CN107609198B

CN107609198B - Recommendation method and device and computer readable storage medium

Info

Publication number: CN107609198B
Application number: CN201710986686.XA
Authority: CN
Inventors: 邹剑波
Original assignee: China Mobile Communications Group Co Ltd; MIGU Interactive Entertainment Co Ltd
Current assignee: China Mobile Communications Group Co Ltd; MIGU Interactive Entertainment Co Ltd
Priority date: 2017-10-20
Filing date: 2017-10-20
Publication date: 2020-06-12
Anticipated expiration: 2037-10-20
Also published as: CN107609198A

Abstract

The invention discloses a recommendation method, which comprises the following steps: acquiring related information capable of determining a scene; determining a target recommendation scene by using the related information; determining a recommendation strategy corresponding to the target recommendation scene based on a preset corresponding relation between each recommendation scene and the recommendation strategy; wherein a single said recommendation policy comprises: a set of results to be recommended corresponding to at least two types of digital content, respectively; in the recommendation strategies contained in the preset corresponding relation, information recommendation algorithms according to which at least two recommendation strategies respectively correspond to a result set to be recommended are different; and recommending information to the user according to the result set to be recommended, which is included in the determined recommendation strategy. The invention also discloses a recommendation device and a computer readable storage medium.

Description

Recommendation method and device and computer readable storage medium

Technical Field

The present invention relates to the field of intelligent recommendation, and in particular, to a recommendation method, apparatus, and computer-readable storage medium.

Background

At present, with the development of computer technology, the amount of information is larger and larger, and it is more and more common to determine effective information through recommendation. During actual recommendation, in the field of intelligent recommendation, particularly in the field of digital content recommendation, a set of recommendation algorithm is deployed for each scene, recommendation algorithm iteration is performed according to the recommendation effect, and the recommendation algorithm is continuously optimized to achieve the recommendation purpose.

However, this recommendation method does not consider the problem that the same user faces different scenes, and the preference of the same user in different scenes may be biased.

Disclosure of Invention

In view of this, embodiments of the present invention are intended to provide a recommendation method, an apparatus, and a computer-readable storage medium, which can recommend information to a user based on a scene.

The technical scheme of the embodiment of the invention is realized as follows:

the embodiment of the invention provides a recommendation method, which comprises the following steps:

acquiring related information capable of determining a scene;

determining a target recommendation scene by using the related information;

determining a recommendation strategy corresponding to the target recommendation scene based on a preset corresponding relation between each recommendation scene and the recommendation strategy; wherein a single said recommendation policy comprises: a set of results to be recommended corresponding to at least two types of digital content, respectively; in the recommendation strategies contained in the preset corresponding relation, information recommendation algorithms according to which at least two recommendation strategies respectively correspond to a result set to be recommended are different;

and recommending information to the user according to the result set to be recommended, which is included in the determined recommendation strategy.

In the foregoing solution, the acquiring related information capable of determining a scene includes:

judging whether a preset operation strategy exists or not;

and when the preset operation strategy does not exist, acquiring relevant information capable of determining the scene.

In the above scheme, the method further comprises:

when the existence of the preset operation strategy is determined, determining a result set to be recommended corresponding to the preset operation strategy based on the preset operation strategy.

In the foregoing solution, the determining a target recommendation scenario by using the relevant information includes:

determining the type of the user by using the related information;

and combining the determined type of the user with the information related to the user to determine a target recommendation scene.

In the above scheme, after determining the recommendation policy corresponding to the target recommendation scenario, the method further includes:

determining the information recommendation proportion of each type of digital content from a recommendation result set included in a recommendation strategy corresponding to the determined target recommendation scene;

and recommending based on the determined information recommendation ratio.

In the above scheme, determining the recommendation strategy corresponding to the target recommendation scene based on the preset corresponding relationship between each recommendation scene and the recommendation strategy includes:

judging whether the determined target recommendation scene is the same as the historical scene or not;

when the historical scenes are determined to be the same and the recommendation is determined to exist aiming at the historical scenes, searching the historical recommendation strategies corresponding to the historical scenes from the corresponding relation between the historical scenes and the historical recommendation strategies, and taking the searched historical recommendation strategies as the recommendation strategies of the determined target recommendation scenes; and when the recommendation strategies are different, determining the recommendation strategies corresponding to the target recommendation scenes based on the preset corresponding relation between each recommendation scene and the recommendation strategy.

In the above scheme, the method further comprises:

acquiring historical feedback information of a user;

judging whether to determine a recommendation strategy based on a scene preferentially or not by utilizing the historical feedback information;

when determining the recommendation strategy based on the scenes, determining the recommendation strategy based on the preset corresponding relation between each recommendation scene and the recommendation strategy;

when the recommendation strategy is determined by using the historical feedback information, a recommendation algorithm is determined based on the historical feedback information, a recommendation result is obtained by using the determined recommendation algorithm, and the recommendation strategy is obtained based on the recommendation result.

acquiring a result set to be recommended, which is included in a recommendation strategy corresponding to a target recommendation scene; the recommendation strategy corresponding to the target recommendation scene is determined by the first server based on the preset corresponding relation between each recommendation scene and the recommendation strategy;

determining the digital content corresponding to the result to be recommended included in the result set to be recommended by using the corresponding relation between the recommendation result and the digital content;

determining whether the determined digital content is available;

and when the determined digital content is available, sending the determined digital content to a client for displaying.

In the above scheme, the method further comprises:

acquiring digital content data or user data;

generating a recommended digital content list according to the type of the digital content by using the acquired digital content data or user data;

storing the generated recommended digital content list in a first database; the recommended digital content list is used for the first server to determine a recommendation result of the corresponding scene.

In the above scheme, when generating the recommended digital content list, the method further includes:

generating a recommendation result by using the digital content data or the user data;

and corresponding the generated recommendation result with a key value to generate a recommended digital content list of the type of the digital content.

An embodiment of the present invention provides a recommendation apparatus, where the apparatus includes:

the system comprises a first acquisition module, a second acquisition module and a third acquisition module, wherein the first acquisition module is used for acquiring related information capable of determining a scene;

the first determining module is used for determining a target recommendation scene by utilizing the related information; determining a recommendation strategy corresponding to the target recommendation scene based on a preset corresponding relation between each recommendation scene and the recommendation strategy; wherein a single said recommendation policy comprises: a set of results to be recommended corresponding to at least two types of digital content, respectively; in the recommendation strategies contained in the preset corresponding relation, information recommendation algorithms according to which at least two recommendation strategies respectively correspond to a result set to be recommended are different;

and the recommending module is used for recommending information to the user according to the result set to be recommended, which is included in the determined recommending strategy.

In the above scheme, the first obtaining module is specifically configured to determine whether a preset operation policy exists;

In the foregoing scheme, the first obtaining module is further configured to, when it is determined that a preset operation policy exists, determine, based on the preset operation policy, a result set to be recommended that corresponds to the preset operation policy.

In the foregoing solution, the first determining module is specifically configured to determine the type of the user by using the relevant information; and combining the determined type of the user with the information related to the user to determine a target recommendation scene.

In the above scheme, the recommendation module is further configured to determine an information recommendation ratio of each type of digital content from a recommendation result set included in a recommendation policy corresponding to the determined target recommendation scenario; and recommending based on the determined information recommendation ratio.

In the above scheme, the first determining module is specifically configured to determine whether the determined target recommended scene is the same as the historical scene; when the historical scenes are determined to be the same and the recommendation is determined to exist aiming at the historical scenes, searching the historical recommendation strategies corresponding to the historical scenes from the corresponding relation between the historical scenes and the historical recommendation strategies, and taking the searched historical recommendation strategies as the recommendation strategies of the determined target recommendation scenes; and when the recommendation strategies are different, determining the recommendation strategies corresponding to the target recommendation scenes based on the preset corresponding relation between each recommendation scene and the recommendation strategy.

In the above scheme, the device further comprises a first judging module,

the first judging module is used for acquiring historical feedback information of a user; judging whether to determine a recommendation strategy based on a scene preferentially or not by utilizing the historical feedback information; when determining the recommendation strategy based on the scenes, determining the recommendation strategy based on the preset corresponding relation between each recommendation scene and the recommendation strategy; when the recommendation strategy is determined by using the historical feedback information, a recommendation algorithm is determined based on the historical feedback information, a recommendation result is obtained by using the determined recommendation algorithm, and the recommendation strategy is obtained based on the recommendation result.

the second obtaining module is used for obtaining a result set to be recommended, wherein the result set to be recommended comprises recommendation strategies corresponding to the target recommendation scene; the recommendation strategy corresponding to the target recommendation scene is determined by the first server based on the preset corresponding relation between each recommendation scene and the recommendation strategy;

the second determining module is used for determining the digital content corresponding to the result to be recommended included in the result set to be recommended by using the corresponding relation between the recommendation result and the digital content;

a second judging module for judging whether the determined digital content is available; and when the determined digital content is available, sending the determined digital content to a client for displaying.

An embodiment of the present invention provides a computer-readable storage medium, on which a computer program is stored, where the computer program, when executed by a processor, implements the steps of any of the recommendation methods described above.

An embodiment of the present invention provides a recommendation apparatus, including: a processor and a memory for storing a computer program capable of running on the processor,

wherein the processor is configured to execute the steps of any of the above-mentioned recommended methods when running the computer program.

The recommendation method, the recommendation device and the computer-readable storage medium provided by the embodiment of the invention can be used for acquiring relevant information capable of determining a scene; determining a target recommendation scene by using the related information; based on a preset corresponding relation between scenes and recommendation results, determining recommendation strategies corresponding to the target recommendation scenes according to the target recommendation scenes and the preset corresponding relation between each recommendation scene and a recommendation strategy; wherein a single said recommendation policy comprises: a set of results to be recommended corresponding to at least two types of digital content, respectively; in the recommendation strategies contained in the preset corresponding relation, information recommendation algorithms according to which at least two recommendation strategies respectively correspond to a result set to be recommended are different; and recommending information to the user according to the result set to be recommended, which is included in the determined recommendation strategy. In the embodiment of the invention, a target recommendation scene is determined by utilizing the related information; and determining a recommendation strategy corresponding to the target recommendation scene based on the preset corresponding relation between each recommendation scene and the recommendation strategy, and recommending information to the user according to a result set to be recommended, wherein the result set to be recommended is included in the determined recommendation strategy. Therefore, information recommendation to the user based on the scene can be realized.

Drawings

FIG. 1 is a first schematic flow chart illustrating an implementation of a recommendation method according to an embodiment of the present invention;

FIG. 2 is a schematic diagram of a second implementation flow of the recommendation method according to the embodiment of the present invention;

FIG. 3 is a first schematic diagram illustrating a first exemplary embodiment of a recommendation device;

FIG. 4 is a schematic diagram of a second exemplary embodiment of a recommendation device;

FIG. 5 is a third schematic view of a component structure of a recommendation device according to an embodiment of the present invention;

FIG. 6 is a flowchart illustrating a preferred implementation of an embodiment of the present invention;

fig. 7 is a schematic structural diagram of components for implementing digital content recommendation by interaction between a first server and a second server according to an embodiment of the present invention.

Detailed Description

In the embodiment of the invention, the related information capable of determining the scene is obtained; determining a target recommendation scene by using the related information; determining a recommendation strategy corresponding to the target recommendation scene according to the target recommendation scene and a preset corresponding relation between each recommendation scene and the recommendation strategy; wherein a single said recommendation policy comprises: a set of results to be recommended corresponding to at least two types of digital content, respectively; in the recommendation strategies contained in the preset corresponding relation, information recommendation algorithms according to which at least two recommendation strategies respectively correspond to a result set to be recommended are different; and recommending information to the user according to the result set to be recommended, which is included in the determined recommendation strategy.

So that the manner in which the features and aspects of the embodiments of the present invention can be understood in detail, a more particular description of the embodiments of the invention, briefly summarized above, may be had by reference to the embodiments, some of which are illustrated in the appended drawings.

As shown in fig. 1, a method for recommending an embodiment of the present invention is described in detail, and the method for recommending an embodiment of the present invention is applied to a first server side, and includes the following steps:

step 101: related information capable of determining a scene is acquired.

In one embodiment, the obtaining can determine that the scene includes:

judging whether a preset operation strategy exists or not; and when the preset operation strategy does not exist, acquiring relevant information capable of determining the scene.

In an embodiment, the method further comprises:

Specifically, the result set to be recommended may be determined according to a significant festival (e.g., a national celebration topic, an intermediate autumn topic, etc.) or a significant event (e.g., a local concert, a drama performance, etc.).

Step 102: and determining a target recommendation scene by using the related information.

In an embodiment, the determining a target recommendation scenario by using the relevant information includes: determining the type of the user by using the related information; and combining the determined type of the user with the information related to the user to determine a target recommendation scene.

Here, the related information may include: the user's natural attributes (e.g., gender, age, etc.) and access records, etc. In actual application, the first server classifies users into a certain category (such as deep search type users or browsing type users) according to the natural attributes and access records of the users; if the user's natural attributes and access records cannot be obtained, the user may be classified into a default category (e.g., a browsing-type user).

Table 1 shows the determined target recommended scene, and as shown in table 1, if the user type is a deep search type user, the determined target recommended scene is scene a; if the user type is a browsing type user, determining that the target recommendation scene is a scene B; if the user type is a browsing type user, the user related information comprises time information (such as night) and location information (such as user home), the user type and the user related information are combined, and the determined target recommendation scene is a scene C.

Deep search type user	Scene A
		Browsing type user	Scene B
Browsing type user + evening + user's home	Scene C

TABLE 1

Step 103: and determining a recommendation strategy corresponding to the target recommendation scene based on the preset corresponding relation between each recommendation scene and the recommendation strategy.

Wherein a single said recommendation policy comprises: a set of results to be recommended corresponding to at least two types of digital content, respectively; in the recommendation strategies contained in the preset corresponding relation, information recommendation algorithms according to which at least two recommendation strategies respectively correspond to the result sets to be recommended are different.

In actual application, a result set to be recommended corresponding to the target recommendation scene may also be determined from the recommended digital content list acquired from the first database.

Table 2 shows the preset correspondence between each recommended scene and the recommendation policy, where as shown in table 2, the recommended scene includes a scene a, a scene B, and a scene C, and the recommendation policy includes a recommendation policy a, a recommendation policy B, and a recommendation result C. Taking the recommendation policy a as an example, the recommendation policy a includes 3 recommendation results of a video class (i.e., recommendation result 1, recommendation result 2, and recommendation result 3), 2 recommendation results of a game class (i.e., recommendation result 1, recommendation result 2), 3 recommendation results of an animation class (i.e., recommendation result 1, recommendation result 2, and recommendation result 3), 2 recommendation results of a music class (i.e., recommendation result 1, recommendation result 2), and 1 recommendation result of a book class (i.e., recommendation result 1).

TABLE 2

For example, scenario a may be: the user watches the pay video at night; the recommendation result 1 of the video class corresponding to the scene a may be: the most recently purchased video. Scenario B may be: a deep search type user; the recommendation corresponding to scenario B includes: recommendation 9 for video category (guess you like video) + recommendation 1 for video category (surprised video). Scenario C may be: a browsing-type user; the recommendation corresponding to scenario B includes: the result of the recommendation for music is 6 (guessing the music you like) +4 (surprise the recommended music).

Here, there are at least two information recommendation algorithms according to which the result sets to be recommended respectively correspond to the recommendation strategies are different. For example, the information recommendation algorithms according to which the result sets to be recommended respectively correspond to the recommendation policy a and the recommendation policy B are different.

In practical application, the key values corresponding to the recommendation results in the set of results to be recommended and the recommendation results can be correspondingly stored in the second database. The second database may be a key-value database, that is, a database storing data in key-value pairs.

In an embodiment, after determining the recommendation policy corresponding to the target recommendation scenario, the method further includes: determining the information recommendation proportion of each type of digital content from a recommendation result set included in a recommendation strategy corresponding to the determined target recommendation scene; and recommending based on the determined information recommendation ratio.

In practical application, the information recommendation proportion of each type of digital content can be determined from the recommendation result set included in the recommendation strategy corresponding to the determined target recommendation scene by combining the acquired terminal information. The terminal information may include: the client side comprises an installation list of various applications in the client side, the percentage of various cached data types, a network scene (such as whether the mobile terminal is in a wireless local area network or the network speed of the mobile terminal), request time (namely the response time of the mobile terminal and used for reflecting the processing speed of a processor of the mobile terminal) and the like.

Specifically, when the user opens the client, the client may obtain an installation list of various applications, may acquire a video file and a music file, etc. cached in the SD card of the user under the permission of the user, and may further obtain a real-time network status. The client reports the terminal information to the first server in real time through the message interface, and the first server determines the information recommendation proportion of each type of digital content from the recommendation result set included in the recommendation strategy corresponding to the determined target recommendation scene according to the acquired terminal information.

For example, if 90% of the local cache data is analyzed according to the terminal information as video and the network is 4G, it can be determined that the video accounts for 80% of the recommended types (game, cartoon, music, video, book) and the game accounts for 20% of the recommended types (game, cartoon, music, video, book). When information recommendation is made, 4 short videos in the video type, 1 mini game in the game type, may be recommended.

In an embodiment, determining a recommendation policy corresponding to the target recommendation scenario based on a preset correspondence between each recommendation scenario and the recommendation policy includes: judging whether the determined target recommendation scene is the same as the historical scene or not; when the historical scenes are determined to be the same and the recommendation is determined to exist aiming at the historical scenes, searching the historical recommendation strategies corresponding to the historical scenes from the corresponding relation between the historical scenes and the historical recommendation strategies, and taking the searched historical recommendation strategies as the recommendation strategies of the determined target recommendation scenes; and when the recommendation strategies are different, determining the recommendation strategies corresponding to the target recommendation scenes based on the preset corresponding relation between each recommendation scene and the recommendation strategy.

In an embodiment, the method further comprises: acquiring historical feedback information of a user; judging whether to determine a recommendation strategy based on a scene preferentially or not by utilizing the historical feedback information; when determining the recommendation strategy based on the scenes, determining the recommendation strategy based on the preset corresponding relation between each recommendation scene and the recommendation strategy; when the recommendation strategy is determined by using the historical feedback information, a recommendation algorithm is determined based on the historical feedback information, a recommendation result is obtained by using the determined recommendation algorithm, and the recommendation strategy is obtained based on the recommendation result.

In actual application, feedback information (such as likes, dislikes, likes, favorites, collections and the like) of the user for the recommended digital content is synchronized to the first server in real time through the KAFKA message interface. Since each recommended type of digital content carries a batch number, the first server can query a recommendation result and a recommendation algorithm corresponding to the recommended digital content according to the batch number.

For example, if the user performs an operation on a certain type of recommended digital content (e.g., clicks on like, dislike, like, favorite, etc. buttons on the screen for the digital content), the first server may determine, according to the operation of the user, a recommendation result and a recommendation algorithm that the user likes in a scene in a comparative manner. And generating an algorithm inclination value according to the hit rate, and if the hit rate reaches 20%, 100% of the digital content corresponding to the recommendation result is inclined to be recommended.

Step 104: and recommending information to the user according to the result set to be recommended, which is included in the determined recommendation strategy.

In actual application, the first server can also send the key values corresponding to the recommendation results in the set of results to be recommended to the second server; and the key value is used for the second server to obtain a recommendation result corresponding to the scene from the second database and determine the digital content corresponding to the scene by using the recommendation result.

According to the recommendation method provided by the embodiment of the invention, the relevant information capable of determining the scene is obtained; determining a target recommendation scene by using the related information; determining a recommendation strategy corresponding to the target recommendation scene based on a preset corresponding relation between each recommendation scene and the recommendation strategy; wherein a single said recommendation policy comprises: a set of results to be recommended corresponding to at least two types of digital content, respectively; in the recommendation strategies contained in the preset corresponding relation, information recommendation algorithms according to which at least two recommendation strategies respectively correspond to a result set to be recommended are different; and recommending information to the user according to the result set to be recommended, which is included in the determined recommendation strategy. Obviously, information recommendation can be made to the user based on the scene.

As shown in fig. 2, the method for recommending an embodiment of the present invention is described in detail in the embodiment of the present invention, and the method for recommending an embodiment of the present invention is applied to the second server side, and includes the following steps:

step 201: and acquiring a result set to be recommended, which is included in the recommendation strategy corresponding to the target recommendation scene.

And the recommendation strategy corresponding to the target recommendation scene is determined by the first server based on the preset corresponding relation between each recommendation scene and the recommendation strategy.

In actual application, the result set to be recommended can be obtained from the second database according to the key value by using the key value corresponding to the recommendation result in the result set to be recommended returned by the first server. Wherein the second database may be a key-value database; the key-value database is a database that stores data in key-value pairs. The second database may be stored on the cluster.

Step 202: and determining the digital content corresponding to the result to be recommended included in the result set to be recommended by utilizing the corresponding relation between the recommendation result and the digital content.

In actual application, the recommendation result can be a short video; the corresponding digital content may be digital content formed with short videos according to an update time. Here, the recommendation result may be an ID of the digital content. The second server may store the correspondence of the digital content ID and the digital content in the first database.

Step 203: determining whether the determined digital content is available; and when the determined digital content is available, sending the determined digital content to a client for displaying.

Here, the determining whether the determined digital content is available may specifically include: judging whether the copyright is expired, and if so, making the digital content unavailable; otherwise the digital content is available.

In an embodiment, the method further comprises: acquiring digital content data or user data; generating a recommended digital content list according to the type of the digital content by using the acquired digital content data or user data; and storing the generated recommended digital content list in a first database. Here, the recommended digital content list may be used for the first server to determine a result set to be recommended included in the recommendation policy of the corresponding scenario.

In one embodiment, when generating the recommended digital content list, the method further comprises: generating a recommendation result by using the digital content data or the user data; and corresponding the generated recommendation result with a key value to generate a recommended digital content list of the type of the digital content.

Specifically, the latest recommendation result (recommendation result 1) may be generated according to the own attribute of the digital content; generating the hottest recommendation result (recommendation result 2) and the latest purchased recommendation result (recommendation result 3) according to the behavior of the user; guess-like recommendation results (recommendation results 4) are generated based on the user feature vectors obtained from the user behaviors and the similarity between the digital contents. The first server can write recommendation results 1-4 into the REDIS cluster according to the category of the digital content for the second server to call. Wherein the types of the digital content include: games, animations, music, videos and books.

Here, including but not limited to the above four recommendation results, more recommendation results may also be generated according to other recommendation algorithms, and the recommendation algorithm for determining the recommendation result is a recommendation algorithm common in the art.

Based on the recommendation method provided in each embodiment of the present application, the present application further provides a recommendation apparatus, which may be disposed on the first server, as shown in fig. 3, where the apparatus includes: a first obtaining module 31, a first determining module 32 and a recommending module 33; wherein the content of the first and second substances,

a first obtaining module 31, configured to obtain relevant information capable of determining a scene.

A first determining module 32, configured to determine a target recommendation scenario by using the relevant information; and determining a recommendation strategy corresponding to the target recommendation scene based on the preset corresponding relation between each recommendation scene and the recommendation strategy.

Wherein a single said recommendation policy comprises: a set of results to be recommended corresponding to at least two types of digital content, respectively; in the recommendation strategies contained in the preset corresponding relation, information recommendation algorithms according to which at least two recommendation strategies respectively correspond to a result set to be recommended are different;

and the recommending module 33 is configured to recommend information to the user according to the result set to be recommended included in the determined recommending policy.

In an embodiment, the first obtaining module 31 is specifically configured to determine whether a preset operation policy exists; and when the preset operation strategy does not exist, acquiring relevant information capable of determining the scene.

In an embodiment, the first obtaining module 31 is further configured to, when it is determined that a preset operation policy exists, determine, based on the preset operation policy, a result set to be recommended that corresponds to the preset operation policy.

In an embodiment, the first determining module 32 is specifically configured to determine the type of the user by using the relevant information; and combining the determined type of the user with the information related to the user to determine a target recommendation scene.

Here, the related information may include: the user's natural attributes (e.g., gender, age, etc.) and access records, etc. In actual application, classifying users into a certain category (such as deep search type users or browsing type users) according to the natural attributes and access records of the users; if the user's natural attributes and access records cannot be obtained, the user may be classified into a default category (e.g., a browsing-type user).

In an embodiment, the recommending module 33 is further configured to determine an information recommendation ratio of each type of digital content from a recommendation result set included in a recommendation policy corresponding to the determined target recommendation scenario; and recommending based on the determined information recommendation ratio.

In an embodiment, the first determining module 32 is specifically configured to determine whether the determined target recommendation scene is the same as the historical scene; when the historical scenes are determined to be the same and the recommendation is determined to exist aiming at the historical scenes, searching the historical recommendation strategies corresponding to the historical scenes from the corresponding relation between the historical scenes and the historical recommendation strategies, and taking the searched historical recommendation strategies as the recommendation strategies of the determined target recommendation scenes; and when the recommendation strategies are different, determining the recommendation strategies corresponding to the target recommendation scenes based on the preset corresponding relation between each recommendation scene and the recommendation strategy.

In one embodiment, the apparatus further comprises a first determining module;

It should be noted that: in the recommendation apparatus provided in the above embodiment, only the division of the program modules is exemplified when performing recommendation, and in practical applications, the above processing may be distributed to different program modules according to needs, that is, the internal structure of the apparatus may be divided into different program modules to complete all or part of the above-described processing. In addition, the recommendation device and the recommendation method provided by the above embodiments belong to the same concept, and specific implementation processes thereof are described in the method embodiments in detail and are not described herein again.

In practical applications, the first obtaining module 31 is implemented by a network interface located on the recommending device; the first determining module 32, the first recommending module 33, and the first determining module may be implemented by a Central Processing Unit (CPU), a microprocessor Unit (MPU), a Digital Signal Processor (DSP), a Field Programmable Gate Array (FPGA), or the like on the recommending apparatus.

Based on the recommendation method provided by each embodiment of the present application, the present application further provides a recommendation apparatus, which may be disposed on a second server, as shown in fig. 4, where the apparatus includes: a second obtaining module 41, a second determining module 42, and a second judging module 43; wherein the content of the first and second substances,

the second obtaining module 41 is configured to obtain a result set to be recommended, where the recommendation policy corresponding to the target recommendation scenario includes the result set. And the recommendation strategy corresponding to the target recommendation scene is determined by the first server based on the preset corresponding relation between each recommendation scene and the recommendation strategy.

A second determining module 42, configured to determine, by using the correspondence between the recommendation result and the digital content, the digital content corresponding to the to-be-recommended result included in the to-be-recommended result set;

a second judging module 43, configured to judge whether the determined digital content is available; and when the determined digital content is available, sending the determined digital content to a client for displaying.

In one embodiment, the apparatus further comprises a generation module,

the generating module is used for acquiring digital content data or user data; generating a recommended digital content list according to the type of the digital content by using the acquired digital content data or user data; storing the generated recommended digital content list in a first database; the recommended digital content list is used for the first server to determine a recommendation result of the corresponding scene.

In an embodiment, the generating module is specifically configured to generate a recommendation result by using the digital content data or the user data; and corresponding the generated recommendation result with a key value to generate a recommended digital content list of the type of the digital content.

In practical applications, the second obtaining module 41 is implemented by a network interface located on the recommending device; the second determining module 42, the second judging module 43, and the generating module may be implemented by a CPU, a microprocessor MPU, a DSP, or an FPGA located on the recommending apparatus.

Fig. 5 is a schematic structural diagram of a recommendation device of the present invention, and the recommendation device 500 shown in fig. 5 includes: at least one processor 501, memory 502, user interface 503, at least one network interface 504. The various components in the recommendation device 500 are coupled together by a bus system 505. It is understood that the bus system 505 is used to enable connection communications between these components. The bus system 505 includes a power bus, a control bus, and a status signal bus in addition to a data bus. For clarity of illustration, however, the various buses are labeled as bus system 505 in FIG. 5.

The user interface 503 may include a display, a keyboard, a mouse, a trackball, a click wheel, a key, a button, a touch pad, a touch screen, or the like, among others.

It will be appreciated that the memory 502 can be either volatile memory or nonvolatile memory, and can include both volatile and nonvolatile memory. Among them, the nonvolatile Memory may be a Read Only Memory (ROM), a Programmable Read Only Memory (PROM), an Erasable Programmable Read-Only Memory (EPROM), an Electrically Erasable Programmable Read-Only Memory (EEPROM), a magnetic random access Memory (FRAM), a Flash Memory (Flash Memory), a magnetic surface Memory, an optical disk, or a Compact Disc Read-Only Memory (CD-ROM); the magnetic surface storage may be disk storage or tape storage. Volatile memory can be Random Access Memory (RAM), which acts as external cache memory. By way of illustration and not limitation, many forms of RAM are available, such as Static Random Access Memory (SRAM), Synchronous Static Random Access Memory (SSRAM), Dynamic Random Access Memory (DRAM), Synchronous Dynamic Random Access Memory (SDRAM), Double Data Rate Synchronous Dynamic Random Access Memory (DDRSDRAM), Enhanced Synchronous Dynamic Random Access Memory (ESDRAM), Enhanced Synchronous Dynamic Random Access Memory (Enhanced DRAM), Synchronous Dynamic Random Access Memory (SLDRAM), Direct Memory (DRmb Access), and Random Access Memory (DRAM). The memory 502 described in connection with the embodiments of the invention is intended to comprise, without being limited to, these and any other suitable types of memory.

The memory 502 in embodiments of the present invention is used to store various types of data to support the operation of the recommendation device 500. Examples of such data include: any computer programs for operating on the recommendation device 500, such as an operating system 5021 and application programs 5022; the operating system 5021 includes various system programs, such as a framework layer, a core library layer, a driver layer, and the like, for implementing various basic services and processing hardware-based tasks. The application programs 5022 may contain various application programs for implementing various application services. The program for implementing the method according to the embodiment of the present invention may be included in the application program 5022.

The method disclosed by the above-mentioned embodiments of the present invention may be applied to the processor 501, or implemented by the processor 501. The processor 501 may be an integrated circuit chip having signal processing capabilities. In implementation, the steps of the above method may be performed by integrated logic circuits of hardware or instructions in the form of software in the processor 501. The processor 501 described above may be a general purpose processor, a digital signal processor, or other programmable logic device, discrete gate or transistor logic device, discrete hardware components, or the like. Processor 501 may implement or perform the methods, steps, and logic blocks disclosed in embodiments of the present invention. A general purpose processor may be a microprocessor or any conventional processor or the like. The steps of the method disclosed by the embodiment of the invention can be directly implemented by a hardware decoding processor, or can be implemented by combining hardware and software modules in the decoding processor. The software modules may be located in a storage medium located in the memory 502, and the processor 501 reads the information in the memory 502 and performs the steps of the aforementioned recommended method in combination with its hardware.

Based on the recommendation method provided in the embodiments of the present application, the present application further provides a computer-readable storage medium, as shown in fig. 5, the computer-readable storage medium may include: a memory 502 for storing a computer program executable by the processor 501 of the recommendation device 500 for performing the steps of the aforementioned recommendation method. The computer readable storage medium may be Memory such as FRAM, ROM, PROM, EPROM, EEPROM, Flash Memory, magnetic surface Memory, optical disk, or CD-ROM.

The following description will explain the implementation and principles of the present invention in practical applications by taking recommendations for implementing digital content as specific examples.

Fig. 6 is a schematic diagram of a specific implementation flow of digital content recommendation according to an embodiment of the present invention, and fig. 7 is a schematic diagram of a composition structure of a first server and a second server for implementing digital content recommendation by interaction, where as shown in fig. 7, a policy engine corresponds to the first server, a computing platform corresponds to the second server, and a first database corresponds to the REDIS cluster 1. The specific implementation process comprises the following steps:

step 601: the computing platform generates different recommendation results according to different recommendation algorithms for different types of digital content, and stores all recommendation results in the REDIS cluster 1 according to the types of the digital content.

The REDIS cluster 1 is a key-value-based storage system, i.e., a key-value database.

Specifically, digital content data related to the digital content and user data related to the behavior of the user are input to the computing platform; the computing platform classifies the digital content according to types, and the digital content is divided into five types of games, cartoons, music, videos and books; generating a latest recommendation result (such as recommendation result 1) according to digital content data (such as the self attribute of the digital content) aiming at each type of digital content; generating the hottest recommendation result (such as recommendation result 2) and the latest purchased recommendation result (such as recommendation result 3) according to the user data (data related to the behavior of the user); guess-like recommendation results (such as recommendation result 4) are generated according to the user feature vectors obtained by the user data and the similarity between the digital contents. And the computing platform writes all recommendation results (such as recommendation results 1-4) in categories (such as games, cartoons, music, videos and books) into the REDIS cluster 1 for the strategy engine to call.

The correspondence of the digital content ID and the digital content may also be stored in the REDIS cluster 1 of the computing platform. When a certain digital content is searched, the computing platform may search the corresponding digital content in the REDIS cluster 1 according to the digital content ID.

Step 602: the strategy engine prestores the corresponding relation between the recommended scenes and the recommended strategies.

For example, the policy engine may store a recommendation policy corresponding to a scene (e.g., a recommendation scene for a user to watch a pay video at night) and a recommendation (e.g., a most recently purchased recommendation in a video category).

Step 603: and the strategy engine determines a target recommendation scene according to the relevant information which is input by the computing platform and can determine the scene.

The computing platform may be a server that performs different services, such as a mini running server, a mini video server, a mini reading server, and so on.

The method comprises the steps of judging whether operation intervention exists or not in advance before determining a target recommendation scene according to relevant information which is input by a computing platform and can determine the scene; if yes, determining a recommendation result according to a preset operation strategy (e.g. a rule that the recommendation result is preset according to a significant festival (such as a national festival topic, a mid-autumn topic, and the like) or a significant event (such as a local concert, a drama performance, and the like)), correspondingly storing a KEY value corresponding to the determined recommendation result and the recommendation result in the REDIS cluster 2, and simultaneously returning KEY value (KEY value) information to the computing platform, and the computing platform determining the recommended digital content according to the KEY value; if not, go to step 603.

Step 604: the policy engine determines a to-be-recommended result set corresponding to a target recommended scene from a recommended digital content list acquired from the REDIS cluster 1 based on the corresponding relationship between the scene and the recommended result, writes a KEY value corresponding to the recommended result included in the to-be-recommended result set and the recommended result corresponding to the scene into the REDIS cluster 2, and simultaneously returns KEY value information (KEY value) to the computing platform for the computing platform to read the recommended result corresponding to the calling scene from the REDIS cluster 2.

Step 605: and the computing platform acquires a recommendation result corresponding to the scene from the REDIS cluster 2 according to the KEY value returned by the strategy engine, and forms the acquired recommendation result into digital content corresponding to the scene. And checking whether the formed digital content is available, and when the formed digital content is available, displaying the digital content of the corresponding scene at the client.

The recommendation result obtained from the REDIS cluster 2 corresponds to the digital content ID, and cannot be directly displayed at the client, so that the digital content corresponding to the digital content ID needs to be obtained from the computing platform by using the digital content ID. And meanwhile, judging whether the formed digital content is available in the current state (if the copyright is judged to be expired, if the copyright is expired, the formed digital content is not available, otherwise, the formed digital content is available), and displaying the formed digital content at the client when the formed digital content is available. When the user operates the recommended digital content at the client, the generated operation data is returned to the strategy engine in real time through the KAFKA message interface, and the strategy engine takes the fed-back data as the basis of the next recommendation.

The above description is only a preferred embodiment of the present invention, and is not intended to limit the scope of the present invention.

Claims

1. A recommendation method, characterized in that the method comprises:

acquiring related information capable of determining a scene;

determining a target recommendation scene by using the related information;

recommending information to the user according to the result set to be recommended, which is included in the determined recommendation strategy;

before determining a target recommendation scene by using the relevant information, the method further includes:

judging whether operation intervention exists; when operation intervention is determined, determining a to-be-recommended result set corresponding to a preset operation strategy according to the preset operation strategy, and sending a key value corresponding to a recommended result in the to-be-recommended result set to a second server; the key value is used for the second server to obtain a recommendation result corresponding to the scene from the second database and determine digital content which corresponds to the scene and is recommended to the user by using the recommendation result; when it is determined that no operation intervention exists, determining a target recommendation scene by using the relevant information;

the information recommendation to the user according to the result set to be recommended included in the determined recommendation strategy comprises the following steps:

sending the key value corresponding to the recommendation result in the set of results to be recommended to a second server; and the key value is used for the second server to obtain a recommendation result corresponding to the scene from the second database and determine the digital content which is corresponding to the scene and is recommended to the user by using the recommendation result.

2. The method of claim 1, wherein the determining a target recommendation scenario using the relevant information comprises:

determining the type of the user by using the related information;

3. The method of claim 1, wherein after determining the recommendation policy corresponding to the target recommendation scenario, the method further comprises:

and recommending based on the determined information recommendation ratio.

4. The method according to claim 1, wherein determining the recommendation strategy corresponding to the target recommendation scenario based on a preset correspondence between each recommendation scenario and the recommendation strategy comprises:

5. The method of claim 1, further comprising:

acquiring historical feedback information of a user;

6. A recommendation method, characterized in that the method comprises:

determining whether the determined digital content is available;

when the digital content is determined to be available, sending the determined digital content to a client for display;

the method for acquiring the result set to be recommended, which is included in the recommendation strategy corresponding to the target recommendation scene, comprises the following steps:

when operation intervention exists, obtaining a key value corresponding to a recommendation result in a result set to be recommended, wherein the result set to be recommended comprises a preset operation strategy; the preset operation strategy is a strategy pre-stored in the first server; when operation intervention does not exist, obtaining a key value corresponding to a recommendation result in a result set to be recommended, wherein the recommendation strategy corresponding to a target recommendation scene comprises the recommendation result set; the key value is used for obtaining a recommendation result corresponding to the scene from a second database and determining digital content which is corresponding to the scene and is recommended to the user by using the recommendation result.

7. The method of claim 6, further comprising:

acquiring digital content data or user data;

8. The method of claim 7, wherein when generating the list of recommended digital content, the method further comprises:

9. A recommendation device, characterized in that the device comprises:

the recommendation module is used for recommending information to the user according to the result set to be recommended, which is included in the determined recommendation strategy;

the device further comprises:

the judging module is used for judging whether operation intervention exists or not; when operation intervention is determined, determining a to-be-recommended result set corresponding to a preset operation strategy according to the preset operation strategy, and sending a key value corresponding to a recommended result in the to-be-recommended result set to a second server; the key value is used for the second server to obtain a recommendation result corresponding to the scene from the second database and determine digital content which corresponds to the scene and is recommended to the user by using the recommendation result; when it is determined that no operation intervention exists, determining a target recommendation scene by using the relevant information;

the recommending module is specifically configured to send a key value corresponding to a recommending result in the set of results to be recommended to a second server; and the key value is used for the second server to obtain a recommendation result corresponding to the scene from the second database and determine the digital content which is corresponding to the scene and is recommended to the user by using the recommendation result.

10. The apparatus of claim 9,

the first determining module is specifically configured to determine the type of the user by using the relevant information; and combining the determined type of the user with the information related to the user to determine a target recommendation scene.

11. The apparatus of claim 9,

the recommendation module is further used for determining the information recommendation proportion of each type of digital content from a recommendation result set included in the recommendation strategy corresponding to the determined target recommendation scene; and recommending based on the determined information recommendation ratio.

12. The apparatus of claim 9,

the first determining module is specifically configured to determine whether the determined target recommendation scene is the same as the historical scene; when the historical scenes are determined to be the same and the recommendation is determined to exist aiming at the historical scenes, searching the historical recommendation strategies corresponding to the historical scenes from the corresponding relation between the historical scenes and the historical recommendation strategies, and taking the searched historical recommendation strategies as the recommendation strategies of the determined target recommendation scenes; and when the recommendation strategies are different, determining the recommendation strategies corresponding to the target recommendation scenes based on the preset corresponding relation between each recommendation scene and the recommendation strategy.

13. The apparatus of claim 9, further comprising a first determining module,

14. A recommendation device, characterized in that the device comprises:

a second judging module for judging whether the determined digital content is available; when the digital content is determined to be available, sending the determined digital content to a client for display;

the second obtaining module is specifically configured to, when there is operation intervention, obtain a key value corresponding to a recommendation result in a to-be-recommended result set included in a preset operation policy; the preset operation strategy is a strategy pre-stored in the first server; when operation intervention does not exist, obtaining a key value corresponding to a recommendation result in a result set to be recommended, wherein the recommendation strategy corresponding to a target recommendation scene comprises the recommendation result set; the key value is used for obtaining a recommendation result corresponding to the scene from a second database and determining digital content which is corresponding to the scene and is recommended to the user by using the recommendation result.

15. A computer-readable storage medium, on which a computer program is stored, which, when being executed by a processor, carries out the steps of the method of any one of claims 1 to 5, or carries out the steps of the method of any one of claims 6 to 8.

16. A recommendation device, comprising: a processor and a memory for storing a computer program capable of running on the processor,

wherein the processor is adapted to perform the steps of the method of any one of claims 1 to 5 when running the computer program.

17. A recommendation device, comprising: a processor and a memory for storing a computer program capable of running on the processor,

wherein the processor is adapted to perform the steps of the method of any one of claims 6 to 8 when running the computer program.