CN110879865A - Recommendation method and device for nuclear products - Google Patents
Recommendation method and device for nuclear products Download PDFInfo
- Publication number
- CN110879865A CN110879865A CN201911052966.9A CN201911052966A CN110879865A CN 110879865 A CN110879865 A CN 110879865A CN 201911052966 A CN201911052966 A CN 201911052966A CN 110879865 A CN110879865 A CN 110879865A
- Authority
- CN
- China
- Prior art keywords
- user
- recommendation
- core
- product
- value
- 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
- 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
- G06F16/9538—Presentation of query results
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F21/00—Security arrangements for protecting computers, components thereof, programs or data against unauthorised activity
- G06F21/30—Authentication, i.e. establishing the identity or authorisation of security principals
- G06F21/31—User authentication
-
- 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
- G06N3/08—Learning methods
Abstract
The embodiment of the specification discloses a recommendation method and device for a nuclear product. One embodiment of the present description discloses a method comprising: respectively inputting the attribute characteristics of the user into the first recommendation models before and after iteration to determine first recommendation values and second recommendation values of the multiple core body products for the user, determining a final recommendation value according to the first recommendation value and the second recommendation value, and recommending at least one core body product to the user according to the final recommendation value; the first recommendation model is a reinforcement learning model.
Description
Technical Field
The present disclosure relates to the field of identity verification, and more particularly, to a method and apparatus for recommending a personal identification number.
Background
With the development of intelligent terminals and the development of network applications, users can access various network applications through various application clients installed on the terminals, for example, social instant messaging applications, shopping applications, and the like. The body-core product is an interactive product for verifying the identity of a user. During the process of accessing the network application, the identity of the user may need to be verified by the core product, so as to allow the user to use the corresponding function after the identity verification is passed. For example, for shopping applications, a core product is used to verify the identity of the user when paying for a purchase. For example, resetting the user's login password requires verifying the user's identity using the authentication product.
The verification mode of the nuclear product can be various, such as fingerprint identification, password verification, face identification, iris identification, short message dynamic code and question-answer verification. Under the condition that various core products are available for selection, a core product recommendation scheme is needed to be provided, so that a server can recommend the core products to a user, and the user experience is improved.
Disclosure of Invention
The embodiments disclosed herein provide recommendations for core products.
According to a first aspect disclosed in the present specification, there is provided a recommendation method for a nuclear product, comprising the steps of:
acquiring attribute characteristics of a user, wherein the attribute characteristics of the user at least comprise characteristics generated based on a historical record of the user using a nuclear product;
inputting attribute characteristics of a user into a first recommendation model before iteration to determine first recommendation values of a plurality of core products for the user;
inputting the attribute characteristics of the user into the iterated first recommendation model to determine second recommendation values of the plurality of nuclear products for the user;
determining final recommended values of a plurality of core products for the user, including: determining a final recommended value of the core product for the user according to the first recommended value and the second recommended value of the core product for the user;
recommending at least one core product to the user according to the final recommended values of the plurality of core products for the user;
the first recommendation model is a reinforcement learning model and is configured to output a recommendation value reflecting the probability of the user using the core product according to the attribute characteristics of the user and iterate according to a preset iteration cycle based on the reinforcement learning attribute of the first recommendation model.
Optionally or preferably, the first recommendation model is iterated, comprising:
obtaining feedback data, wherein the feedback data comprises real records of using the core body products after a plurality of users receive the core body product recommendation;
generating reward values according to the core product recommendations and the feedback data received by the users;
iterating based on the feedback data and the reward value.
Optionally or preferably, the method further comprises:
selecting part of users as target users;
and adjusting the final recommended value of one or more core products to the target user so as to recommend the core products which are not originally recommended to the target user.
Optionally or preferably, the method further comprises:
selecting part of users as target users;
and setting a final recommendation value of the new core product to the target user for the new core product so as to recommend the new core product to the target user.
Optionally or preferably, the determining a final recommended value of the core product for the user according to the first recommended value and the second recommended value of the core product for the user includes:
and carrying out weighted average on the first recommended value and the second recommended value of the core body product for the user so as to determine the final recommended value of the core body product for the user.
Optionally or preferably, the method further comprises:
inputting the attribute characteristics of the user into a second recommendation model to determine a third recommendation value of a plurality of core products for the user; the second recommendation model is a non-reinforcement learning model;
the determining of the final recommended values of the plurality of core products for the user comprises: and determining a final recommended value of the core product for the user according to the first recommended value, the second recommended value and the third recommended value of the core product for the user.
Optionally or preferably, the determining a final recommended value of the core product for the user according to the first recommended value, the second recommended value and the third recommended value of the core product for the user includes:
and carrying out weighted average on the first recommended value, the second recommended value and the third recommended value of the core body product for the user so as to determine the final recommended value of the core body product for the user.
According to a second aspect disclosed in the present specification, there is provided a recommendation apparatus for a nuclear product, comprising the following modules:
the attribute feature acquisition module is used for acquiring attribute features of a user, wherein the attribute features of the user at least comprise features generated based on a historical record of the user using the nuclear body product;
the first recommendation value determining module is used for inputting the attribute characteristics of the user into a first recommendation model before iteration so as to determine first recommendation values of the plurality of nuclear body products for the user;
the second recommendation value determining module is used for inputting the attribute characteristics of the user into the iterated first recommendation model so as to determine second recommendation values of the plurality of nuclear body products for the user;
the final recommendation value determining module is used for determining the final recommendation values of a plurality of nuclear products for the user, and comprises the following steps: determining a final recommended value of the core product for the user according to the first recommended value and the second recommended value of the core product for the user;
the recommending module is used for recommending at least one core product to the user according to the final recommended values of the core products for the user;
the first recommendation model is a reinforcement learning model and is configured to output a recommendation value reflecting the probability of the user using the core product according to the attribute characteristics of the user and iterate according to a preset iteration cycle based on the reinforcement learning attribute of the first recommendation model.
Optionally or preferably, the first recommendation model is iterated, comprising:
obtaining feedback data, wherein the feedback data comprises real records of using the core body products after a plurality of users receive the core body product recommendation;
generating reward values according to the core product recommendations and the feedback data received by the users;
iterating based on the feedback data and the reward value.
Optionally or preferably, the apparatus further comprises a first final recommendation adjustment module:
the first final recommendation value adjusting module is used for selecting part of users as target users, adjusting the final recommendation values of one or more core products to the target users, and recommending the core products which cannot be originally recommended to the target users.
Optionally or preferably, the apparatus further comprises a second final recommendation adjustment module:
and the second final recommendation value adjusting module is used for selecting part of users as target users, setting the final recommendation value of the new core product to the target users for the new core product, and recommending the new core product to the target users.
Optionally or preferably, the apparatus further comprises a third recommendation value determining module:
the third recommendation value determining module is used for inputting the attribute characteristics of the user into a second recommendation model so as to determine a third recommendation value of a plurality of nuclear products for the user; the second recommendation model is a non-reinforcement learning model;
the determining of the final recommended values of the plurality of core products for the user comprises: and determining a final recommended value of the core product for the user according to the first recommended value, the second recommended value and the third recommended value of the core product for the user.
According to a third aspect of the present disclosure, there is provided a recommendation apparatus for a core product, comprising a processor and a memory, the memory storing a computer program, the computer program implementing the method of any one of the preceding claims when executed by the processor.
Features of embodiments of the present specification and advantages thereof will become apparent from the following detailed description of exemplary embodiments thereof, which proceeds with reference to the accompanying drawings.
Drawings
The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments of the specification and together with the description, serve to explain the principles of the embodiments of the specification.
FIG. 1 is a block diagram of a core recommendation system provided in one embodiment of the present description;
FIG. 2 is a diagram of a reinforcement learning model provided in one embodiment of the present description;
FIG. 3 is a flow diagram of an iterative process of a first recommendation model provided by one embodiment of the present description;
FIG. 4 is a flow chart of a method for recommending a core product according to an embodiment of the present disclosure;
FIG. 5 is a flow chart of a method for recommending a core product according to an embodiment of the present disclosure;
FIG. 6 is a block diagram of a recommendation device for a core product according to one embodiment of the present disclosure;
FIG. 7 is a block diagram of a recommendation device for a core product according to one embodiment of the present disclosure;
FIG. 8 is a block diagram of a recommendation device for a core product provided in one embodiment of the present disclosure;
FIG. 9 is a block diagram of a recommendation device for a core product according to one embodiment of the present disclosure;
fig. 10 is a block diagram of a recommendation device for a core product according to an embodiment of the present disclosure.
Detailed Description
Various exemplary embodiments of the present specification will now be described in detail with reference to the accompanying drawings.
The following description of at least one exemplary embodiment is merely illustrative in nature and is in no way intended to limit the embodiments, their application, or uses.
It should be noted that: like reference numbers and letters refer to like items in the following figures, and thus, once an item is defined in one figure, further discussion thereof is not required in subsequent figures.
< body-building System >
Fig. 1 is a block diagram of a core system provided in an embodiment of the present specification. As shown in fig. 1, the electronic system includes a server 101 providing a core service and a terminal device 103 of a mass of users. The server 101 and the terminal device 103 may be communicatively connected via a network 102.
The configuration of the server 101 may include, but is not limited to: processor 1011, memory 1012, interface 1013, communication device 1014, input device 1015, output device 1016. The processor 1011 may include, but is not limited to, a central processing unit CPU, a microprocessor MCU, or the like. The memory 1012 may include, but is not limited to, a ROM (read only memory), a RAM (random access memory), a nonvolatile memory such as a hard disk, and the like. Interface device 1013 may include, but is not limited to, a USB interface, a serial interface, a parallel interface, and the like. The communication device 1014 is capable of wired or wireless communication, for example, and may specifically include WiFi communication, bluetooth communication, 2G/3G/4G/5G communication, and the like. Input devices 1015 include, but are not limited to, a keyboard, a mouse, a touch screen, a microphone, and the like. Output devices 1016 include, but are not limited to, a display screen, speakers, and the like. The configuration of the server 101 may include only some of the above devices.
The terminal device 103 may be, for example, an electronic device installed with an intelligent operating system (e.g., android, IOS, Windows, Linux, etc.) including, but not limited to, a laptop, a desktop computer, a mobile phone, a tablet computer, etc. Configurations of terminal equipment 103 include, but are not limited to, processor 1031, memory 1032, interface device 1033, communication device 1034, GPU 1035, display device 1036, input device 1037, speaker 1038, microphone 1039, and camera 1030. The processor 1031 includes, but is not limited to, a central processing unit CPU, a microprocessor MCU, and the like. The memory 1032 includes, but is not limited to, a ROM (read only memory), a RAM (random access memory), a nonvolatile memory such as a hard disk, and the like. Interface device 1033 includes, but is not limited to, a USB interface, a serial interface, a parallel interface, and the like. The communication device 1034 is capable of wired or wireless communication, for example, and specifically may include WiFi communication, bluetooth communication, 2G/3G/4G/5G communication, and the like. The GPU 1035 is used to process the image. The display device 1036 includes, but is not limited to, a liquid crystal screen, a touch screen, and the like. Input devices 1037 include, but are not limited to, a keyboard, a mouse, a touch screen, and the like. The configuration of the terminal device 103 may include only some of the above-described apparatuses.
In one embodiment applied to the present description, a user may access various network applications through various application clients installed on the terminal device 103, and the network applications may be, for example, social instant messaging applications, shopping applications, and the like. In the process of accessing the network application, the identity of the user may need to be verified by using the core product, in this case, the server 101 may send a core product recommendation to the terminal device 103 of the user, and the user may select a suitable core product for authentication after receiving the recommendation. During the authentication process, the user may interact with the server 101 through the terminal device 103. For example, the user inputs a password and a fingerprint on the terminal device 103 to provide the password and the fingerprint to the server 101 for authentication; after the server 101 performs authentication, the result of successful or failed authentication is fed back to the user, for example, a "password error" (corresponding to the case of failed authentication) or a "payment success" (corresponding to the case of successful authentication) is fed back to the user.
In some embodiments, the web application itself has a core functionality. For example, some instant messaging applications, shopping applications and payment applications require a user to log in an account for use, and the server of these network applications also serves as a server for providing the core service.
In some embodiments, the server 101 may be a server dedicated to providing the core services for a variety of web applications, for example, for a variety of different shopping applications, car calling applications.
The core system shown in fig. 1 is merely illustrative and is in no way intended to limit the embodiments of the present description, their application, or uses. It should be understood by those skilled in the art that although a plurality of devices of the server and the terminal equipment are described in the foregoing, the embodiments of the present specification may refer to only some of the devices. For example, the server may relate only to the processor, the memory and the communication means, and the terminal device may relate only to the processor, the memory, the communication means, the display screen and the speaker. Those skilled in the art can design instructions based on the disclosed embodiments of the present specification. How the instructions control the operation of the processor is well known in the art and will not be described in detail herein.
< method for recommending personal care products >
First, referring to fig. 2 and 3, a first recommendation model used in a recommendation method for a core product according to an embodiment of the present disclosure will be described.
In one embodiment, the first recommendation model is a reinforcement learning model, and is configured to output a recommendation value of the core product according to the attribute characteristics of the user and iterate according to a preset iteration cycle based on the reinforcement learning attribute of the user.
For any user, the attribute features at least include features generated based on the user's history of use of the core product. In one embodiment, the attribute features of the user may also include other features, for example, the attribute features of the user may also include: user information including gender, age, city, registration duration, card binding, grade, academic calendar, occupation, etc.; user preferences including network preferences, wireless preferences, etc.; user equipment including brand, version information, system type, etc.; user scenarios include home, public, company, hotel, taxi, code scanning, account transfer, etc. The user attribute features mentioned in the present embodiment are also applicable to the second recommendation model referred to later.
In one embodiment, when the user needs to perform authentication, the server 101 providing the core product service may recommend the core product to the user using the first recommendation model. Specifically, the server 101 obtains the attribute characteristics of the user, inputs the attribute characteristics of the user into a first recommendation model, and the first recommendation model outputs a recommendation value of the core product for the user. For any one of the nuclear products, the recommendation value of the nuclear product to the user reflects the probability of using the nuclear product by the user. In a specific example, the first recommendation model may be represented as y ═ f (a, x), x is an attribute feature of the user, a represents a certain core product, y represents a recommendation value of the core product a for the user, and f represents a recommendation policy of the first recommendation model, that is, recommendation values of a plurality of core products for the user may be determined by inputting the attribute feature of the user a plurality of times and setting a different core product each time. In another example, a may represent a set including a plurality of core products, f may represent a recommendation policy of the first recommendation model, and y may represent a set of recommendation values of the plurality of core products for the user.
The following describes the process of training to obtain the first recommendation model by using a large amount of sample data: acquiring a historical record of the core product used by the user and other information of the user, and obtaining a sample of the attribute characteristics of the user according to the historical record of the core product used by the user and the other information of the user; counting the times of using each nuclear product by a user and the total times of using all the nuclear products by the user according to the historical record of using the nuclear products by the user, and taking the ratio of the times of using a certain nuclear product by the user to the total times of using all the nuclear products by the user as a sample of the recommended value of the nuclear product to the user; and training the basic model by using a large number of samples of the attribute characteristics of the user and samples corresponding to the recommended values of the user to obtain a first recommended model.
The specific architecture and recommendation strategy of the first recommendation model and the specific process of obtaining the first recommendation model through training of a large amount of sample data can be set according to actual conditions, and are not described too much here.
In this embodiment, the first recommendation model is a reinforcement learning model, and may be a Markov Decision Process (MDP) model, for example.
Fig. 2 is a schematic diagram of a reinforcement learning model, which can be generally understood as a main body of interaction between an agent and an environment. An Agent is a main body for making a decision, and an environment (environment) is a main body for information feedback. Reinforcement learning refers to the fact that the agent 21 can sense the state (state) of the environment 22 and reward (reward) fed back by the environment 22, and learn and make decisions based on the sensed state and reward. That is, agent 21 has dual functions of learning and decision making. The decision function of the agent 21 means that the agent 21 can perform different actions (actions) according to the state (state) and policy (policy) of the environment 22. The learning function of the agent 21 means that the agent 21 can sense the state (state) of the external environment 22 and reward (reward) for feedback, and learn and improve the policy (policy) based on the sensed state (state) and reward (reward). The reinforcement learning embodies the process of continuous interaction of the intelligent agent and the environment, and the three elements of the state, the action and the reward are the key of the reinforcement learning.
In the first recommendation model provided in this embodiment, the agent makes the recommendation of the nuclear product (corresponding to "action") according to the attribute characteristics (corresponding to "state") of the user and the policy (corresponding to "policy"). The environment awards (corresponding to "reward") according to the core product recommendation of the agent and the use condition of the core product after the user receives the core product recommendation. The environment feeds back the condition and reward of using the core product after the user receives the recommendation of the core product to the intelligent agent, so that the intelligent agent can sense the change of the state of the environment and the reward fed back by the environment, the strategy is improved, and the iteration of the first recommendation model is realized. That is, the first recommendation model automatically iterates itself based on the reinforcement learning properties of the first recommendation model itself.
Referring to fig. 3, an iterative process of the first recommendation model provided in this embodiment is described, which includes the following steps:
302. and obtaining feedback data, wherein the feedback data comprises real records of using the core body products after the plurality of users receive the core body product recommendation.
304. And generating reward values according to the core product recommendations and the feedback data received by the users.
A specific example is used for explaining a generation mechanism of the reward value, and if the user uses the recommended core product to perform core and the core succeeds, the reward value is a larger positive value; if the user only attempts the recommended core product, last used or other core products, the reward value is a small positive value; if the user does not try the recommended core product and directly selects other core products, the reward value is 0; if the user uses the recommended core product but the user identity is stolen, the reward value is a negative value. This is merely an illustrative simple example, and in other examples, the mechanism for generating the prize value may be set according to the actual situation.
306. Iterating based on the feedback data and the reward value.
The first recommendation model makes the core product recommendation for the user according to the recommendation strategy based on the condition that the core product is used by the user in the past and other information of the user. When the first recommendation model performs self-iteration according to a preset period, a reward value is generated based on the condition of the actual core product of the user after recommendation, and a recommendation strategy is updated by using the reward value and the condition of the actual core product used by the user. The user receives the core action performed after the core product recommendation, and the recommendation strategy is influenced, so that the subsequent core product recommendation is influenced.
< example one >
Referring to fig. 4, a method for recommending a core product according to an embodiment of the present disclosure is described, including:
Referring to the foregoing, the first recommendation model is a reinforcement learning model, and is configured to output a recommendation value of the core product according to the attribute feature of the user and iterate according to a preset iteration cycle based on the reinforcement learning attribute of the user. In this example, the iteration cycle is one day, that is, the first recommendation model is updated and iterated once a day, and the specific iteration mode may be as in the foregoing step 302-306, and iteration is performed by obtaining the core product recommendation provided to the user on the day and the actual situation of using the core product after the user receives the core product recommendation.
In step 204, attribute characteristics of the user may be input into the first recommendation model after the latest iteration and the first recommendation model before the iteration, so as to obtain second recommendation values and first recommendation values of the plurality of core products for the user, respectively.
And step 206, determining a final recommended value of the core product for the user according to the first recommended value and the second recommended value of the core product for the user.
In a first specific example, the first recommended value and the second recommended value of the core product for the user may be summed, and the sum may be used as a final recommended value of the core product for the user.
In a second specific example, the first recommended value and the second recommended value of the core product for the user may be weighted and averaged, and the weighted average may be used as the final recommended value of the core product for the user. In a specific example, a higher weight may be set for the first recommendation value and a lower weight may be set for the second recommendation value, i.e. the weight of the recommendation value output by the first recommendation model before the iteration is higher.
For example, there are 4 recommended core body products, including a password core body product, a short message dynamic code core body product, a face recognition core body product, and a question and answer verification core body product, and the corresponding first recommended value and second recommended value may be referred to in the following table.
In a first specific example, the final recommended value of the password core product is 5.0+4.0, the final recommended value of the short message dynamic code core product is 4.0+3.8, the final recommended value of the face recognition core product is 2+2.5, and the final recommended value of the question and answer verification core product is 1.5+ 1.0.
In a second specific example, if the weight of the first recommended value is set to 2 and the weight of the second recommended value is set to 1, the final recommended value of the password core product is (5.0 × 2+4.0)/3, the final recommended value of the short message dynamic code core product is (4.0 × 2+3.8)/3, the final recommended value of the face recognition core product is (2.0 × 2+2.5)/3, and the final recommended value of the question and answer verification core product is (1.5 × 2+ 1.0)/3.
In other examples, the final recommended value may be calculated in other manners, which are not limited herein.
And 208, recommending at least one core product to the user according to the final recommended values of the plurality of core products for the user.
In a specific example, the final recommendation values of the core products for the user may be sorted from high to low, and one or more core products with the highest final recommendation value may be recommended to the user. If a plurality of core products are recommended to the user, the core products can be presented to the user in a form of a list according to the final recommended value from high to low.
For example, the final recommended values are sorted from high to low, and the password body checking product, the short message dynamic code body checking product, the face recognition body checking product and the question and answer checking body checking product are sequentially arranged. In one example, a password authentication product may be recommended to the user. In another example, the password authentication product and the short message dynamic code authentication product can be recommended to the user together in a list form.
The recommendation method for the core body product provided by one embodiment of the specification can provide personalized core body product recommendation for a user, the recommendation effect is more in line with individual requirements and preferences of the user, and the recommendation effect of 'thousands of people and thousands of faces' is achieved.
The recommendation method of the nuclear product provided by one embodiment of the specification has the functions of reinforcement learning and self iteration.
The recommendation method for the nuclear product provided by one embodiment of the specification is more stable and reliable.
< example two >
Referring to fig. 5, a method for recommending a core product according to an embodiment of the present disclosure is described, including:
And step 406, determining a final recommended value of the core product for the user according to the first recommended value, the second recommended value and the third recommended value of the core product for the user.
In a first specific example, the first recommended value, the second recommended value and the third recommended value of the core product for the user may be summed, and the sum may be used as a final recommended value of the core product for the user.
In a second specific example, the first recommended value, the second recommended value and the third recommended value of the core product for the user may be weighted and averaged, and the weighted average may be used as the final recommended value of the core product for the user. In a specific example, a higher weight may be set for the first recommendation value and a lower weight may be set for the second recommendation value, i.e. the weight of the recommendation value output by the first recommendation model before the iteration is higher than the weight of the recommendation value output by the first recommendation model after the iteration.
In other examples, the final recommended value may be calculated in other manners, which are not limited herein.
And step 408, recommending at least one core product to the user according to the final recommended values of the plurality of core products for the user.
Step 402 and step 408 in example two are similar to step 202 and step 208 in example one, respectively, and are not described again here.
The main difference between example two and example one is that the recommendation method of example two also uses a second recommendation learning model. In a specific example, the second recommended model is a non-reinforcement learning model, and may be set not to be automatically updated iteratively. Or, in a specific example, the second recommendation model is a non-reinforcement learning model, and the iteration cycle is much higher than that of the first recommendation model; for example, the iteration cycle of the first recommendation model is one day, namely, the first recommendation model is updated in a mode of "T + 1"; the iteration period of the second recommendation model is three months, namely, the second recommendation model is updated every three months; the iteration period of the second recommendation model is much higher than that of the first recommendation model.
The recommendation method for the core body product provided by one embodiment of the specification can provide personalized core body product recommendation for a user, the recommendation effect is more in line with individual requirements and preferences of the user, and the recommendation effect of 'thousands of people and thousands of faces' is achieved.
The recommendation method of the nuclear product provided by one embodiment of the specification has the functions of reinforcement learning and self iteration.
The recommendation method for the nuclear product provided by one embodiment of the specification is more stable and reliable.
< example III >
For some users, the user may not actively try other core products, as they may have been used to using a certain core product or products before, resulting in recommended core products that are also only used by the user. For some users, since many core products have not been tried, the core history data may not be comprehensive, and the first recommendation model trained based on the core history data is one-sided. These conditions affect the global optimality of the recommendation.
In a specific example, a portion of the users are selected as target users. And for the target user, after the final recommended value is determined, the final recommended value of one or more core products for the target user is adjusted, so that the core products which are not originally recommended to the target user are recommended to the target user.
For example: the final recommended value of the password core product is 5.0+ 4.0-9.0, the final recommended value of the short message dynamic code core product is 4.0+ 3.8-7.8, the final recommended value of the face recognition core product is 2+ 2.5-4.5, and the final recommended value of the question and answer verification core product is 1.5+ 1.0-2.5. According to the scheme of the example one, the password body-checking product can be recommended to the target user, and the face recognition body-checking product cannot be recommended to the target user. According to the scheme of the third example, the final recommendation value of the face recognition core body product is adjusted to 9.5, and the face recognition core body product which cannot be originally recommended to the target user is recommended to the target user. After receiving the recommendation of the face recognition core product, the target user may use the face recognition core product or select other core products, and both the cases become the real records of the target user using the core product. Based on the reinforcement learning attributes of the first recommendation model, these real records are used as feedback data, and are iterated in the manner of steps 302 and 306.
The recommendation method for the nuclear product provided by one embodiment of the specification has the function of trying to explore the deep preference of the user, and is favorable for realizing the global optimization of the recommendation effect.
< example four >
Sometimes, the server 101 may wish to bring on-line a completely new core product, in which case the first recommendation model cannot support the recommendation of the new core product since there is no case where the user has used the core product.
In a specific example, a portion of the users are selected as target users. And for the target user, after determining the final recommended value, setting the final recommended value of the new core product for the target user for the new core product after determining the final recommended value of the core product for the target user according to the first recommended value and the second recommended value of the core product for the target user, so as to recommend the new core product to the target user.
For example: the originally available recommended core body products comprise password core body products, short message dynamic code core body products, face recognition core body products and question and answer checking core body products. The new core product is a fingerprint identification core product. The final recommended value of the password core product is 5.0+ 4.0-9.0, the final recommended value of the short message dynamic code core product is 4.0+ 3.8-7.8, the final recommended value of the face recognition core product is 2+ 2.5-4.5, and the final recommended value of the question and answer verification core product is 1.5+ 1.0-2.5. According to the scheme of the example one, the password authentication product is recommended to the target user. According to the scheme of the fourth example, the final recommendation value of the fingerprint identification core product to the target user is set to be 9.5, so that the fingerprint identification core product is recommended to the target user. After receiving the recommendation of the fingerprint identification core product, the target user may use the fingerprint identification core product or select other core products, and both the two conditions become the real records of the target user using the core product. These real records are used as feedback data based on the reinforcement learning properties of the first recommendation model, and are iterated in the manner of steps 302 and 306 described above. Finally, a first recommendation model supporting fingerprint identification core product recommendation is obtained.
The recommendation method of the nuclear body product provided by one embodiment of the specification has an attempt exploration function of a new nuclear body product.
The method for recommending the core body product provided by one embodiment of the specification can realize the cold start of a new core body product.
According to the recommendation method for the core body product, which is provided by one embodiment of the specification, when a new core body product needs to be put on line, the user can be prevented from being disturbed as much as possible.
It should be noted that the methods of example three and example four may be used in combination with the methods of example one and example two, and are not described here again.
< recommendation apparatus for personal care products >
< example one >
Referring to fig. 6, a recommendation device for a nuclear product according to an embodiment of the present disclosure will be described. The recommendation device 501 for the core product comprises:
an attribute feature obtaining module 510, configured to obtain attribute features of a user, where the attribute features of the user include at least features generated based on a history of usage of the core product by the user.
The first recommendation value determining module 521 is configured to input the attribute features of the user into a first recommendation model before iteration to determine first recommendation values of the plurality of core products for the user.
A second recommendation value determining module 522, configured to input the attribute features of the user into the iterated first recommendation model to determine second recommendation values of the plurality of nuclear products for the user.
A final recommendation value determining module 530, configured to determine final recommendation values of a plurality of core products for the user, including: and determining a final recommended value of the core product for the user according to the first recommended value and the second recommended value of the core product for the user.
The recommending module 590 is configured to recommend at least one core product to the user according to the final recommended values of the plurality of core products for the user.
In a specific example, the first recommendation model is a reinforcement learning model, and is configured to output a recommendation value reflecting the probability of using the core-body product by the user according to the attribute characteristics of the user and iterate according to a preset iteration cycle based on the reinforcement learning attribute of the user.
< example two >
Referring to fig. 7, a recommendation apparatus for a core product 502 according to an embodiment of the present disclosure is described, including:
an attribute feature obtaining module 510, configured to obtain attribute features of a user, where the attribute features of the user include at least features generated based on a history of usage of the core product by the user.
The first recommendation value determining module 521 is configured to input the attribute features of the user into a first recommendation model before iteration to determine first recommendation values of the plurality of core products for the user.
A second recommendation value determining module 522, configured to input the attribute features of the user into the iterated first recommendation model to determine second recommendation values of the plurality of nuclear products for the user.
A final recommendation value determining module 530, configured to determine final recommendation values of a plurality of core products for the user, including: and determining a final recommended value of the core product for the user according to the first recommended value and the second recommended value of the core product for the user.
The recommending module 590 is configured to recommend at least one core product to the user according to the final recommended values of the plurality of core products for the user.
The recommendation apparatus 502 for core products further includes a first final recommendation adjustment module 540.
The first final recommendation value adjusting module 540 is configured to select a part of users as target users, and adjust the final recommendation values of one or more core products for the target users, so that the recommending module recommends core products that would not be recommended to the target users.
In a specific example, the first recommendation model is a reinforcement learning model, and is configured to output a recommendation value reflecting the probability of using the core-body product by the user according to the attribute characteristics of the user and iterate according to a preset iteration cycle based on the reinforcement learning attribute of the user.
< example III >
Referring to fig. 8, a recommendation device for a nuclear product according to an embodiment of the present disclosure will be described. The recommendation device 503 for a core product includes:
an attribute feature obtaining module 510, configured to obtain attribute features of a user, where the attribute features of the user include at least features generated based on a history of usage of the core product by the user.
The first recommendation value determining module 521 is configured to input the attribute features of the user into a first recommendation model before iteration to determine first recommendation values of the plurality of core products for the user.
A second recommendation value determining module 522, configured to input the attribute features of the user into the iterated first recommendation model to determine second recommendation values of the plurality of nuclear products for the user.
A final recommendation value determining module 530, configured to determine final recommendation values of a plurality of core products for the user, including: and determining a final recommended value of the core product for the user according to the first recommended value and the second recommended value of the core product for the user.
The recommending module 590 is configured to recommend at least one core product to the user according to the final recommended values of the plurality of core products for the user.
The recommendation apparatus 502 for core products further includes a second final recommendation adjustment module 550.
And a second final recommendation value adjusting module 550, configured to select a part of the users as target users, and set a final recommendation value of the new core product for the target users for the new core product, so that the recommending module recommends the new core product to the target users.
In a specific example, the first recommendation model is a reinforcement learning model, and is configured to output a recommendation value reflecting the probability of using the core-body product by the user according to the attribute characteristics of the user and iterate according to a preset iteration cycle based on the reinforcement learning attribute of the user.
< example four >
Referring to fig. 9, a recommendation device for a nuclear product according to an embodiment of the present disclosure will be described. The recommendation device 600 for the nuclear product comprises:
an attribute feature obtaining module 610, configured to obtain attribute features of a user, where the attribute features of the user at least include features generated based on a history of usage of a core product by the user;
the first recommendation value determining module 621 is configured to input the attribute features of the user into a first recommendation model before iteration to determine first recommendation values of the plurality of core products for the user;
a second recommendation value determining module 622, configured to input the attribute features of the user into the iterated first recommendation model to determine second recommendation values of the plurality of nuclear products for the user;
a third recommendation value determining module 623, configured to input the attribute features of the user into a second recommendation model to determine a third recommendation value of the plurality of nuclear products for the user;
a final recommendation value determining module 630, configured to determine final recommendation values of a plurality of core products for the user, including: determining a final recommended value of the core product for the user according to the first recommended value and the second recommended value of the core product for the user;
a recommending module 690, configured to recommend at least one core product to the user according to the final recommended values of the plurality of core products for the user;
in a specific example, the first recommendation model is a reinforcement learning model, and is configured to output a recommendation value reflecting the probability of using the core-body product by the user according to the attribute characteristics of the user and iterate according to a preset iteration cycle based on the reinforcement learning attribute of the user. In a specific example, the second recommendation model is a non-reinforcement learning model.
The recommendation device for a core product provided in an embodiment of the present specification may include an attribute feature obtaining module, a first recommendation value determining module, a second recommendation value determining module, a final recommendation value determining module, a recommendation module, a first final recommendation value adjusting module, and a second final recommendation value adjusting module at the same time.
The recommendation device for a nuclear product provided in an embodiment of the present specification may include an attribute feature obtaining module, a first recommendation value determining module, a second recommendation value determining module, a third recommendation value determining module, a final recommendation value determining module, a recommendation module, and a first final recommendation value adjusting module at the same time.
The apparatus for recommending a core product provided in an embodiment of the present specification may include an attribute feature obtaining module, a first recommended value determining module, a second recommended value determining module, a third recommended value determining module, a final recommended value determining module, a recommending module, and a second final recommended value adjusting module at the same time.
The recommendation device for a nuclear product provided in an embodiment of the present specification may include an attribute feature obtaining module, a first recommendation value determining module, a second recommendation value determining module, a third recommendation value determining module, a final recommendation value determining module, a recommendation module, a first final recommendation value adjusting module, and a second final recommendation value adjusting module at the same time.
< example five >
Referring to fig. 10, a recommendation apparatus for a core product according to an embodiment of the present disclosure is described, where the recommendation apparatus 700 includes a memory 702 and a processor 701.
The memory 702 stores a computer program, which when executed by the processor 701 implements the method for recommending a core product disclosed in any of the foregoing embodiments.
The recommendation device for the core body product provided by one embodiment of the specification can provide personalized core body product recommendation for a user, the recommendation effect is more in line with individual requirements and preferences of the user, and the recommendation effect of 'thousands of people and thousands of faces' is achieved.
The recommendation device for the nuclear product provided by one embodiment of the specification has the functions of reinforcement learning and self iteration.
The recommendation device for the nuclear products provided by one embodiment of the specification is more stable and reliable.
The recommendation device for the nuclear product provided by one embodiment of the specification has a function of trying to explore the deep preference of the user, and is favorable for realizing global optimization of a recommendation effect.
The recommendation method of the nuclear body product provided by one embodiment of the specification has an attempt exploration function of a new nuclear body product.
The method for recommending the core body product provided by one embodiment of the specification can realize the cold start of a new core body product.
According to the recommendation method for the core body product, which is provided by one embodiment of the specification, when a new core body product needs to be put on line, the user can be prevented from being disturbed as much as possible.
< computer-readable Medium >
The embodiment of the present specification further provides a computer readable medium, on which a computer program is stored, and when the computer program is executed by a processor, the method for recommending the nuclear product disclosed in any one of the foregoing embodiments is implemented.
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, as for the device and apparatus embodiments, since they are substantially similar to the method embodiments, the description is relatively simple, and reference may be made to some descriptions of the method embodiments for relevant points.
The foregoing description has been directed to specific embodiments of this disclosure. Other embodiments are within the scope of the following claims. In some cases, the actions or steps recited in the claims may be performed in a different order than in the embodiments and still achieve desirable results. In addition, the processes depicted in the accompanying figures do not necessarily require the particular order shown, or sequential order, to achieve desirable results. In some embodiments, multitasking and parallel processing may also be possible or may be advantageous.
Embodiments of the present description may be systems, methods, and/or computer program products. The computer program product may include a computer-readable storage medium having computer-readable program instructions embodied thereon for causing a processor to implement aspects of embodiments of the specification.
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.
Computer program instructions for carrying out operations for embodiments of the present description may be assembly instructions, Instruction Set Architecture (ISA) instructions, machine related instructions, microcode, firmware instructions, state setting data, or source code 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, an electronic circuit, such as a programmable logic circuit, a Field Programmable Gate Array (FPGA), or a Programmable Logic Array (PLA), can execute computer-readable program instructions to implement various aspects of embodiments of the present specification by utilizing state information of the computer-readable program instructions to personalize the electronic circuit.
Aspects of embodiments of the present specification 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 specification. 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 processor 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 processor 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 description. 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. It is well known to those skilled in the art that implementation by hardware, by software, and by a combination of software and hardware are equivalent.
The foregoing description of the embodiments of the present specification has been presented for purposes of illustration and description, but is not intended to be exhaustive or limited to the embodiments disclosed. Many modifications and variations will be apparent to those of ordinary skill in the art without departing from the scope of the described embodiments. The terms used herein were chosen in order to best explain the principles of the embodiments, the practical application, or technical improvements to the techniques in the marketplace, or to enable others of ordinary skill in the art to understand the embodiments disclosed herein.
Claims (13)
1. A recommendation method of a nuclear product comprises the following steps:
acquiring attribute characteristics of a user, wherein the attribute characteristics of the user at least comprise characteristics generated based on a historical record of the user using a nuclear product;
inputting attribute characteristics of a user into a first recommendation model before iteration to determine first recommendation values of a plurality of core products for the user;
inputting the attribute characteristics of the user into the iterated first recommendation model to determine second recommendation values of the plurality of nuclear products for the user;
determining final recommended values of a plurality of core products for the user, including: determining a final recommended value of the core product for the user according to the first recommended value and the second recommended value of the core product for the user;
recommending at least one core product to the user according to the final recommended values of the plurality of core products for the user;
the first recommendation model is a reinforcement learning model and is configured to output a recommendation value reflecting the probability of the user using the core product according to the attribute characteristics of the user and iterate according to a preset iteration cycle based on the reinforcement learning attribute of the first recommendation model.
2. The method of claim 1, the first recommendation model iterates, comprising:
obtaining feedback data, wherein the feedback data comprises real records of using the core body products after a plurality of users receive the core body product recommendation;
generating reward values according to the core product recommendations and the feedback data received by the users;
iterating based on the feedback data and the reward value.
3. The method of claim 2, further comprising:
selecting part of users as target users;
and adjusting the final recommended value of one or more core products to the target user so as to recommend the core products which are not originally recommended to the target user.
4. The method of claim 2, further comprising:
selecting part of users as target users;
and setting a final recommendation value of the new core product to the target user for the new core product so as to recommend the new core product to the target user.
5. The method of claim 1, wherein determining a final recommendation value of a core product for the user based on the first and second recommendations of the core product for the user comprises:
and carrying out weighted average on the first recommended value and the second recommended value of the core body product for the user so as to determine the final recommended value of the core body product for the user.
6. The method of claim 1, further comprising:
inputting the attribute characteristics of the user into a second recommendation model to determine a third recommendation value of a plurality of core products for the user; the second recommendation model is a non-reinforcement learning model;
the determining of the final recommended values of the plurality of core products for the user comprises: and determining a final recommended value of the core product for the user according to the first recommended value, the second recommended value and the third recommended value of the core product for the user.
7. The method of claim 6, wherein determining a final recommendation value of a core product for the user based on the first, second, and third recommendations of the core product for the user comprises:
and carrying out weighted average on the first recommended value, the second recommended value and the third recommended value of the core body product for the user so as to determine the final recommended value of the core body product for the user.
8. A recommendation device for a nuclear product comprises the following modules:
the attribute feature acquisition module is used for acquiring attribute features of a user, wherein the attribute features of the user at least comprise features generated based on a historical record of the user using the nuclear body product;
the first recommendation value determining module is used for inputting the attribute characteristics of the user into a first recommendation model before iteration so as to determine first recommendation values of the plurality of nuclear body products for the user;
the second recommendation value determining module is used for inputting the attribute characteristics of the user into the iterated first recommendation model so as to determine second recommendation values of the plurality of nuclear body products for the user;
the final recommendation value determining module is used for determining the final recommendation values of a plurality of nuclear products for the user, and comprises the following steps: determining a final recommended value of the core product for the user according to the first recommended value and the second recommended value of the core product for the user;
the recommending module is used for recommending at least one core product to the user according to the final recommended values of the core products for the user;
the first recommendation model is a reinforcement learning model and is configured to output a recommendation value reflecting the probability of the user using the core product according to the attribute characteristics of the user and iterate according to a preset iteration cycle based on the reinforcement learning attribute of the first recommendation model.
9. The apparatus of claim 8, the first recommendation model iterates, comprising:
obtaining feedback data, wherein the feedback data comprises real records of using the core body products after a plurality of users receive the core body product recommendation;
generating reward values according to the core product recommendations and the feedback data received by the users;
iterating based on the feedback data and the reward value.
10. The apparatus of claim 9, further comprising a first final recommendation adjustment module:
the first final recommendation value adjusting module is used for selecting part of users as target users, adjusting the final recommendation values of one or more core products to the target users, and recommending the core products which cannot be originally recommended to the target users.
11. The apparatus of claim 9, further comprising a second final recommendation adjustment module to:
and the second final recommendation value adjusting module is used for selecting part of users as target users, setting the final recommendation value of the new core product to the target users for the new core product, and recommending the new core product to the target users.
12. The apparatus of claim 8, further comprising a third recommendation determination module to:
the third recommendation value determining module is used for inputting the attribute characteristics of the user into a second recommendation model so as to determine a third recommendation value of a plurality of nuclear products for the user; the second recommendation model is a non-reinforcement learning model;
the determining of the final recommended values of the plurality of core products for the user comprises: and determining a final recommended value of the core product for the user according to the first recommended value, the second recommended value and the third recommended value of the core product for the user.
13. A recommendation device for a core product comprising a processor and a memory, said memory being stored with a computer program which, when executed by said processor, implements the method of any one of claims 1-7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911052966.9A CN110879865B (en) | 2019-10-31 | 2019-10-31 | Recommendation method and device for nuclear products |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911052966.9A CN110879865B (en) | 2019-10-31 | 2019-10-31 | Recommendation method and device for nuclear products |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110879865A true CN110879865A (en) | 2020-03-13 |
| CN110879865B CN110879865B (en) | 2022-08-12 |
Family
ID=69728334
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201911052966.9A Active CN110879865B (en) | 2019-10-31 | 2019-10-31 | Recommendation method and device for nuclear products |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110879865B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112925210A (en) * | 2021-05-11 | 2021-06-08 | 北京三快在线科技有限公司 | Method and device for model training and unmanned equipment control |
Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20020099780A1 (en) * | 1999-04-02 | 2002-07-25 | Scheussler Robert W. | System and method for identifying users in a distributed network |
| CN106407349A (en) * | 2016-09-06 | 2017-02-15 | 北京三快在线科技有限公司 | Product recommendation method and device |
| CN108230058A (en) * | 2016-12-09 | 2018-06-29 | 阿里巴巴集团控股有限公司 | Products Show method and system |
| CN108763251A (en) * | 2018-04-02 | 2018-11-06 | 阿里巴巴集团控股有限公司 | The personalized recommendation method and device and electronic equipment of core body product |
| CN109359972A (en) * | 2018-08-15 | 2019-02-19 | 阿里巴巴集团控股有限公司 | The push of core body product and core body method and system |
| CN110060188A (en) * | 2019-03-06 | 2019-07-26 | 阿里巴巴集团控股有限公司 | Core body mode recommended method, device and electronic equipment |
| CN110059457A (en) * | 2018-11-05 | 2019-07-26 | 阿里巴巴集团控股有限公司 | One seed nucleus body method and device |
| CN110071915A (en) * | 2019-04-10 | 2019-07-30 | 阿里巴巴集团控股有限公司 | A kind of identity verification product method for pushing, device, equipment and system architecture |
| CN110263618A (en) * | 2019-04-30 | 2019-09-20 | 阿里巴巴集团控股有限公司 | The alternative manner and device of one seed nucleus body model |
-
2019
- 2019-10-31 CN CN201911052966.9A patent/CN110879865B/en active Active
Patent Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20020099780A1 (en) * | 1999-04-02 | 2002-07-25 | Scheussler Robert W. | System and method for identifying users in a distributed network |
| CN106407349A (en) * | 2016-09-06 | 2017-02-15 | 北京三快在线科技有限公司 | Product recommendation method and device |
| CN108230058A (en) * | 2016-12-09 | 2018-06-29 | 阿里巴巴集团控股有限公司 | Products Show method and system |
| CN108763251A (en) * | 2018-04-02 | 2018-11-06 | 阿里巴巴集团控股有限公司 | The personalized recommendation method and device and electronic equipment of core body product |
| CN109359972A (en) * | 2018-08-15 | 2019-02-19 | 阿里巴巴集团控股有限公司 | The push of core body product and core body method and system |
| CN110059457A (en) * | 2018-11-05 | 2019-07-26 | 阿里巴巴集团控股有限公司 | One seed nucleus body method and device |
| CN110060188A (en) * | 2019-03-06 | 2019-07-26 | 阿里巴巴集团控股有限公司 | Core body mode recommended method, device and electronic equipment |
| CN110071915A (en) * | 2019-04-10 | 2019-07-30 | 阿里巴巴集团控股有限公司 | A kind of identity verification product method for pushing, device, equipment and system architecture |
| CN110263618A (en) * | 2019-04-30 | 2019-09-20 | 阿里巴巴集团控股有限公司 | The alternative manner and device of one seed nucleus body model |
Non-Patent Citations (2)
| Title |
|---|
| KAI SHU 等: "User Identity Linkage across Online Social Networks: A Review", 《SIGKDD EXPLORATIONS》 * |
| 邵悦: "基于android手机的支付系统设计与实现", 《中国优秀硕士学位论文全文数据库(信息科技辑)》 * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112925210A (en) * | 2021-05-11 | 2021-06-08 | 北京三快在线科技有限公司 | Method and device for model training and unmanned equipment control |
| CN112925210B (en) * | 2021-05-11 | 2021-09-07 | 北京三快在线科技有限公司 | Method and device for model training and unmanned equipment control |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110879865B (en) | 2022-08-12 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US11836453B2 (en) | Virtual assistant domain functionality | |
| US10509837B2 (en) | Modeling actions for entity-centric search | |
| US11151324B2 (en) | Generating completed responses via primal networks trained with dual networks | |
| US10785045B2 (en) | Socially enabled consensus blockchain summarization | |
| US11049169B2 (en) | System, computer program product, and method for automated gift determination and delivery | |
| TWI717673B (en) | Method, device and electronic equipment for resetting payment password | |
| US11875166B2 (en) | User interface modality switching for transaction management | |
| US11348172B2 (en) | User interfaces that differentiate payment instruments having a trusted beneficiary | |
| CN111368973A (en) | Method and apparatus for training a hyper-network | |
| US11436596B2 (en) | Eligibility determination for delegation exemption to strong authentication requirements | |
| US11281867B2 (en) | Performing multi-objective tasks via primal networks trained with dual networks | |
| WO2023283128A1 (en) | Authentication question topic exclusion based on response hesitation | |
| US20210065170A1 (en) | Selecting exemptions to strong authentication requirements | |
| CN110879865B (en) | Recommendation method and device for nuclear products | |
| JP2024508502A (en) | Methods and devices for pushing information | |
| US11709579B2 (en) | Systems and methods for customized navigation menus | |
| US11227224B2 (en) | Information and data collaboration among multiple artificial intelligence (AI) systems | |
| US11831807B2 (en) | Systems and methods for generating customized customer service menu | |
| US11037554B1 (en) | Network of domain knowledge based conversational agents | |
| KR102650587B1 (en) | Method, device and system for providing evaluation and matching solutions for contract brokers considering quantitative and qualitative criteria | |
| CN113010784B (en) | Method, apparatus, electronic device and medium for generating prediction information | |
| US20240095327A1 (en) | Computer authentication using knowledge of former devices | |
| US20230370459A1 (en) | Verifier-built profiles for dating application | |
| US20240028923A1 (en) | Extracting enriched target-oriented common sense from grounded graphs to support next step decision making | |
| Csiki et al. | Software Application for Disabled Students |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| REG | Reference to a national code |
Ref country code: HK Ref legal event code: DE Ref document number: 40029270 Country of ref document: HK |
|
| GR01 | Patent grant | ||
| GR01 | Patent grant |