CN108572984A - A kind of active user interest recognition methods and device - Google Patents
A kind of active user interest recognition methods and device Download PDFInfo
- Publication number
- CN108572984A CN108572984A CN201710146012.9A CN201710146012A CN108572984A CN 108572984 A CN108572984 A CN 108572984A CN 201710146012 A CN201710146012 A CN 201710146012A CN 108572984 A CN108572984 A CN 108572984A
- Authority
- CN
- China
- Prior art keywords
- interest
- user
- time
- real
- score
- 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.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 60
- 239000013598 vector Substances 0.000 claims abstract description 127
- 230000002123 temporal effect Effects 0.000 claims abstract description 3
- 230000006399 behavior Effects 0.000 claims description 212
- 238000004364 calculation method Methods 0.000 claims description 29
- 239000011159 matrix material Substances 0.000 claims description 13
- 230000005540 biological transmission Effects 0.000 claims description 3
- 230000002452 interceptive effect Effects 0.000 claims description 2
- 230000000694 effects Effects 0.000 abstract description 2
- 230000003542 behavioural effect Effects 0.000 abstract 6
- 238000010586 diagram Methods 0.000 description 18
- 238000004422 calculation algorithm Methods 0.000 description 12
- 238000004891 communication Methods 0.000 description 7
- 230000003993 interaction Effects 0.000 description 6
- 238000012549 training Methods 0.000 description 6
- 230000007774 longterm Effects 0.000 description 5
- 230000006870 function Effects 0.000 description 4
- 238000012545 processing Methods 0.000 description 4
- 230000009471 action Effects 0.000 description 3
- 238000012986 modification Methods 0.000 description 3
- 230000004048 modification Effects 0.000 description 3
- 230000008569 process Effects 0.000 description 3
- 230000006978 adaptation Effects 0.000 description 2
- 238000004458 analytical method Methods 0.000 description 2
- 238000004590 computer program Methods 0.000 description 2
- 238000013507 mapping Methods 0.000 description 2
- 238000013459 approach Methods 0.000 description 1
- 238000013528 artificial neural network Methods 0.000 description 1
- 238000007418 data mining Methods 0.000 description 1
- 230000007423 decrease Effects 0.000 description 1
- 238000009792 diffusion process Methods 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 230000005055 memory storage Effects 0.000 description 1
- 238000003672 processing method Methods 0.000 description 1
- 230000000750 progressive effect Effects 0.000 description 1
- 230000004044 response Effects 0.000 description 1
- 230000002441 reversible effect Effects 0.000 description 1
Landscapes
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
A kind of active user interest recognition methods of the embodiment of the present application offer and equipment, the method includes：Obtain active user behavioral data, the feature vector for obtaining the active user behavioral data obtains the real-time interest score of user using the feature vector of the temporal characteristics of active user behavioral data, time attenuation parameter, interest relationship weight, the active user behavioral data；According to user's history behavioral data, user's history interest score is obtained；According to user's history behavioral data, the point of interest vector of user is obtained, the similarity score of point of interest is obtained according to the point of interest vector of the user；The interest recognition result of the user is determined according to the similarity score of the active user interest score, the user's history interest score and the point of interest.The embodiment of the present application can identify that accuracy is high, real-time by the real-time Activity recognition user interest of user.
Description
Technical Field
The embodiment of the application relates to the technical field of computers, in particular to a real-time user interest identification method and device.
Background
The user interest identification algorithm is a technique for identifying interest categories of a user through data mining analysis. In the prior art, user interest identification generally adopts an offline method, and analysis and calculation are performed according to historical behavior data of a user, so that interest preference of the user is obtained. However, the historical behavior data of the user often represents the interest point of the user at a certain moment in the history, the interest of the user may be met or changed with the time attenuation, and if the interest of the user is still identified by using the historical behavior data, the accuracy is not high. In the prior art, a user interest identification method with strong real-time performance and high accuracy does not exist.
Disclosure of Invention
The embodiment of the application provides a real-time user interest identification method and device, which can identify user interests through real-time behaviors of users, and are high in identification accuracy and strong in real-time performance.
Therefore, the embodiment of the application provides the following technical scheme:
in a first aspect, an embodiment of the present application provides a real-time user interest identification method, including: acquiring real-time user behavior data and acquiring a feature vector of the real-time user behavior data; obtaining a user real-time interest score by utilizing the time characteristics, the time attenuation parameters, the interest relation weight and the characteristic vector of the real-time user behavior data; and determining an interest identification result of the user according to the real-time user interest score.
In some embodiments, the temporal characteristic of the real-time user behavior data is a difference between an occurrence time of the real-time user behavior and a current time.
In some embodiments, the interest relationship weight is embodied as a parameter matrix of a hierarchical network of interest points.
In some embodiments, the method further comprises: obtaining a user historical interest score according to the user historical behavior data; the determining the interest recognition result of the user according to the real-time user interest score comprises: and determining an interest identification result of the user according to the real-time user interest score and the historical user interest score.
In some embodiments, the method further comprises: obtaining an interest point vector of a user according to historical behavior data of the user; obtaining a similarity score of the interest points according to the interest point vectors of the users; the determining the interest recognition result of the user according to the real-time user interest score comprises: determining an interest identification result of the user according to the real-time user interest score and the similarity score of the interest points; or determining the interest identification result of the user according to the real-time user interest score, the user historical interest score and the similarity score of the interest points.
In a second aspect, an embodiment of the present application provides a real-time user interest identification method, including: acquiring real-time user behavior data, acquiring a feature vector of the real-time user behavior data, and acquiring a user real-time interest score by using a time feature, a time attenuation parameter, an interest relation weight and the feature vector of the real-time user behavior data; obtaining a user historical interest score according to the user historical behavior data; obtaining interest point vectors of users according to historical behavior data of the users, and obtaining similarity scores of the interest points according to the interest point vectors of the users; and determining an interest identification result of the user according to the real-time user interest score, the historical user interest score and the similarity score of the interest points.
In a third aspect, an embodiment of the present application provides a device for identifying a real-time user interest, including: the real-time data acquisition unit is used for acquiring real-time user behavior data and acquiring a feature vector of the real-time user behavior data; the real-time interest score calculating unit is used for obtaining a user real-time interest score by utilizing the time characteristics, the time attenuation parameters, the interest relation weight and the characteristic vector of the real-time user behavior data; and the interest identification result determining unit is used for determining the interest identification result of the user according to the real-time user interest score.
In a fourth aspect, an embodiment of the present application provides a device for identifying a real-time user interest, including: the real-time interest score calculation module is used for acquiring real-time user behavior data, acquiring a feature vector of the real-time user behavior data, and acquiring a user real-time interest score by utilizing the time feature, the time attenuation parameter, the interest relation weight and the feature vector of the real-time user behavior data; the historical interest score calculating module is used for acquiring the historical interest score of the user according to the historical behavior data of the user; the interest point similarity score calculating module is used for acquiring an interest point vector of the user according to the historical behavior data of the user and obtaining the similarity score of the interest point according to the interest point vector of the user; and the interest identification result determining module is used for determining the interest identification result of the user according to the real-time user interest score, the historical interest score of the user and the similarity score of the interest points.
In a fifth aspect, an embodiment of the present application provides an apparatus for real-time user interest identification, comprising a memory, and one or more programs, wherein the one or more programs are stored in the memory and configured to be executed by the one or more processors, the one or more programs including instructions for: acquiring real-time user behavior data and acquiring a feature vector of the real-time user behavior data; obtaining a user real-time interest score by utilizing the time characteristics, the time attenuation parameters, the interest relation weight and the characteristic vector of the real-time user behavior data; and determining an interest identification result of the user according to the real-time user interest score.
In a sixth aspect, embodiments of the present application provide an apparatus for real-time user interest identification, comprising a memory, and one or more programs, wherein the one or more programs are stored in the memory and configured to be executed by the one or more processors, the one or more programs including instructions for: acquiring real-time user behavior data, acquiring a feature vector of the real-time user behavior data, and acquiring a user real-time interest score by using a time feature, a time attenuation parameter, an interest relation weight and the feature vector of the real-time user behavior data; obtaining a user historical interest score according to the user historical behavior data; obtaining interest point vectors of users according to historical behavior data of the users, and obtaining similarity scores of the interest points according to the interest point vectors of the users; and determining an interest identification result of the user according to the real-time user interest score, the historical user interest score and the similarity score of the interest points.
In a seventh aspect, an embodiment of the present application provides a real-time user interest identification system, including a real-time computing server, an offline computing server, and a result output device, where: the real-time computing server is used for acquiring real-time user behavior data, acquiring a feature vector of the real-time user behavior data, and acquiring a user real-time interest score by utilizing the time feature, the time attenuation parameter, the interest relation weight and the feature vector of the real-time user behavior data; the off-line computing server is used for acquiring a user historical interest score according to the user historical behavior data; and/or obtaining interest point vectors of the users according to the historical behavior data of the users, and obtaining similarity scores of the interest points according to the interest point vectors of the users; the real-time computing server is also used for determining an interest identification result of the user according to the real-time interest score of the user; or determining an interest identification result of the user according to the real-time interest score and the historical interest score of the user; or determining an interest identification result of the user according to the real-time interest score of the user and the similarity score of the interest points; or determining an interest identification result of the user according to the real-time user interest score, the user historical interest score and the similarity score of the interest points; and the result output device is used for outputting the interest identification result of the user.
In an eighth aspect, an embodiment of the present application provides a real-time computing server, including: the real-time user behavior acquisition device is used for acquiring real-time user behavior data; the real-time computing node is used for obtaining the feature vector of the real-time user behavior data and obtaining a user real-time interest score by utilizing the time feature, the time attenuation parameter, the interest relation weight and the feature vector of the real-time user behavior data; determining an interest identification result of the user according to the real-time interest score of the user; or determining an interest identification result of the user according to the real-time interest score and the historical interest score of the user; or determining an interest identification result of the user according to the real-time interest score of the user and the similarity score of the interest points; or determining an interest identification result of the user according to the real-time user interest score, the historical user interest score and the similarity score of the interest points; and calculating the user historical interest score and/or the similarity score of the interest points by an offline calculation server.
In a ninth aspect, an embodiment of the present application provides an offline computing server, including: the user historical interest score calculation node is used for acquiring a user historical interest score according to the user historical behavior data; the similarity score calculation node of the interest points is used for acquiring the interest point vectors of the users according to the historical behavior data of the users and obtaining the similarity scores of the interest points according to the interest point vectors of the users; and the output device is used for outputting the historical interest scores of the users and/or the similarity scores of the interest points to a real-time computing server so as to obtain the user interest identification results.
In a tenth aspect, an embodiment of the present application provides an interaction apparatus, including: the input module is used for inputting real-time user behavior data and historical user behavior data; the transmission module is used for transmitting the real-time user behavior data input by the input module to a real-time calculation server and transmitting the historical user behavior data input by the input module to an offline calculation server; the output module is used for outputting a user interest identification result; and the user interest identification result is calculated by the real-time calculation server.
In an eleventh aspect, an embodiment of the present application provides a real-time user interest identification method, which is applied to a real-time computing server, and includes: acquiring real-time user behavior data; obtaining a feature vector of the real-time user behavior data, and obtaining a user real-time interest score by utilizing the time feature, the time attenuation parameter, the interest relation weight and the feature vector of the real-time user behavior data; determining an interest identification result of the user according to the real-time interest score of the user; or determining an interest identification result of the user according to the real-time interest score and the historical interest score of the user; or determining an interest identification result of the user according to the real-time interest score of the user and the similarity score of the interest points; or determining an interest identification result of the user according to the real-time user interest score, the historical user interest score and the similarity score of the interest points; and calculating the user historical interest score and/or the similarity score of the interest points by an offline calculation server.
In a twelfth aspect, an embodiment of the present application provides a user interest identification method, applied to an offline computing server, including: obtaining a user historical interest score according to the user historical behavior data; obtaining interest point vectors of users according to historical behavior data of the users, and obtaining similarity scores of the interest points according to the interest point vectors of the users; and outputting the historical interest scores and/or the similarity scores of the interest points of the users to a real-time computing server to obtain the user interest identification results.
The real-time user interest identification method and device provided by the embodiment of the application can acquire real-time user behavior data, consider the influence of time attenuation factors on the interest point identification result, obtain the real-time interest score of a user by utilizing the time characteristics, the time attenuation parameters, the interest relation weight and the characteristic vector of the real-time user behavior data, and obtain the user interest identification result according to the real-time interest score of the user, so that the obtained result is more reliable, accurate and strong in real-time performance.
Drawings
In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings needed to be used in the description of the embodiments or the prior art will be briefly described below, it is obvious that the drawings in the following description are only some embodiments described in the present application, and other drawings can be obtained by those skilled in the art without creative efforts.
FIG. 1 is a schematic diagram of a user interest identification system according to an embodiment of the present application;
FIG. 2 is a flowchart of a method for identifying user interests in real time according to an embodiment of the present application;
FIG. 3 is a flow chart of a method for real-time user interest identification according to another embodiment of the present application;
FIG. 4 is a schematic diagram illustrating an algorithm structure of a real-time user interest identification method according to an embodiment of the present application;
FIG. 5 is a schematic diagram of a real-time user interest recognition apparatus according to an embodiment of the present application;
FIG. 6 is a schematic diagram of a real-time user interest identification apparatus according to yet another embodiment of the present application;
FIG. 7 is a block diagram illustrating an apparatus for real-time user interest identification in accordance with an exemplary embodiment;
FIG. 8 is a block diagram illustrating an apparatus for real-time user interest identification in accordance with an exemplary embodiment;
FIG. 9 is a schematic diagram of a real-time computing server provided by an embodiment of the present application;
FIG. 10 is a schematic diagram of an offline computing server provided by an embodiment of the present application;
fig. 11 is a schematic diagram of an interaction device according to an embodiment of the present application.
Detailed Description
The embodiment of the application provides a real-time user interest identification method and device, which can identify user interests through real-time behaviors of users, and are high in identification accuracy and strong in real-time performance.
In order to make those skilled in the art better understand the technical solutions in the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application, and it is obvious that the described embodiments are only a part of the embodiments of the present application, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present application.
The terminology used in the embodiments of the present application is for the purpose of describing particular embodiments only and is not intended to be limiting of the application. As used in the examples of this application and the appended claims, the singular forms "a", "an", and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise. It should also be understood that the term "and/or" as used herein refers to and encompasses any and all possible combinations of one or more of the associated listed items.
Referring to fig. 1, a schematic diagram of a user interest identification system provided in the embodiment of the present application is shown. The method provided by the embodiment of the present application may be applied to a user interest recognition system as shown in fig. 1, which may include a real-time calculation server 900, an offline calculation server 1000, and a result output device 1100. The result output device can be an independent device and is respectively connected with the real-time computing server and the off-line computing server; the result output device may also exist as part of a real-time computing server. The real-time computing server 900 is configured to obtain real-time user behavior data, obtain a feature vector of the real-time user behavior data, and obtain a user real-time interest score by using a time feature, a time decay parameter, an interest relationship weight, and the feature vector of the real-time user behavior data. The offline computing server 1000 is configured to obtain a user historical interest score according to the user historical behavior data; and/or obtaining the interest point vector of the user according to the historical behavior data of the user, and obtaining the similarity score of the interest point according to the interest point vector of the user. The real-time computing server 900 is further configured to determine an interest recognition result of the user according to the real-time interest score of the user; or determining an interest identification result of the user according to the real-time interest score and the historical interest score of the user; or determining an interest identification result of the user according to the real-time interest score of the user and the similarity score of the interest points; or determining an interest identification result of the user according to the real-time user interest score, the user historical interest score and the similarity score of the interest points; the result output device 1100 is configured to output an interest recognition result of the user.
Of course, the embodiments of the present invention may also be applied to other scenarios, which are not limited herein. It should be noted that the above application scenarios are only presented to facilitate understanding of the present invention, and the embodiments of the present invention are not limited in any way in this respect. Rather, embodiments of the present invention may be applied to any scenario where applicable.
The real-time user interest identification method shown in the exemplary embodiment of the present application will be described with reference to fig. 2 to 4.
It should be noted that, in the internet field, several indexes, namely large scale, real-time performance and accuracy, are generally considered when processing data. The data processing method can support massive users in a large scale, and the larger the user amount is, the more abundant the obtained information is; the real-time response time of the service is short, and the timeliness and the value of the message can be ensured only when the algorithm is more real-time; the accuracy is realized in that the algorithm can provide correct category division and personalized service for each of the mass users, and the more accurate the algorithm is, the greater the profit is. However, most algorithms cannot simultaneously consider large scale, real-time performance and accuracy. Therefore, most of the methods generally involve a corresponding compromise among three factors, for example, in order to realize large-scale data processing, it is a common practice to provide an off-line algorithm to trade off real-time performance and accuracy. In the prior art, user interest identification is based on offline processing, so that certain limitations exist in real-time performance and accuracy. The application aims to provide an accurate large-scale real-time user interest identification algorithm, which can help sellers in the E-commerce field to make reasonable user interest classification and real-time accurate marketing, and the application of the algorithm is not limited to the method.
Referring to fig. 2, a flowchart of a real-time user interest identification method according to an embodiment of the present application is provided. As shown in fig. 2, may include:
s201, acquiring real-time user behavior data and acquiring a feature vector of the real-time user behavior data.
In specific implementation, real-time user behavior data can be collected through the real-time user behavior log. The user behavior data may include, but is not limited to, occurrence time of the user behavior, category of the user behavior, information of the user, and commodity information corresponding to the user behavior. For example, the category of user behavior may include types of behavior of a user regarding browsing, collecting, adding a shopping cart, purchasing, etc. of a certain item. The information of the user may include information of a terminal where the user is located, user login duration, a commodity sequence browsed by the user before the real-time user behavior data occurs, duration of real-time behavior of the user, and the like. The commodity information corresponding to the user behavior may include information such as a brand of the commodity, a price of the commodity, and a category to which the commodity belongs. It should be noted that one or more pieces of information included in the user behavior data may be extracted as a feature vector of the user behavior data. The feature vector of the user behavior data may be a vector X abstractly expressed in one m-dimension. In some embodiments, the feature vector X of the user behavior data at least includes information of the user, commodity information corresponding to the user behavior, and a user behavior category. The application does not limit the types of the user features included in the feature vector X of the user behavior data.
S202, obtaining a user real-time interest score by utilizing the time characteristics, the time attenuation parameters, the interest relation weight and the characteristic vector of the real-time user behavior data.
The applicant has found that the accuracy of the user's interest recognition results is affected by a time decay factor. For example, the degree of memory of each user on the behavior of the commodity can be represented by an Ebingos memory curve:
Y＝1-0.56X0.06
wherein Y represents the memory level and X represents the time (unit may be hour). From the above disclosure, it can be seen that the memory level of the user for the merchandise tends to decline with the increase of time. Therefore, the user interest identification result is obtained by collecting the real-time user behaviors, and the influence of time factors is considered.
In specific implementation, the time characteristic of the real-time user behavior data is specifically a difference value between occurrence time of the real-time user behavior and current time. For example, Δ t may be used to represent the difference between the current time and the real-time user behavior time. The time decay parameters may include a time decay factor and a time decay coefficient. The interest relation weight is specifically a parameter matrix of a hierarchical interest point network.
In some embodiments, the user real-time interest score may be obtained by the following formularealtime：
Where Δ t is the difference between the current time and the occurrence time of the real-time user behavior, σ is the time decay factor, and λ is the time decay coefficient. WiRepresenting interest relationship weight, WiWill be updated on-line in real time, XTIs an input parameter, i.e. a user behavior feature vector. One point to be described is Wi，XTAre all matrices. When embodied, WiSpecifically, the method is a parameter matrix of the hierarchical interest point network, and N represents the number of layers of the hierarchical interest point network model.
In some embodiments, the user real-time interest score may be obtained by building a hierarchical point of interest network model. As shown in fig. 4, the hierarchical point of interest network model may be a model built in a multi-level network structure in a real-time or offline manner. As shown in fig. 4, the parameters of the hierarchical point of interest network include two parts: each circle in the network represents an interest point, the network has increasingly strong interest abstraction degree from left to right, and the rightmost end is an interest output end and is a highly abstract interest category; another parameter is the edge between layers, representing the interest relationship weight. Both parameters of the network can be updated in real time and offline. In specific implementation, when a hierarchical interest point network model is constructed, interest points in the network can be trained layer by layer according to a vector X with m-dimension characteristics of an input layer, so that a parameter vector W of each interest point is obtained, and n interest points in the same layer form a matrix W of a matrix m X n. And after the first layer is trained, training the second layer network on the basis of the first layer until the last layer. W in formula (1)iThen the parameter matrix representing each layer network, which is the behaviorAnd mapping relation matrixes of the entities and the interest points. It should be noted that the model establishment and the parameter update may be based on offline data or real-time data.
After a hierarchical interest point network model is constructed, a real-time interest score can be obtained by using the formula (1). Each user behavior can be represented by a feature vector X, and the interest score of the real-time behavior of the user can be obtained according to the public expression (1). Wherein, scorerealtimeThe dimension of (1) is the number of interest points of the last layer of network, and each value of the vector represents a score of a certain interest.
In some embodiments, the interest recognition result of the user may be determined directly by the real-time user interest score. For example, assume scorerealtimeIs a 2-dimensional vector, and 2 interest points can be taken as the interest identification result. Of course, the interest category with a high score may be used as the final recognition result. The present application does not limit how the interest recognition result of the user is determined according to the real-time user interest score.
In some embodiments, the user historical interest score may be obtained through an offline historical point of interest model, taking into account the impact of the user's long-term preferences on the user's interest recognition results. For example, a user historical interest score may be obtained based on the user historical behavior data. When determining the interest recognition result of the user, the interest recognition result of the user may be determined according to the real-time user interest score and the user history interest score. For example, the real-time user interest score and the user historical interest score are given different weights to obtain the final interest recognition result.
In some embodiments, in order to obtain the potential interest of the user, the similarity score of the interest point may also be obtained according to the interest similarity model. For example, the interest point vector of the user can be obtained according to the historical behavior data of the user; and obtaining the similarity score of the interest points according to the interest point vector of the user. When determining the interest recognition result of the user, the interest recognition result of the user may be determined according to the real-time user interest score and the similarity score of the interest point. For example, the real-time user interest score and the similarity score of the interest point are given different weights to obtain the final interest recognition result. Of course, in some embodiments, the influence of the 3 factors may be comprehensively considered, and the interest recognition result of the user is determined according to the real-time user interest score, the user historical interest score and the similarity score of the interest points, so as to obtain a more accurate result.
In order to facilitate those skilled in the art to more clearly understand the embodiments of the present application in a specific context, the following describes the embodiments of the present application with a specific example. It should be noted that the specific example is only to make the present application more clearly understood by those skilled in the art, but the embodiments of the present application are not limited to the specific example.
Referring to fig. 3, a flowchart of a real-time user interest identification method according to an embodiment of the present application is provided. As shown in fig. 3, may include:
s301, acquiring real-time user behavior data, acquiring a feature vector of the real-time user behavior data, and acquiring a user real-time interest score by using the time feature, the time attenuation parameter, the interest relation weight and the feature vector of the real-time user behavior data.
And S302, obtaining a user historical interest score according to the user historical behavior data.
S303, obtaining the interest point vector of the user according to the historical behavior data of the user, and obtaining the similarity score of the interest point according to the interest point vector of the user.
S304, determining an interest identification result of the user according to the real-time user interest score, the user historical interest score and the similarity score of the interest points.
It should be noted that S301, S302, and S303 do not necessarily have a sequential execution order, and the order may be executed in reverse or in parallel.
Fig. 3 is described below with reference to fig. 4. Referring to fig. 4, a schematic diagram of an algorithm structure of a real-time user interest identification method provided in an embodiment of the present application is shown. The algorithm structure of the application can comprise 3 parts, namely a real-time user interest model, an offline user interest point model and an interest similarity model. The real-time user interest model ensures real-time accuracy of the algorithm, the offline user interest point model provides long-term interest preference of the user, and the interest similarity model provides potential interest of the user. And weighting the three parts to obtain the real-time interest score. Wherein the real-time interest score is calculated by the following formula:
Score＝α*scorerealtime+β*scoreoffline+γ*scoresimilarity(2)
α+β+γ＝1 (3)
the method provided by the embodiment shown in fig. 2 is provided when α ═ 1, β ═ 0, and γ ═ 0.
Wherein, scorerealtimeIs a real-time interest user model score, scoreofflineIs an offline user interest score (also referred to as a user historical interest score), scoresimilarityIs a user interest similarity score (which may also be referred to as a similarity score for a point of interest).
The real-time user model includes two parts: one part is a real-time user behavior sequence, and the other part is a hierarchical interest point model. The real-time behavior sequence records the time stamps and behavior types of users and behavior subjects, and the hierarchical interest point model is a model established by adopting a multi-level network structure in a real-time or off-line mode. The parameters of the network comprise two parts, each circle represents an interest point, the network has increasingly strong interest abstraction degree from left to right, and the rightmost end is an interest output end and is a highly abstract interest category; another parameter is the edge between layers, representingInterest relationship weights. Both parameters of the network can be updated in real time and offline. In specific implementation, when a hierarchical interest point network model is constructed, interest points in the network can be trained layer by layer according to a vector X with m-dimension characteristics of an input layer, so that a parameter vector W of each interest point is obtained, and n interest points in the same layer form a matrix W of a matrix m X n. And after the first layer is trained, training the second layer network on the basis of the first layer until the last layer. W in formula (1)iA parameter matrix representing each layer network is a mapping relation matrix of the behavior entities and the interest points. It should be noted that the model establishment and the parameter update may be based on offline data or real-time data.
After a hierarchical interest point network model is constructed, a real-time interest score can be obtained by using the formula (1). Each user behavior can be represented by a feature vector X, and the interest score of the real-time behavior of the user can be obtained according to the public expression (1). Wherein, scorerealtimeThe dimension of (1) is the number of interest points of the last layer of network, and each value of the vector represents a score of a certain interest.
In consideration of the influence of the long-term preference of the user on the user interest, the method and the device obtain the long-term preference of the user through the offline model. Offline models play a significant role, particularly when real-time behavior is not available. The offline model may model points of interest in both a behavior-based and a content-based manner. The behavior-based interest point model can be in a hierarchical clustering mode, and the content-based model can be in a title embedding mode. Specifically, the offline modeling approach may utilize a user's point of interest samples to perform multi-class training using a model (e.g., dnn), and output an offline model. The offline model can score historical behaviors in multiple categories to obtain historical interest scores scoreofflineThe historical interestingness score is a vector, and each dimension is a point of interest score.
In order to predict the potential interest of the user, the interest similarity score of the user can be obtained through an interest similarity model. For example, behavior data of a user can be acquired from a log system, behaviors of the user are sorted according to time for a specific user, and an object commodity with a behavior function is abstracted to be a point of interest C. The following sequence of behaviors is then obtained:
wherein u isiRepresenting users i, cijPoints of interest representing the j-th activity of user i.
In a specific implementation, the behavior sequences of the users are regarded as a piece of 'natural' text, and the 'behavior texts' of all the users are regarded as final training 'corpora'. The 'text' is trained by using a neural network-based language model, and Word Embedding (Word Embedding, which is a vectorization expression method of words) encoding is carried out on the interest points. Here, the coding result of the interest point is marked asWith the vector encoding of the points of interest, the similarity of the points of interest can be calculated based on the vectors. Specifically, cosine similarity may be used as the similarity of the evaluation points of interest. Through the calculation, the similarity scores of any 2 interest points can be obtained, and the scores are used as the basis for interest point diffusion.
Wherein, the objective function of Embedding is:
here, the training dataset is C, and context (C) is the context of the commodity (or point of interest) C in the training dataset. The similarity of the commodity clusters (interest points) can also be obtained through the Embedding of the commodities.
Thereby, the potential interest of the user can be obtained. For example, assuming that the interest of the user in the brand Nike is determined through a real-time user interest model or an offline interest model, whether the user is interested in a similar sports brand adidas can be determined through the interest similarity score, and therefore the potential interest category of the user can be predicted.
In the embodiment, the influence of real-time user interest, long-term preference of the user and potential interest of the user on the user interest recognition result is comprehensively considered, so that the obtained recognition result is more accurate.
It should be noted that the real-time user interest identification method provided in the present application is introduced from the perspective of the method performed by the real-time user interest identification system as a whole. Correspondingly, the embodiment of the application also provides a method executed by the real-time computing server side and a method executed by the off-line computing server. In some embodiments, the present application further provides a real-time user interest identification method applied to a real-time computing server, including the following steps: A. acquiring real-time user behavior data; B. obtaining a feature vector of the real-time user behavior data, and obtaining a user real-time interest score by utilizing the time feature, the time attenuation parameter, the interest relation weight and the feature vector of the real-time user behavior data; C. determining an interest identification result of the user according to the real-time interest score of the user; or determining an interest identification result of the user according to the real-time interest score and the historical interest score of the user; or determining an interest identification result of the user according to the real-time interest score of the user and the similarity score of the interest points; or determining an interest identification result of the user according to the real-time user interest score, the historical user interest score and the similarity score of the interest points; and calculating the user historical interest score and/or the similarity score of the interest points by an offline calculation server. Specific implementations can be made with reference to the methods shown in fig. 2 to 4.
In some embodiments, the present application further provides a user interest identification method applied to an offline computing server, including the following steps: a', obtaining a user historical interest score according to user historical behavior data; b', obtaining interest point vectors of the users according to historical behavior data of the users, and obtaining similarity scores of the interest points according to the interest point vectors of the users; and C', outputting the historical interest scores of the users and/or the similarity scores of the interest points to a real-time computing server to obtain the user interest identification results. Specific implementations can be made with reference to the methods shown in fig. 2 to 4.
The following describes a device corresponding to the method provided by the embodiment of the present application.
Referring to fig. 5, a schematic diagram of a device for identifying a user's interest in real time according to an embodiment of the present application is shown.
A real-time user interest recognition apparatus 500, comprising:
the real-time data obtaining unit 501 is configured to obtain real-time user behavior data and obtain a feature vector of the real-time user behavior data.
The real-time interest score calculating unit 502 is configured to obtain a real-time interest score of the user by using the time characteristic of the real-time user behavior data, the time decay parameter, the interest relationship weight, and the feature vector of the real-time user behavior data.
An interest recognition result determining unit 503, configured to determine an interest recognition result of the user according to the real-time user interest score.
In some embodiments, the real-time interest score calculating unit 502 is specifically configured to obtain the real-time interest score of the user by using a difference between the occurrence time of the real-time user behavior and the current time, a time decay parameter, an interest relationship weight, and a feature vector of the real-time user behavior data.
In some embodiments, the real-time interest score calculating unit 502 is specifically configured to obtain the real-time interest score of the user by using the time characteristic of the user behavior data, the time decay parameter, the parameter matrix of the hierarchical interest point network, and the feature vector of the real-time user behavior data.
In some embodiments, the apparatus further comprises:
the historical interest score calculating unit is used for acquiring the historical interest score of the user according to the historical behavior data of the user;
the interest recognition result determining unit 503 is specifically configured to determine an interest recognition result of the user according to the real-time user interest score and the user historical interest score.
In some embodiments, the apparatus further comprises:
the interest similarity score calculating unit is specifically used for acquiring an interest point vector of the user according to the historical behavior data of the user; obtaining a similarity score of the interest points according to the interest point vectors of the users;
the interest recognition result determining unit 503 is specifically configured to determine an interest recognition result of the user according to the real-time user interest score and the similarity score of the interest point; or determining the interest identification result of the user according to the real-time user interest score, the user historical interest score and the similarity score of the interest points.
Referring to fig. 6, a schematic diagram of a device for identifying a user's interest in real time according to an embodiment of the present application is shown.
A real-time user interest recognition apparatus 600, comprising:
a real-time interest score calculating module 601, configured to obtain real-time user behavior data, obtain a feature vector of the real-time user behavior data, and obtain a user real-time interest score by using a time feature, a time decay parameter, an interest relationship weight, and the feature vector of the real-time user behavior data;
a historical interest score calculating module 602, configured to obtain a historical interest score of the user according to the historical behavior data of the user;
the interest point similarity score calculating module 603 is configured to obtain an interest point vector of the user according to the historical behavior data of the user, and obtain a similarity score of the interest point according to the interest point vector of the user;
an interest recognition result determining module 604, configured to determine an interest recognition result of the user according to the real-time user interest score, the user historical interest score, and the similarity score of the interest point.
Referring to fig. 7, a block diagram of a device for data communication according to another embodiment of the present application is shown. The method comprises the following steps: at least one processor 701 (e.g., CPU), a memory 702, and at least one communication bus 703 for enabling communications among the devices. The processor 701 is adapted to execute executable modules, such as computer programs, stored in the memory 702. The Memory 702 may comprise a high-speed Random Access Memory (RAM) and may also include a non-volatile Memory (non-volatile Memory), such as at least one disk Memory. One or more programs are stored in the memory and configured to be executed by the one or more processors 701, including instructions for:
acquiring real-time user behavior data and acquiring a feature vector of the real-time user behavior data;
obtaining a user real-time interest score by utilizing the time characteristics, the time attenuation parameters, the interest relation weight and the characteristic vector of the real-time user behavior data;
and determining an interest identification result of the user according to the real-time user interest score.
In some embodiments, processor 701 is specifically configured to execute the one or more programs including instructions for: and obtaining the real-time interest score of the user by utilizing the difference value between the occurrence time of the real-time user behavior and the current time, the time attenuation parameter, the interest relation weight and the feature vector of the real-time user behavior data.
In some embodiments, processor 701 is specifically configured to execute the one or more programs including instructions for: and obtaining the real-time interest score of the user by utilizing the time characteristic of the user behavior data, the time attenuation parameter, the parameter matrix of the hierarchical interest point network and the characteristic vector of the real-time user behavior data.
In some embodiments, processor 701 is specifically configured to execute the one or more programs including instructions for: obtaining a user historical interest score according to the user historical behavior data; and determining an interest identification result of the user according to the real-time user interest score and the historical user interest score.
In some embodiments, processor 701 is specifically configured to execute the one or more programs including instructions for: obtaining an interest point vector of a user according to historical behavior data of the user; obtaining a similarity score of the interest points according to the interest point vectors of the users; determining an interest identification result of the user according to the real-time user interest score and the similarity score of the interest points; or determining the interest identification result of the user according to the real-time user interest score, the user historical interest score and the similarity score of the interest points.
Referring to fig. 8, a block diagram of a device for data communication according to another embodiment of the present application is shown. The method comprises the following steps: at least one processor 801 (e.g., CPU), memory 802, and at least one communication bus 803 for enabling communications among the devices. The processor 801 is used to execute executable modules, such as computer programs, stored in the memory 802. The Memory 802 may comprise a high-speed Random Access Memory (RAM) and may further comprise a non-volatile Memory (non-volatile Memory), such as at least one disk Memory. One or more programs are stored in the memory and configured to be executed by the one or more processors 801 include instructions for: acquiring real-time user behavior data, acquiring a feature vector of the real-time user behavior data, and acquiring a user real-time interest score by using a time feature, a time attenuation parameter, an interest relation weight and the feature vector of the real-time user behavior data; obtaining a user historical interest score according to the user historical behavior data; obtaining interest point vectors of users according to historical behavior data of the users, and obtaining similarity scores of the interest points according to the interest point vectors of the users; and determining an interest identification result of the user according to the real-time user interest score, the historical user interest score and the similarity score of the interest points.
The arrangement of each unit or module of the apparatus of the present application can be implemented by referring to the methods shown in fig. 2 to 4, which are not described herein again.
Referring to fig. 9, a schematic diagram of a real-time computing server according to an embodiment of the present application is provided. The real-time computing server 900 may include a real-time user behavior acquisition apparatus 901 and a real-time computing node 902. The real-time user behavior obtaining device 901 is configured to obtain real-time user behavior data. The real-time computing node 902 is configured to obtain a feature vector of the real-time user behavior data, and obtain a user real-time interest score by using a time feature, a time decay parameter, an interest relationship weight of the real-time user behavior data and the feature vector of the real-time user behavior data; determining an interest identification result of the user according to the real-time interest score of the user; or determining an interest identification result of the user according to the real-time interest score and the historical interest score of the user; or determining an interest identification result of the user according to the real-time interest score of the user and the similarity score of the interest points; or determining an interest identification result of the user according to the real-time user interest score, the historical user interest score and the similarity score of the interest points; and calculating the user historical interest score and/or the similarity score of the interest points by an offline calculation server.
Referring to fig. 10, a schematic diagram of an offline computing server according to an embodiment of the present application is provided. The offline calculation server 1000 may include a user history interest score calculation node 1001, a similarity score calculation node 1002 for points of interest, and an output device 1003. The user history interest score calculating node 1001 is configured to obtain a user history interest score according to the user history behavior data. The interest point similarity score calculation node 1002 is configured to obtain an interest point vector of a user according to historical behavior data of the user, and obtain a similarity score of the interest point according to the interest point vector of the user. The output device 1003 is configured to output the user historical interest score and/or the similarity score of the interest points to a real-time computing server to obtain a user interest identification result.
Referring to fig. 11, a schematic diagram of an interaction apparatus according to an embodiment of the present application is provided.
An interaction device 1200 comprises:
an input module 1201, configured to input real-time user behavior data and user historical behavior data;
a transmission module 1202, configured to transmit the real-time user behavior data input by the input module to a real-time computing server, and transmit the user historical behavior data input by the input module to an offline computing server;
an output module 1203, configured to output a user interest identification result; and the user interest identification result is calculated by the real-time calculation server.
It should be noted that, in this embodiment, the interaction device 1200 is used for interacting with a user using the interaction device 1200, the input data is real-time user behavior data and user historical behavior data, and the output result is a user interest identification result. The interactive apparatus 1200 may further be connected to a real-time computing server and an offline computing server, and configured to transmit the real-time user behavior data input by the input module to the real-time computing server and transmit the historical user behavior data input by the input module to the offline computing server, respectively. The real-time computing server may be specifically configured as shown in fig. 9 and may implement the same functions. Correspondingly, the offline computation server may specifically have a structure shown in fig. 10 and implement the same functions.
Other embodiments of the present application will be apparent to those skilled in the art from consideration of the specification and practice of the invention disclosed herein. This application is intended to cover any variations, uses, or adaptations of the invention following, in general, the principles of the application and including such departures from the present disclosure as come within known or customary practice in the art to which the invention pertains. It is intended that the specification and examples be considered as exemplary only, with a true scope and spirit of the application being indicated by the following claims.
It will be understood that the present application is not limited to the precise arrangements described above and shown in the drawings and that various modifications and changes may be made without departing from the scope thereof. The scope of the application is limited only by the attached claims
The above description is only exemplary of the present application and should not be taken as limiting the present application, as any modification, equivalent replacement, or improvement made within the spirit and principle of the present application should be included in the protection scope of the present application.
It is noted that, herein, relational terms such as first and second, and the like may be used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions. Also, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising an … …" does not exclude the presence of other identical elements in a process, method, article, or apparatus that comprises the element. The application may be described in the general context of computer-executable instructions, such as program modules, being executed by a computer. Generally, program modules include routines, programs, objects, components, data structures, etc. that perform particular tasks or implement particular abstract data types. The application may also be practiced in distributed computing environments where tasks are performed by remote processing devices that are linked through a communications network. In a distributed computing environment, program modules may be located in both local and remote computer storage media including memory storage devices.
The embodiments in the present specification are described in a progressive manner, and the same and similar parts among the embodiments are referred to each other, and each embodiment focuses on the differences from the other embodiments. In particular, for the apparatus embodiment, since it is substantially similar to the method embodiment, it is relatively simple to describe, and reference may be made to some descriptions of the method embodiment for relevant points. The above-described embodiments of the apparatus are merely illustrative, and the units described as separate parts may or may not be physically separate, and parts displayed as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of the present embodiment. One of ordinary skill in the art can understand and implement it without inventive effort. The foregoing is directed to embodiments of the present application and it is noted that numerous modifications and adaptations may be made by those skilled in the art without departing from the principles of the present application and are intended to be within the scope of the present application.
Claims (16)
1. A real-time user interest identification system is characterized by comprising a real-time calculation server, an off-line calculation server and a result output device, wherein:
the real-time computing server is used for acquiring real-time user behavior data, acquiring a feature vector of the real-time user behavior data, and acquiring a user real-time interest score by utilizing the time feature, the time attenuation parameter, the interest relation weight and the feature vector of the real-time user behavior data;
the off-line computing server is used for acquiring a user historical interest score according to the user historical behavior data; and/or obtaining interest point vectors of the users according to the historical behavior data of the users, and obtaining similarity scores of the interest points according to the interest point vectors of the users;
the real-time computing server is also used for determining an interest identification result of the user according to the real-time interest score of the user; or determining an interest identification result of the user according to the real-time interest score and the historical interest score of the user; or determining an interest identification result of the user according to the real-time interest score of the user and the similarity score of the interest points; or determining an interest identification result of the user according to the real-time user interest score, the user historical interest score and the similarity score of the interest points;
and the result output device is used for outputting the interest identification result of the user.
2. A real-time computing server, comprising:
the real-time user behavior acquisition device is used for acquiring real-time user behavior data;
the real-time computing node is used for obtaining the feature vector of the real-time user behavior data and obtaining a user real-time interest score by utilizing the time feature, the time attenuation parameter, the interest relation weight and the feature vector of the real-time user behavior data; determining an interest identification result of the user according to the real-time interest score of the user; or determining an interest identification result of the user according to the real-time interest score and the historical interest score of the user; or determining an interest identification result of the user according to the real-time interest score of the user and the similarity score of the interest points; or determining an interest identification result of the user according to the real-time user interest score, the historical user interest score and the similarity score of the interest points; and calculating the user historical interest score and/or the similarity score of the interest points by an offline calculation server.
3. An offline computing server, comprising:
the user historical interest score calculation node is used for acquiring a user historical interest score according to the user historical behavior data;
the similarity score calculation node of the interest points is used for acquiring the interest point vectors of the users according to the historical behavior data of the users and obtaining the similarity scores of the interest points according to the interest point vectors of the users;
and the output device is used for outputting the historical interest scores of the users and/or the similarity scores of the interest points to a real-time computing server so as to obtain the user interest identification results.
4. An interactive apparatus, comprising:
the input module is used for inputting real-time user behavior data and historical user behavior data;
the transmission module is used for transmitting the real-time user behavior data input by the input module to a real-time calculation server and transmitting the historical user behavior data input by the input module to an offline calculation server;
the output module is used for outputting a user interest identification result; and the user interest identification result is calculated by the real-time calculation server.
5. A real-time user interest identification method is applied to a real-time computing server and comprises the following steps:
acquiring real-time user behavior data;
obtaining a feature vector of the real-time user behavior data, and obtaining a user real-time interest score by utilizing the time feature, the time attenuation parameter, the interest relation weight and the feature vector of the real-time user behavior data;
determining an interest identification result of the user according to the real-time interest score of the user; or determining an interest identification result of the user according to the real-time interest score and the historical interest score of the user; or determining an interest identification result of the user according to the real-time interest score of the user and the similarity score of the interest points; or determining an interest identification result of the user according to the real-time user interest score, the historical user interest score and the similarity score of the interest points; and calculating the user historical interest score and/or the similarity score of the interest points by an offline calculation server.
6. A user interest identification method is applied to an offline computing server and comprises the following steps:
obtaining a user historical interest score according to the user historical behavior data;
obtaining interest point vectors of users according to historical behavior data of the users, and obtaining similarity scores of the interest points according to the interest point vectors of the users;
and outputting the historical interest scores and/or the similarity scores of the interest points of the users to a real-time computing server to obtain the user interest identification results.
7. A real-time user interest identification method is characterized by comprising the following steps:
acquiring real-time user behavior data and acquiring a feature vector of the real-time user behavior data;
obtaining a user real-time interest score by utilizing the time characteristics, the time attenuation parameters, the interest relation weight and the characteristic vector of the real-time user behavior data;
and determining an interest identification result of the user according to the real-time user interest score.
8. The method of claim 7, wherein the temporal characteristic of the real-time user behavior data is a difference between a time of occurrence of the real-time user behavior and a current time.
9. The method according to claim 7, wherein the interest relation weight is a parameter matrix of a hierarchical interest point network.
10. The method of claim 7, further comprising:
obtaining a user historical interest score according to the user historical behavior data;
the determining the interest recognition result of the user according to the real-time user interest score comprises:
and determining an interest identification result of the user according to the real-time user interest score and the historical user interest score.
11. The method according to claim 7 or 10, characterized in that the method further comprises:
obtaining an interest point vector of a user according to historical behavior data of the user; obtaining a similarity score of the interest points according to the interest point vectors of the users;
the determining the interest recognition result of the user according to the real-time user interest score comprises:
determining an interest identification result of the user according to the real-time user interest score and the similarity score of the interest points; or,
and determining an interest identification result of the user according to the real-time user interest score, the historical user interest score and the similarity score of the interest points.
12. A real-time user interest identification method is characterized by comprising the following steps:
acquiring real-time user behavior data, acquiring a feature vector of the real-time user behavior data, and acquiring a user real-time interest score by using a time feature, a time attenuation parameter, an interest relation weight and the feature vector of the real-time user behavior data;
obtaining a user historical interest score according to the user historical behavior data;
obtaining interest point vectors of users according to historical behavior data of the users, and obtaining similarity scores of the interest points according to the interest point vectors of the users;
and determining an interest identification result of the user according to the real-time user interest score, the historical user interest score and the similarity score of the interest points.
13. A real-time user interest recognition apparatus, comprising:
the real-time data acquisition unit is used for acquiring real-time user behavior data and acquiring a feature vector of the real-time user behavior data;
the real-time interest score calculating unit is used for obtaining a user real-time interest score by utilizing the time characteristics, the time attenuation parameters, the interest relation weight and the characteristic vector of the real-time user behavior data;
and the interest identification result determining unit is used for determining the interest identification result of the user according to the real-time user interest score.
14. A real-time user interest recognition apparatus, comprising:
the real-time interest score calculation module is used for acquiring real-time user behavior data, acquiring a feature vector of the real-time user behavior data, and acquiring a user real-time interest score by utilizing the time feature, the time attenuation parameter, the interest relation weight and the feature vector of the real-time user behavior data;
the historical interest score calculating module is used for acquiring the historical interest score of the user according to the historical behavior data of the user;
the interest point similarity score calculating module is used for acquiring an interest point vector of the user according to the historical behavior data of the user and obtaining the similarity score of the interest point according to the interest point vector of the user;
and the interest identification result determining module is used for determining the interest identification result of the user according to the real-time user interest score, the historical interest score of the user and the similarity score of the interest points.
15. An apparatus for real-time user interest identification, comprising a memory, and one or more programs, wherein the one or more programs are stored in the memory and configured to be executed by the one or more processors, the one or more programs comprising instructions for:
acquiring real-time user behavior data and acquiring a feature vector of the real-time user behavior data;
obtaining a user real-time interest score by utilizing the time characteristics, the time attenuation parameters, the interest relation weight and the characteristic vector of the real-time user behavior data;
and determining an interest identification result of the user according to the real-time user interest score.
16. An apparatus for real-time user interest identification, comprising a memory, and one or more programs, wherein the one or more programs are stored in the memory and configured to be executed by the one or more processors, the one or more programs comprising instructions for:
acquiring real-time user behavior data, acquiring a feature vector of the real-time user behavior data, and acquiring a user real-time interest score by using a time feature, a time attenuation parameter, an interest relation weight and the feature vector of the real-time user behavior data;
obtaining a user historical interest score according to the user historical behavior data;
obtaining interest point vectors of users according to historical behavior data of the users, and obtaining similarity scores of the interest points according to the interest point vectors of the users;
and determining an interest identification result of the user according to the real-time user interest score, the historical user interest score and the similarity score of the interest points.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710146012.9A CN108572984A (en) | 2017-03-13 | 2017-03-13 | A kind of active user interest recognition methods and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710146012.9A CN108572984A (en) | 2017-03-13 | 2017-03-13 | A kind of active user interest recognition methods and device |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN108572984A true CN108572984A (en) | 2018-09-25 |
Family
ID=63578424
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710146012.9A Pending CN108572984A (en) | 2017-03-13 | 2017-03-13 | A kind of active user interest recognition methods and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108572984A (en) |
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109460514A (en) * | 2018-11-02 | 2019-03-12 | 北京京东尚科信息技术有限公司 | Method and apparatus for pushed information |
| CN109685561A (en) * | 2018-12-17 | 2019-04-26 | 北京字节跳动网络技术有限公司 | Electronic certificate method for pushing, device and electronic equipment based on user behavior |
| CN110119762A (en) * | 2019-04-15 | 2019-08-13 | 华东师范大学 | Human behavior dependency analysis method based on cluster |
| CN110738047A (en) * | 2019-09-03 | 2020-01-31 | 华中科技大学 | Microblog user interest mining method and system based on graphic data and time effect |
| CN110909627A (en) * | 2019-11-04 | 2020-03-24 | 中国科学院深圳先进技术研究院 | Region POI configuration visualization method and system |
| CN113159184A (en) * | 2021-04-23 | 2021-07-23 | 武汉陌青春创意科技有限公司 | User interest feature identification method and device |
| CN114416246A (en) * | 2021-12-31 | 2022-04-29 | 北京五八信息技术有限公司 | Data processing method and device, electronic equipment and storage medium |
| CN114742569A (en) * | 2021-01-08 | 2022-07-12 | 广州视源电子科技股份有限公司 | User life stage prediction method and device, computer equipment and storage medium |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20100023506A1 (en) * | 2008-07-22 | 2010-01-28 | Yahoo! Inc. | Augmenting online content with additional content relevant to user interests |
| CN101694652B (en) * | 2009-09-30 | 2012-11-28 | 西安交通大学 | Network resource personalized recommended method based on ultrafast neural network |
| CN104572734A (en) * | 2013-10-23 | 2015-04-29 | 腾讯科技（深圳）有限公司 | Question recommendation method, device and system |
| CN104679743A (en) * | 2013-11-26 | 2015-06-03 | 阿里巴巴集团控股有限公司 | Method and device for determining preference model of user |
| CN105574216A (en) * | 2016-03-07 | 2016-05-11 | 达而观信息科技（上海）有限公司 | Personalized recommendation method and system based on probability model and user behavior analysis |
-
2017
- 2017-03-13 CN CN201710146012.9A patent/CN108572984A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20100023506A1 (en) * | 2008-07-22 | 2010-01-28 | Yahoo! Inc. | Augmenting online content with additional content relevant to user interests |
| CN101694652B (en) * | 2009-09-30 | 2012-11-28 | 西安交通大学 | Network resource personalized recommended method based on ultrafast neural network |
| CN104572734A (en) * | 2013-10-23 | 2015-04-29 | 腾讯科技（深圳）有限公司 | Question recommendation method, device and system |
| CN104679743A (en) * | 2013-11-26 | 2015-06-03 | 阿里巴巴集团控股有限公司 | Method and device for determining preference model of user |
| CN105574216A (en) * | 2016-03-07 | 2016-05-11 | 达而观信息科技（上海）有限公司 | Personalized recommendation method and system based on probability model and user behavior analysis |
Non-Patent Citations (1)
| Title |
|---|
| 任星怡 等: "基于用户签到行为的兴趣点推荐", 《计算机学报》 * |
Cited By (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109460514A (en) * | 2018-11-02 | 2019-03-12 | 北京京东尚科信息技术有限公司 | Method and apparatus for pushed information |
| CN109460514B (en) * | 2018-11-02 | 2024-06-14 | 北京京东尚科信息技术有限公司 | Method and device for pushing information |
| CN109685561A (en) * | 2018-12-17 | 2019-04-26 | 北京字节跳动网络技术有限公司 | Electronic certificate method for pushing, device and electronic equipment based on user behavior |
| CN110119762A (en) * | 2019-04-15 | 2019-08-13 | 华东师范大学 | Human behavior dependency analysis method based on cluster |
| CN110119762B (en) * | 2019-04-15 | 2023-09-26 | 华东师范大学 | Clustering-based human behavior dependency analysis method |
| CN110738047A (en) * | 2019-09-03 | 2020-01-31 | 华中科技大学 | Microblog user interest mining method and system based on graphic data and time effect |
| CN110738047B (en) * | 2019-09-03 | 2021-04-20 | 华中科技大学 | Microblog user interest mining method and system based on graphic data and time effect |
| CN110909627A (en) * | 2019-11-04 | 2020-03-24 | 中国科学院深圳先进技术研究院 | Region POI configuration visualization method and system |
| CN110909627B (en) * | 2019-11-04 | 2022-04-26 | 中国科学院深圳先进技术研究院 | Regional POI configuration visualization method and system |
| CN114742569A (en) * | 2021-01-08 | 2022-07-12 | 广州视源电子科技股份有限公司 | User life stage prediction method and device, computer equipment and storage medium |
| CN113159184A (en) * | 2021-04-23 | 2021-07-23 | 武汉陌青春创意科技有限公司 | User interest feature identification method and device |
| CN114416246A (en) * | 2021-12-31 | 2022-04-29 | 北京五八信息技术有限公司 | Data processing method and device, electronic equipment and storage medium |
| CN114416246B (en) * | 2021-12-31 | 2024-03-19 | 北京五八信息技术有限公司 | Data processing method and device, electronic equipment and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN113626719B (en) | Information recommendation method, device, equipment, storage medium and computer program product | |
| CN111784455B (en) | Article recommendation method and recommendation equipment | |
| Ghose et al. | Modeling consumer footprints on search engines: An interplay with social media | |
| CN109191240B (en) | Method and device for recommending commodities | |
| CN108572984A (en) | A kind of active user interest recognition methods and device | |
| CN111400613A (en) | Article recommendation method, device, medium and computer equipment | |
| CN106251174A (en) | Information recommendation method and device | |
| CN104090919A (en) | Advertisement recommending method and advertisement recommending server | |
| CN105469263A (en) | Commodity recommendation method and device | |
| CN110008397B (en) | Recommendation model training method and device | |
| CN111626832A (en) | Product recommendation method and device and computer equipment | |
| US20190080352A1 (en) | Segment Extension Based on Lookalike Selection | |
| WO2020221022A1 (en) | Service object recommendation method | |
| CN114511387A (en) | Product recommendation method and device, electronic equipment and storage medium | |
| CN118043802A (en) | Recommendation model training method and device | |
| CN111680213A (en) | Information recommendation method, data processing method and device | |
| CN117455621A (en) | Personalized recommendation method and device, storage medium and computer equipment | |
| US20230316106A1 (en) | Method and apparatus for training content recommendation model, device, and storage medium | |
| CN115982463A (en) | Resource recommendation method, device, equipment and storage medium | |
| CN111178987B (en) | Method and device for training user behavior prediction model | |
| Cao et al. | Distributed design and implementation of SVD++ algorithm for e-commerce personalized recommender system | |
| Diwandari et al. | Research Methodology for Analysis of E-Commerce User Activity Based on User Interest using Web Usage Mining. | |
| KR102546328B1 (en) | Method, device and system for providing content information monitoring and content planning automation solution for online marketing | |
| CN116975426A (en) | Service data processing method, device, equipment and medium | |
| TWM573493U (en) | System for predicting conversion probability by visitors' browsing paths |
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 | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20180925 |
|
| RJ01 | Rejection of invention patent application after publication |