CN112100489A - Object recommendation method, device and computer storage medium - Google Patents
Object recommendation method, device and computer storage medium Download PDFInfo
- Publication number
- CN112100489A CN112100489A CN202010879004.7A CN202010879004A CN112100489A CN 112100489 A CN112100489 A CN 112100489A CN 202010879004 A CN202010879004 A CN 202010879004A CN 112100489 A CN112100489 A CN 112100489A
- Authority
- CN
- China
- Prior art keywords
- target object
- user
- target
- degree
- threshold
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/95—Retrieval from the web
- G06F16/953—Querying, e.g. by the use of web search engines
- G06F16/9535—Search customisation based on user profiles and personalisation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/22—Matching criteria, e.g. proximity measures
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Databases & Information Systems (AREA)
- General Physics & Mathematics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Computation (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Computational Linguistics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Health & Medical Sciences (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Evolutionary Biology (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
According to an embodiment of the present disclosure, a method, an apparatus, and a computer storage medium for object recommendation are provided, which may be used for information recommendation. A method of object recommendation, comprising: determining at least one reference user selecting a target object, the attention of which is less than a threshold attention; determining a first degree of match between the at least one reference user and the target user based on the characteristics of the at least one reference user and the characteristics of the target user, the characteristics being associated with historical selection of the sample object by the reference user; and recommending the target object to the target user if the first matching degree is greater than the first threshold matching degree. According to the scheme disclosed by the invention, the new resources can be accurately represented, so that the new resources can be more accurately recommended to the user, and the resource distribution rate and the user experience are improved.
Description
Technical Field
Embodiments of the present disclosure relate to the field of information recommendation, and more particularly, to a method, apparatus, and computer storage medium 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 interested by users under the condition of information overload, so that the user experience is improved, and the resource distribution efficiency is improved. New resources and cold resources, especially some short video resources, are often not properly recommended to the user because they are difficult to learn and represent adequately. In view of the importance of information recommendation in various applications, there is a need for a method of correctly representing resources and recommending resources.
Disclosure of Invention
According to an embodiment of the present disclosure, a solution for object recommendation is provided.
In a first aspect of the present disclosure, there is provided a method of object recommendation, comprising: determining at least one reference user selecting a target object, the attention of which is less than a threshold attention; determining a first degree of match between the at least one reference user and the target user based on the characteristics of the at least one reference user and the characteristics of the target user, the characteristics being associated with historical selection of the sample object by the reference user; and recommending the target object to the target user if the first matching degree is greater than the first threshold matching degree.
In a second aspect of the present disclosure, there is provided an apparatus for object recommendation, comprising: a first reference user determination module configured to determine at least one reference user selecting a target object, a degree of attention of the target object being less than a threshold degree of attention; a first matching degree determination module configured to determine a first matching degree between the at least one reference user and the target user based on the features of the at least one reference user and the features of the target user, the features being associated with historical selection of the sample object by the reference user; and a target user recommending module configured to recommend the target object to the target user if the first matching degree is greater than the first threshold matching degree.
In a third aspect of the disclosure, an electronic device is provided that includes 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, and wherein:
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 determining and updating an initial tag according to some embodiments of the present disclosure;
FIG. 4 illustrates a flow diagram of a process of determining a final label according to some embodiments of the present disclosure;
FIG. 5 illustrates a flow diagram of a process for object recommendation, according to some embodiments of the present disclosure;
FIG. 6 shows a schematic block diagram of an apparatus for object recommendation, in accordance with some embodiments of the present disclosure; and
FIG. 7 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 the interesting tags, the features of the user a and the user B are relatively close to each other in space. 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 and recommending resources. In the current resource recommendation scheme, there are mainly three modes:
(1) and fitting click information through historical point showing training of the user to obtain vector representation of the user and the resource, and then identifying the matching degree of the user and the resource piece through the inner product of the vectors of the user and the resource to recommend the corresponding resource to the user. The method has the disadvantages that a large amount of new resources are poured in every day, and the model cannot well and accurately represent the vectors of the new resources because the historical point spread information of the new resources is less. Resulting in inefficient distribution of the new resources.
(2) Recommending resources liked by other users with similar interests to the user, firstly establishing a resource clicking user list, and training the co-occurrence relationship of the user by clicking the user so as to obtain user vector representation. Then, the similarity between each user and the target user is calculated according to the user similarity, and the similarities are ranked to obtain a set of a plurality of users similar to the user interests. And finally recommending a plurality of user favorite resources similar to the interests of the users to the users, finding the resources which are favorite by the users but not watched by the current users, then calculating the matching degree between the users and the resources, and sequencing the resources according to the matching degree so as to recommend the corresponding resources to the users. The disadvantage of this method is that new resources are not efficiently recommended to users who have the same preference as the new resources, since the new resources are not efficiently distributed, resulting in the resources that users prefer not being well distributed.
(3) The user is recommended resources similar to the resources that he has been interested in the past. Firstly, a resource list clicked by a user is established, and the co-occurrence relation of resources is trained by clicking the resources, so that the resource vector expression is obtained. Several resources are then selected from the resources of interest that the user has already seen. And then calculating the matching degree of the resources which are never seen by the user and the plurality of resources, and recommending the resources with higher matching degree to the user according to the sorting of the matching degree. The method has the disadvantage that the resource vector of the new resource is not sufficiently learned, so that the calculation of the new resource vector and the resource vector of the previous favorite resource is not accurate enough, and the distribution efficiency of the new resource is reduced.
Therefore, a method for efficiently and accurately recommending new resources and accurately describing new resources is needed.
According to an embodiment of the disclosure, an object recommendation scheme is provided. In this scenario, a user who clicked on, shared with, or published a new resource matches the current user to determine whether to recommend the new resource. In some embodiments, at least one reference user that has historically selected a target object, such as a new or cold resource, may first be determined. And then determining the matching degree between the target user and the reference user through the characteristics of the at least one reference user and the characteristics of the target user. The characteristics of the reference user may be pre-trained by the reference user's historical selection of sample objects. If the matching degree is greater than the threshold matching degree, the target object can be recommended to the target user.
In the early stage of resource release (or the resource is a cold resource), the characteristics of the new resource are expressed by the characteristics of the fully-learned reference users (core users and seed users), so that the new resource can be recommended to other users more accurately and efficiently, the Click-Through-Rate (Click-Through-Rate) of the resource 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, the environment 100 includes a target user 110, a target object 120, a computing device 140, at least one reference user 150, a sample object 160, and historical selections 130 of the sample object 160 by the at least one reference user 150.
The target user 110 and the at least one reference user 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 target objects 120, sample objects 160 may be merchandise, live rooms, short videos, pictures, music, persona information, etc. in the above-described applications including recommendation systems, the target objects may be newly released or cold resources, the sample objects may be hot resources, the sample objects may be data for referencing user clicks to get click history to get a characterization for input, the sample objects may be historical hot objects in the application or a subset thereof. The target user 110 and the at least one reference user 150 receive recommended video, pictures, text, speech, or a combination thereof related to the target object 120, the sample object 160 in the above-described application. For example, after entering a news application, a user receives a cover page picture, news headline information, or video information of recommended news in a display interface.
Note that the reference user and target user may be interchanged for different sample objects 160 and target objects 120, which is 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 may derive characteristics of the sample object 160 and at least one reference user 150 based on historical selections 130 of the sample object 160 by the reference users 150, such as behavior of reference users clicking on, forwarding, publishing different sample objects 160. This will be described in detail below.
The computing device 140 may match the target object 120 and the incoming application target users 110 based on the above-described features, thereby recommending the target object 120 to the target users 110 that need it. The computing device 140 may also utilize the characteristics of the target user 110 to update the target object 120, and/or ultimately determine the characteristics of the target object 120, at different stages. And recommends the target object 120 using the final characteristics.
Although the computing device 140 is shown as including the historical selections 130, the computing device 140 may also be an entity other than the historical selections 130. 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 150 that selects a target object 120, the target object 120 having a degree of attention that is less than a threshold degree of attention. For example, the computing device 140 may view historical operational information of the reference user 150, and may determine the reference user 120 or the set of reference users 120 that selected the target object 120 if the reference user 150 historically clicks on a fan that viewed the target object 120, shared the target object 120, or is a publisher of the target object 120.
In one embodiment, the computing device 140 may predetermine a degree of interest for an object indicating a degree to which the object is "popped" in a corresponding application, such as a frequency with which the object is selected, a number of times the object is recommended, a number of times the object is selected, and a time at which the object is created. The computing device may obtain the information to determine whether the object is a newly released object or a cold object, and the target object 120 may be a newly released object or a long-tailed object without too much interaction information.
In one example, the computing device 140 may re-determine the attention of the object at intervals and determine a newly published object or a cold object as the target object 120, and the computing device may store the set of target objects 120 in a database for subsequent use.
In one embodiment, after the computing device 140 determines the set of reference users 120 that have selected the target object 120, i.e., the cold object, the computing device may further determine the core reference users in the set of reference users 120. In one example, the at least one reference user 120 most associated with the target object may be further determined by removing edge users that select too many types of objects from the set of reference users 120.
At block 220, the computing device 140 determines a first degree of match between the at least one reference user 150 and the target user 110 based on the characteristics of the at least one reference user 150 and the characteristics of the target user 110, the characteristics being associated with historical selections of the sample object by the reference user. For example, when the computing device 140 detects that the target user 110 enters the application, the target user 110 may be determined to be the set of clicked target objects 120, and then at least one reference user 150 that selected at least one target object in the set of target objects 120 is obtained. The degree of match between the at least one reference user 150 and the target user 110 is then determined by the characteristics of the at least one reference user.
In one embodiment, the computing device 140 may obtain the feature vectors of the reference users 150 through historical selection of the sample object 160 by at least one reference user 150. In one example, the computing device first obtains operational behavior of the plurality of sample objects 160 by the plurality of reference users 150 that records user clicks, shares, publishes, etc. of the objects. The computing device 140 then constructs a node graph based on the user's operation behavior, for example, there are two nodes in the graph, i.e., a user node and a resource node, and when the user operates an object, an edge is created between the two nodes. For example, user node A, B, C, D and object nodes 1, 2, 3, 4, a click through 1, 2, respectively, connecting between node a and nodes 1, 2, respectively, resulting in an edge from node a to nodes 1, 2. The computing device then traverses the nodes in the graph randomly walking 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 the skip-gram model, resulting in the characteristics of the reference user 150.
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 derived characteristics of the reference user 150 to more accurately represent the user from that person.
In one embodiment, the matching degree between the at least one reference user 150 and the target user 110 may be obtained by calculating the distance between the above-obtained feature vectors. For example, the distance between features of two users or the distance of features between one user and a plurality of users may be calculated by euclidean distance. Of course, cosine similarity may also be used for calculation, and the present disclosure is not limited thereto, and existing or future developed techniques may be used to calculate the distance seen by the feature vector. In one example, a close distance may indicate a high degree of match and a far distance may indicate a low degree of match.
At block 230, the computing device 140 determines whether the first degree of match is greater than a first threshold degree of match. For example, if the computing device 140 determines that the distance between the feature vectors of the two users is a, then the matching degree corresponding to the distance a in the lookup table between the recorded distance and the matching degree is 75%, and if the threshold matching degree is 70%, the computing device 140 determines that the first matching degree is greater than the first threshold matching degree. Note that the lookup table and threshold matching degree may be determined according to different user and application scenarios, and the above numbers are exemplary and not intended to limit the scope of the present disclosure.
At block 240, the computing device 140 determines yes, i.e., in response to the first degree of match being greater than a first threshold degree of match, to recommend the target object 120 to the target user 110. For example, the computing device 140 may send video, images, text, or speech representing the target object 120 to the user, such as presenting a picture of news and text information to the target user 120, or presenting an image cover formed by a series of live room frames to the user. In one embodiment, if the determination is negative, i.e., the first degree of match is not greater than the first threshold degree of match, the computing device 140 continues to look for other target objects 120 billion in number for recommendation to the user.
The characteristics of the users are used for representing the new resources to carry out matching among the users, so that the new resources and the cold resources required by the users can be recommended to the users more accurately, modeling of the resources is more accurate, and user experience is enhanced.
The above discusses a scheme of how to recommend a resource during a "cold start" of the resource, however, it is also important to update in real time the user information that is constantly clicked on during the "cold start" to select the new resource, which enables the system to collect in real time the input features required by the recommendation system model, enabling the recommendation system to always make predictions and recommendations using the latest features. How to determine characteristics of a new resource in real time for recommendation to a user will be described below in conjunction with fig. 3-5.
FIG. 3 illustrates a flow diagram of a process of determining and updating an initial tag according to some embodiments of the present disclosure.
At block 310, the computing device 140 determines all reference users that select the target object 120. For example, the computing device 140 may query to determine all reference users that clicked on, shared with the target object 120 during "cold start," and then store the information of all reference users, including the feature vectors, into a feature database (such as redis).
At block 320, the computing device 140 determines an initial label for the target object 120 based on the characteristics of all of the reference users. For example, the computing device 140 may determine the initial label of the target object 120 by averaging all of the reference user's different features associated with the target object 120 in the feature database described above through an averaging posing layer. The initial tag may also be a feature of the target object 120.
In one example, the computing device may further perform a weighted average operation according to different weights given to features of each user by different target users, obtain a weighted average feature vector, input the weighted average feature vector into the softmax layer, so as to obtain a weight of each feature through gradient back propagation, and then determine the initial label of the target object 120, that is, the initial feature of the target object, using the weights.
At block 330, the computing device 140 determines whether the target user 110 selected the recommended target object 120. For example, the computing device may determine whether the target user 110 selects the target object 120 recommended at block 230.
At block 340, the computing device 140 determines yes, i.e., in response to the target user 110 selecting the recommended target object 120, the initial label for the target object 120 is updated based on the characteristics of the target user 110. For example, if the target user selects the target object, the computing device may obtain the characteristics of the target user 110 and then update the initial label of the target object 120 with the characteristics and the obtained initial label to obtain an updated label (characteristic). The updating method may be similar to the averaging or weighting operation described above, and will not be described herein.
Fig. 4 illustrates a flow diagram of a process of determining a final label according to some embodiments of the present disclosure.
At block 410, the computing device 140 determines whether any of the following is satisfied: a threshold time is reached; a first threshold number of times that target object 120 is recommended is reached; a second threshold number of times target object 120 is selected is reached. The computing device 140 may determine whether the target object 120 is a still cold resource to determine whether to continue updating the characterization of the target object 120. For example, the computing device 140 determines that the target object 120 has been recommended 1 ten thousand times, more than 8000 first threshold times, or that the target object 120 has been selected 5 ten thousand times, more than 4 ten thousand second threshold times, or that the predetermined recommendation and selection times have not been reached, but 24 hours have elapsed. The above numbers are exemplary and not intended to be limiting. The number of recommendations, number of selections, and threshold time may be determined based on the target object, the age of the user, and the type of application.
At block 420, the computing device 140 determines yes, i.e., stops updating in response to the condition being satisfied. For example, when the computing device 140 determines that the target object 120 is not already a cold resource or that the target object is still a cold resource but has elapsed a sufficient amount of time, the computing device 140 no longer updates the target object 120.
In one embodiment, if the computing device 140 determines "no," the click information of the target object 120 continues to be monitored and obtained.
At block 430, the computing device 140 determines the updated initial label as the final label for the target object 120. After stopping the update, the computing device may determine the last updated label as the final label, i.e., the final feature representation of the target object 120.
FIG. 5 illustrates a flow diagram of a process for object recommendation, according to some embodiments of the present disclosure.
At block 510, the computing device 140 determines a second degree of match based on the characteristics of the other users and the final label of the target object 120. The computing device 140 may determine the distance between the feature of the other user and the final tag of the target object 120 according to the euclidean distance method or the cosine similarity method, and then determine the second matching degree according to the lookup table corresponding to the distance and the matching degree. For a specific process, reference may be made to the above example of determining the first matching degree, which is not described herein again. At this point, the characterization of target object 120 has been accurately determined, i.e., the label covering the target object has been sufficiently learned with the characteristics of the user who selected it.
At block 520, the computing device 140 determines whether the second degree of match is greater than a second threshold degree of match. The computing device 140 may determine whether the determined degree of match is greater than a predetermined degree of match.
At block 530, the computing device 140 determines yes, i.e., in response to the second degree of match being greater than the second threshold degree of match, the target object 120 is recommended to the other user. In one embodiment, the computing device 140 determines that the set of target objects 120 that satisfy the threshold is sent to a filtering module, which may be, for example, a pre-trained neural network, that further recommends to the user one or more target objects 120 from the set of target objects 120 that are determined to be most relevant based on the information of the other users.
In one example, if the computing device 140 determines no, then other target objects 120 continue to be sought for recommendation.
In the initial stage of resource publishing, resources are represented by fully-learned user characteristics, then resource representation is further perfected by continuously updating a user characteristic vector library, and finally, the characteristic vectors of the resources are represented by averaging and/or weighting the characteristic vectors of the users in a short time, so that the characteristics of new resources can be accurately represented.
Fig. 6 illustrates a schematic block diagram of an apparatus for object recommendation, in accordance with some embodiments of the present disclosure. The apparatus 600 may be included in the computing device 140 of fig. 1 or implemented as the computing device 140.
As shown in fig. 6, the apparatus 600 includes a first reference user determination module 610 configured to determine at least one reference user selecting a target object, the target object having a degree of attention less than a threshold degree of attention; a first degree of match determination module 620 configured to determine a first degree of match between the at least one reference user and the target user based on the features of the at least one reference user and the features of the target user, the features being associated with historical selections of the sample object by the reference user; and a target user recommending module 630 configured to recommend the target object to the target user if the first matching degree is greater than the first threshold matching degree.
In some embodiments, the apparatus 600 may further comprise: a second reference user determination module configured to determine all reference users selecting the target object; and an initial label determination module configured to determine an initial label of the target object based on the characteristics of all the reference users.
In some embodiments, the apparatus 600 may further comprise: an initial tag update module configured to update an initial tag of the target object based on a characteristic of the target user if the target user selects the recommended target object.
In some embodiments, the apparatus 600 may further comprise: : a stop update module configured to stop updating if any of the following is satisfied: a threshold time is reached; reaching a first threshold number of times that the target object is recommended; reaching a second threshold number of times that the target object is selected; and a final tag determination module configured to determine the updated initial tag as a final tag of the target object.
In some embodiments, the apparatus 600 may further comprise: a second matching degree determination module configured to determine a second matching degree based on the features of the other users and the final label of the target object; and the user recommending module is configured to recommend the target object to other users if the second matching degree is greater than the second threshold matching degree.
In some embodiments, the target user recommendation module 630 may include: an object transmission module configured to transmit video, images, text or voice representing the target object to the user.
In some embodiments, wherein the degree of attention is determined based on at least one of: a frequency with which the target object is selected, a number of times the target object is recommended, a number of times the target object is selected, and a time at which the target object is created.
Fig. 7 illustrates a schematic block diagram of an example device 700 that may be used to implement embodiments of the present disclosure. For example, the computing device 140 in the example environment 100 shown in FIG. 1 may be implemented by the device 700. As shown, device 700 includes a Central Processing Unit (CPU)701 that may perform various appropriate actions and processes in accordance with computer program instructions stored in a Read Only Memory (ROM)702 or computer program instructions loaded from a storage unit 708 into a Random Access Memory (RAM) 703. In the RAM 703, various programs and data required for the operation of the device 700 can also be stored. The CPU 701, the ROM 702, and the RAM 703 are connected to each other via a bus 704. An input/output (I/O) interface 705 is also connected to bus 704.
Various components in the device 700 are connected to the I/O interface 705, including: an input unit 706 such as a keyboard, a mouse, or the like; an output unit 707 such as various types of displays, speakers, and the like; a storage unit 708 such as a magnetic disk, optical disk, or the like; and a communication unit 709 such as a network card, modem, wireless communication transceiver, etc. The communication unit 709 allows the device 700 to exchange information/data with other devices via a computer network, such as the internet, and/or various telecommunication networks.
The various processes and processes described above, such as methods 200, 300, 400, and 500, may be performed by processing unit 701. For example, in some embodiments, methods 200, 300, 400, and 500 may be implemented as a computer software program tangibly embodied in a machine-readable medium, such as storage unit 708. In some embodiments, part or all of a computer program may be loaded onto and/or installed onto device 700 via ROM 702 and/or communications unit 709. When loaded into RAM 703 and executed by CPU 701, may perform one or more of the actions of methods 200, 300, 400, and 500 described above.
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, 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 (16)
1. A method of object recommendation, comprising:
determining at least one reference user selecting a target object, the target object having a degree of attention less than a threshold degree of attention;
determining a first degree of match between the at least one reference user and a target user based on features of the at least one reference user and features of the target user, the features being associated with historical selection of a sample object by the reference user; and
and if the first matching degree is greater than a first threshold matching degree, recommending the target object to the target user.
2. The method of claim 1, further comprising:
determining all reference users selecting the target object; and
determining an initial label of the target object based on the characteristics of all the reference users.
3. The method of claim 2, further comprising:
updating the initial label of the target object based on the characteristics of the target user if the target user selects the recommended target object.
4. The method of claim 3, further comprising:
stopping the updating if any of the following is satisfied:
a threshold time is reached;
reaching a first threshold number of times that the target object is recommended;
reaching a second threshold number of times the target object is selected; and
determining the updated initial label as a final label of the target object.
5. The method of claim 4, further comprising:
determining a second matching degree based on the characteristics of other users and the final label of the target object; and
and if the second matching degree is greater than a second threshold matching degree, recommending the target object to the other users.
6. The method of claim 1, wherein recommending the target object to the target user if the degree of match is greater than a threshold comprises: and sending the video, the image, the text or the voice representing the target object to the user.
7. The method of claim 1, wherein the degree of attention is determined based on at least one of: a frequency with which the target object is selected, a number of times the target object is recommended, a number of times the target object is selected, and a time at which the target object is created.
8. An apparatus for object recommendation, comprising:
a first reference user determination module configured to determine at least one reference user selecting a target object having a degree of attention less than a threshold degree of attention;
a first degree of match determination module configured to determine a first degree of match between the at least one reference user and a target user based on features of the at least one reference user and features of the target user, the features being associated with historical selection of a sample object by the reference user; and
a target user recommending module configured to recommend the target object to the target user if the first matching degree is greater than a first threshold matching degree.
9. The apparatus of claim 8, further comprising:
a second reference user determination module configured to determine all reference users selecting the target object; and
an initial label determination module configured to determine an initial label of the target object based on the characteristics of all reference users.
10. The apparatus of claim 9, further comprising:
an initial tag update module configured to update the initial tag of the target object based on the characteristics of the target user if the target user selects the recommended target object.
11. The apparatus of claim 10, further comprising:
a stop update module configured to stop the update if any of the following is satisfied:
a threshold time is reached;
reaching a first threshold number of times that the target object is recommended;
reaching a second threshold number of times the target object is selected; and
a final tag determination module configured to determine the updated initial tag as a final tag of the target object.
12. The apparatus of claim 11, further comprising:
a second matching degree determination module configured to determine a second matching degree based on the features of the other users and the final label of the target object; and
and the user recommending module is configured to recommend the target object to the other users if the second matching degree is greater than a second threshold matching degree.
13. The apparatus of claim 8, wherein the target user recommendation module comprises:
an object sending module configured to send a video, image, text or voice representing the target object to the user.
14. The apparatus of claim 8, wherein the degree of attention is determined based on at least one of: a frequency with which the target object is selected, a number of times the target object is recommended, a number of times the target object is selected, and a time at which the target object is created.
15. An electronic device, the device comprising:
one or more processors; and
storage means for storing 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 any one of claims 1-8.
16. A computer-readable storage medium, on which a computer program is stored which, when being executed by a processor, carries out the method according to any one of claims 1-8.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010879004.7A CN112100489B (en) | 2020-08-27 | 2020-08-27 | Object recommendation method, device and computer storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010879004.7A CN112100489B (en) | 2020-08-27 | 2020-08-27 | Object recommendation method, device and computer storage medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112100489A true CN112100489A (en) | 2020-12-18 |
| CN112100489B CN112100489B (en) | 2022-07-15 |
Family
ID=73757987
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010879004.7A Active CN112100489B (en) | 2020-08-27 | 2020-08-27 | Object recommendation method, device and computer storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112100489B (en) |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112765478A (en) * | 2021-03-12 | 2021-05-07 | 北京百度网讯科技有限公司 | Method, apparatus, device, medium, and program product for recommending content |
| CN112989186A (en) * | 2021-02-25 | 2021-06-18 | 腾讯科技（深圳）有限公司 | Information recommendation model training method and device, electronic equipment and storage medium |
| CN113220989A (en) * | 2021-04-29 | 2021-08-06 | 北京百度网讯科技有限公司 | Method and apparatus for object management |
| CN113301442A (en) * | 2021-05-20 | 2021-08-24 | 北京百度网讯科技有限公司 | Method, apparatus, medium, and program product for determining live broadcast resource |
| CN113722523A (en) * | 2021-09-01 | 2021-11-30 | 北京百度网讯科技有限公司 | Object recommendation method and device |
| CN116485560A (en) * | 2023-06-21 | 2023-07-25 | 杭州大鱼网络科技有限公司 | Target user screening method and system based on feedback mechanism |
Citations (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103778145A (en) * | 2012-10-22 | 2014-05-07 | 中国移动通信集团广东有限公司 | Information recommendation method and system |
| CN105426550A (en) * | 2015-12-28 | 2016-03-23 | Tcl集团股份有限公司 | Collaborative filtering tag recommendation method and system based on user quality model |
| CN105913296A (en) * | 2016-04-01 | 2016-08-31 | 北京理工大学 | Customized recommendation method based on graphs |
| CN106991522A (en) * | 2017-03-16 | 2017-07-28 | 西南交通大学 | A kind of new model score in predicting method based on content |
| CN107526850A (en) * | 2017-10-12 | 2017-12-29 | 燕山大学 | Social networks friend recommendation method based on multiple personality feature mixed architecture |
| CN107590243A (en) * | 2017-09-14 | 2018-01-16 | 中国人民解放军信息工程大学 | The personalized service recommendation method to be sorted based on random walk and diversity figure |
| CN107608743A (en) * | 2017-09-04 | 2018-01-19 | 维沃移动通信有限公司 | A kind of operating system customizing method, server and mobile terminal |
| CN110851729A (en) * | 2019-11-19 | 2020-02-28 | 深圳前海微众银行股份有限公司 | Resource information recommendation method, device, equipment and computer storage medium |
| CN110910215A (en) * | 2019-11-20 | 2020-03-24 | 深圳前海微众银行股份有限公司 | Product recommendation method, device, equipment and computer-readable storage medium |
| US20200201844A1 (en) * | 2017-11-17 | 2020-06-25 | Alibaba Group Holding Limited | Cluster-based random walk processing |
| CN111414548A (en) * | 2020-05-09 | 2020-07-14 | 中国工商银行股份有限公司 | Object recommendation method and device, electronic equipment and medium |
-
2020
- 2020-08-27 CN CN202010879004.7A patent/CN112100489B/en active Active
Patent Citations (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103778145A (en) * | 2012-10-22 | 2014-05-07 | 中国移动通信集团广东有限公司 | Information recommendation method and system |
| CN105426550A (en) * | 2015-12-28 | 2016-03-23 | Tcl集团股份有限公司 | Collaborative filtering tag recommendation method and system based on user quality model |
| CN105913296A (en) * | 2016-04-01 | 2016-08-31 | 北京理工大学 | Customized recommendation method based on graphs |
| CN106991522A (en) * | 2017-03-16 | 2017-07-28 | 西南交通大学 | A kind of new model score in predicting method based on content |
| CN107608743A (en) * | 2017-09-04 | 2018-01-19 | 维沃移动通信有限公司 | A kind of operating system customizing method, server and mobile terminal |
| CN107590243A (en) * | 2017-09-14 | 2018-01-16 | 中国人民解放军信息工程大学 | The personalized service recommendation method to be sorted based on random walk and diversity figure |
| CN107526850A (en) * | 2017-10-12 | 2017-12-29 | 燕山大学 | Social networks friend recommendation method based on multiple personality feature mixed architecture |
| US20200201844A1 (en) * | 2017-11-17 | 2020-06-25 | Alibaba Group Holding Limited | Cluster-based random walk processing |
| CN110851729A (en) * | 2019-11-19 | 2020-02-28 | 深圳前海微众银行股份有限公司 | Resource information recommendation method, device, equipment and computer storage medium |
| CN110910215A (en) * | 2019-11-20 | 2020-03-24 | 深圳前海微众银行股份有限公司 | Product recommendation method, device, equipment and computer-readable storage medium |
| CN111414548A (en) * | 2020-05-09 | 2020-07-14 | 中国工商银行股份有限公司 | Object recommendation method and device, electronic equipment and medium |
Non-Patent Citations (2)
| Title |
|---|
| MA TIE-MIN，等: "Social Network Event Recommendation Algorithms Based on User Similarity Random Walk", 《JOURNAL OF NORTHEASTERN UNIVERSITY (NATURAL SCIENCE)》 * |
| 时宇岑，等: "基于多开发者社区的用户推荐算法", 《软件学报》 * |
Cited By (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112989186A (en) * | 2021-02-25 | 2021-06-18 | 腾讯科技（深圳）有限公司 | Information recommendation model training method and device, electronic equipment and storage medium |
| CN112989186B (en) * | 2021-02-25 | 2024-01-12 | 深圳市雅阅科技有限公司 | Information recommendation model training method and device, electronic equipment and storage medium |
| CN112765478A (en) * | 2021-03-12 | 2021-05-07 | 北京百度网讯科技有限公司 | Method, apparatus, device, medium, and program product for recommending content |
| CN113220989A (en) * | 2021-04-29 | 2021-08-06 | 北京百度网讯科技有限公司 | Method and apparatus for object management |
| CN113220989B (en) * | 2021-04-29 | 2023-08-29 | 北京百度网讯科技有限公司 | Method and apparatus for object management |
| CN113301442A (en) * | 2021-05-20 | 2021-08-24 | 北京百度网讯科技有限公司 | Method, apparatus, medium, and program product for determining live broadcast resource |
| CN113301442B (en) * | 2021-05-20 | 2022-10-21 | 北京百度网讯科技有限公司 | Method, device, medium, and program product for determining live broadcast resource |
| CN113722523A (en) * | 2021-09-01 | 2021-11-30 | 北京百度网讯科技有限公司 | Object recommendation method and device |
| CN113722523B (en) * | 2021-09-01 | 2023-11-17 | 北京百度网讯科技有限公司 | Object recommendation method and device |
| CN116485560A (en) * | 2023-06-21 | 2023-07-25 | 杭州大鱼网络科技有限公司 | Target user screening method and system based on feedback mechanism |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112100489B (en) | 2022-07-15 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN112100489B (en) | Object recommendation method, device and computer storage medium | |
| CN110297848B (en) | Recommendation model training method, terminal and storage medium based on federal learning | |
| CN109086439B (en) | Information recommendation method and device | |
| CN108776676B (en) | Information recommendation method and device, computer readable medium and electronic device | |
| US10747771B2 (en) | Method and apparatus for determining hot event | |
| CN108520470B (en) | Method and apparatus for generating user attribute information | |
| CN109471978B (en) | Electronic resource recommendation method and device | |
| CN110413867B (en) | Method and system for content recommendation | |
| CN110020162B (en) | User identification method and device | |
| CN114036398B (en) | Content recommendation and ranking model training method, device, equipment and storage medium | |
| CN112182360A (en) | Personalized recommender with limited data availability | |
| CN111368973A (en) | Method and apparatus for training a hyper-network | |
| CN112765478A (en) | Method, apparatus, device, medium, and program product for recommending content | |
| CN110059172B (en) | Method and device for recommending answers based on natural language understanding | |
| CN110992127A (en) | Article recommendation method and device | |
| CN112749323A (en) | Method and device for constructing user portrait | |
| WO2024051707A1 (en) | Recommendation model training method and apparatus, and resource recommendation method and apparatus | |
| CN114119123A (en) | Information pushing method and device | |
| CN112950321A (en) | Article recommendation method and device | |
| US8971645B1 (en) | Video categorization using heterogeneous signals | |
| CN110727853B (en) | Presenting controlled heterogeneous digital content to a user | |
| CN112100507B (en) | Object recommendation method, computing device and computer-readable storage medium | |
| WO2021204039A1 (en) | Method and apparatus for pushing information | |
| CN113365090B (en) | Object recommendation method, object recommendation device, electronic equipment and readable storage medium | |
| CN115329864A (en) | Method and device for training recommendation model and electronic equipment |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |