CN112100507B

CN112100507B - Object recommendation method, computing device and computer-readable storage medium

Info

Publication number: CN112100507B
Application number: CN202011242763.9A
Authority: CN
Inventors: 常建龙; 赵玮; 张辛宇; 姜浩然
Original assignee: Nanjing Yanli Technology Co ltd
Current assignee: Shanghai shouqianba Internet Technology Co.,Ltd.
Priority date: 2020-11-10
Filing date: 2020-11-10
Publication date: 2021-02-12
Anticipated expiration: 2040-11-10
Also published as: CN112100507A

Abstract

According to an embodiment of the present disclosure, a method of object recommendation, an electronic device, and a computer-readable storage medium are provided. A method of object recommendation, comprising: determining at least one reference user associated with the target user based on historical selection records of the target user and the plurality of reference users for the plurality of online objects and the plurality of offline objects; determining the characteristics of the user and the object based on the historical selection record; associating the at least one offline object with the at least one online object based on the characteristics of the objects; updating the historical selection record of the at least one reference user for the associated at least one online object based on the historical selection record; and determining at least one online object for recommendation to the target user based on the updated historical selection record and the characteristics of the user. According to the scheme disclosed by the invention, the object can be accurately recommended to the user, and the user experience is improved.

Description

Object recommendation method, computing device and computer-readable storage medium

Technical Field

Embodiments of the present disclosure relate to the field of information recommendation, and more particularly, to object recommendation methods, computing devices, and computer-readable storage media for object recommendation.

Background

With the rapid development of the internet, the information received by people is also increased explosively, and a recommendation system needs to recommend resources and commodities which are interested by users under the condition of information overload, so that the user experience is improved, and the distribution efficiency of the resources and the purchase rate of the commodities are improved. In conventional object recommendation schemes, a computing device typically makes a recommendation for a target object based on obtaining historical operational behavior data on-line by a user. However, when there is a lack of online behavior of the user (e.g., online purchase recording and online video viewing recording), or there are a large number of target objects that are not operated, resources and goods are often not recommended to the user appropriately.

Therefore, the conventional object recommendation scheme has disadvantages in that: when the online operation behavior data of the user is very sparse, the target object cannot be accurately recommended to the user.

Therefore, a method for efficiently and accurately recommending an object to a user who lacks an online selection record is needed.

Disclosure of Invention

According to the embodiments of the present disclosure, an object recommendation method, a computing device and a computer-readable storage medium are provided, which can efficiently and accurately recommend a matching target object to a user even in the absence of online operation behavior data.

In a first aspect of the present disclosure, there is provided a method of object recommendation, comprising: determining at least one reference user of the plurality of reference users associated with the target user based on historical selection records of the target user and the plurality of reference users for the plurality of online objects and the plurality of offline objects, the historical selection records indicating at least a number of historical selections for the object; determining a first feature of a target user, a first feature of at least one reference user, a first feature of a plurality of online objects, and a first feature of a plurality of offline objects based on a historical selection record; associating the at least one offline object with the at least one online object based on the first features of the plurality of online objects and the first features of the plurality of offline objects; updating the historical selection record of the at least one reference user for the associated at least one online object based on the historical selection record of the at least one reference user for the at least one offline object; and determining at least one online object for recommendation to the target user based on the updated historical selection record, the first characteristic of the target user, and the first characteristic of the at least one reference user.

In a second aspect of the present disclosure, there is provided an apparatus for object recommendation, comprising: determining at least one reference user of the plurality of reference users associated with the target user based on historical selection records of the target user and the plurality of reference users for the plurality of online objects and the plurality of offline objects, the historical selection records indicating at least a number of historical selections for the object; determining a first feature of a target user, a first feature of at least one reference user, a first feature of a plurality of online objects, and a first feature of a plurality of offline objects based on a historical selection record; associating the at least one offline object with the at least one online object based on the first features of the plurality of online objects and the first features of the plurality of offline objects; updating the historical selection record of the at least one reference user for the associated at least one online object based on the historical selection record of the at least one reference user for the at least one offline object; and determining at least one online object for recommendation to the target user based on the updated historical selection record, the first characteristic of the target user, and the first characteristic of the at least one reference user.

In a third aspect of the disclosure, a computing device is provided, comprising one or more processors; and storage means for storing the one or more programs which, when executed by the one or more processors, cause the one or more processors to carry out the method according to the first aspect of the disclosure.

In a fourth aspect of the present disclosure, a computer-readable storage medium is provided, on which a computer program is stored which, when executed by a processor, implements a method according to the first aspect of the present disclosure.

It should be understood that the statements herein reciting aspects are not intended to limit the critical or essential features of the embodiments of the present disclosure, nor are they intended to limit the scope of the present disclosure. Other features of the present disclosure will become apparent from the following description.

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. In the drawings, like or similar reference characters designate like or similar elements.

FIG. 1 illustrates a schematic diagram of an example environment in which embodiments of the present disclosure can be implemented.

FIG. 2 illustrates a flow diagram of a process for object recommendation, according to some embodiments of the present disclosure.

FIG. 3 illustrates a flow diagram of a process of updating a historical selection record, according to some embodiments of the present disclosure.

FIG. 4 illustrates a flow diagram of a process of determining a recommendation for an object to a user according to some embodiments of the present disclosure.

FIG. 5 illustrates a block diagram of a computing device capable of implementing various embodiments of the present disclosure.

Detailed Description

Preferred embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. While the preferred embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure may be embodied in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the disclosure to those skilled in the art.

The term "include" and variations thereof as used herein is meant to be inclusive in an open-ended manner, i.e., "including but not limited to". Unless specifically stated otherwise, the term "or" means "and/or". The term "based on" means "based at least in part on". The terms "one example embodiment" and "one embodiment" mean "at least one example embodiment". The term "another embodiment" means "at least one additional embodiment". The terms "first," "second," and the like may refer to different or the same object. Other explicit and implicit definitions are also possible below.

In the description of embodiments of the present disclosure, the term "model" may learn from training data the associations between respective inputs and outputs, such that after training is completed, a given input is processed based on a trained set of parameters to generate a corresponding output. The "model" may also sometimes be referred to as a "neural network", "learning model", "learning network", or "network". These terms are used interchangeably herein.

The term "feature" is a vector whose values represent an object or user through a low dimension. The nature of the feature vectors allows objects corresponding to vectors that are close in distance to have similar meanings. For example, if the two objects, i.e., the car and the digital product, belong to the scientific class, the feature vectors of the car and the feature vectors of the digital product are relatively close to each other in space. For another example, if the user a and the user B simultaneously select entertainment information as a tag of interest or the user a and the user B have consumed at a coffee shop at the same time, the features of the user a and the user B are spatially close to each other. The characteristic that the object can be coded by using the concept of 'characteristic' through a low-dimensional vector and the meaning of the object can be kept is very suitable for deep learning.

As mentioned above, there is a need for a method of correctly representing objects and recommending objects. In the current object recommendation scheme, there are mainly the following modes:

(1) and matching degree between the user and the object piece is identified through the inner product of the vectors of the user and the object, so that the corresponding object is recommended to the user. The disadvantage of this approach is that when the on-line historical click-through information of the user is sparse, the model does not represent the vector representation of the user and the object well and accurately. Resulting in a lower push conversion rate for that user.

(2) Recommending to the user objects similar to the objects of interest to the user in the past. Firstly, an object list clicked by a user is established, and the co-occurrence relation of objects is trained by clicking the objects, so that the object vector representation is obtained. Several objects are then selected from the objects of interest that the user has already seen. And then calculating the matching degree of the clicked object and the plurality of objects by the user, and recommending the object with higher matching degree to the user according to the sorting of the matching degree. The method has the disadvantage that when the online exhibition information of the user is less, enough objects which are interested by the user cannot be found, and further, objects which are not clicked by the user cannot be accurately recommended to the user.

According to an embodiment of the disclosure, an object recommendation scheme is provided. In the scheme, the selection of the on-line object actually selected by the user and the selection of the on-line object mapped according to the corresponding relation are combined to be used as the final selection of the on-line object by the user, and the object is recommended by using the combined data. In some embodiments, at least one reference user similar to the target user is first determined. The characteristics of the target user, the reference user and their selected online and offline objects are then determined. And then establishing a corresponding relation between the online object and the offline object so as to convert the selection record of the reference user for the offline object into the selection record of the online object. The actual on-line recording is then merged with the converted on-line recording. And finally, recommending the object according to the matching degree between the target user and the reference user and the combined record.

Under the condition that the selection records of the online object by the user are sparse, the selection records of the online object by the user are converted into the selection records of the online object, so that the online object can be recommended to the user more accurately and efficiently, the pushing conversion rate can be improved, and the user experience is improved.

The basic principles and several example implementations of the present disclosure are explained below with reference to the drawings.

Fig. 1 illustrates a schematic diagram of an example environment 100 in which various embodiments of the present disclosure can be implemented. It should be understood that the environment 100 shown in FIG. 1 is merely exemplary and should not be construed as limiting in any way the functionality or scope of the implementations described in this disclosure. As shown in FIG. 1, environment 100 includes a target user 110, a computing device 140, a plurality of reference users 150, an offline object 120, an online object 160, and a historical selection 130 of the offline object 120 and the online object 160 by the plurality of reference users 150.

The target user 110 and the plurality of reference users 150 may be users of various types of applications, which may be applications including recommendation systems, including but not limited to shopping applications, short video applications, music applications, dating applications, news applications, cafe applications, cloud storage applications, search applications, and the like. The present disclosure is not limited thereto.

The online object 160 may be a commodity, a live room, a short video, a picture, music, character information, etc. in the above-described application including the recommendation system, and the offline object 120 may be an entity industry such as an offline fast food restaurant, a coffee shop, a bookstore, a video store, a clothing store, etc. The target user 110 and the plurality of reference users 150 receive recommended videos, pictures, text, voices or combinations thereof related to the online object 160, the offline object 120 in the above-described application. For example, after the user enters the shopping application, text information or video information of a recommended product is received in the display interface.

In one embodiment, the historical selection times of the target user 110 and the plurality of reference users 150 for the plurality of online objects 160 is less than the historical selection times of the plurality of offline objects 120, i.e., the target user has more records of selections of offline items than online items.

Note that the reference user and target user may be interchanged for different online objects 160 and offline objects 120, which are exemplary only in fig. 1, and there may be one or more reference users as well as target users, and the present disclosure is not intended to be limiting.

The computing device 140 obtains the characteristics of the target user 110 and the plurality of reference users 150 from historical selection records 130 of the plurality of online objects 160 and the plurality of offline objects 120 by the target user 110 and the plurality of reference users 150, such as the behavior of the target user and the reference users clicking on, forwarding, publishing different sample objects 160, or the behavior of the target user and the reference users selecting different items in an offline brick and mortar store. This will be described in detail below.

The computing device 140 may select at least one reference user 150 that is most similar to the target user 110 based on characteristics of the target user 110 and the plurality of reference users 150. The computing device 140 may also determine their respective characteristics by the most similar at least one reference user 150 selection relationship of the plurality of online objects 160 and the plurality of offline objects 120 by the target user 110. The final feature is then used to recommend the target object 120.

Although the computing device 140 is shown as including the historical selection record 130, the computing device 140 may also obtain the historical selection record 130 from other entities. Computing device 140 may be any device with computing capabilities. By way of non-limiting example, the computing device 140 may be any type of stationary, mobile, or portable computing device, including but not limited to a desktop computer, laptop computer, notebook computer, netbook computer, tablet computer, multimedia computer, mobile phone, or the like; all or a portion of the components of the computing device 140 may be distributed in the cloud. The computing device 140 contains at least a processor, memory, and other components typically found in a general purpose computer to implement the functions of computing, storage, communication, control, and the like.

The detailed object recommendation process is further described below in conjunction with fig. 2 through 5. FIG. 2 illustrates a flow diagram of a method 200 of object recommendation in accordance with an embodiment of the present disclosure. The method 200 may be implemented by the computing device 140 of fig. 1. For ease of description, the method 200 will be described with reference to fig. 1.

At block 210, the computing device 140 determines at least one reference user of the plurality of reference users 150 associated with the target user 110 based on historical selection records of the target user 110 and the plurality of reference users 150 for the plurality of online objects 160 and the plurality of offline objects 120, the historical selection record 130 indicating at least a historical number of selections for the object. For example, the computing device 140 determines the target user 110 and the at least one reference user similar thereto by the user's online or offline record of purchases of items, record of views of videos.

The computing device 140 may determine a second feature of the target user and a second feature of the plurality of reference users, determine a fourth degree of match between the at least one reference user and the target user based on the second feature of the target user and the second feature of the at least one reference user; and associating the determined at least one reference user with the target user if the fourth degree of match is greater than the second degree of match threshold.

The computing device 140 may first convert the plurality of inline objects 160 and the plurality of offline objects 120 into a word vector vec characterizing them, using principles such as word2 vec. For example, the computing device 140 may convert "fast food" to vec (fast food), "coffee shop" to vec (coffee shop), "car" to vec (car), and so on. The word2vec principle is that words are mapped into a new space and expressed by multi-dimensional continuous real number vectors, and association attributes among the words can be mined by using the word2vec principle, so that the accuracy on vector semantics is improved. This is merely an example and users and objects may be represented in other ways, and the disclosure is not limited thereto.

In one embodiment, the vector dimension (vec _ dim) of the word vector may be adjusted according to the precision, coverage, and computational resource. The larger vec _ dim is, the more user information is kept, the more accurate the recommendation result is, but the higher the required computing resources are, different vector representation dimensions can be computed according to different recommendation scenes, and recommendation can be flexibly and accurately performed.

The computing device 140 may then represent the characteristics of the user based on the historical selection records 130 of the plurality of online objects 160 and the plurality of offline objects 120 by the target user 110 and the plurality of reference users 150 stored therein or in the cloud server.

In one example, the computing device 140 may calculate weights for features corresponding to the plurality of inline objects 160 and the plurality of offline objects 120. In one embodiment, the computing device 140 may determine an inverse document rate (IDF) of the word vector, where IDF is used to characterize the general importance of a word, and the word vector vec is in a one-to-one correspondence, and is added as a weight to the word vector vec when computing the features of the user. The IDF can be calculated according to the following equation:

inverse document rate (IDF) = log (number of users to be trained/(number of users including the word +1)) (1).

Next, the computing device 140 may determine a characteristic (which may also be referred to as a second characteristic hereinafter) of a user based on historical selections of offline objects and/or online objects by the user in the historical selection record 130. Taking target user 110 as an example, the IDF of all word vectors vec associated with target user 110 may be weighted averaged, e.g., if the target user has selected items at a fast food restaurant and a coffee shop, the characteristics of the target user are represented as: (vec (fast food) + vec (cafe) + idf (cafe))/2. The second characteristics of the multiple reference users can be determined in a similar manner, which is not described herein. This is merely an example, and other algorithms may be applied to calculate the weights or the weights may not be calculated in order to save computational resources, which is not limited by the present disclosure.

Finally, the computing device 140 determines a plurality of reference users similar to the target user 110 based on the second characteristics of the users determined above. In one embodiment, similar users may be determined using a Faiss repository in a search style that includes clusters, and the computing device 140 may determine similar users by determining the following parameters: clustering center number nlist: determined from the training data, e.g., about 20% of the training data; and adjusting the number k of the similar users according to the coverage rate, the precision rate and the calculation resources of the result. The computing device 140 finally obtains a similarity score table (table 1, shown below) including three columns of "target user", "similar user", and "matching degree (which may also be referred to as a fourth matching degree hereinafter) (user _ similarity _ table) between the users.

In one embodiment, the computing device 140 may compare the obtained fourth matching degree with a matching degree threshold (which may also be referred to as a second matching degree threshold hereinafter), for example, the second matching degree threshold is 0.6, and the computing device regards users B, C and D with the fourth matching degree greater than 0.6 as reference users of a.

In an alternative embodiment, the users with the fourth matching degree ranking 50 percent can be used as the reference users.

Note that the number of users, the matching degree, and the threshold described above are merely exemplary and are not limiting. Any suitable number and number may be used to determine similar users. Further, the names of the second feature of the users, the fourth matching degree between the users, and the second matching degree threshold appearing above are merely exemplary, and may be the same as or different from other features (e.g., the first feature), other matching degrees (e.g., the first, second, third, and fifth matching degrees), or other matching degree thresholds (e.g., the first and third matching degree thresholds) appearing in the context, and the present disclosure is not intended to be limited.

At block 220, the computing device 140 determines, based on the historical selection record 130, a first feature of the target user 110, a first feature of the at least one reference user, a first feature of the plurality of online objects 160, and a first feature of the plurality of offline objects 120. For example, the computing device 140 may look up the historical selections of the online object and the offline object by the target user and the plurality of reference users determined at 210 in the historical selection record 130.

In one embodiment, the computing device 140 builds a relationship graph of users and objects based on historical selection records of the target user 110 and a plurality of reference users 150, for example, there are two nodes in the graph, namely user node and object node, and when a user selects an object, an edge is created between the two nodes (e.g., user A selects coffee at a coffee shop, sees a movie at a movie theater, selects an educational category of books at a book shop, and then an edge is created between user A and the "coffee shop", "movie theater", "book shop", "educational" node). In the following, for the sake of simplicity, the user is represented by letters, the online object and the offline object are represented by numbers, there are user node A, B, C, D and object nodes 1, 2, 3, 4, a has selected 1, 2, and a is connected between node a and nodes 1, 2, respectively, resulting in edges from node a to nodes 1, 2. The computing device 140 then traverses the nodes of the graph randomly walks in the graph, starting from the user node and the object node, respectively, to generate two sequences of walks (e.g., A1B2D4C3A, 1C3D4B2A, etc.) starting from the user and starting from the resource. The computing device then enters the sequence into a skip-gram model, which yields the characteristics of the user and the object.

In one example, the characteristics of the reference user 150 may be more accurately obtained by adjusting the random walk weights described above. For example, in the walking process, the calculation sets different weights for edges between nodes by setting homogeneity (homophily) and structure (structural equality) between the user node and the object node. In particular, "homogeneity" of nodes means that features from nearby nodes should be as close as possible, and "structural" means that features of structurally similar nodes should be as close as possible. The computing device 140 may derive a more accurate sequence of walks based on edges having different weights set to optimize the resulting characteristics of the user and object to more accurately represent the user and object.

In one embodiment, the computing device 140 may increase the weight associated with the frequency of selection between nodes to reduce the impact of high frequency selection on the results. For example, the weight may be calculated by the following equation (2):

the weight of a certain edge = number of consumption of the user in a certain industry/total number of consumption of the user (2).

At block 230, the computing device 140 associates the at least one offline object with the at least one online object based on the first characteristics of the plurality of online objects 160 and the first characteristics of the plurality of offline objects 120. For example, the computing device may determine a degree of match (which may also be referred to as a fifth degree of match below) between the features of the offline object and the features of the online object determined at 220.

The computing device 140 may determine a fifth degree of match between the at least one offline object and the at least one online object based on the first feature of the at least one offline object and the first feature of the at least one online object; and associating the at least one offline object with the at least one online object if the fifth degree of match is greater than the third degree of match threshold.

In one embodiment, the computing device 140 may obtain a fifth matching degree between the at least one offline object and the at least one online object by calculating a distance between the above-obtained feature vectors. As an example, the feature of the object on the line is vector a, and the feature of the object under the line is vector B, the cosine distance (i.e. the fifth matching degree) therebetween can be calculated according to the following equation (3):

（3）。

in an alternative embodiment, the computing device 140 may also calculate the distance between the features of the two objects by euclidean distance. The present disclosure is not limited herein and existing or future developed techniques may be utilized to calculate the distance between feature vectors. In one example, a close distance may indicate a high degree of match and a far distance may indicate a low degree of match.

The computing device 140 obtains an association table (table 2) (mapping _ table) between the online object and the offline object based on the distance calculation, where the table includes three columns, "online object", "offline object", and "fifth matching degree (i.e., cosine distance)".

The computing device 140 may compare the obtained fifth matching degree with a third matching degree threshold, for example, the third matching degree threshold is 0.8, and the computing device determines the offline object b with the fifth matching degree greater than 0.8 as the object associated with the online object a. Note that the number of objects, the matching degree, and the threshold value described above are merely exemplary and are not limiting. Any suitable number and number may be employed to determine the associated object.

In one embodiment, the computing device 140 may control the output parameters and number of parameters of table 2 above by adjusting the threshold of the degree of match, the number of users, and the like.

At block 240, the computing device 140 updates the historical selection record of the at least one reference user for the associated at least one online object based on the historical selection record of the at least one reference user for the at least one offline object. For example, the computing device 140 translates the more dense offline data of the user into online data for more accurate recommendations. This will be described in detail below with reference to fig. 3.

At block 310, the computing device 140 determines a first degree of match between the at least one offline object and the at least one online object based on the first features of the at least one offline object and the first features of the at least one online object. For example, the computing device 140 may determine that the degree of match between the associated online object a and offline object b is 0.8. The calculation method of the matching degree can refer to the above contents and equations, and is not described herein again.

At block 320, the computing device 140 determines a second historical number of selections based on the first degree of match and the first historical number of selections of the at least one reference user for the at least one offline object. For example, the computing device 140 may obtain the first historical selection number of the reference user 150 for the line object b as 10, and then multiply the 10 by the matching degree 0.8 to obtain the second historical selection number of 8, where 8 indicates the historical selection number of the reference user 150 for the line object a derived from the associated mapping relationship. Note that the above calculation method and numbers are merely exemplary, and the number of times or frequency of the selection of the on-line object a by the reference user may be derived in other ways.

At block 330, the computing device 140 updates the historical number of selections of the associated at least one inline object by the at least one reference user based on the second historical number of selections.

In one embodiment, the at least one reference user may not actually have selected the online object a online, and the computing device 140 directly takes 8 times as the number of selections. Or the at least one reference user selects the on-line object a1 times, and then 9 times are taken as the selection times.

Through the above calculations, the computing device 140 may derive the consolidated data table 3 that references the actual selection of the online object by the user and the mapped selection of the online object.

Continuing back to the description of FIG. 2, at block 250, the computing device 140 determines the at least one online object for recommendation to the target user 110 based on the updated historical selection record, the first characteristic of the target user 110, and the first characteristic of the at least one reference user. This process will be further described in conjunction with fig. 4.

At block 410, the computing device 140 determines a second degree of match between the target user 110 and the at least one reference user based on the first characteristics of the target user 110 and the first characteristics of the at least one reference user. For example, the computing device 140 may determine a matching degree (similarity) between the target user and the reference user, and the specific process may refer to the above equation and content, which is not described herein again.

In one embodiment, the computing device 140 may look up the degree of match between users in table 1.

At block 420, the computing device 140 determines a third degree of match between the target user 110 and the at least one offline object based on the second degree of match, the sum of the historical selections of the plurality of offline objects by the at least one reference user, and the updated historical selection record.

In one example, the computing device 140 may determine the number of selections (num) of the reference user for at least one online object and the sum of the number of selections (sum) of all online objects from table 3 above, and then calculate the degree of match between the target user and the offline object based on the following equation:

degree of matching = num similarity/sum equation (4).

From the above equation, the computing device 140 may determine a table of matches between the computing target user and the online object (Table 4).

At block 430, the computing device 140 determines whether the third degree of match is greater than the first degree of match threshold. For example, the computing device 140 may predetermine the first threshold of degree of match to be 0.5. The computing device 140 may determine that the third degree of match between the target user 110 and the online objects a, b, and c is greater than 0.5.

At block 440, the computing device 140 determines the at least one online object for recommendation to the target user if it determines that the third degree of match is greater than the first degree of match threshold. For example, the computing device 140 may determine to recommend the online objects a, b, and c to the target user.

In one embodiment, if the target user selects the recommended at least one online object, the computing device 140 may update the characteristics of the target user based on the first characteristics of the at least one online object. For example, if the target user selects a recommended online object a, b, or c, the target user's feature representation may be updated by the features of the online object a, b, or c. This may further increase the consistency of the online data, which in turn may more accurately represent the target user, increasing the accuracy of the recommendation.

In one embodiment, the number of historical selections of the plurality of inline objects by the target user is less than the number of historical selections of the plurality of inline objects by the target user.

In one embodiment, the computing device 140 may also send the determined at least one online object to the target user, the online object including at least one of video, image, text, and speech.

And under the condition that the online historical selection records are extremely sparse, the online selection records are supplemented by mapping of the offline historical selection records and the online historical selection, so that the coverage of the final result is improved on the premise of ensuring the quality. The method determines the corresponding relation between the offline object and the online object, and maps part of offline object records into online object records according to the corresponding relation, so as to achieve the aim of reasonably increasing the data of the online object, thereby enabling the algorithm to cover more users under the condition of ensuring the recommendation quality.

FIG. 5 illustrates a schematic block diagram of an example electronic device 500 that can be used to implement embodiments of the present disclosure. For example, the electronic device 500 may be used to implement the computing device 140 shown in fig. 1. As shown, electronic device 500 includes a Central Processing Unit (CPU) 501 that may perform various appropriate actions and processes according to computer program instructions stored in a Read Only Memory (ROM) 502 or loaded from a storage unit 508 into a Random Access Memory (RAM) 503. In the RAM, various programs and data required for the operation of the electronic device 500 may also be stored. The CPU, ROM, and RAM are connected to each other by a bus 504. An input/output (I/O) interface 505 is also connected to bus 504.

A number of components in the electronic device 500 are connected to the I/O interface 505, including: an input unit 506 such as a keyboard, a mouse, or the like; an output unit 507 such as various types of displays, speakers, and the like; a storage unit 508, such as a magnetic disk, optical disk, or the like; and a communication unit 509 such as a network card, modem, wireless communication transceiver, etc. The communication unit 509 allows the electronic device 500 to exchange information/data with other devices through a computer network such as the internet and/or various telecommunication networks.

The central processing unit 501 performs the various methods and processes described above, such as any of the methods 200 through 400. For example, in some embodiments, any of the methods 200-400 may be implemented as a computer software program or computer program product that is tangibly embodied in a machine-readable medium, such as the storage unit 508. In some embodiments, part or all of the computer program may be loaded and/or installed onto the electronic device 500 via the ROM and/or the communication unit 509. When loaded into RAM and executed by a CPU, a computer program may perform one or more steps of any of the methods 200 to 400 described above. Alternatively, in other embodiments, the CPU may be configured to perform any of the above methods by any other suitable means (e.g., by means of firmware).

The present disclosure may be methods, apparatus, systems, and/or computer program products. The computer program product may include a computer-readable storage medium having computer-readable program instructions embodied thereon for carrying out various aspects of the present disclosure.

The computer readable storage medium may be a tangible device that can hold and store the instructions for use by the instruction execution device. The computer readable storage medium may be, for example, but not limited to, an electronic memory device, a magnetic memory device, an optical memory device, an electromagnetic memory device, a semiconductor memory device, any non-transitory memory device, or any suitable combination of the foregoing. More specific examples (a non-exhaustive list) of the computer readable storage medium would include the following: 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), a Static Random Access Memory (SRAM), a portable compact disc read-only memory (CD-ROM), a Digital Versatile Disc (DVD), a memory stick, a floppy disk, a mechanical coding device, such as punch cards or in-groove projection structures having instructions stored thereon, and any suitable combination of the foregoing. Computer-readable storage media as used herein is not to be construed as transitory signals per se, such as radio waves or other freely propagating electromagnetic waves, electromagnetic waves propagating through a waveguide or other transmission medium (e.g., optical pulses through a fiber optic cable), or electrical signals transmitted through electrical wires.

The computer-readable program instructions described herein may be downloaded from a computer-readable storage medium to a respective computing/processing device, or to an external computer or external storage device via a network, such as the internet, a local area network, a wide area network, and/or a wireless network. The network may include copper transmission cables, fiber optic transmission, wireless transmission, routers, firewalls, switches, gateway computers and/or edge servers. The network adapter card or network interface in each computing/processing device receives computer-readable program instructions from the network and forwards the computer-readable program instructions for storage in a computer-readable storage medium in the respective computing/processing device.

The computer program instructions for carrying out operations of the present disclosure may be assembler instructions, Instruction Set Architecture (ISA) instructions, machine-related instructions, microcode, firmware instructions, state setting data, or source or object code written in any combination of one or more programming languages, including an object oriented programming language such as Smalltalk, C + + or the like and conventional procedural programming languages, such as the "C" programming language or similar programming languages. The computer-readable program instructions 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). In some embodiments, the electronic circuitry that can execute the computer-readable program instructions implements aspects of the present disclosure by utilizing the state information of the computer-readable program instructions to personalize the electronic circuitry, such as a programmable logic circuit, a Field Programmable Gate Array (FPGA), or a Programmable Logic Array (PLA).

Various aspects of the present disclosure are described herein with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems) and computer program products according to embodiments of the disclosure. It will be understood that each block of the flowchart illustrations and/or block diagrams, and combinations of blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer-readable program instructions.

These computer-readable program instructions may be provided to a processing unit of a general purpose computer, special purpose computer, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processing unit of the computer or other programmable data processing apparatus, create means for implementing the functions/acts specified in the flowchart and/or block diagram block or blocks. These computer-readable program instructions may also be stored in a computer-readable storage medium that can direct a computer, programmable data processing apparatus, and/or other devices to function in a particular manner, such that the computer-readable medium storing the instructions comprises an article of manufacture including instructions which implement the function/act specified in the flowchart and/or block diagram block or blocks.

The computer readable program instructions may also be loaded onto a computer, other programmable data processing apparatus, or other devices to cause a series of operational steps to be performed on the computer, other programmable apparatus or other devices to produce a computer implemented process such that the instructions which execute on the computer, other programmable apparatus or other devices implement the functions/acts specified in the flowchart and/or block diagram block or blocks.

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 instructions, which comprises one or more executable instructions for implementing the specified logical function(s). 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.

Having described embodiments of the present disclosure, the foregoing description is intended to be exemplary, not exhaustive, and not limited to the disclosed embodiments. Many modifications and variations will be apparent to those of ordinary skill in the art without departing from the scope and spirit of the described embodiments. The terminology used herein was chosen in order to best explain the principles of the embodiments, the practical application, or improvements made to the technology in the marketplace, or to enable others of ordinary skill in the art to understand the embodiments disclosed herein.

Claims

1. A method of object recommendation, comprising:

determining at least one reference user of the plurality of reference users associated with a target user based on historical selection records of the target user and the plurality of reference users for a plurality of online objects and a plurality of offline objects, the historical selection records indicating at least a historical number of selections for the object;

determining, based on the historical selection record, a first feature of the target user, a first feature of the at least one reference user, a first feature of the plurality of online objects, and a first feature of the plurality of offline objects;

associating at least one offline object with at least one online object based on the first features of the plurality of online objects and the first features of the plurality of offline objects;

updating the historical selection record of the at least one reference user for the associated at least one online object based on the historical selection record of the at least one reference user for the at least one offline object; and

determining the at least one online object for recommendation to the target user based on the updated historical selection record, the first characteristic of the target user, and the first characteristic of the at least one reference user,

wherein the number of historical selections of the plurality of online objects by the target user is less than the number of historical selections of the plurality of offline objects by the target user.

2. The method of claim 1, wherein updating the historical selection record of the at least one reference user for the associated at least one online object comprises:

determining a first degree of match between the at least one offline object and the at least one online object based on the first features of the at least one offline object and the first features of the at least one online object; and

determining a second historical selection number based on the first matching degree and the first historical selection number of the at least one reference user for the at least one offline object; and

updating the historical selection times of the at least one reference user for the associated at least one online object based on the second historical selection times.

3. The method of claim 1, wherein determining the at least one online object for recommendation to the target user comprises:

determining a second degree of match between the target user and the at least one reference user based on the first characteristics of the target user and the first characteristics of the at least one reference user;

determining a third degree of match between the target user and the at least one online object based on the second degree of match, a sum of the historical selection times of the plurality of offline objects by the at least one reference user, and the updated historical selection record; and

and if the third matching degree is larger than a first matching degree threshold value, determining the at least one online object for recommending to the target user.

4. The method of claim 1, wherein determining at least one reference user associated with a target user comprises:

determining a second characteristic of the target user and a second characteristic of the plurality of reference users;

determining a fourth degree of matching between the at least one reference user and the target user based on the second features of the target user and the second features of the at least one reference user; and

associating the determined at least one reference user with the target user if the fourth degree of match is greater than a second degree of match threshold.

5. The method of claim 1, wherein associating at least one offline object with at least one online object comprises:

determining a fifth degree of match between the at least one offline object and the at least one online object based on the first features of the at least one offline object and the first features of the at least one online object; and

associating the at least one offline object with the at least one online object if the fifth degree of match is greater than a third degree of match threshold.

6. The method of claim 1, further comprising:

updating the characteristics of the target user based on the first characteristics of the at least one online object if the target user selects the recommended at least one online object.

7. The method of claim 1, further comprising:

and sending the determined at least one online object to the target user, wherein the online object comprises at least one of video, image, text and voice.

8. A computing device, comprising:

at least one processing unit; and

at least one memory coupled to the at least one processing unit and storing instructions for execution by the at least one processing unit, the instructions when executed by the at least one processing unit causing the computing device to perform the steps of the method of any of claims 1-7.

9. A computer readable storage medium having stored thereon computer program code which, when executed, performs the method according to any of claims 1 to 7.