CN115082844A - Similar crowd extension method and device, electronic equipment and readable storage medium - Google Patents
Similar crowd extension method and device, electronic equipment and readable storage medium Download PDFInfo
- Publication number
- CN115082844A CN115082844A CN202110278343.4A CN202110278343A CN115082844A CN 115082844 A CN115082844 A CN 115082844A CN 202110278343 A CN202110278343 A CN 202110278343A CN 115082844 A CN115082844 A CN 115082844A
- Authority
- CN
- China
- Prior art keywords
- recall
- user
- user sample
- sample
- target scene
- 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 63
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 101
- 238000001914 filtration Methods 0.000 claims abstract description 85
- 230000006399 behavior Effects 0.000 claims description 63
- 230000015654 memory Effects 0.000 claims description 11
- 230000003542 behavioural effect Effects 0.000 claims description 8
- 230000008569 process Effects 0.000 description 11
- 238000004590 computer program Methods 0.000 description 10
- 238000010586 diagram Methods 0.000 description 8
- 230000002159 abnormal effect Effects 0.000 description 7
- 230000006870 function Effects 0.000 description 7
- 238000004891 communication Methods 0.000 description 6
- 239000013598 vector Substances 0.000 description 6
- 238000004364 calculation method Methods 0.000 description 5
- 230000000694 effects Effects 0.000 description 5
- 238000012545 processing Methods 0.000 description 5
- 238000001514 detection method Methods 0.000 description 4
- 238000012360 testing method Methods 0.000 description 4
- 238000010276 construction Methods 0.000 description 2
- 238000011156 evaluation Methods 0.000 description 2
- 239000008267 milk Substances 0.000 description 2
- 210000004080 milk Anatomy 0.000 description 2
- 235000013336 milk Nutrition 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 230000003287 optical effect Effects 0.000 description 2
- 238000013450 outlier detection Methods 0.000 description 2
- 238000004806 packaging method and process Methods 0.000 description 2
- 239000000843 powder Substances 0.000 description 2
- 239000000758 substrate Substances 0.000 description 2
- 108010001267 Protein Subunits Proteins 0.000 description 1
- 230000003044 adaptive effect Effects 0.000 description 1
- 238000013528 artificial neural network Methods 0.000 description 1
- 238000012512 characterization method Methods 0.000 description 1
- 238000013136 deep learning model Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 235000015219 food category Nutrition 0.000 description 1
- 238000007429 general method Methods 0.000 description 1
- 238000002955 isolation Methods 0.000 description 1
- 239000004973 liquid crystal related substance Substances 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 230000000717 retained effect Effects 0.000 description 1
- 230000002441 reversible effect Effects 0.000 description 1
- 238000012216 screening Methods 0.000 description 1
- 239000004065 semiconductor Substances 0.000 description 1
- 239000000126 substance Substances 0.000 description 1
- 230000002123 temporal effect Effects 0.000 description 1
- 238000012549 training Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q30/00—Commerce
- G06Q30/02—Marketing; Price estimation or determination; Fundraising
- G06Q30/0201—Market modelling; Market analysis; Collecting market data
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q30/00—Commerce
- G06Q30/06—Buying, selling or leasing transactions
- G06Q30/0601—Electronic shopping [e-shopping]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q30/00—Commerce
- G06Q30/06—Buying, selling or leasing transactions
- G06Q30/0601—Electronic shopping [e-shopping]
- G06Q30/0631—Item recommendations
Landscapes
- Business, Economics & Management (AREA)
- Finance (AREA)
- Accounting & Taxation (AREA)
- Engineering & Computer Science (AREA)
- Development Economics (AREA)
- Strategic Management (AREA)
- General Business, Economics & Management (AREA)
- Economics (AREA)
- Marketing (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Entrepreneurship & Innovation (AREA)
- Game Theory and Decision Science (AREA)
- Data Mining & Analysis (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The present disclosure provides a similar population expansion method, comprising: firstly, selecting a target scene, and determining a filtering algorithm corresponding to the target scene through historical behavior data of a user. And carrying out exception filtering on the original user sample through the filtering algorithm to obtain a seed user sample. And finally, under a target scene, performing similar population expansion on the seed user sample to obtain an expanded user sample. The disclosure also provides a similar crowd extension device, an electronic device and a computer readable storage medium.
Description
Technical Field
The present disclosure relates to the field of internet technologies, and in particular, to a method and an apparatus for extending similar people, an electronic device, and a readable storage medium.
Background
Similar population extension (Look-align) refers to a technique for finding populations similar to or potentially associated with seed users based on a set of seed users through label rules or an algorithm model. The prior art mainly comprises two categories of people and algorithm models which are selected by portrait labels: the former selects people by manually selecting labels such as age, gender and the like, and the latter outputs related people by inputting user portrait and behavior characteristics into a machine learning or deep learning model.
In the course of implementing the disclosed concept, the inventor finds that at least the following problems exist in the related art, because it is difficult to discover the internal relation between users by manually selecting the label by frame, and it is also difficult to utilize the potential features of the users who have the seeds. Meanwhile, the quality of the seed user also affects the overall quality of the similar users obtained by expansion.
Disclosure of Invention
In view of the above, the present disclosure provides a similar population expansion method and apparatus.
One aspect of the present disclosure provides a similar population expansion method, including: selecting a target scene; determining a filtering algorithm corresponding to the target scene; performing exception filtering on the original user sample through the filtering algorithm to obtain a seed user sample; and under the target scene, performing similar population expansion on the seed user sample to obtain an expanded user sample.
According to an embodiment of the present disclosure, the determining a filtering algorithm corresponding to the target scene includes: acquiring a first user sample from a candidate user sample set based on preset behavior characteristics; respectively carrying out exception filtering on the first user sample through a plurality of preset algorithms to obtain a plurality of second user samples; in the target scene, recalling each second user sample in the candidate user sample set through a recall algorithm to obtain a plurality of recalled user samples; respectively detecting the recall quality of each recall user sample based on the preset behavior characteristics; and determining a filtering algorithm corresponding to the target scene based on the recall quality, wherein the filtering algorithm is one of the preset algorithms.
According to an embodiment of the present disclosure, the detecting recall quality of each of the recall user samples based on the preset behavior feature includes: respectively counting the number of designated users in each recalled user sample, wherein the designated users have preset behavior characteristics; and comparing the number of all the designated users, and determining an optimal recall user sample, wherein the optimal recall user sample is a recall user sample which contains the largest number of designated users in all the recall user samples.
According to an embodiment of the present disclosure, the detecting recall quality of each of the recall user samples based on the preset behavior feature further includes: counting the number of characteristic users in the candidate user sample set except the first user sample, wherein the characteristic users have preset behavior characteristics; calculating a recall ratio of each of the recalled user samples, the recall ratio being a ratio of the number of the designated users to the number of the feature users; comparing all of the recall rates; determining an optimal recall user sample, the optimal recall user sample being a recall user sample having a greatest recall rate of all of the recall user samples.
According to an embodiment of the present disclosure, the determining a filtering algorithm corresponding to the target scene based on the recall quality includes: determining a preset algorithm corresponding to the optimal recall user sample; and determining the preset algorithm as a filtering algorithm corresponding to the target scene.
According to an embodiment of the present disclosure, the determining a filtering algorithm corresponding to the target scene further includes: under the target scene, recalling the first user sample in the candidate user sample set through a recall algorithm to obtain an original recall user sample; detecting an original recall quality of the original recall user sample based on the preset behavioral characteristics; comparing the original recall quality to recall qualities of a plurality of the recalled user samples; if the original recall quality is better than the recall quality of all the recall user samples, the target scene has no corresponding filtering algorithm.
According to an embodiment of the present disclosure, the user sample is a set of user feature data, which includes user portrait information and user behavior information.
Another aspect of the present disclosure provides a similar population extending device comprising: a selection module for selecting a target scene; the determining module is used for determining a filtering algorithm corresponding to the target scene; the filtering module is used for carrying out exception filtering on the original user sample through the filtering algorithm to obtain a seed user sample; and the expansion module is used for performing similar population expansion on the seed user sample in the target scene to obtain an expanded user sample.
Another aspect of the present disclosure provides an electronic device comprising one or more processors; and memory for storing one or more programs, wherein the one or more programs, when executed by the one or more processors, cause the one or more processors to implement the method of any of the preceding claims.
Another aspect of the present disclosure provides a computer-readable storage medium storing computer-executable instructions for implementing the method as described above when executed.
According to the embodiment of the disclosure, because the technical means of filtering the seed users through the specified filtering algorithm to obtain the extended population with the preset behavior characteristics in the target scene is adopted, the technical problem of how to reduce the influence of the noise users in the seed user sample on the extension of the similar population is at least partially overcome, and the technical effect of increasing the effectiveness of the extended population is further achieved.
Drawings
The above and other objects, features and advantages of the present disclosure will become more apparent from the following description of embodiments of the present disclosure with reference to the accompanying drawings, in which:
FIG. 1 schematically illustrates an exemplary system architecture to which the similar population expansion method and apparatus of the present disclosure may be applied;
FIG. 2 schematically illustrates a flow chart of a similar population expansion method according to an embodiment of the present disclosure;
FIG. 3A schematically illustrates a flow chart of a similar population expansion method according to another embodiment of the present disclosure;
FIG. 3B schematically illustrates a flow chart of a similar population expansion method according to another embodiment of the present disclosure;
FIG. 3C schematically illustrates a flow chart of a similar population expansion method according to another embodiment of the present disclosure;
FIG. 3D schematically illustrates a flow chart of a similar population expansion method according to another embodiment of the present disclosure;
FIG. 3E schematically illustrates a flow chart of a similar population expansion method according to another embodiment of the present disclosure;
FIG. 4 schematically illustrates a block diagram of a similar population expansion device according to an embodiment of the present disclosure; and
fig. 5 schematically illustrates a block diagram of an electronic device suitable for implementing a similar population extending apparatus according to an embodiment of the present disclosure.
Detailed Description
Hereinafter, embodiments of the present disclosure will be described with reference to the accompanying drawings. It should be understood that the description is illustrative only and is not intended to limit the scope of the present disclosure. In the following detailed description, for purposes of explanation, numerous specific details are set forth in order to provide a thorough understanding of the embodiments of the disclosure. It may be evident, however, that one or more embodiments may be practiced without these specific details. Moreover, in the following description, descriptions of well-known structures and techniques are omitted so as to not unnecessarily obscure the concepts of the present disclosure.
The terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the disclosure. The terms "comprises," "comprising," and the like, as used herein, specify the presence of stated features, steps, operations, and/or components, but do not preclude the presence or addition of one or more other features, steps, operations, or components.
All terms (including technical and scientific terms) used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs, unless otherwise defined. It is noted that the terms used herein should be interpreted as having a meaning that is consistent with the context of this specification and should not be interpreted in an idealized or overly formal sense.
Where a convention analogous to "at least one of A, B and C, etc." is used, in general such a construction is intended in the sense one having skill in the art would understand the convention (e.g., "a system having at least one of A, B and C" would include but not be limited to systems that have a alone, B alone, C alone, a and B together, a and C together, B and C together, and/or A, B, C together, etc.). Where a convention analogous to "A, B or at least one of C, etc." is used, in general such a construction is intended in the sense one having skill in the art would understand the convention (e.g., "a system having at least one of A, B or C" would include but not be limited to systems that have a alone, B alone, C alone, a and B together, a and C together, B and C together, and/or A, B, C together, etc.).
The embodiment of the disclosure provides a similar crowd expansion method and device. The method comprises the following steps: and selecting a target scene, and determining a filtering algorithm corresponding to the target scene according to the historical behavior data of the user. Performing exception filtering on an original user sample through a filtering algorithm to obtain a seed user sample; and under the target scene, performing similar population expansion on the sub-user sample to obtain an expanded user sample.
Fig. 1 schematically illustrates an exemplary system architecture 100 to which the similar population expansion method and apparatus may be applied, according to an embodiment of the present disclosure. It should be noted that fig. 1 is only an example of a system architecture to which the embodiments of the present disclosure may be applied to help those skilled in the art understand the technical content of the present disclosure, and does not mean that the embodiments of the present disclosure may not be applied to other devices, systems, environments or scenarios.
As shown in fig. 1, the system architecture 100 according to this embodiment may include terminal devices 101, 102, 103, a network 104 and a server 105. The network 104 serves as a medium for providing communication links between the terminal devices 101, 102, 103 and the server 105. Network 104 may include various connection types, such as wired and/or wireless communication links, and so forth.
The user may use the terminal devices 101, 102, 103 to interact with the server 105 via the network 104 to receive or send messages or the like. The terminal devices 101, 102, 103 may have various messaging client applications installed thereon, such as shopping applications, web browser applications, search applications, instant messaging tools, mailbox clients, and/or social platform software, to name a few examples.
The terminal devices 101, 102, 103 may be various electronic devices having a display screen and supporting web browsing, including but not limited to smart phones, tablet computers, laptop portable computers, desktop computers, and the like.
The server 105 may be a server providing various services, such as a background management server (for example only) providing support for websites browsed by users using the terminal devices 101, 102, 103. The background management server may analyze and perform other processing on the received data such as the user request, and feed back a processing result (e.g., a webpage, information, or data obtained or generated according to the user request) to the terminal device.
It should be noted that the similar population expansion method provided by the embodiment of the present disclosure can be generally executed by the server 105. Accordingly, similar population expansion means provided by the embodiments of the present disclosure may be generally disposed in the server 105. Similar population expansion methods provided by embodiments of the present disclosure may also be performed by servers or server clusters that are different from server 105 and that are capable of communicating with terminal devices 101, 102, 103 and/or server 105. Accordingly, similar population expansion means provided by the embodiments of the present disclosure may also be provided in a server or a server cluster different from the server 105 and capable of communicating with the terminal devices 101, 102, 103 and/or the server 105. Alternatively, the similar crowd extension method provided by the embodiment of the present disclosure may also be executed by the terminal device 101, 102, or 103, or may also be executed by another terminal device different from the terminal device 101, 102, or 103. Accordingly, the similar population expansion apparatus provided by the embodiment of the present disclosure may also be disposed in the terminal device 101, 102, or 103, or in another terminal device different from the terminal device 101, 102, or 103.
For example, the user sample may be originally stored in any of the terminal devices 101, 102, or 103 (e.g., but not limited to terminal device 101), or stored on an external storage device and may be imported into terminal device 101. Then, the terminal device 101 may locally perform the similar population expansion method provided by the embodiment of the present disclosure, or send the user sample to another terminal device, server, or server cluster, and the similar population expansion method provided by the embodiment of the present disclosure is performed by another terminal device, server, or server cluster that receives the user sample.
It should be understood that the number of terminal devices, networks, and servers in fig. 1 is merely illustrative. There may be any number of terminal devices, networks, and servers, as desired for an implementation.
Fig. 2 schematically illustrates a flow chart of a similar population expansion method according to an embodiment of the present disclosure.
As shown in fig. 2, the method includes operations S201 to S204.
In operation S201, a target scene is selected.
Specifically, the target scenario may be understood as a business marketing scenario of a merchant, that is, a sales scenario of a certain product class by the merchant. For example, the target scenario may specifically be that a merchant performs a promotion activity on a living product, or a merchant performs a promotion activity on a mother-infant product, and the like. Product categories include, but are not limited to, category 3C products, jewelry categories, apparel categories, food categories, and the like.
In operation S202, a filtering algorithm corresponding to the target scene is determined.
In operation S203, the original user sample is filtered for exception through a filtering algorithm, so as to obtain a seed user sample.
In the similar population expansion process, the original user sample may be provided by an external partner or an internal database. In the similar population expansion process, the original user sample is used as a reference user sample, and a user sample with the same behavior characteristics as the original user sample can be obtained. However, the quality of the known original user sample cannot be guaranteed, that is, the users contained in the original user sample do not necessarily contain the behavior characteristics expected by the developer. If the original user sample is directly used for similar population expansion without performing exception filtering on the original user sample, the overall quality of similar users recalled is likely to be reduced due to the influence of part of the abnormal user samples.
Meanwhile, some deviations may occur in the process of acquiring the original user sample, resulting in poor quality of the user sample. For example, the system determines to acquire a user sample with a characteristic of browsing a notebook computer in a target scene related to a 3C product, but there is a phenomenon of acquisition error in the acquisition process, so that the acquired user sample includes a user sample with a characteristic of browsing only clothing. In order to ensure the quality of the user sample, the original user sample needs to be filtered by a filtering algorithm.
It should be noted that different filtering algorithms have certain adaptability to different service scenarios. Different filtering algorithms are different in the field of good business scenes, so that after a developer selects a target scene from a plurality of business scenes, the filter algorithm most suitable for the target scene needs to be selected to ensure that the filtered seed user samples can be expanded to more accurate crowds.
In particular, the filtering algorithms include, but are not limited to, outlier detection algorithms, isolated forest algorithms, and self-encoding algorithms.
The Local Outlier Factor detection (Local Outlier Factor) is an anomaly detection method based on density, an Outlier Factor of the data is obtained by performing a clustering Factor calculation on each data point, whether the data is an Outlier is judged by the Outlier Factor, and if LOF > 1, the data is the Outlier.
Isolated Forest (Isolation Forest) is a learning algorithm that integrates multiple isolated class trees. The algorithm has the characteristic of no need of calculation based on density and distance indexes, so that the calculation complexity of the algorithm is small. The simultaneous calculation is based on ensemble learning, and thus the temporal complexity is linear. Each classification tree is independently generated, and the classification trees are convenient to be deployed in a distributed system for calculation.
An auto encoder (AutoEncoder) is an unsupervised learning model that generates its low-dimensional representation from higher-dimensional inputs via a neural network. However, because only the most information characteristic is retained during training, when the decoder reconstructs the original data, the abnormal point cannot be restored well, and the error is larger.
In operation S204, in the target scene, similar population expansion is performed on the seed user sample to obtain an expanded user sample.
In the embodiment of the present disclosure, the user sample is a set of user feature data, and in actual use, the user feature data includes user portrait information and user behavior information. The user profile information includes personal attribute information of the user, such as age, sex, occupation, family members, personal preference, and the like. The user behavior information is business behavior information which the user has in a certain time period, such as behavior information of browsing, clicking and purchasing articles.
Generally, the e-commerce platform needs to reasonably guess the future business behavior of the user according to the user-specific portrait information and behavior information, and provide corresponding business services to improve the use feeling of the user, for example, provide services such as "recommend for you".
Specifically, the e-commerce platform acquires a mobile phone number, an IMEI number, an IFA number, and the like of a certain user, binds the mobile phone number, the IMEI number, and the IFA number of the user with an account of the user, and uses the bound mobile phone number, IMEI number, and IFA number as a specific identity of the user. The e-commerce platform can send corresponding messages to the user account, collect characteristic information related to the user account, and use the collected characteristic information as characteristic data of the user.
According to the embodiment of the disclosure, according to different requirements, a corresponding target scene is selected, a filtering algorithm which is most adaptive to the target scene is determined, an original user sample is subjected to abnormal filtering, noise users which do not meet conditions in the original user sample are filtered, a seed user sample is obtained, similar population expansion is performed on the seed user sample in the target scene, a similar population which has the same behavior characteristics as the seed user sample is obtained, and the technical effect of improving the validity of expanded data is achieved.
The method shown in fig. 2 is further described with reference to fig. 3A-3D in conjunction with specific embodiments.
Fig. 3A schematically illustrates a flow chart of a similar population expansion method according to another embodiment of the present disclosure.
As shown in fig. 3A, determining a filtering algorithm corresponding to a target scene further includes operations S301 to S305.
In operation S301, a first user sample is obtained from a candidate user sample set based on a preset behavior feature.
The sample set of candidate users is the user corpus. Specifically, the user set may be a user set composed of all users in the database, or a user set composed of part of user data. In the whole process of determining the corresponding relation between the target scene and the filtering algorithm, the selection and recall of the user sample are finished in the candidate user sample set. That is, the users of the user sample obtained or recalled are all the users in the candidate user sample set.
The preset behavior characteristics may be behavior characteristics of browsing a commodity, purchasing a commodity in a shopping cart, purchasing a commodity and the like. The users in the first user sample have at least one preset behavior characteristic. For example, in the target scenario of 3C category, the first user sample may be composed of all users who have a feature of browsing a notebook computer, may be composed of all users who have a feature of purchasing a notebook computer, may be composed of both users who have a feature of browsing a notebook computer and users who have a feature of purchasing a notebook computer, may be composed of both users who have features of browsing a notebook computer and purchasing a notebook computer, or may be composed of both users who have features of browsing a notebook computer and purchasing a notebook computer and only users who have a feature of composing a browsing notebook computer. The present application does not limit the specific way in which the feature types of the users of the first user sample are composed.
The first user sample is a certain number of users with preset behavior characteristics selected from the candidate sample set, but due to the fact that abnormal phenomena may occur in the selection process, not all the users in the first user sample actually obtained have the preset behavior characteristics.
In operation S302, the first user sample is filtered through a plurality of preset algorithms to obtain a plurality of second user samples.
The plurality of pre-set algorithms include, but are not limited to, outlier detection algorithms, isolated forest algorithms, and self-encoding algorithms. And obtaining a second user sample corresponding to each filtering algorithm after exception filtering.
In operation S303, in the target scenario, each second user sample is recalled in the candidate user sample set through a recall algorithm, so as to obtain a plurality of recalled user samples.
In the disclosed embodiment, the second user sample may be understood as a seed user sample, and the recall algorithm is specifically a Faiss recall algorithm. In the Faiss recall algorithm, the seed user sample is represented by a user embedding vector (user embedding vector) trained by the user portrait in combination with recent behavior features. The Faiss recall algorithm specifically determines N users closest to each seed user in a recall vector space, wherein N is determined by developers according to actual conditions, and N is the number of final recalled users.
In operation S304, recall quality of each of the recalled user samples is respectively detected based on the preset behavior feature.
In the actual recall result, not every user in the sample of recall users has a preset behavior profile. Therefore, the recall quality of each sample of recalled users needs to be checked to ensure that the optimal filtering algorithm corresponding to the target scene is determined.
In operation S305, a filtering algorithm corresponding to the target scene is determined based on the recall quality, where the filtering algorithm is one of a plurality of preset algorithms.
Practice shows that the distribution characteristics of the embedded characterization vectors generated by the user portrait and the behavior sequence show different characteristics with different behavior types (such as browsing or shopping) and target commodities (such as electronic products or daily necessities). In the prior art, noise users are filtered only by using a general abnormal point detection method, and specific behavior characteristics and target commodities cannot be screened. On the premise of user embedded vector heterogeneity, simply using a general method to remove the abnormal seed user may have the effect of deviating from the business target.
Therefore, the relevance between the target scene and the preset algorithm is determined, and the filtering capabilities of different filtering algorithms in different feature fields are reflected. Determining the correspondence between a particular business scenario and a filtering algorithm may improve the accuracy of the expansion.
Fig. 3B schematically illustrates a flow chart of a similar population expansion method according to another embodiment of the present disclosure.
As shown in fig. 3B, the detecting the recall quality of each of the recalled user samples respectively based on the preset behavior feature includes operations S306 to S307.
In operation S306, the number of designated users in each recalled user sample is counted, respectively, and the designated users have preset behavior characteristics.
In operation S307, the numbers of all the designated users are compared, and an optimal recall user sample is determined, where the optimal recall user sample is a recall user sample including the largest number of designated users among all the recall user samples.
In the embodiment of the disclosure, the recall quality of the recalled user sample is determined by counting the number of users meeting the preset conditions in the recalled user sample.
For example, on the premise of selecting a target scene of a 3C category, a first user sample a is obtained from the candidate sample set, and the first user sample a is composed of users having a feature of browsing a notebook computer. Filtering the first user sample A through different filtering algorithms (outlier factor detection algorithm, isolated forest algorithm and self-coding algorithm) to obtain different second user samples which are respectively marked as S 1 、S 2 And S 3 . For the second user sample S under the target scene of the 3C category 1 、S 2 And S 3 Respectively carrying out similar population expansion to obtain a sample R with a recall user 1 、R 2 And R 3 Wherein the user sample R is recalled 1 、R 2 And R 3 The user of (2) has a feature of at least one of browsing, purchasing and purchasing the notebook computer.
Specifically, in the recall vector space, a second user sample S is determined 1 100 users with the nearest distance are taken as a recall user sample R 1 Determining a second user sample S 2 100 users with the nearest distance are taken as a recall user sample R 2 Determining a second user sample S 3 Sample R of 100 users closest as recalling users 3 . In the ideal case, recall the user sample R 1 、R 2 And R 3 Has a feature of at least one of browsing, shopping, and purchasing the notebook computer. For example, recall user sample R 1 、R 2 And R 3 The user of (1) includes a user who has only the feature of browsing the notebook computer, a user who has only the feature of purchasing the notebook computer, a user who has two features of browsing the notebook computer and purchasing the notebook computer, or a user who has three features of browsing the notebook computer, purchasing the notebook computer and purchasing the notebook computer.
Recall the user sample R in the actual recall result 1 、R 2 And R 3 There are users who do not have any of the features of browsing, buying, and purchasing a notebook computer. Screening recall usersSample R 1 、R 2 And R 3 The user having the feature of browsing at least one of the notebook computer, the additional purchasing of the notebook computer and the purchasing of the notebook computer is taken as a qualified user, that is, a designated user. The number of eligible users is counted, and the recall quality of each sample of recall users is determined based on the number of eligible users.
And determining the recall user sample of the user with the most conforming conditions as the optimal recall user sample, namely the recall quality of the recall user sample is the best.
It should be noted that, the present application does not specifically limit the relationship between the behavior characteristics of the first user sample and the behavior characteristics of the recalled user sample, and only needs to ensure that the behavior characteristics of the first user sample and the behavior characteristics of the recalled user sample belong to the same target scenario.
For example, in the above example, the user of the first user sample A has only a view of a laptop computer feature, and the user sample R is recalled 1 、R 2 And R 3 The user may have a feature of at least one of browsing, shopping for and purchasing a notebook computer. Or the user of the first user sample A only has the characteristic of browsing the notebook computer and selects to recall the user sample R 1 、R 2 And R 3 The user of (2) also has only a browsing notebook computer feature. Or the user of the first user sample A has the characteristic of browsing at least one of the notebook computer, the additional purchasing notebook computer and the purchasing notebook computer, and the user sample R is selected to be recalled 1 、R 2 And R 3 May also have the feature of at least one of browsing, shopping for and purchasing a notebook computer. Or the user of the first user sample A only has the characteristic of browsing the notebook computer, and the user sample R is selected to be recalled 1 、R 2 And R 3 The user of (2) may have a feature of at least one of browsing, shopping, and purchasing the cell phone. Browsing a notebook computer, browsing a mobile phone, browsing an earphone, browsing a sound box and the like are preset behavior characteristics belonging to the 3C category. Other first user sample toolsSome examples of the relationship between the behavior characteristics and the behavior characteristics of the recalled user sample are similar to the above examples, and are not described in detail here.
In particular, fig. 3C schematically illustrates a flow chart of a similar population expansion method according to another embodiment of the present disclosure.
As shown in fig. 3C, separately detecting the recall quality of each of the recalled user samples for each of the preset features includes operations S308 to S310.
In operation S308, the number of feature users in the candidate user sample set except the first user sample is counted, where the feature users have preset behavior features.
In operation S309, a recall rate of each sample of recalled users is calculated, the recall rate being a ratio of the number of designated users to the number of feature users.
In operation S310, all recall rates are compared to determine an optimal recall user sample, which is a recall user sample having the greatest recall rate among all recall user samples.
In the embodiment of the present disclosure, in the whole process of determining the corresponding relationship between the target scene and the filtering algorithm, both the selection and recall processes of the user sample are completed in the candidate user sample set. That is, the users of the user sample obtained or recalled are all the users in the candidate user sample set. When the number of users with preset behavior characteristics in the candidate user sample set is small, the number of specified users meeting the conditions in the recall user sample is correspondingly reduced.
When the number of the specified users meeting the condition in the recall user sample is counted to be too small, the reason is that a recall exception exists in the recall or the number of the users having the preset behavior characteristics in the candidate user sample set is small. In order to avoid the shortage of the number of the specified users meeting the conditions in the recall user sample caused by the recall abnormity existing in the recall, the recall quality can be further evaluated in a recall rate mode.
The recall ratio is the ratio of the number of designated users to the number of feature users. The characteristic user is a user with preset behavior characteristics in the candidate user sample set except the first user sample. The users with the preset behavior characteristics in the candidate user sample set can be considered as a set M, a preset number of users are selected from the set M to form a first user sample a, and the rest users in the set M form a test user sample B. Understandably, in the recall process, the users recalled by the second user sample do not include the users included in the second user sample, and the recall of the user sample is completed in the candidate user sample set, and the users of the user sample obtained after the recall are all the users in the candidate user sample set. The feature users recalled in the recalled user sample are users in the test user sample B.
Therefore, when the number of users in the selected first user sample is much larger than that of the users in the test user sample, the calculated recall ratio may be a very small value if the users with the preset behavior characteristics in the first user sample are not removed in the statistical characteristic user number in calculating the recall ratio.
Specifically, the number of users in the test user sample B is N B The number of feature users in the recall user sample R is N (R∩B) And finally calculating the Recall rate Recall.
Wherein the content of the first and second substances,
the recall user sample with the greatest recall rate is the optimal recall user sample with the best recall quality.
It should be noted that, in this embodiment, the relationship between the behavior feature of the first user sample, the behavior feature of the feature user, and the behavior feature of the recall user sample is not specifically limited, and it is only required to ensure that the behavior feature of the first user sample, the behavior feature of the feature user, and the behavior feature of the recall user sample belong to the same target scene, and a person skilled in the art selects the behavior features according to actual needs. The specific corresponding relationship is exemplified in the foregoing, and is not described in detail herein. It should be emphasized, however, that to ensure consistencyThe category of the behavior feature of the feature user is the same as the category of the behavior feature of the recall user sample. For example, the user of the first user sample A has only a browsing notebook feature, and chooses to recall the user sample R 1 、R 2 And R 3 The user may only have a feature of browsing the notebook computer, then the statistical feature user should also be the user who only has a feature of browsing the notebook computer.
Fig. 3D schematically illustrates a flow chart of a similar population expansion method according to another embodiment of the present disclosure.
As shown in fig. 3D, determining the filtering algorithm corresponding to the target scene based on the recall quality includes operations S311 to S313.
In operation S311, a preset algorithm corresponding to the optimal recall user sample is determined.
In operation S312, it is determined that the preset algorithm is the filtering algorithm corresponding to the target scene.
The optimal recall user sample having the optimal recall quality is obtained through different criteria in operation S307 and operation S310, respectively. When the recall quality can be accurately judged through the number designated in the recall user sample in operation S307, a corresponding preset algorithm can be determined from the optimal recall user sample obtained in operation S307, and a filtering algorithm corresponding to the target scene is further determined. If the recall quality cannot be accurately judged through the number designated in the recall user sample in operation S307, an optimal recall user sample is further determined in operation S310 through calculating the recall rate, and a filtering algorithm corresponding to the target scene, that is, a most suitable filtering algorithm, is further determined.
Fig. 3E schematically illustrates a flow chart of a similar population expansion method according to another embodiment of the present disclosure.
As shown in fig. 3E, determining the filtering algorithm corresponding to the target scene further includes operations S313 to S316.
In operation S313, in the target scenario, the first user sample is recalled within the candidate user sample set through a recall algorithm, so as to obtain an original recall user sample.
In operation S314, an original recall quality of the original recall user sample is detected based on the preset behavior feature.
In operation S315, the original recall quality is compared to recall qualities of a plurality of recalled user samples.
In operation S316, if the original recall quality is better than the recall qualities of all the recalled user samples, the target scene has no corresponding filtering algorithm.
In order to further compare the influence of the filtering algorithm on the recall quality, the first user sample which is not subjected to the filtering algorithm is recalled to obtain an original recall user sample, and the recall quality of the original recall user sample is detected. And comparing the recall quality of the user sample as a reference value through a filtering algorithm to obtain the recall quality of the recall user sample. In the actual filtering process, the preset algorithm which may be selected is not suitable for the target scene, and the quality of the first user sample is reduced after the first user sample is subjected to abnormal filtering, so that the original recall user sample is taken as a reference, and if the original recall quality is superior to the recall quality of all the recall user samples obtained after filtering, it is determined that the target scene does not have a corresponding filtering algorithm, that is, in the actual similar crowd extension, the original user sample does not need to be filtered through the filtering algorithm in the target scene.
It should be further noted that, in the present disclosure, only an embodiment of selecting a certain service scene as a target scene for similar population expansion is provided, but in actual application, the service scenes are diverse, and in order to ensure the accuracy of the similar population expansion result for each service scene, the same evaluation criteria needs to be determined to determine the optimal filtering algorithm corresponding to each service scene. The most critical evaluation criterion is to determine that the relationship among the behavior characteristics of the first user sample, the behavior characteristics of the feature user and the behavior characteristics of the recall user sample is the same in each service scenario.
For example, in determining the quality of the recall, the relationship between the behavioral characteristics of the first sample of users, the behavioral characteristics of the feature users, and the behavioral characteristics of the sample of recall users has the following relationship:
in a 3C category business scenario: users of the first user sample aRecall user samples R with a view notebook only feature 1 、R 2 And R 3 And feature users also have only the feature of browsing a laptop.
Under the business scene of the clothing category: the users of the first user sample A have only a browse women's dress feature, recall the user sample R 1 、R 2 And R 3 And feature users also have only a browse suit-dress feature.
Under the service scene of mother and infant article class: the users of the first user sample A have only a feature of browsing milk powder, recall the user sample R 1 、R 2 And R 3 And the feature user also only has the feature of browsing the milk powder.
That is, when determining the filtering algorithms corresponding to different service scenarios, the criteria of the lateral contrast should be consistent.
Fig. 4 schematically illustrates a block diagram of a similar population extending device according to an embodiment of the present disclosure.
As shown in fig. 4, the similar population expansion device 400 includes a selection module 410, a definite reception module 420, a filtering module 430, and an expansion module 440.
A selection module 410 for selecting a target scene;
a determining module 420, configured to determine a filtering algorithm corresponding to the target scene;
the filtering module 430 is configured to perform exception filtering on the original user sample through a filtering algorithm to obtain a seed user sample;
and the expanding module 440 is configured to perform similar population expansion on the seed user sample in the target scene to obtain an expanded user sample.
Any number of modules, sub-modules, units, sub-units, or at least part of the functionality of any number thereof according to embodiments of the present disclosure may be implemented in one module. Any one or more of the modules, sub-modules, units, and sub-units according to the embodiments of the present disclosure may be implemented by being split into a plurality of modules. Any one or more of the modules, sub-modules, units, sub-units according to embodiments of the present disclosure may be implemented at least in part as a hardware circuit, such as a Field Programmable Gate Array (FPGA), a Programmable Logic Array (PLA), a system on a chip, a system on a substrate, a system on a package, an Application Specific Integrated Circuit (ASIC), or may be implemented in any other reasonable manner of hardware or firmware by integrating or packaging a circuit, or in any one of or a suitable combination of software, hardware, and firmware implementations. Alternatively, one or more of the modules, sub-modules, units, sub-units according to embodiments of the disclosure may be at least partially implemented as a computer program module, which when executed may perform the corresponding functions.
For example, any number of the selecting module 410, the determining module 420, the filtering module 430, and the expanding module 440 may be combined in one module/unit/subunit to be implemented, or any one of the modules/units/subunits may be split into a plurality of modules/units/subunits. Alternatively, at least part of the functionality of one or more of these modules/units/sub-units may be combined with at least part of the functionality of other modules/units/sub-units and implemented in one module/unit/sub-unit. According to an embodiment of the present disclosure, at least one of the identity module 410, the determination module 420, the filtering module 430, and the expansion module 440 may be implemented at least partially as a hardware circuit, such as a Field Programmable Gate Array (FPGA), a Programmable Logic Array (PLA), a system on a chip, a system on a substrate, a system on a package, an Application Specific Integrated Circuit (ASIC), or may be implemented in any other reasonable manner of integrating or packaging a circuit, as hardware or the same, or in any one of three implementations of software, hardware, and firmware, or in any suitable combination of any of them. Alternatively, at least one of the selecting module 410, the determining module 420, the filtering module 430 and the expanding module 440 may be at least partially implemented as a computer program module, which when executed, may perform a corresponding function.
It should be noted that, in the embodiment of the present disclosure, the similar population expansion device portion corresponds to the similar population expansion method portion in the embodiment of the present disclosure, and the description of the similar population expansion device portion specifically refers to the similar population expansion method portion, and is not repeated herein.
Fig. 5 schematically shows a block diagram of an electronic device adapted to implement the above described method according to an embodiment of the present disclosure. The electronic device shown in fig. 5 is only an example, and should not bring any limitation to the functions and the scope of use of the embodiments of the present disclosure.
As shown in fig. 5, an electronic device 500 according to an embodiment of the present disclosure includes a processor 501 that can perform various appropriate actions and processes according to a program stored in a Read Only Memory (ROM)502 or a program loaded from a storage section 508 into a Random Access Memory (RAM) 503. The processor 501 may comprise, for example, a general purpose microprocessor (e.g., a CPU), an instruction set processor and/or associated chipset, and/or a special purpose microprocessor (e.g., an Application Specific Integrated Circuit (ASIC)), among others. The processor 501 may also include onboard memory for caching purposes. Processor 501 may include a single processing unit or multiple processing units for performing different actions of a method flow according to embodiments of the disclosure.
In the RAM 503, various programs and data necessary for the operation of the system 500 are stored. The processor 501, the ROM502, and the RAM 503 are connected to each other by a bus 504. The processor 501 performs various operations of the method flows according to the embodiments of the present disclosure by executing programs in the ROM502 and/or the RAM 503. Note that the programs may also be stored in one or more memories other than the ROM502 and the RAM 503. The processor 501 may also perform various operations of method flows according to embodiments of the present disclosure by executing programs stored in the one or more memories.
According to an embodiment of the present disclosure, system 500 may also include an input/output (I/O) interface 505, input/output (I/O) interface 505 also being connected to bus 504. The system 500 may also include one or more of the following components connected to the I/O interface 505: an input portion 506 including a keyboard, a mouse, and the like; an output portion 507 including a display such as a Cathode Ray Tube (CRT), a Liquid Crystal Display (LCD), and the like, and a speaker; a storage portion 508 including a hard disk and the like; and a communication section 509 including a network interface card such as a LAN card, a modem, or the like. The communication section 509 performs communication processing via a network such as the internet. The driver 510 is also connected to the I/O interface 505 as necessary. A removable medium 511 such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, or the like is mounted on the drive 510 as necessary, so that a computer program read out therefrom is mounted into the storage section 508 as necessary.
According to embodiments of the present disclosure, method flows according to embodiments of the present disclosure may be implemented as computer software programs. For example, embodiments of the present disclosure include a computer program product comprising a computer program embodied on a computer readable storage medium, the computer program containing program code for performing the method illustrated by the flow chart. In such an embodiment, the computer program may be downloaded and installed from a network through the communication section 509, and/or installed from the removable medium 511. The computer program, when executed by the processor 501, performs the above-described functions defined in the system of the embodiments of the present disclosure. The systems, devices, apparatuses, modules, units, etc. described above may be implemented by computer program modules according to embodiments of the present disclosure.
The present disclosure also provides a computer-readable storage medium, which may be embodied in the device/apparatus/system described in the above embodiments; or may exist alone without being assembled into the device/apparatus/system. The computer-readable storage medium carries one or more programs which, when executed, implement the method according to an embodiment of the disclosure.
According to an embodiment of the present disclosure, the computer-readable storage medium may be a non-volatile computer-readable storage medium. Examples may include, but are not limited to: 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 portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing. In the present disclosure, a computer readable storage medium may be any tangible medium that can contain, or store a program for use by or in connection with an instruction execution system, apparatus, or device.
For example, according to embodiments of the present disclosure, a computer-readable storage medium may include ROM502 and/or RAM 503 and/or one or more memories other than ROM502 and RAM 503 described above.
The flowchart and block diagrams in the figures illustrate the architecture, functionality, and operation of possible implementations of systems, methods and computer program products according to various embodiments of the present disclosure. In this regard, each block in the flowchart or block diagrams may represent a module, segment, or portion of code, which comprises one or more executable instructions for implementing the specified logical function(s). It should also be noted that, 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 or flowchart illustration, and combinations of blocks in the block diagrams 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.
Those skilled in the art will appreciate that various combinations and/or combinations of features recited in the various embodiments and/or claims of the present disclosure can be made, even if such combinations or combinations are not expressly recited in the present disclosure. In particular, various combinations and/or combinations of the features recited in the various embodiments and/or claims of the present disclosure may be made without departing from the spirit or teaching of the present disclosure. All such combinations and/or associations are within the scope of the present disclosure.
The embodiments of the present disclosure have been described above. However, these examples are for illustrative purposes only and are not intended to limit the scope of the present disclosure. Although the embodiments are described separately above, this does not mean that the measures in the embodiments cannot be used advantageously in combination. The scope of the disclosure is defined by the appended claims and equivalents thereof. Various alternatives and modifications can be devised by those skilled in the art without departing from the scope of the present disclosure, and such alternatives and modifications are intended to be within the scope of the present disclosure.
Claims (10)
1. A method of similar population expansion, comprising:
selecting a target scene;
determining a filtering algorithm corresponding to the target scene;
performing exception filtering on the original user sample through the filtering algorithm to obtain a seed user sample;
and under the target scene, performing similar population expansion on the seed user sample to obtain an expanded user sample.
2. The method of claim 1, wherein the determining the filtering algorithm corresponding to the target scene comprises:
acquiring a first user sample from a candidate user sample set based on preset behavior characteristics;
respectively carrying out exception filtering on the first user sample through a plurality of preset algorithms to obtain a plurality of second user samples;
in the target scene, recalling each second user sample in the candidate user sample set through a recall algorithm to obtain a plurality of recalled user samples;
respectively detecting the recall quality of each recall user sample based on the preset behavior characteristics;
and determining a filtering algorithm corresponding to the target scene based on the recall quality, wherein the filtering algorithm is one of the preset algorithms.
3. The method of claim 2, wherein said separately detecting recall quality for each of said recalled user samples based on said preset behavioral characteristics comprises:
respectively counting the number of designated users in each recall user sample, wherein the designated users have preset behavior characteristics;
and comparing the number of all the designated users, and determining an optimal recall user sample, wherein the optimal recall user sample is a recall user sample which contains the largest number of designated users in all the recall user samples.
4. The method of claim 3, wherein said separately detecting a recall quality of each of said recalled user samples based on said preset behavioral characteristics, further comprises:
counting the number of characteristic users in the candidate user sample set except the first user sample, wherein the characteristic users have preset behavior characteristics;
calculating a recall ratio of each of the recalled user samples, the recall ratio being a ratio of the number of the designated users to the number of the feature users;
and comparing all the recall rates to determine an optimal recall user sample, wherein the optimal recall user sample is a recall user sample with the maximum recall rate in all the recall user samples.
5. The method of claim 3 or 4, wherein the determining a filtering algorithm corresponding to the target scene based on the recall quality comprises:
determining a preset algorithm corresponding to the optimal recall user sample;
and determining the preset algorithm as a filtering algorithm corresponding to the target scene.
6. The method of claim 2, wherein the determining the filtering algorithm corresponding to the target scene further comprises:
under the target scene, recalling the first user sample in the candidate user sample set through a recall algorithm to obtain an original recall user sample;
detecting an original recall quality of the original recall user sample based on the preset behavioral characteristics;
comparing the original recall quality to recall qualities of a plurality of the recalled user samples;
if the original recall quality is better than the recall quality of all the recall user samples, the target scene has no corresponding filtering algorithm.
7. The method of claim 1, wherein the user sample is a collection of user characteristic data, the user characteristic data including user portrait information and user behavioral information.
8. A similar population extension device comprising:
a selection module for selecting a target scene;
the determining module is used for determining a filtering algorithm corresponding to the target scene;
the filtering module is used for carrying out exception filtering on the original user sample through the filtering algorithm to obtain a seed user sample;
and the expansion module is used for performing similar population expansion on the seed user sample in the target scene to obtain an expanded user sample.
9. An electronic device, comprising:
one or more processors;
a memory for storing one or more programs,
wherein the one or more programs, when executed by the one or more processors, cause the one or more processors to implement the method of any of claims 1-7.
10. A computer readable storage medium having stored thereon executable instructions which, when executed by a processor, cause the processor to carry out the method of any one of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110278343.4A CN115082844A (en) | 2021-03-15 | 2021-03-15 | Similar crowd extension method and device, electronic equipment and readable storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110278343.4A CN115082844A (en) | 2021-03-15 | 2021-03-15 | Similar crowd extension method and device, electronic equipment and readable storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115082844A true CN115082844A (en) | 2022-09-20 |
Family
ID=83245676
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110278343.4A Pending CN115082844A (en) | 2021-03-15 | 2021-03-15 | Similar crowd extension method and device, electronic equipment and readable storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115082844A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117172855A (en) * | 2023-09-20 | 2023-12-05 | 南通捷米科技有限公司 | Elevator advertisement playing method and system based on face recognition |
-
2021
- 2021-03-15 CN CN202110278343.4A patent/CN115082844A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117172855A (en) * | 2023-09-20 | 2023-12-05 | 南通捷米科技有限公司 | Elevator advertisement playing method and system based on face recognition |
| CN117172855B (en) * | 2023-09-20 | 2024-05-14 | 南通捷米科技有限公司 | Elevator advertisement playing method and system based on face recognition |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US11847663B2 (en) | Subscription churn prediction | |
| CN110647921B (en) | User behavior prediction method, device, equipment and storage medium | |
| CN107808314B (en) | User recommendation method and device | |
| CN110363604B (en) | Page generation method and device | |
| CN111523964A (en) | Clustering-based recall method and apparatus, electronic device and readable storage medium | |
| US20210103957A1 (en) | Generating personalized banner images using machine learning | |
| CN110473042B (en) | Method and device for acquiring information | |
| CN112598472A (en) | Product recommendation method, device, system, medium and program product | |
| CN110245684B (en) | Data processing method, electronic device, and medium | |
| US9201968B2 (en) | System and method for finding mood-dependent top selling/rated lists | |
| CN115082844A (en) | Similar crowd extension method and device, electronic equipment and readable storage medium | |
| CN111340566B (en) | Commodity classification method and device, electronic equipment and storage medium | |
| KR20220122414A (en) | Method of providing curating service on subscription platform using my data of artificial intelligence-based users and sellers | |
| CN106204163B (en) | Method and device for determining user attribute characteristics | |
| CN111177564A (en) | Product recommendation method and device | |
| WO2020150597A1 (en) | Systems and methods for entity performance and risk scoring | |
| CN112330382A (en) | Item recommendation method and device, computing equipment and medium | |
| CN107357847B (en) | Data processing method and device | |
| CN110827044A (en) | Method and device for extracting user interest mode | |
| CN110689032A (en) | Data processing method and system, computer system and computer readable storage medium | |
| CN110750238B (en) | Method and device for determining product demand and electronic equipment | |
| US10904346B2 (en) | Weighted digital image object tagging | |
| CN113822691A (en) | User account identification method, device, system and medium | |
| CN113792952A (en) | Method and apparatus for generating a model | |
| CN113516524A (en) | Method and device for pushing information |
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 |