CN111340522A - Resource recommendation method, device, server and storage medium - Google Patents
Resource recommendation method, device, server and storage medium Download PDFInfo
- Publication number
- CN111340522A CN111340522A CN201911398736.8A CN201911398736A CN111340522A CN 111340522 A CN111340522 A CN 111340522A CN 201911398736 A CN201911398736 A CN 201911398736A CN 111340522 A CN111340522 A CN 111340522A
- Authority
- CN
- China
- Prior art keywords
- candidate resource
- resource object
- feature
- features
- objects
- 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
- 238000000034 method Methods 0.000 title claims abstract description 41
- 238000013528 artificial neural network Methods 0.000 claims description 17
- 238000004590 computer program Methods 0.000 claims description 14
- 230000010354 integration Effects 0.000 claims description 4
- 238000000605 extraction Methods 0.000 claims description 3
- 238000007781 pre-processing Methods 0.000 claims description 2
- 239000013598 vector Substances 0.000 description 16
- 239000010410 layer Substances 0.000 description 15
- 238000010586 diagram Methods 0.000 description 14
- 238000012545 processing Methods 0.000 description 10
- 230000006870 function Effects 0.000 description 9
- 230000008569 process Effects 0.000 description 9
- 230000000694 effects Effects 0.000 description 5
- 238000012986 modification Methods 0.000 description 5
- 230000004048 modification Effects 0.000 description 5
- 230000006399 behavior Effects 0.000 description 4
- 238000012549 training Methods 0.000 description 4
- 230000005540 biological transmission Effects 0.000 description 3
- 238000012163 sequencing technique Methods 0.000 description 3
- 238000011156 evaluation Methods 0.000 description 2
- 230000036541 health Effects 0.000 description 2
- 230000006872 improvement Effects 0.000 description 2
- 239000011159 matrix material Substances 0.000 description 2
- 230000004913 activation Effects 0.000 description 1
- 230000003542 behavioural effect Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 239000004566 building material Substances 0.000 description 1
- 238000004422 calculation algorithm Methods 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 230000002452 interceptive effect Effects 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 239000008267 milk Substances 0.000 description 1
- 230000002093 peripheral effect Effects 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 230000004044 response Effects 0.000 description 1
- 239000002356 single layer Substances 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
- G06Q30/0241—Advertisements
- G06Q30/0251—Targeted advertisements
- G06Q30/0269—Targeted advertisements based on user profile or attribute
- G06Q30/0271—Personalized advertisement
-
- 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
- G06Q30/0207—Discounts or incentives, e.g. coupons or rebates
- G06Q30/0224—Discounts or incentives, e.g. coupons or rebates based on user history
-
- 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
- G06Q30/0207—Discounts or incentives, e.g. coupons or rebates
- G06Q30/0239—Online discounts or incentives
-
- 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
- G06Q30/0241—Advertisements
- G06Q30/0251—Targeted advertisements
- G06Q30/0255—Targeted advertisements based on user history
-
- 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
- G06Q30/0241—Advertisements
- G06Q30/0251—Targeted advertisements
- G06Q30/0259—Targeted advertisements based on store location
-
- 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
- G06Q30/0241—Advertisements
- G06Q30/0251—Targeted advertisements
- G06Q30/0261—Targeted advertisements based on user location
Landscapes
- Business, Economics & Management (AREA)
- Engineering & Computer Science (AREA)
- Accounting & Taxation (AREA)
- Development Economics (AREA)
- Strategic Management (AREA)
- Finance (AREA)
- Game Theory and Decision Science (AREA)
- Entrepreneurship & Innovation (AREA)
- Economics (AREA)
- Marketing (AREA)
- Physics & Mathematics (AREA)
- General Business, Economics & Management (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The embodiment of the specification provides a resource recommendation method, a resource recommendation device, a server and a storage medium: performing feature intersection on the resource object features and the user features of the candidate resource objects through a combined pre-estimation model to obtain combined features containing multi-order intersection features, performing user click rate pre-estimation on the candidate resource objects according to the combined features to obtain CTR pre-estimation results of the candidate resource objects, determining target resource objects from the M candidate resource objects according to the CTR pre-estimation results of each candidate resource object in the M candidate resource objects, and recommending the determined target resource objects to target users.
Description
Technical Field
The embodiment of the specification relates to the technical field of internet, in particular to a resource recommendation method, a resource recommendation device, a server and a storage medium.
Background
In the field of computers, there are various electronic resources that can be used by corresponding objects to implement the functions of the electronic resources. The interests and hobbies of different users are different, and thus the preferences for electronic resources are different. Such as: shared memory, electronic data (such as video, audio, documents), electronic coupons, and the like.
For example, by taking coupons as an example, electronic commerce, O2O and other online commerce are vigorously generated, and various platforms build abundant basic capability for supporting better operation of merchants. The e-commerce platform will typically stimulate the consumer to consume through coupons. And the E-commerce platform issues the coupon activity to the online merchant to attract the merchants. The online merchant applies for providing the coupon through online activities. Different users have different interests and hobbies, so different commodity preferences exist, and different coupons need to be recommended to different users in an individualized manner so as to recommend the most suitable coupons to the users.
Disclosure of Invention
Embodiments of the present specification provide a resource recommendation method, apparatus, server, and storage medium to improve accuracy and rationality of resource recommendation, and further avoid repeated resource recommendation to a user due to inaccurate resource recommendation, thereby saving network transmission resources.
In a first aspect, an embodiment of the present specification provides a resource recommendation method, including: acquiring user characteristics of a target user and resource object characteristics of M candidate resource objects, wherein M is a positive integer; performing CTR estimation on the M candidate resource objects respectively to obtain a CTR estimation result of each candidate resource object in the M candidate resource objects; wherein the performing CTR pre-estimation on the M candidate resource objects respectively comprises: for each candidate resource object, performing feature intersection on resource object features of the candidate resource object and user features through a combined pre-estimation model to obtain combined features of the candidate resource object, and performing CTR pre-estimation on the candidate resource object according to the combined features to obtain a CTR pre-estimation result of the candidate resource object, wherein the combined features comprise linear features, in-group intersection features and inter-group intersection features obtained according to a multi-layer factorization machine submodel of the combined pre-estimation model, and high-order intersection features obtained according to a neural network submodel of the combined pre-estimation model; and determining a target resource object from the M candidate resource objects according to the CTR estimation result of each candidate resource object in the M candidate resource objects, and recommending the target resource object to the target user.
In a second aspect, an embodiment of the present specification provides a resource recommendation device, including: the characteristic extraction unit is used for acquiring the user characteristics of a target user and the resource object characteristics of M candidate resource objects, wherein M is a positive integer; the CTR pre-estimation unit is used for respectively carrying out CTR pre-estimation on the M candidate resource objects to obtain a CTR pre-estimation result of each candidate resource object in the M candidate resource objects; wherein the performing CTR pre-estimation on the M candidate resource objects respectively comprises: for each candidate resource object, performing feature intersection on resource object features of the candidate resource object and user features through a combined pre-estimation model to obtain combined features of the candidate resource object, and performing CTR pre-estimation on the candidate resource object according to the combined features to obtain a CTR pre-estimation result of the candidate resource object, wherein the combined features comprise linear features, in-group intersection features and inter-group intersection features obtained according to a multi-layer factorization machine submodel of the combined pre-estimation model, and high-order intersection features obtained according to a neural network submodel of the combined pre-estimation model; and the resource recommending unit is used for determining a target resource object from the M candidate resource objects according to the CTR estimation result of each candidate resource object in the M candidate resource objects and recommending the target resource object to the target user.
In a third aspect, an embodiment of the present specification provides a server, including a memory, a processor, and a computer program stored on the memory and executable on the processor, where the processor implements the steps of any one of the resource recommendation methods when executing the program.
In a fourth aspect, the present specification provides a computer-readable storage medium, on which a computer program is stored, where the computer program, when executed by a processor, implements the steps of any of the resource recommendation methods described above.
The technical scheme provided by the embodiment of the specification at least realizes the following beneficial effects or advantages:
performing feature crossing on resource object features and user features of the candidate resource objects through a combined estimation model, wherein the obtained combined features comprise linear features, in-group crossing features and inter-group crossing features obtained according to a multi-layer factorization model of the combined estimation model, and high-order crossing features obtained according to a neural network sub-model of the combined estimation model; performing CTR prediction according to the combination characteristics of the candidate resource objects to obtain CTR prediction results of the candidate resource objects, and recommending target resource objects to target users from target resource objects in the M candidate resource objects according to the CTR prediction results of each candidate resource object in the M candidate resource objects. By the technical scheme, the CTR prediction of the candidate resource object can be simultaneously influenced by the linear feature of the first order, the cross feature in the group of the second order, the cross feature between the groups of the third order and the fourth order, and the combination feature of the higher order. The characteristic information for predicting the click rate of the candidate resource object is enriched, and the characteristics of the candidate resource object can be reflected by more-order cross characteristics, so that the click rate of the candidate resource object can be predicted more accurately, the accuracy and the reasonability of recommending resources to a user can be improved, the resources can be recommended to the user accurately at one time, the resources are prevented from being recommended to the same user for multiple times, and the network transmission resources are saved.
Drawings
FIG. 1 is a diagram of a system architecture to which a resource recommendation method in an embodiment of the present specification is applied;
FIG. 2 is a flowchart illustrating a resource recommendation method according to a first aspect of an embodiment of the present disclosure;
FIG. 3 is a diagram illustrating an architecture of an FM submodel used in an embodiment of a resource recommendation method in the present disclosure;
FIG. 4 is a schematic structural diagram of a resource recommendation device in a second aspect of the embodiments of the present specification;
fig. 5 is a schematic structural diagram of a resource recommendation server in the third aspect of the embodiment of the present specification.
Detailed Description
In order to better understand the technical solutions, the technical solutions of the embodiments of the present specification are described in detail below with reference to the drawings and specific embodiments, and it should be understood that the specific features of the embodiments and embodiments of the present specification are detailed descriptions of the technical solutions of the embodiments of the present specification, and are not limitations of the technical solutions of the present specification, and the technical features of the embodiments and embodiments of the present specification may be combined with each other without conflict.
Referring to fig. 1, fig. 1 is a system architecture diagram of a resource recommendation method applied in an embodiment of the present specification, where the resource recommendation method includes:
the merchant device 10, the merchant device 10 includes a plurality of merchant devices (2 merchant devices 10 are illustrated in fig. 1), and each merchant device 10 provides a candidate resource object. The user equipment 20, the user equipment 20 includes a plurality of (fig. 1 illustrates schematically as 4 user equipments), and each user equipment 20 acquires the candidate resource object and uses the candidate resource object. The server 30 may be a server of an e-commerce platform. The server 30 obtains a candidate resource object for the target user, performs feature crossing according to the user feature of the target user and the resource object feature of the candidate resource object to obtain a combined feature, performs CTR evaluation on the candidate resource object according to the combined feature, and determines whether to recommend the candidate resource object to the target user according to the CTR evaluation result of the candidate resource object.
The merchant device 10 and the user device 20 in this embodiment may specifically be entity devices of smart phones, desktop computers, tablet computers, notebook computers, digital assistants, smart wearable devices, and the like, and the operating systems that run may include, but are not limited to, an android system, an IOS system, linux, windows, and the like. The server 30 may comprise a server operating independently, or a distributed server, or a server cluster consisting of a plurality of servers.
In embodiments of the present specification, the terms "comprises," "comprising," and "having," and any variations thereof, are intended to cover non-exclusive inclusions, such that a process, method, system, article, or server that comprises a list of steps or elements is not necessarily limited to those steps or elements expressly listed, but may include other steps or elements not expressly listed or inherent to such process, method, article, or apparatus.
In a first aspect, the embodiments of the present specification provide a resource recommendation method, which is applied in the server 30 of the system architecture diagram shown in fig. 1. Next, a resource recommendation method provided in an embodiment of the present specification is described with reference to fig. 2.
As shown in fig. 2, the resource recommendation method provided in the embodiment of the present specification includes the following steps:
s201, obtaining user characteristics of a target user and resource object characteristics of M candidate resource objects, wherein M is a positive integer.
In this specification embodiment, the M candidate resource objects are resource objects recalled for a target user. Specifically, the one or more candidate resource objects provided to the target user may be recalled Based on an LBS (Location Based Service) recall policy. The number M of the recall candidate resource objects is determined by the number of the resource objects around the current real-time position of the target user and a distance value preset in the LBS recall strategy. For example, it may be possible to recall candidate resource objects within 10km of the current real-time location of the target user.
Specifically, M candidate resource objects are obtained by: and according to the LBS recall strategy, recalling the candidate resource objects provided by the merchant equipment within the preset distance range of the target user to obtain M candidate resource objects for providing for the target user.
In the embodiment of the present specification, different application scenarios correspond to different candidate resource objects. For a merchant recommendation scenario in which the application scenario is an APP of an offline merchant recommendation type, the candidate resource object may specifically be: and the equity resource objects provided by the offline merchants to the users, such as coupons, redemption tickets, discount tickets and the like provided by the offline merchants to the users. The coupons offered by merchant devices within a preset distance range of the target user's current real-time location are recalled. For example, coupons offered by merchant devices within 1000m of the target user's current real-time location are recalled.
Of course, the candidate resource object may also be an introduction information file provided by an offline merchant, and the introduction information file sent by more than one merchant device within a preset distance range of the current real-time location of the target user is recalled. For the case that the application scenario is the on-map function point recommendation of the map APP, the candidate resource objects are some type or multiple types of function points within a preset distance range of the target user. Such as gas stations, high-speed service stations, supermarkets, hotels, etc., within a preset distance range of the target user. For the short video recommendation scene, if the candidate resource object is a short video, recalling the short video published by other users within the preset example range of the current real-time position of the target user.
In the specific implementation process, the multidimensional user characteristics of the target user include: age characteristics, gender characteristics, and user preference characteristics of the target user, among others. Specifically, the user preference characteristics of the target user are obtained by processing the historical behavior record data of the target user using the APP. Wherein the historical behavior record data of the target user comprises: and recording historical behaviors of the target user for getting and verifying resource objects and the like in a historical time period, and regarding the coupons, the historical behaviors of the target user for getting and verifying the coupons in the historical time period. By way of example, historical behavioral record data includes: the recorded video types, browsing durations, etc. of short videos browsed by the target user in a historical period of time.
And age characteristics, gender, etc. are derived from user information stored on the server. The application scenarios are different, and the types of the corresponding user preference features are completely or partially different. For an APP with an application scenario of an offline merchant recommendation type, the user preference information may be a consumption amount level (which segment of a plurality of preset consumption amount segments belongs to), and a consumption type preference (one or more of consumption types such as food, movie show, medical health, leisure and entertainment, shopping, sports and fitness, and living trip). Regarding map-based APPs in an application scenario, the user preference information is specifically an activity trajectory preference (the activity trajectory preference is one or more of the activity trajectory preferences of food, movie performance, medical health, leisure and entertainment, shopping, sports and fitness, life and travel, and the like).
Under different application scenarios, the resource object characteristics of the obtained candidate resource object may be completely or partially different, taking the candidate resource object as a coupon as an example, the resource object characteristics include: the coupon belongs to the industry, such as clothing, food, building materials or digital products; the coupon has the advantages of discount amount, discount value such as discount, discount threshold, general degree of discount and the like. It should be noted that the offer amount and the discount represent the amount or the proportion of the actual discount of the offer, and the discount threshold represents the difficulty of using the coupon, for example: can be determined by the threshold condition of the full discount coupon and the original price of the commodity for exchanging the coupon. For example, if a full discount coupon is full 100 minus 10, the offer threshold is 100. For example, a redemption ticket is redeemed for goods under 100 dollars, and the coupon threshold is 100 dollars. The goodness of popularity characterizes the size of the application scope of the coupon, such as: the whole-field general ticket is applicable to all commodities, but the egg-milk exchange ticket is only applicable to specified commodities.
The resource object characteristics when the candidate resource object is taken as the coupon are only used for illustration. When the candidate resource object is of another type, what the resource object features include can be determined according to the actual resource attribute, and for the sake of brevity of the description, details are not repeated here.
S202, performing CTR estimation on the M candidate resource objects respectively to obtain a CTR estimation result of each candidate resource object in the M candidate resource objects.
Specifically, the performing CTR prediction on the M candidate resource objects respectively includes: for each candidate resource object, CTR prediction is performed through the following steps S2021 to S2022, so as to obtain M CTR prediction results corresponding to M candidate resource objects:
s2021, performing feature crossing on resource object features of the candidate resource objects and user features through a combined prediction model to obtain combined features of the candidate resource objects, wherein the combined features comprise linear features, in-group crossing features and inter-group crossing features which are obtained according to a multi-layer factorization machine submodel of the combined prediction model, and high-order crossing features which are obtained according to a neural network submodel of the combined prediction model.
In the embodiment of the present specification, an improved combination model obtained by performing hierarchical improvement on an original combination model includes: a multi-layer factorization machine submodel and a neural network submodel. The original combination model is a factorization machine submodel which comprises a neural network submodel and supports single-layer feature intersection. And training the improved combination model obtained through the hierarchical improvement based on training sample data to adjust each weight matrix of the improved combination model to obtain a combination estimation model. The multi-layer factorization machine submodel comprises at least two layers of feature intersections. In a specific implementation process, the combined pre-estimation model may be: and the combination estimation model is obtained by training through a Deep Factorization (Deep Factorization) model or an improved combination model obtained by improving the Xdeep FM in a layering way. The pre-estimation model obtained by training combines the advantages of the breadth model and the depth model.
Regarding the combined estimation model obtained by hierarchically improving Deep FM, the combined estimation model is a parallel structure formed by an FM (factitio Networks) submodel and a DNN (Deep Neural Networks) submodel, and the FM submodel and the DNN submodel share the same input: the embedding feature vectors are used after grouping and embedding processing is carried out on the user features and the resource object features of the M candidate resource objects. The multilayer FM submodel forms a first order linear feature and second, third, and fourth order cross features by performing feature crossing (multiplication or cartesian product) on the features of the N original feature groups. The DNN submodel forms higher order cross features (fifth, sixth order cross features, etc.) by feature crossing N original feature groupings.
The process of constructing cross features and inter features within a group based on a multi-layer factorization machine submodel in the embodiments of the present description is described in more detail below with reference to fig. 3 (one circle in the figure represents a one-dimensional feature): the method specifically comprises the following steps of 1-4:
step 1, grouping and embedding the user characteristics and the resource object characteristics of the candidate resource object to obtain N original characteristic groups of the candidate resource object, wherein N is a positive integer.
Specifically, the features belonging to the same type in the obtained multi-dimensional user features and multi-dimensional resource object features of the candidate resource object are divided into the same original feature group, so that N original feature groups are obtained, and the number of the original feature groups is determined by the number of the feature types. The following m-dimensional raw features field _0, field _1, …, field _ m are grouped according to the feature type of the m-dimensional raw features, i.e. various user features and various resource object features. The following N feature groupings are formed: group _0, group _1, …, group _ N, and embeddings (i.e. embedding processing) are performed on the original features of each dimension in each feature group, so as to correspondingly obtain N original feature groups. Through the processing of the embeddings algorithm, dimension reduction is carried out on m-dimensional original features which are sparse features, specifically, the dimension is reduced through weight matrix calculation of an embedding layer, and an embedding feature vector in each original feature group in N original feature groups is obtained.
And 2, performing in-group feature crossing on the N original feature groups through a multi-layer factorization machine submodel to obtain N in-group crossing results of the candidate resource object.
Specifically, performing intra-group feature crossing on embedding feature vectors in N original feature groups to obtain N groups of intra-group crossing results, that is: intersecting the embedding characteristic vectors in the group _0 to obtain a first group of in-group intersection results; intersecting the embedding characteristic vectors in gropu _1 to obtain a second group of in-group intersection results; …, intersecting the embedding feature vectors in the group _ m to obtain a third group inner intersection result. It should be noted that feature crossing is performed on the embedding feature vectors in each of the N original feature groups, and the obtained crossing results in the N groups of groups are all second-order crossing features.
And 3, performing feature addition, interclass feature crossing and integration of original features in the N original feature groups on the N groups of interclass crossing results and the N original feature groups of the candidate resource object through a multilayer factorization machine submodel to generate linear features, interclass crossing features and interclass crossing features of the candidate resource object.
Specifically, step 3 above is described in more detail in conjunction with fig. 3:
and 3A, embedding the N groups of in-group crossing results of the candidate resource object, and embedding the original features of the N original feature groups to form N intermediate feature groups of the candidate resource object.
Specifically, embedding processing is performed on each group of intra-group crossing results through an intermediate embedding layer, an intermediate feature group is formed by an embedding feature vector obtained after the intra-group crossing results are subjected to the embedding and an embedding feature vector in an original feature group generating the intra-group crossing results, and therefore N intermediate feature groups are formed by corresponding N original feature groups and N original feature groups. It can be seen that the same intermediate feature group contains both the embedding feature vector of the cross feature (second order) in the group and the embedding feature vector of the original feature (first order). The formed embedding feature vectors in the N intermediate feature groups are features for interclass crossing.
And 3B, performing interclass feature crossing on the N intermediate feature groups of the candidate resource object to form interclass crossing features of the candidate resource object. Specifically, interclass feature crossing is carried out on the embedding feature vectors of N intermediate feature groups, the interclass feature of the third order and the interclass feature of the fourth order are generated,
and step 3C: and adding the characteristics of the N intermediate characteristic groups of the candidate resource object to obtain the cross characteristics in the group of the candidate resource object.
And then, integrating the second-order interclass cross feature, the third-order interclass cross feature and the fourth-order interclass cross feature with the first-order imbedding feature vector. And enabling the first-order linear features (including the original features and the linear combination features), the second-order in-group cross features and the third-order and fourth-order inter-group cross features generated by the final multilayer factorization machine submodel.
And 4, performing feature crossing on the N original feature groups through the neural network submodel to construct high-order crossing features of the candidate resource object.
In the embodiment of the present specification, for each candidate resource object, the high-order cross feature of the candidate resource object is obtained through a neural network submodel, and the specific process includes the following steps:
inputting the embedding feature vectors in the N original feature groups into a neural network sub-model (such as a DNN model), and transmitting the embedding feature vectors downwards layer by layer through the features to obtain the high-order cross features of the candidate resource object. In the embodiment of the present specification, the high-order cross features include five orders and more, and specifically, the high-order cross features of several orders are used, and the structure of the neural network submodel may be set according to actual requirements.
S2022, performing CTR (Click Through Rate) estimation on the candidate resource object according to the combination characteristics to obtain a CTR estimation result of the candidate resource object.
Outputting the multi-level combination characteristics obtained by the two sub-models through an Output unit (Output Units) to obtain the CTR estimation result of the obtained candidate resource object, namely: and inputting the multi-order combined characteristics into an output unit, and obtaining a CTR (computer to computer) estimation result based on the sidmoid activation function.
S203, determining a target resource object from the M candidate resource objects according to the CTR estimation result of each candidate resource object in the M candidate resource objects, and recommending the target resource object to a target user.
It should be noted that the CTR estimation result of each candidate resource object in the M candidate resource objects includes a click rate estimation result of each candidate resource object. In an optional implementation manner, resource sorting is performed on the M candidate resource objects according to the CTR prediction result of each candidate resource object in the M candidate resource objects. And determining a plurality of candidate resource objects ranked in the front according to the resource ranking result and the number of the display pit positions, and recommending the determined plurality of candidate resource objects to the target user.
In another optional embodiment, the method specifically comprises the following steps: acquiring priority intervention information of the M candidate resource objects; integrating the priority intervention information of the M candidate resource objects and the CTR estimation result of each candidate resource object in the M candidate resource objects to obtain the final recommendation sequence of the M candidate resource objects; and determining a target resource object from the M candidate resource objects according to the number of the display pit positions and the final recommended sequence.
Specifically, the priority intervention information is to improve the recommendation priority of the specific resource object, and in the implementation process, the specific resource object may be a coupon of a merchant who has delivered the advertisement, that is, an advertisement coupon. In the case of a map scene, a particular resource object may be a user-specified type of function point, such as a gas station. For example, taking an advertising coupon as an example: if the advertisement coupon is in the recommendation sequencing before the recommendation priority of the advertisement coupon is not improved, improving the recommendation priority of the advertisement coupon to the first sequencing to obtain the final recommendation sequencing; if the advertising coupon is not within the recommendation ranking before the recommendation priority of the advertising coupon is not increased, the recommendation priority of the advertising coupon is increased to the end within the recommendation ranking to arrive at a final recommendation ranking.
In the following, an interactive embodiment is given by taking a coupon as an example: in the process of using the offline merchant APP, the user A recalls all coupons provided by all merchants within 1000m of the current real-time location (mall B) of the user A as candidate coupons in response to the search operation of the user A or without any user operation trigger when arriving at the mall B, and if so, recalls 30 candidate coupons provided by 20 offline merchants. And extracting the coupon characteristics of all the recalled candidate coupons and the user characteristics of the user A, performing characteristic intersection on the coupon characteristics of each candidate coupon and the user characteristics of the user A through the combined estimation model to obtain the combined characteristics of the candidate coupons including first-order linear characteristics and multi-order intersection characteristics, and performing click rate estimation on the candidate coupons according to the respective combined characteristics of the candidate coupons to obtain the click rate estimation result of each candidate coupon in 30 candidate coupons. And finally, sorting the 30 candidate coupons according to the click rate, and recommending 3 coupons sorted in the top 3 to the user A if 5 pit positions are displayed.
By the technical scheme, the CTR prediction of the candidate resource object can be simultaneously influenced by the linear feature of the first order, the cross feature in the group of the second order, the cross feature between the groups of the third order and the fourth order, and the combination feature of the higher order. The characteristic information is enriched, so that the characteristics of resources can be better reflected, the click rate of each recalled candidate coupon for a target user can be more accurately predicted, a plurality of coupons ranked in the front are recommended to the user based on the click rate ranking result, the accuracy and the reasonability of recommending the resources to the user can be improved, the success of recommending the resources to the user at one time is ensured, and the repeated invalid recommendation is avoided.
In a second aspect, based on the same inventive concept as the resource recommendation method in the foregoing embodiments, an embodiment of this specification provides a resource recommendation device, which is shown with reference to fig. 4 and includes:
a feature extraction unit 401, configured to obtain a user feature of a target user and resource object features of M candidate resource objects, where M is a positive integer;
a CTR predicting unit 402, configured to perform CTR prediction on the M candidate resource objects, respectively, to obtain a CTR prediction result of each candidate resource object in the M candidate resource objects:
wherein the performing CTR pre-estimation on the M candidate resource objects respectively includes: for each candidate resource object, performing feature intersection on resource object features of the candidate resource object and user features through a combined pre-estimation model to obtain combined features of the candidate resource object, and performing CTR pre-estimation on the candidate resource object according to the combined features to obtain a CTR pre-estimation result of the candidate resource object, wherein the combined features comprise linear features, in-group intersection features and inter-group intersection features obtained according to a multi-layer factorization machine submodel of the combined pre-estimation model, and high-order intersection features obtained according to a neural network submodel of the combined pre-estimation model;
a resource recommending unit 403, configured to determine a target resource object from the M candidate resource objects according to the CTR estimation result of each candidate resource object in the M candidate resource objects, and recommend the target resource object to the target user.
In an optional implementation manner, the CTR predicting unit includes:
the characteristic preprocessing subunit is used for grouping and embedding the user characteristics and the resource object characteristics of the candidate resource object to obtain N original characteristic groups of the candidate resource object, wherein N is a positive integer;
the intra-group cross subunit is used for performing intra-group feature cross on the N original feature groups through the multilayer factorization machine submodel to obtain N groups of intra-group cross results of the candidate resource object;
an interclass crossing subunit, configured to perform feature addition, interclass feature crossing, and integration of original features in the N original feature groups on the N interclass crossing results and the N original feature groups of the candidate resource object through the multi-layer factorization machine submodel, and generate a linear feature, an interclass crossing feature, and an interclass crossing feature of the candidate resource object;
and the high-order cross subunit is used for performing feature cross on the N original feature groups through the neural network submodel to construct high-order cross features of the candidate resource object.
In an optional implementation manner, the inter-group cross subunit is specifically configured to:
embedding the N groups of in-group crossing results of the candidate resource object, and embedding the original features of the N original feature groups to form N intermediate feature groups of the candidate resource object;
performing interclass feature crossing on the N intermediate feature groups of the candidate resource object to form interclass crossing features of the candidate resource object;
and adding the characteristics of the N intermediate characteristic groups of the candidate resource object to obtain the cross characteristics in the group of the candidate resource object.
In an optional implementation manner, the apparatus further includes a resource recall unit, configured to obtain the M candidate resource objects by:
and recalling candidate resource objects provided by the merchant equipment within a preset distance range of the real-time position of the target user according to a location-based service LBS recall strategy to obtain the M candidate resource objects.
In an optional implementation manner, the resource recommending unit 403 is specifically configured to:
acquiring priority intervention information of the M candidate resource objects;
integrating the priority intervention information of the M candidate resource objects and the CTR pre-estimation result of each candidate resource object in the M candidate resource objects to obtain the final recommendation sequence of the M candidate resource objects;
and determining the target resource object from the M candidate resource objects according to the number of the display pit bits and the final recommended sequence.
In a third aspect, based on the same inventive concept as the resource recommendation method in the foregoing embodiments, an embodiment of this specification further provides a server, as shown in fig. 5, including a memory 504, a processor 502, and a computer program stored on the memory 504 and executable on the processor 502, where the processor 502 executes the computer program to implement the steps of any of the embodiments of the resource recommendation method.
Where in fig. 5 a bus architecture (represented by bus 500) is shown, bus 500 may include any number of interconnected buses and bridges, and bus 500 links together various circuits including one or more processors, represented by processor 502, and memory, represented by memory 504. The bus 500 may also link together various other circuits such as peripherals, voltage regulators, power management circuits, and the like, which are well known in the art, and therefore, will not be described any further herein. A bus interface 506 provides an interface between the bus 500 and the receiver 501 and transmitter 503. The receiver 501 and the transmitter 503 may be the same element, i.e. a transceiver, providing a means for communicating with various other apparatus over a transmission medium. The processor 502 is responsible for managing the bus 500 and general processing, and the memory 504 may be used for storing data used by the processor 502 in performing operations.
In a fourth aspect, based on the inventive concept of resource recommendation in the foregoing embodiments, this specification embodiment further provides a computer-readable storage medium, on which a computer program is stored, and the computer program, when executed by a processor, implements the steps of any of the foregoing embodiments of the resource recommendation method.
The description has been presented with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the description. It will be understood that each flow and/or block of the flow diagrams and/or block diagrams, and combinations of flows and/or blocks in the flow diagrams and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, 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 specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
While preferred embodiments of the present specification have been described, additional variations and modifications in those embodiments may occur to those skilled in the art once they learn of the basic inventive concepts. Therefore, it is intended that the appended claims be interpreted as including preferred embodiments and all changes and modifications that fall within the scope of the specification.
It will be apparent to those skilled in the art that various changes and modifications may be made in the present specification without departing from the spirit and scope of the specification. Thus, if such modifications and variations of the present specification fall within the scope of the claims of the present specification and their equivalents, the specification is intended to include such modifications and variations.
Claims (12)
1. A resource recommendation method, comprising:
acquiring user characteristics of a target user and resource object characteristics of M candidate resource objects, wherein M is a positive integer;
performing CTR estimation on the M candidate resource objects respectively to obtain a CTR estimation result of each candidate resource object in the M candidate resource objects;
wherein the performing CTR pre-estimation on the M candidate resource objects respectively comprises: for each candidate resource object, performing feature intersection on resource object features of the candidate resource object and user features through a combined pre-estimation model to obtain combined features of the candidate resource object, and performing CTR pre-estimation on the candidate resource object according to the combined features to obtain a CTR pre-estimation result of the candidate resource object, wherein the combined features comprise linear features, in-group intersection features and inter-group intersection features obtained according to a multi-layer factorization machine submodel of the combined pre-estimation model, and high-order intersection features obtained according to a neural network submodel of the combined pre-estimation model;
and determining a target resource object from the M candidate resource objects according to the CTR estimation result of each candidate resource object in the M candidate resource objects, and recommending the target resource object to the target user.
2. The method of claim 1, wherein the performing feature intersection on the resource object feature of the candidate resource object and the user feature through the combined predictive model to obtain the combined feature of the candidate resource object comprises:
grouping and embedding the user characteristics and the resource object characteristics of the candidate resource object to obtain N original characteristic groups of the candidate resource object, wherein N is a positive integer;
performing intra-group feature crossing on the N original feature groups through the multilayer factorization machine submodel to obtain N groups of intra-group crossing results of the candidate resource object;
performing feature addition, inter-group feature intersection and integration on the N groups of inter-group intersection results of the candidate resource object and the N original feature groups through the multi-layer factorization machine submodel to generate linear features, inter-group intersection features and inter-group intersection features of the candidate resource object;
and performing feature intersection on the N original feature groups through the neural network submodel to construct a high-order intersection feature of the candidate resource object.
3. The method of claim 2, said performing feature addition and inter-group feature intersection on said N groups of intra-group intersection results of said candidate resource object and said N original feature groups by said multi-layer factorization machine submodel, comprising:
embedding the N groups of in-group crossing results of the candidate resource object, and embedding the original features of the N original feature groups to form N intermediate feature groups of the candidate resource object;
performing interclass feature crossing on the N intermediate feature groups of the candidate resource object to form interclass crossing features of the candidate resource object;
and adding the characteristics of the N intermediate characteristic groups of the candidate resource object to obtain the cross characteristics in the group of the candidate resource object.
4. The method of claim 1, the step of obtaining the M candidate resource objects comprising:
and recalling candidate resource objects provided by the merchant equipment within a preset distance range of the real-time position of the target user according to a location-based service LBS recall strategy to obtain the M candidate resource objects.
5. The method according to any of claims 1-4, wherein said determining a target resource object from said M candidate resource objects based on the CTR predictor of each of said M candidate resource objects comprises:
acquiring priority intervention information of the M candidate resource objects;
integrating the priority intervention information of the M candidate resource objects and the CTR pre-estimation result of each candidate resource object in the M candidate resource objects to obtain the final recommendation sequence of the M candidate resource objects;
and determining the target resource object from the M candidate resource objects according to the number of the display pit bits and the final recommended sequence.
6. A resource recommendation device, comprising:
the characteristic extraction unit is used for acquiring the user characteristics of a target user and the resource object characteristics of M candidate resource objects, wherein M is a positive integer;
a CTR prediction unit, configured to perform CTR prediction on the M candidate resource objects, respectively, to obtain a CTR prediction result of each candidate resource object in the M candidate resource objects:
wherein the performing CTR pre-estimation on the M candidate resource objects respectively comprises: for each candidate resource object, performing feature intersection on resource object features of the candidate resource object and user features through a combined pre-estimation model to obtain combined features of the candidate resource object, and performing CTR pre-estimation on the candidate resource object according to the combined features to obtain a CTR pre-estimation result of the candidate resource object, wherein the combined features comprise linear features, in-group intersection features and inter-group intersection features obtained according to a multi-layer factorization machine submodel of the combined pre-estimation model, and high-order intersection features obtained according to a neural network submodel of the combined pre-estimation model;
and the resource recommending unit is used for determining a target resource object from the M candidate resource objects according to the CTR estimation result of each candidate resource object in the M candidate resource objects and recommending the target resource object to the target user.
7. The apparatus of claim 6, the CTR prediction unit comprising:
the characteristic preprocessing subunit is used for grouping and embedding the user characteristics and the resource object characteristics of the candidate resource object to obtain N original characteristic groups of the candidate resource object, wherein N is a positive integer;
the intra-group cross subunit is used for performing intra-group feature cross on the N original feature groups through the multilayer factorization machine submodel to obtain N groups of intra-group cross results of the candidate resource object;
an interclass crossing subunit, configured to perform feature addition, interclass feature crossing, and integration of original features in the N original feature groups on the N interclass crossing results and the N original feature groups of the candidate resource object through the multi-layer factorization machine submodel, and generate a linear feature, an interclass crossing feature, and an interclass crossing feature of the candidate resource object;
and the high-order cross subunit is used for performing feature cross on the N original feature groups through the neural network submodel to construct high-order cross features of the candidate resource object.
8. The apparatus of claim 7, the interclass crossing subunit to be specifically configured to:
embedding the N groups of in-group crossing results of the candidate resource object, and embedding the original features of the N original feature groups to form N intermediate feature groups of the candidate resource object;
performing interclass feature crossing on the N intermediate feature groups of the candidate resource object to form interclass crossing features of the candidate resource object;
and adding the characteristics of the N intermediate characteristic groups of the candidate resource object to obtain the cross characteristics in the group of the candidate resource object.
9. The apparatus of claim 6, further comprising a resource recall unit to obtain the M candidate resource objects by:
and recalling candidate resource objects provided by the merchant equipment within a preset distance range of the real-time position of the target user according to a location-based service LBS recall strategy to obtain the M candidate resource objects.
10. The apparatus according to any of claims 6-9, wherein the resource recommendation unit is specifically configured to:
acquiring priority intervention information of the M candidate resource objects;
integrating the priority intervention information of the M candidate resource objects and the CTR pre-estimation result of each candidate resource object in the M candidate resource objects to obtain the final recommendation sequence of the M candidate resource objects;
and determining the target resource object from the M candidate resource objects according to the number of the display pit bits and the final recommended sequence.
11. A server comprising a memory, a processor and a computer program stored on the memory and executable on the processor, the processor implementing the steps of the method of any one of claims 1 to 5 when executing the program.
12. A computer-readable storage medium, on which a computer program is stored which, when being executed by a processor, carries out the steps of the method of any one of claims 1 to 5.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911398736.8A CN111340522B (en) | 2019-12-30 | 2019-12-30 | Resource recommendation method, device, server and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911398736.8A CN111340522B (en) | 2019-12-30 | 2019-12-30 | Resource recommendation method, device, server and storage medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111340522A true CN111340522A (en) | 2020-06-26 |
| CN111340522B CN111340522B (en) | 2024-03-08 |
Family
ID=71179609
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201911398736.8A Active CN111340522B (en) | 2019-12-30 | 2019-12-30 | Resource recommendation method, device, server and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111340522B (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113157951A (en) * | 2021-03-26 | 2021-07-23 | 北京达佳互联信息技术有限公司 | Multimedia resource processing method, device, server and storage medium |
| CN113204697A (en) * | 2021-04-29 | 2021-08-03 | 五八有限公司 | Searching method, searching device, electronic equipment and storage medium |
| CN113327131A (en) * | 2021-06-03 | 2021-08-31 | 太原理工大学 | Click rate estimation model for feature interactive selection based on three-branch decision theory |
Citations (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106446005A (en) * | 2015-08-07 | 2017-02-22 | 谷歌公司 | Factorized models |
| WO2017084344A1 (en) * | 2015-11-16 | 2017-05-26 | 北京百度网讯科技有限公司 | Information recommendation method and apparatus |
| WO2018090793A1 (en) * | 2016-11-18 | 2018-05-24 | 腾讯科技（深圳）有限公司 | Multimedia recommendation method and device |
| CN108132963A (en) * | 2017-11-23 | 2018-06-08 | 广州优视网络科技有限公司 | Resource recommendation method and device, computing device and storage medium |
| WO2018212710A1 (en) * | 2017-05-19 | 2018-11-22 | National University Of Singapore | Predictive analysis methods and systems |
| WO2018212711A1 (en) * | 2017-05-19 | 2018-11-22 | National University Of Singapore | Predictive analysis methods and systems |
| CN109408724A (en) * | 2018-11-06 | 2019-03-01 | 北京达佳互联信息技术有限公司 | Multimedia resource estimates the determination method, apparatus and server of clicking rate |
| CN110008399A (en) * | 2019-01-30 | 2019-07-12 | 阿里巴巴集团控股有限公司 | A kind of training method and device, a kind of recommended method and device of recommended models |
| CN110020156A (en) * | 2017-12-07 | 2019-07-16 | 深圳市腾讯计算机系统有限公司 | Information recommendation method, front end implementation method, device, equipment and storage medium |
| CN110046952A (en) * | 2019-01-30 | 2019-07-23 | 阿里巴巴集团控股有限公司 | A kind of training method and device, a kind of recommended method and device of recommended models |
| WO2019141072A1 (en) * | 2018-01-22 | 2019-07-25 | 阿里巴巴集团控股有限公司 | Method, device, and client for recommending store information |
| CN110162693A (en) * | 2019-03-04 | 2019-08-23 | 腾讯科技（深圳）有限公司 | A kind of method and server of information recommendation |
| CN110263243A (en) * | 2019-01-23 | 2019-09-20 | 腾讯科技（深圳）有限公司 | Media information recommending method, apparatus, storage medium and computer equipment |
-
2019
- 2019-12-30 CN CN201911398736.8A patent/CN111340522B/en active Active
Patent Citations (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106446005A (en) * | 2015-08-07 | 2017-02-22 | 谷歌公司 | Factorized models |
| WO2017084344A1 (en) * | 2015-11-16 | 2017-05-26 | 北京百度网讯科技有限公司 | Information recommendation method and apparatus |
| WO2018090793A1 (en) * | 2016-11-18 | 2018-05-24 | 腾讯科技（深圳）有限公司 | Multimedia recommendation method and device |
| WO2018212710A1 (en) * | 2017-05-19 | 2018-11-22 | National University Of Singapore | Predictive analysis methods and systems |
| WO2018212711A1 (en) * | 2017-05-19 | 2018-11-22 | National University Of Singapore | Predictive analysis methods and systems |
| CN108132963A (en) * | 2017-11-23 | 2018-06-08 | 广州优视网络科技有限公司 | Resource recommendation method and device, computing device and storage medium |
| CN110020156A (en) * | 2017-12-07 | 2019-07-16 | 深圳市腾讯计算机系统有限公司 | Information recommendation method, front end implementation method, device, equipment and storage medium |
| WO2019141072A1 (en) * | 2018-01-22 | 2019-07-25 | 阿里巴巴集团控股有限公司 | Method, device, and client for recommending store information |
| CN109408724A (en) * | 2018-11-06 | 2019-03-01 | 北京达佳互联信息技术有限公司 | Multimedia resource estimates the determination method, apparatus and server of clicking rate |
| CN110263243A (en) * | 2019-01-23 | 2019-09-20 | 腾讯科技（深圳）有限公司 | Media information recommending method, apparatus, storage medium and computer equipment |
| CN110046952A (en) * | 2019-01-30 | 2019-07-23 | 阿里巴巴集团控股有限公司 | A kind of training method and device, a kind of recommended method and device of recommended models |
| CN110008399A (en) * | 2019-01-30 | 2019-07-12 | 阿里巴巴集团控股有限公司 | A kind of training method and device, a kind of recommended method and device of recommended models |
| CN110162693A (en) * | 2019-03-04 | 2019-08-23 | 腾讯科技（深圳）有限公司 | A kind of method and server of information recommendation |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113157951A (en) * | 2021-03-26 | 2021-07-23 | 北京达佳互联信息技术有限公司 | Multimedia resource processing method, device, server and storage medium |
| CN113204697A (en) * | 2021-04-29 | 2021-08-03 | 五八有限公司 | Searching method, searching device, electronic equipment and storage medium |
| CN113327131A (en) * | 2021-06-03 | 2021-08-31 | 太原理工大学 | Click rate estimation model for feature interactive selection based on three-branch decision theory |
Also Published As
| Publication number | Publication date |
|---|---|
| CN111340522B (en) | 2024-03-08 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Cui et al. | Personalized travel route recommendation using collaborative filtering based on GPS trajectories | |
| CN111242748B (en) | Method, apparatus, and storage medium for recommending items to a user | |
| CN110162693A (en) | A kind of method and server of information recommendation | |
| US8732015B1 (en) | Social media pricing engine | |
| EP4181026A1 (en) | Recommendation model training method and apparatus, recommendation method and apparatus, and computer-readable medium | |
| CN111798280B (en) | Multimedia information recommendation method, device and equipment and storage medium | |
| EP4242955A1 (en) | User profile-based object recommendation method and device | |
| CN111340522B (en) | Resource recommendation method, device, server and storage medium | |
| CN112070577A (en) | Commodity recommendation method, system, equipment and medium | |
| CN112288554B (en) | Commodity recommendation method and device, storage medium and electronic device | |
| CN112380449B (en) | Information recommendation method, model training method and related device | |
| CN111008335A (en) | Information processing method, device, equipment and storage medium | |
| CN109034853A (en) | Similar users method, apparatus, medium and electronic equipment are found based on seed user | |
| CN111967924A (en) | Commodity recommendation method, commodity recommendation device, computer device, and medium | |
| CN113781139A (en) | Item recommendation method, item recommendation device, equipment and medium | |
| CN114511387A (en) | Product recommendation method and device, electronic equipment and storage medium | |
| CN111489196B (en) | Prediction method and device based on deep learning network, electronic equipment and medium | |
| CN114429384A (en) | Intelligent product recommendation method and system based on e-commerce platform | |
| CN113159877A (en) | Data processing method, device, system and computer readable storage medium | |
| CN114596109A (en) | Method and device for determining recommendation information, electronic equipment and storage medium | |
| CN113822734A (en) | Method and apparatus for generating information | |
| CN112561000A (en) | Classification method, device, equipment and storage medium based on combined model | |
| CN110766478A (en) | Method and device for improving user connectivity | |
| Liu et al. | Customer preference analysis towards online shopping decisions based on optimized feature extraction | |
| Smirnov et al. | Multicriteria Context-Driven Recommender Systems: Model and Method |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |