form of content. The value ranges of the author quality index, the user feedback index and the timeliness index can be (0, 1).

The author quality index can be obtained by the following formula:

where W may be an author quality indicator for the second form of content. C may be the number of user fans.

It should be noted that only the author who has the output of the second form of content is calculated. The content weight may be processed down for authors who have content output in the second form but who do not have fans.

The user feedback index can be obtained by the following formula:

wherein,

the effective playing is performed when the playing time is longer than or equal to 10 seconds, and the playing mentioned in the formula is statistically and effectively performed. The executing agent may grant rights to the second form content of a different author. For example, platform operation C1.2>Brand quotient C1.15>Direct seeding C1.1 in combined bin>POP merchant C1.05>Customer manager C.

Wherein the timeliness index can be determined by the following formula:

where Ln is the base 10 logarithm. The release duration is the time interval between the video release date and the current date. The unit may be a day. The time length is calculated according to the natural day, and the time length of less than one day is marked as 1. Here, if the issuance time length is 0, the aging characteristic value is 1. Wherein the more recently released content is recalled earlier.

As an example, the content quality score of the second form content may be found by the following formula:

the content quality score of the second form of content is 20% author quality indicator + 50% user feedback indicator + 30% timeliness indicator.

Step 203, according to the associated item information, matching a second target number of first form contents ranked in the first form content sequence with the second form content sequence to determine the first form contents of the matched second form contents.

In some embodiments, the execution subject may match a second target number of first-form contents ranked first in the first-form content sequence with the second-form content sequence according to the associated item information to determine that there is a first-form content of the matched second-form content. Wherein the second target number may be preset. For example, the second target number may be 50.

As an example, the executing entity may match a second target number of first-form contents ranked in the first-form content sequence with the second-form content sequence by means of database query according to the associated item information to determine the first-form contents of the matched second-form contents.

And 204, replacing the first form content of at least one matched second form content in the first form content sequence with the corresponding second form content to obtain a recommended content sequence.

In some embodiments, the execution subject may replace the first-form content of at least one matching second-form content in the first-form content sequence with the corresponding second-form content, resulting in a recommended content sequence.

As an example, there is at least one matching first-form content for the second-form content, the execution subject may replace a top-ranked first-form content of the at least one matching first-form content in the sequence of first-form content with the second-form content, and the rest of the at least one matching first-form content is not transformed in the sequence of first-form content.

It should be noted that, in response to that the number of the first-form contents of the at least one matching second-form content in the first-form content sequence is greater than the preset first number, the executing entity may replace the second-form content with a second number of the first-form contents with the highest content quality score. Wherein the first number is greater than or equal to the second number.

In response to that the number of the first form contents of at least one matching second form content in the first form content sequence is smaller than the preset first number, the execution subject may directly intersperse the second form content to be replaced to a fixed position in the first form content sequence without replacing the second form content.

As shown in fig. 3, the first form content sequence 301 may be: "first-form content 3011, first-form content 3012, first-form content 3013". The first form content 3011 and the first form content 3013 both correspond to the second form content 302. Since the first-form content 3011 is located before the first-form content 3012 in the first-form content sequence 301, the first-form content 301 is replaced with the second-form content 302, resulting in a recommended content sequence 303. The recommended content sequence 303 may be: "second-form content 302, first-form content 3012, first-form content 3013".

Step 205, pushing the recommended content sequence to the terminal.

In some embodiments, the execution subject may push the recommended content sequence to the terminal.

In some optional implementations of some embodiments, before step 205, the step further includes:

the execution subject may reorder the recommended content sequence so that the obtained rearranged recommended content sequence satisfies at least one of the following constraints:

(1) the contents of any two second forms are not adjacent and the interval between the contents of any two second forms meets the preset condition.

As an example, the interval between any two second form contents may be 6 first form contents.

(2) The article information corresponding to any two adjacent contents is different.

Further, the article information corresponding to two adjacent first form contents may be different, and the article information corresponding to the adjacent first form contents and the adjacent second form contents is also different.

And the pushing the recommended content sequence to the terminal may include:

the execution main body may push the rearranged recommended content sequence to the terminal.

In some optional implementations of some embodiments, the foregoing step further includes:

the first step, in response to receiving a content acquisition request for a second channel sent by the terminal, executing the following content screening operation to obtain a screened content sequence:

and a first substep of dividing the content pool into an attention content pool and an non-attention content pool according to the attention list of the user corresponding to the terminal.

In some embodiments, the execution subject may divide the content pool into an attention content pool and an non-attention content pool according to an attention list of a user corresponding to the terminal. The concerned content pool stores the content of each author in the concerned list of the user. The unapproved content pool stores the content of each author in the user's unapproved list.

As an example, the execution subject may determine the focus list of the user corresponding to the terminal by querying the target database. Further, the execution subject may divide the content pool into a focused content pool and an unapproved content pool.

And a second substep of dividing the execution subject into a first interested content group, a second interested content group and a third interested content group according to the content form for the interested content pool. The first attention content in the first attention content group, the second attention content in the second attention content group and the third attention content in the third attention content group may be different forms of content.

As an example, the first attention content in the above-described first attention content group may be content in a preview live form. The second attention content in the above-mentioned second attention content group may be content in the form of short video. The third-interest content in the third-interest content group may be content in a teletext format.

And a third substep, for the non-attention content pool, dividing the execution main body according to the content form to obtain a first non-attention content group, a second non-attention content group and a third non-attention content group. The first non-concerned content in the first non-concerned content group, the second non-concerned content in the second non-concerned content group and the third non-concerned content in the third non-concerned content group may be different forms of content.

As an example, the first non-noted content in the above-mentioned first non-noted content group may also be content in a preview live form. The second non-notable content in the second non-notable content group may be content in the form of short video. The third non-concerned content in the third non-concerned content group may be a content in a teletext format.

In the fourth sub-step, the executing body may sort each content group in the content groups according to the content quality score to obtain each content sequence.

As an example, the execution subject described above may sort each of the respective content groups in order of the content quality scores from top to bottom.

In the fifth sub-step, the executing agent may filter, from the respective content sequences, content whose correlation score is smaller than a target value, to obtain respective filtered content sequences. Wherein, the target value may be preset. The relevance score of the content may be a score of each tag in the content-corresponding tag set. Wherein the score of the tag may characterize how relevant the tag is to the content.

As an example, when the score of a tag is between (0-0.4), it may be stated that the tag is not related to the content. When the score of the label is between 0.4-0.6), it can be stated that the label is weakly related to the content. When the score of the label is between 0.6-0.8), it can be stated that the label is highly related to the content. When the score of the label is between 0.8-1), it can be stated that the label is strongly related to the content.

And secondly, inserting the content among the content sequences according to a first preset proportion to obtain the screened content sequences.

As an example, the execution main body may intersperse the contents between the content sequences according to a first preset ratio by:

firstly, mutually interleaving each content sequence in the non-attention content pool and each content sequence in the attention content pool according to a first target proportion, so that 1/3 non-attention content in the non-attention content pool exists in the interleaved attention content pool, and part of attention content in the attention content pool exists in the non-attention content pool. Wherein the first target proportion may be preset.

Here, the content of interest that is inserted into the content pool of interest may be the top content of interest in each content sequence in the content pool of interest.

And secondly, performing mutual content interleaving on each content sequence in the interleaved concerned content pool according to a second target proportion and the content quality score to obtain a screened content sequence. Wherein the second target ratio may be preset.

For example, for each content sequence in the interspersed content-of-interest pool: content sequences in a live broadcast form, short video form and graphic form are announced in advance. Wherein, the content sequence of the preview live broadcast form is as follows: the first forecast live broadcast form content and the second forecast live broadcast form content. The quality score of the first preview live form content is higher than that of the second preview live form content. The content sequence of the short video form is as follows: first short video format content, second short video format content. The quality score of the first short video format content is higher than that of the second short video format content. The content sequence of the image-text form is as follows: the first image-text form content and the second image-text form content. The quality score of the first graphic content is higher than that of the second graphic content. Finally, the screening content sequence may be: the content in the form of the first preview live broadcast, the content in the form of the first short video, the content in the form of the first image and text, the content in the form of the second preview live broadcast, the content in the form of the second short video and the content in the form of the second image and text.

Optionally, the relevance score of each content in the content sequences is generated by:

firstly, determining a target label set corresponding to the content.

As an example, the execution subject may receive the target tag set corresponding to the content by tagging by an expert.

For example, the content is content associated with a sports introduction. The target tag set corresponding to the content may include: "basketball" label, "football" label, "table tennis" label.

And secondly, determining score information corresponding to each target label in the target label set.

As an example, the execution subject may score each target tag with a frequency of occurrence of each target tag in the content. For example, in the content, the "basketball" tag appears 5 times. The "football" label appears 3 times. The frequency of occurrence of the "ping-pong ball" label is 2 times. The executing agent may determine the score corresponding to the "basketball" label as 0.5, the score corresponding to the "football" label as 0.3, and the score corresponding to the "table tennis" label as 0.2.

And thirdly, determining the score information corresponding to each target label as the relevance score of the content.

Optionally, the determining score information corresponding to each target tag in the target tag set may include the following steps:

first, a historical content set in a target time period is obtained.

As an example, the execution subject may obtain the historical content set in the target time period from a database storing the browsing content of the user. The target time period may be preset. For example, half a year.

And secondly, performing word segmentation on the historical content set to obtain a word set.

As an example, the execution subject may perform word segmentation on each historical content in the historical content set by using the ending word segmentation to obtain a word set. Wherein, the word set can have a plurality of repeated words.

And thirdly, determining score information corresponding to each target label in the target label set according to the word set.

As an example, the execution subject may determine the frequency of each target word in the word set corresponding to a primary category word, a secondary category word, a tertiary category word, a brand word, and a feature word. Then, the number of occurrences of each target word/the number of words in the current content type word bag is used to obtain a first probability. The first probability may characterize the importance of the target word in the current historical content set. A second probability of the number of documents in which the target word occurs/the total number of documents may then be determined, representing the importance of the word in all documents. And finally, multiplying the first probability corresponding to the target word by the second probability to obtain the probability value of the target word, namely the score information of the target label.

For example, the last half year content may be 2 short videos, 2 teletext.

Wherein, the short video corresponds to 11 target words in total, and the graphics and text correspond to 12 target words in total.

Then the above target words include: combinations of foods, beverages, carbonated beverages, bubbles, tea beverages, coffee, bottles, macleydi, pet life, cat and dog food, cat and dog snacks, deep sea fish, salmon.

Then, for short video content: the first probability corresponding to the union is: association is 1/11, the number of occurrences/short video corresponds to the target word. The first probability for drinking the milk product is: the number of times of the water-drinking milk product appearing in the 2 short videos/the number of the target words corresponding to the short videos is 2/11. Similarly, the first probability of the remaining target words in the short video may be obtained, which is not described herein again.

Similarly, for the graphic content: the first probability of the madx dis correspondence is: the number of occurrences of mady/the number of target words corresponding to a character is 1/12. The first probability for drinking the milk product is: the number of occurrences of the water-drinking milk product/the number of the target words corresponding to the pictures is 1/12.

Then, for short video content: the second probability for the water-in-milk product is: the number of short videos/total number of short videos of the milk drink is 2/2. The second probability corresponding to the union is: the number of short videos/the number of all short videos of union x appears 1/2. Similarly, the first probability of the remaining target words in the short video may be obtained, which is not described herein again.

Similarly, for the graphic content: the second probability for the water-in-milk product is: the number of pictures and texts/the number of all pictures and texts of the water-drinking milk product is 1/2. The second probability corresponding to maxtadi is: the number of images/images in which the image/images appears is 1/2.

Then, for short video content: the associated probability values are: 1/11 × 1/2 ═ 1/22. The probability value corresponding to the water-drinking milk product is as follows: 2/11 × 2/2 ═ 2/11.

Similarly, for the graphic content: the probability values corresponding to maxtadi are: 1/12 × 1/2 ═ 1/24. The probability value corresponding to the water-drinking milk product is as follows: 1/12 × 1/2 ═ 1/24.

With further reference to fig. 4, a flow 400 of further embodiments of a content recommendation method according to the present disclosure is shown. The content recommendation method comprises the following steps:

step 401, in response to receiving a content acquisition request for a first channel sent by a terminal, determining a first-format content sequence corresponding to the content acquisition request.

Step 402, deleting the second-form content that has been actively played in the second channel from the second-form content pool; and/or deleting the second form content of which the associated item information does not meet the preset condition from the second form content pool.

In some embodiments, the executing agent (e.g., the electronic device shown in fig. 1) may delete the second-format content that has been actively played on the second channel from the second-format content pool; and/or deleting the second form content of which the associated item information does not meet the preset condition from the second form content pool. The second type content that has been effectively played on the second channel may be determined according to whether the user corresponding to the terminal has played the second type content once, and the playing duration is greater than or equal to the target duration. For example, the target duration may be 5 seconds. In addition, the second form content of which the associated item information does not satisfy the preset condition may be the second form content corresponding to a special state such as an item off-shelf state of the item information.

In step 403, a first target amount of the second-type content is selected from the second-type content pool based on the content quality score of the second-type content, so as to obtain a second-type content sequence.

And step 404, according to the associated item information, matching a second target number of first form contents ranked at the top in the first form content sequence with the second form content sequence to determine the first form contents of the matched second form contents.

Step 405, replacing the first form content of at least one matching second form content in the first form content sequence with the corresponding second form content to obtain a recommended content sequence.

Step 406, pushing the recommended content sequence to the terminal.

The specific implementation of

steps

401 and 403 and the technical effect thereof can refer to step 201 and 205 in the embodiment corresponding to fig. 2, and will not be described herein again.

As can be seen from fig. 4, the process 400 of the content recommendation method in some embodiments corresponding to fig. 4 highlights the specific steps of processing the second form of content in the second form of content pool compared to the description of some embodiments corresponding to fig. 2. Therefore, the scheme described in the embodiments can make the generated recommended content sequence more likely to be favored by the user terminal, and further guarantee the user traffic.

With further reference to fig. 5, as an implementation of the methods shown in the above figures, the present disclosure provides some embodiments of a content recommendation apparatus, which correspond to those shown in fig. 2, and which may be applied in various electronic devices in particular.

As shown in fig. 5, a content recommendation apparatus 500 includes: a determination unit 501, a selection unit 502, a matching unit 503, a replacement unit 504 and a push unit 505. The determining unit 501 is configured to determine, in response to receiving a content acquisition request for a first channel sent by a terminal, a first-form content sequence corresponding to the content acquisition request, where there is article information associated with first-form content in the first-form content sequence; a selecting unit 502 configured to select a first target amount of second form content from a second form content pool based on a content quality score of the second form content, resulting in a second form content sequence, wherein the second form content has associated item information; a matching unit 503 configured to match a second target number of first format contents ranked in the first format content sequence with the second format content sequence according to the associated item information to determine first format contents of the second format contents having a match; a replacing unit 504, configured to replace the first form content of at least one matching second form content in the first form content sequence with the corresponding second form content, resulting in a recommended content sequence; a pushing unit 505 configured to push the recommended content sequence to the terminal.

In some optional implementations of some embodiments, the content quality score of the second form of content is obtained by determining a weighted sum of an author quality indicator, a user feedback indicator, and a timeliness indicator of the second form of content.

In some optional implementations of some embodiments, the apparatus 500 further includes: a deletion unit (not shown in the figure). Wherein the deletion unit may be configured to: deleting the second-form content which is effectively played on the second channel from the second-form content pool; and/or deleting the second form content of which the associated item information does not meet the preset condition from the second form content pool.

In some optional implementations of some embodiments, the apparatus 500 further includes: a sorting unit (not shown in the figure). Wherein the sorting unit may be configured to: reordering the recommended content sequence so that the obtained reordered recommended content sequence satisfies at least one of the following constraints: the contents of any two second forms are not adjacent and the interval between the contents of any two second forms meets the preset condition; the article information corresponding to any two adjacent contents is different. And the pushing unit 505 may be further configured to: and pushing the rearranged recommended content sequence to the terminal.

In some optional implementations of some embodiments, the apparatus 500 further includes: a screening unit (not shown). Wherein the screening unit may be configured to: in response to receiving the content acquisition request for the second channel sent by the terminal, performing the following content screening operation to obtain a screened content sequence: dividing a content pool into a concerned content pool and an unconcerned content pool according to an concerned list of a user corresponding to the terminal; dividing the attention content pool according to the content form to obtain a first attention content group, a second attention content group and a third attention content group; dividing the non-attention content pool according to content forms to obtain a first non-attention content group, a second non-attention content group and a third non-attention content group; sequencing each content group in each content group according to the content quality score to obtain each content sequence; filtering the content with the correlation score smaller than the target value from each content sequence to obtain each filtered content sequence; and inserting the content among the content sequences according to a first preset proportion to obtain the screened content sequences.

In some optional implementations of some embodiments, the relevance score of each content in the respective content sequences is generated by: determining a target label set corresponding to the content; determining score information corresponding to each target label in the target label set; and determining the score information corresponding to each target label as the relevance score of the content.

In some optional implementations of some embodiments, the apparatus 500 further includes: a score information determination unit. Wherein the score information determination unit may be further configured to: acquiring a historical content set in a target time period; performing word segmentation on the historical content set to obtain a word set; and determining score information corresponding to each target label in the target label set according to the word set.

In some optional implementations of some embodiments, the apparatus 500 further includes: and an object label determination unit. Wherein the target tag determination unit may be further configured to: acquiring a label set of the content; at least one label of the article information related to the content is fused into the label set of the content to obtain a fused label set; and screening out at least one label of the target level from the merged label set to serve as the target label set.

It will be understood that the elements described in the apparatus 500 correspond to various steps in the method described with reference to fig. 2. Thus, the operations, features and resulting advantages described above with respect to the method are also applicable to the apparatus 500 and the units included therein, and are not described herein again.

Referring now to FIG. 6, a block diagram of an electronic device (e.g., the electronic device of FIG. 1) 600 suitable for use in implementing some embodiments of the present disclosure is shown. The electronic device shown in fig. 6 is only an example, and should not bring any limitation to the functions and the scope of use of the embodiments of the present disclosure.

As shown in fig. 6, electronic device 600 may include a processing means (e.g., central processing unit, graphics processor, etc.) 601 that may perform various appropriate actions and processes in accordance with a program stored in a Read Only Memory (ROM)602 or a program loaded from a storage means 608 into a Random Access Memory (RAM) 603. In the RAM603, various programs and data necessary for the operation of the electronic apparatus 600 are also stored. The processing device 601, the ROM 602, and the RAM603 are connected to each other via a bus 604. An input/output (I/O) interface 605 is also connected to bus 604.

Generally, the following devices may be connected to the I/O interface 605: input devices 606 including, for example, a touch screen, touch pad, keyboard, mouse, camera, microphone, accelerometer, gyroscope, etc.; output devices 607 including, for example, a Liquid Crystal Display (LCD), a speaker, a vibrator, and the like; storage 608 including, for example, tape, hard disk, etc.; and a communication device 609. The communication means 609 may allow the electronic device 600 to communicate with other devices wirelessly or by wire to exchange data. While fig. 6 illustrates an electronic device 600 having various means, it is to be understood that not all illustrated means are required to be implemented or provided. More or fewer devices may alternatively be implemented or provided. Each block shown in fig. 6 may represent one device or may represent multiple devices as desired.

In particular, according to some embodiments of the present disclosure, the processes described above with reference to the flow diagrams may be implemented as computer software programs. For example, some embodiments of the present disclosure include a computer program product comprising a computer program embodied on a computer readable medium, the computer program comprising program code for performing the method illustrated in the flow chart. In some such embodiments, the computer program may be downloaded and installed from a network through the communication device 609, or installed from the storage device 608, or installed from the ROM 602. The computer program, when executed by the processing device 601, performs the above-described functions defined in the methods of some embodiments of the present disclosure.

It should be noted that the computer readable medium described above in some embodiments of the present disclosure may be a computer readable signal medium or a computer readable storage medium or any combination of the two. A computer readable storage medium may be, for example, but not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination of the foregoing. More specific examples of the computer readable storage medium may include, but are not limited to: an electrical connection having one or more wires, a portable computer diskette, a hard disk, a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing. In some embodiments of the disclosure, a computer readable storage medium may be any tangible medium that can contain, or store a program for use by or in connection with an instruction execution system, apparatus, or device. In some embodiments of the present disclosure, however, a computer readable signal medium may include a propagated data signal with computer readable program code embodied therein, for example, in baseband or as part of a carrier wave. Such a propagated data signal may take many forms, including, but not limited to, electro-magnetic, optical, or any suitable combination thereof. A computer readable signal medium may also be any computer readable medium that is not a computer readable storage medium and that can communicate, propagate, or transport a program for use by or in connection with an instruction execution system, apparatus, or device. Program code embodied on a computer readable medium may be transmitted using any appropriate medium, including but not limited to: electrical wires, optical cables, RF (radio frequency), etc., or any suitable combination of the foregoing.

In some embodiments, the clients, servers may communicate using any currently known or future developed network Protocol, such as HTTP (HyperText Transfer Protocol), and may interconnect with any form or medium of digital data communication (e.g., a communications network). Examples of communication networks include a local area network ("LAN"), a wide area network ("WAN"), the Internet (e.g., the Internet), and peer-to-peer networks (e.g., ad hoc peer-to-peer networks), as well as any currently known or future developed network.

The computer readable medium may be embodied in the electronic device; or may exist separately without being assembled into the electronic device. The computer readable medium carries one or more programs which, when executed by the electronic device, cause the electronic device to: in response to a content acquisition request which is sent by a terminal and aims at a first channel, determining a first form content sequence corresponding to the content acquisition request, wherein article information associated with first form content exists in the first form content sequence; selecting a first target quantity of second form contents from a second form content pool based on the content quality scores of the second form contents to obtain a second form content sequence, wherein the second form contents have associated item information; according to the associated item information, matching a second target number of first form contents ranked at the top in the first form content sequence with the second form content sequence to determine the first form contents of the matched second form contents; replacing the first form content of at least one matched second form content in the first form content sequence with the corresponding second form content to obtain a recommended content sequence; and pushing the recommended content sequence to the terminal.

Computer program code for carrying out operations for embodiments of the present disclosure may be written in any combination of one or more programming languages, including an object oriented programming language such as Java, Smalltalk, C + +, and conventional procedural programming languages, such as the "C" programming language or similar programming languages. The program code may execute entirely on the user's computer, partly on the user's computer, as a stand-alone software package, partly on the user's computer and partly on a remote computer or entirely on the remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer through any type of network, including a Local Area Network (LAN) or a Wide Area Network (WAN), or the connection may be made to an external computer (for example, through the Internet using an Internet service provider).

The flowchart and block diagrams in the figures illustrate the architecture, functionality, and operation of possible implementations of systems, methods and computer program products according to various embodiments of the present disclosure. In this regard, each block in the flowchart or block diagrams may represent a module, segment, or portion of code, which comprises one or more executable instructions for implementing the specified logical function(s). It should also be noted that, in some alternative implementations, the functions noted in the block may occur out of the order noted in the figures. For example, two blocks shown in succession may, in fact, be executed substantially concurrently, or the blocks may sometimes be executed in the reverse order, depending upon the functionality involved. It will also be noted that each block of the block diagrams and/or flowchart illustration, and combinations of blocks in the block diagrams and/or flowchart illustration, can be implemented by special purpose hardware-based systems which perform the specified functions or acts, or combinations of special purpose hardware and computer instructions.

The units described in some embodiments of the present disclosure may be implemented by software, and may also be implemented by hardware. The described units may also be provided in a processor, and may be described as: a processor comprises a determining unit, a selecting unit, a matching unit, a replacing unit and a pushing unit. The names of these units do not in some cases constitute a limitation on the units themselves, and for example, the determination unit may also be described as "a unit that determines a first-format content sequence corresponding to a content acquisition request for a first channel transmitted by a terminal in response to receiving the content acquisition request.

The functions described herein above may be performed, at least in part, by one or more hardware logic components. For example, without limitation, exemplary types of hardware logic components that may be used include: field Programmable Gate Arrays (FPGAs), Application Specific Integrated Circuits (ASICs), Application Specific Standard Products (ASSPs), systems on a chip (SOCs), Complex Programmable Logic Devices (CPLDs), and the like.

The foregoing description is only exemplary of the preferred embodiments of the disclosure and is illustrative of the principles of the technology employed. It will be appreciated by those skilled in the art that the scope of the invention in the embodiments of the present disclosure is not limited to the specific combination of the above-mentioned features, but also encompasses other embodiments in which any combination of the above-mentioned features or their equivalents is made without departing from the inventive concept as defined above. For example, the above features and (but not limited to) technical features with similar functions disclosed in the embodiments of the present disclosure are mutually replaced to form the technical solution.

Claims

1. A content recommendation method, comprising:

in response to a content acquisition request which is sent by a terminal and aims at a first channel, determining a first form content sequence corresponding to the content acquisition request, wherein article information associated with first form content exists in the first form content sequence;

selecting a first target quantity of second form contents from a second form content pool based on the content quality scores of the second form contents to obtain a second form content sequence, wherein the second form contents have associated item information;

according to the associated item information, matching a second target number of first form contents which are ranked at the top in the first form content sequence with the second form content sequence to determine the first form contents of the matched second form contents;

replacing the first form content of at least one matched second form content in the first form content sequence with the corresponding second form content to obtain a recommended content sequence;

and pushing the recommended content sequence to the terminal.

2. The method of claim 1, wherein the content quality score for the second form of content is derived by determining a weighted sum of an author quality indicator, a user feedback indicator, and a timeliness indicator for the second form of content.

3. The method of claim 1, wherein before the selecting a first target amount of the second-form content from the second-form content pool based on the content quality score of the second-form content to obtain the second-form content sequence, the method further comprises:

deleting second-form content that has been actively played on a second channel from the second-form content pool; and/or

And deleting the second form content of which the associated article information does not meet the preset condition from the second form content pool.

4. The method of claim 1, wherein prior to said pushing the recommended content sequence to the terminal, the method further comprises:

reordering the recommended content sequence such that the resulting reordered recommended content sequence satisfies at least one of the following constraints:

the contents of any two second forms are not adjacent and the interval between the contents of any two second forms meets the preset condition;

the article information corresponding to any two adjacent contents is different; and

the pushing the recommended content sequence to the terminal includes:

and pushing the rearranged recommended content sequence to the terminal.

5. The method of claim 4, wherein the method further comprises:

in response to receiving a content acquisition request for a second channel sent by the terminal, performing the following content screening operations to obtain a screened content sequence:

dividing a content pool into an attention content pool and an non-attention content pool according to an attention list of a user corresponding to the terminal;

dividing the concerned content pool according to content forms to obtain a first concerned content group, a second concerned content group and a third concerned content group;

for the non-attention content pool, dividing according to content forms to obtain a first non-attention content group, a second non-attention content group and a third non-attention content group;

sequencing each content group in each content group according to the content quality score to obtain each content sequence;

filtering the content with the correlation score smaller than the target numerical value from each content sequence to obtain each filtered content sequence;

and inserting the content among the content sequences according to a first preset proportion to obtain the screened content sequences.

6. The method of claim 5, wherein the relevance score for each content in the respective content sequence is generated by:

determining a target label set corresponding to the content;

determining score information corresponding to each target label in the target label set;

and determining the score information corresponding to each target label as the relevance score of the content.

7. The method of claim 6, wherein the determining score information corresponding to each target tag in the set of target tags comprises:

acquiring a historical content set in a target time period;

performing word segmentation on the historical content set to obtain a word set;

and determining the score information corresponding to each target label in the target label set according to the word set.

8. The method of claim 6, wherein the determining a target set of tags to which the content corresponds comprises:

acquiring a tag set of the content;

at least one label of the article information related to the content is fused into a label set of the content to obtain a fused label set;

and screening out at least one label of a target level from the merged label set to serve as the target label set.

9. A content recommendation apparatus comprising:

the device comprises a determining unit, a judging unit and a display unit, wherein the determining unit is configured to respond to a content acquisition request which is sent by a terminal and aims at a first channel, and determine a first form content sequence corresponding to the content acquisition request, and article information associated with first form content exists in the first form content sequence;

the selecting unit is configured to select a first target amount of second form content from a second form content pool based on the content quality score of the second form content to obtain a second form content sequence, wherein the second form content has associated item information;

a matching unit configured to match a second target number of first form contents ranked in the first form content sequence with the second form content sequence according to the associated item information to determine first form contents of the second form contents having a match;

a replacing unit configured to replace a first form content of at least one matching second form content in the first form content sequence with a corresponding second form content, resulting in a recommended content sequence;

a pushing unit configured to push the recommended content sequence to the terminal.

10. An electronic device, comprising:

one or more processors;

a storage device having one or more programs stored thereon,

when executed by the one or more processors, cause the one or more processors to implement the method of any one of claims 1-8.

11. A computer-readable medium, on which a computer program is stored, wherein the program, when executed by a processor, implements the method of any one of claims 1-8.