CN115809374A - Method, system, device and storage medium for correcting mainstream deviation of recommendation system - Google Patents

Method, system, device and storage medium for correcting mainstream deviation of recommendation system Download PDF

Info

Publication number
CN115809374A
CN115809374A CN202310104256.6A CN202310104256A CN115809374A CN 115809374 A CN115809374 A CN 115809374A CN 202310104256 A CN202310104256 A CN 202310104256A CN 115809374 A CN115809374 A CN 115809374A
Authority
CN
China
Prior art keywords
mainstream
user
dynamic
vector
degree
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
Application number
CN202310104256.6A
Other languages
Chinese (zh)
Other versions
CN115809374B (en
Inventor
张海仙
张宽易
谢敏
张懿
谌祖港
黄粱可汗
李欣洋
尚文一
尹腾
杨雨奇
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Sichuan University
Original Assignee
Sichuan University
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Sichuan University filed Critical Sichuan University
Priority to CN202310104256.6A priority Critical patent/CN115809374B/en
Publication of CN115809374A publication Critical patent/CN115809374A/en
Application granted granted Critical
Publication of CN115809374B publication Critical patent/CN115809374B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Landscapes

  • Information Retrieval, Db Structures And Fs Structures Therefor (AREA)

Abstract

The invention discloses a method, a system, equipment and a storage medium for correcting mainstream deviation of a recommendation system, belongs to mainstream deviation correction of the recommendation system in the field of data recommendation, and aims to solve the technical problems of low overall recommendation accuracy and low recommendation fairness caused by the fact that the mainstream deviation is not fully considered by the recommendation system in the prior art. According to the method, the mainstream score of the user is calculated and weighted, then the interaction data of the user after weighting is used as the training data of model training, through the method, the mainstream degree of the user can be considered in the reconstruction of the model, and the phenomenon of excessively recommending popular goods can not occur, so that the influence of mainstream deviation on a recommendation system can be effectively reduced, the effect of the recommendation system on a wider user group is improved, higher fairness is realized, the overall recommendation accuracy of the recommendation system is improved, and the recommendation fairness of the recommendation system is higher.

Description

Method, system, device and storage medium for correcting mainstream deviation of recommendation system
Technical Field
The invention belongs to the technical field of data recommendation, relates to the field of mainstream deviation correction of a recommendation system, and particularly relates to a method, a system, equipment and a storage medium for correcting mainstream deviation of the recommendation system.
Background
With the expansion of the scale of the internet and information systems, the amount of information generated by the internet and information systems also shows an explosive growth situation, and the difficulty of a user in retrieving the content desired by the user from mass information is greatly improved. At present, various websites use a recommendation system in the background, the recommendation system calculates candidate products most suitable for being recommended to a user according to the access characteristics of the user, and then the candidate products are displayed to the user for selection. As a technology capable of effectively solving information overload, the recommendation system can filter out the most interesting part of the user from massive contents according to personalized requirements of different users, so that the recommendation system is widely applied to the fields of e-commerce, video and audio entertainment, accurate advertisement delivery and the like. The collaborative filtering method is a common method in a recommendation system, and can be divided into user-based collaborative filtering and article-based collaborative filtering according to different targets. The collaborative filtering based on the users takes the users as the center, and the idea is to recommend articles similar to the users to the target users; item-based collaborative filtering is item-centric, with the idea being to recommend similar items to a target item to a user who likes the current item. However, due to the influence of the user crowd effect, the platform display mechanism, the quality difference of the articles, and other factors, a phenomenon that the interaction of many users is concentrated on a small part of the articles easily occurs. The traditional collaborative filtering method is easy to learn the deviation during the training process, so that the recommendation of a small part of popular items is prone to be realized, and the recommendation of a large part of items is difficult to obtain the recommendation opportunity, so that the recommendation result cannot reflect the real preference of the user. This causes that the recommendation effect received by the mainstream users who like to pursue the hot spot is often very good, while the recommendation effect received by the rest of the broader user groups is not satisfactory, and the received recommendation effect of different users is greatly different due to different mainstream degrees, which is the mainstream deviation phenomenon in the recommendation system.
Most of the existing deviation correcting schemes start from the perspective of articles, namely, the mainstream deviation is indirectly reduced by correcting the popularity deviation of the articles in a recommendation system. The invention patent application with application number CN202110218946.5 discloses a causal reasoning method for correcting popularity deviation of a recommendation system, which comprises the following steps: acquiring a matching score of a user and an article in a current recommendation system; predicting an item score according to the popularity of the item, and predicting a user score according to the preference of the user; and aggregating the matching scores of the user and the articles, the article scores and the user scores, predicting the matching scores of the user and the articles, and removing the influence caused by the popularity deviation to obtain the final matching scores of the user and the articles. The method is a model-independent counterfactual reasoning framework, can be suitable for various recommendation systems, improves the recommendation performance of the recommendation system by eliminating the popularity deviation, and can provide high-quality and accurate personalized recommendation content for users. The method is the same as other article-based collaborative filtering methods, mainly aims at improving the phenomenon that recommendation is concentrated on a small part of popular articles, reduces the influence of the popular articles on the overall model recommendation decision in the training process by adopting modes such as inverse tendency fraction weighting, and the like, and simultaneously gives higher weight to the long-tail articles to increase the recommendation probability of the long-tail articles.
In recent years, there are also methods for removing mainstream deviations from the perspective of users, such as adjusting weights of different users in a training process, training a model separately for user groups with different preferences, and the like, so as to enhance capturing capability of the model for preferences of a specific user group. The invention patent application with the application number of CN201911056270.3 discloses a recommendation list re-ranking method for improving the diversity of a recommendation system. The method is the same as other collaborative filtering based on the user, different requirements of the user on diversity of the recommendation list can be considered, so that the recommended articles are more fit for real feeling of people, grading deviation of different users on the same article is also considered, the diversity is properly improved on the balance of accuracy and diversity, and the influence on the accuracy is small.
The method for correcting the mainstream deviation of the recommendation system can actually expand the recommendation range of the recommendation system, so that the recommendation system can not be limited to a part of popular articles, but can take care of some long-tail articles, and the fairness problem in article recommendation is solved to a certain extent. However, this does not mean that these long-tailed items can be recommended to the appropriate users, but rather reduces the accuracy of the recommendation system if recommended to mainstream users who prefer to pursue hot spots. Therefore, a method of correcting the deviation of popularity alone does not necessarily play a positive role in correcting the deviation of the mainstream. The existing method for directly correcting the mainstream deviation also has a certain problem, and the effect of the part of users is easily damaged by reducing the weight of the mainstream user in the training process, so that the overall accuracy of the recommendation system is reduced; the method for training different models separately for different user groups also has problems, and the division of the user groups, the training of a plurality of models and the consumption during integration make the method difficult to realize in the actual production environment. In addition, the existing method for correcting the deviation of the main stream does not consider the characteristic of the change of the main stream, a group of users belonging to the main stream at present are not necessarily the main stream users in the past, and a group of users not belonging to the main stream in the past can also become the main stream users in the future due to pursuit of hot spots.
Disclosure of Invention
The invention aims to: in order to solve the technical problems of low overall recommendation accuracy and low recommendation fairness caused by the fact that a recommendation system does not fully consider mainstream deviation in the prior art, the invention provides a method, a system, equipment and a storage medium for correcting the mainstream deviation of the recommendation system.
The invention specifically adopts the following technical scheme for realizing the purpose:
a method for correcting mainstream deviation of a recommendation system comprises the following steps:
step S1, data collection and processing
Obtaining user information, article information and user article interaction information in a recommendation system, and respectively constructing user co-occurrence vectors
Figure SMS_1
Co-occurrence vector of article
Figure SMS_2
；
Step S2, calculating the mainstream score
According to co-occurrence vector of articles
Figure SMS_6
Calculating the total interaction times of the articles
Figure SMS_8
(ii) a According to user co-occurrence vectors
Figure SMS_10
Calculating the total number of interactions of the user
Figure SMS_5
(ii) a According to the total number of interactions of the article
Figure SMS_11
Total number of interactions of user
Figure SMS_13
Item categories, computing users
Figure SMS_14
Dynamic mainstream degree score of
Figure SMS_3
(ii) a According to the user
Figure SMS_7
Dynamic mainstream level score of
Figure SMS_9
Calculating the average value of the dynamic mainstream degree scores of all the users to obtain the global dynamic mainstream degree score
Figure SMS_12
And forming global dynamic mainstream degree scores of all article categories into a global dynamic mainstream degree vector
Figure SMS_4
；
S3, constructing a dynamic mainstream degree characteristic model
Constructing a dynamic mainstream degree characteristic model based on a three-layer perceptron MLP model, wherein the first two layers of the dynamic mainstream degree characteristic model use a ReLU function as an activation function, and the last layer of the dynamic mainstream degree characteristic model uses a softmax activation function; with user information vectors
Figure SMS_15
And the global dynamic mainstream degree vector output by the step S2
Figure SMS_16
Spliced and used as input of dynamic mainstream degree characteristic model
Figure SMS_17
Outputting a dynamic mainstream characteristic hidden vector by the dynamic mainstream degree characteristic model;
s4, constructing a collaborative filtering module
Constructing a collaborative filtering module comprising an encoder and a decoder;
the encoder is constructed by adopting a three-layer perceptron MLP model; user interaction data
Figure SMS_20
Input encoder, encoder calculating user interaction data
Figure SMS_23
And each dimension of (a) and generate separately
Figure SMS_25
Mean value of
Figure SMS_19
Variance, forming a mean vector of the user
Figure SMS_22
Sum variance vector
Figure SMS_24
Wherein the two vectors are both in t dimension, and constitute the mean vector of the user
Figure SMS_26
Sum variance vector
Figure SMS_18
Wherein the two vectors are both in t dimension, and then h dimension user interaction hidden vectors are generated by random sampling
Figure SMS_21
；
The decoder is constructed by adopting a four-layer perceptron MLP model, the first three layers of activation functions of the decoder are tanh functions, and the last layer of activation functions of the decoder are softmax functions; step S3, outputting the dynamic mainstream characteristic hidden vector and the user interaction hidden vector output by the encoder
Figure SMS_27
As input to a decoder, the decoder output reconstructs user interaction data
Figure SMS_28
And reconstructing dynamic mainstream feature vectors
Figure SMS_29
Reconstructing dynamic mainstream feature vectors
Figure SMS_30
For completing the reconstruction of the decoder;
step S5, recommendation result generation
Inputting user interaction data currently observed by the user to be predicted according to the collaborative filtering module obtained by the training completion in the step S4
Figure SMS_31
The encoder outputs a user interaction hidden vector
Figure SMS_32
(ii) a Then the user interaction is hidden
Figure SMS_33
And inputting the dynamic mainstream characteristic hidden vector output by the step S3 into a decoder, and outputting reconstructed user interaction data by the decoder.
Further, in step S1, any user is constructed according to the user item interaction information
Figure SMS_34
User co-occurrence vector with all items
Figure SMS_35
Constructing any article according to the user article interaction information
Figure SMS_36
Co-occurrence vector with all articles
Figure SMS_37
；
Wherein,
Figure SMS_40
which represents the total number of users,
Figure SMS_42
the total number of items is indicated and,
Figure SMS_44
the items are shown as being in the form of objects,
Figure SMS_39
is shown as
Figure SMS_41
The number of the articles is one,
Figure SMS_45
representing a user
Figure SMS_48
And articles
Figure SMS_38
The interaction situation of (a) is,
Figure SMS_43
representing an article
Figure SMS_46
And the user
Figure SMS_47
The interaction scenario of (2).
Further, in step S2, according to the article
Figure SMS_49
Co-occurrence vector with article
Figure SMS_50
Calculating the total number of article interactions
Figure SMS_51
；
According to the user
Figure SMS_52
Co-occurrence vector with user
Figure SMS_53
And the total number of times of interaction of the user is calculated,
Figure SMS_54
；
according to the total number of interaction times of the articles
Figure SMS_56
Total number of user interactions
Figure SMS_58
Item categories, computing users
Figure SMS_63
Dynamic mainstream level score of
Figure SMS_57
(ii) a In the case of insensitive number of interactions, for the category
Figure SMS_59
Article of, user
Figure SMS_62
Dynamic mainstream level score of
Figure SMS_64
Calculating according to the formula (1); in the case of sensitive number of interactions, for the category
Figure SMS_55
Article of, user
Figure SMS_60
Dynamic mainstream degree score of
Figure SMS_61
Calculating according to the formula (2);
Figure SMS_65
（1）
Figure SMS_66
（2）
according to the user
Figure SMS_67
Dynamic mainstream degree score of
Figure SMS_68
Calculating the average value of the dynamic mainstream degree scores of all the users to obtain the global dynamic mainstream degree score
Figure SMS_69
The calculation formula is as follows:
Figure SMS_70
then, the global dynamic mainstream degree scores of all the article categories are divided
Figure SMS_71
Make up a dimension of
Figure SMS_72
Global dynamic mainstream level vector of
Figure SMS_73
Expressed as:
Figure SMS_74
wherein,
Figure SMS_76
、
Figure SMS_80
all of which represent the time of day,
Figure SMS_82
representing the hyper-parameter (for controlling the logarithmic curve),
Figure SMS_77
representing an article
Figure SMS_79
Belong to the category
Figure SMS_81
，
Figure SMS_83
Which represents the total number of users,
Figure SMS_75
a set of all the users is represented,
Figure SMS_78
representing the total number of categories of items.
Furthermore, a co-occurrence vector weighting process is also performed, and the weighting process is described as: user will be
Figure SMS_84
To belong to the category
Figure SMS_85
Article of (2)
Figure SMS_86
Of the interaction situation
Figure SMS_87
Multiplication by the user
Figure SMS_88
In the category of
Figure SMS_89
Dynamic mainstream degree score of
Figure SMS_90
；
For user co-occurrence vectors
Figure SMS_91
Each item in the list is weighted, and after the weighting is finished, the co-occurrence vector of the whole user is obtained
Figure SMS_92
Normalizing by using a softmax function to obtain user interaction data for inputting into the collaborative filtering module
Figure SMS_93
：
Figure SMS_94
Wherein any of the user interaction data
Figure SMS_95
Has a value range of [0,1 ]]。
Further, in step S3, the user information vector
Figure SMS_96
Expressed as:
Figure SMS_97
wherein,
Figure SMS_98
representing a user
Figure SMS_99
The age information of the person to be treated is quantified,
Figure SMS_100
representing a user
Figure SMS_101
The binary gender information of (1);
input of dynamic mainstream degree characteristic model
Figure SMS_102
Expressed as:
Figure SMS_103
；
wherein,
Figure SMS_104
representing a vector stitching operation.
Further, in step S4, the loss function of the filter module is cooperated
Figure SMS_105
Dividing into reconstructed target losses
Figure SMS_106
Distribution approximation loss
Figure SMS_107
And dynamic mainstream eigenvector approximation loss
Figure SMS_108
Three parts, loss function
Figure SMS_109
The calculation formula of (2) is as follows:
Figure SMS_110
reconstructing object losses
Figure SMS_111
The calculation formula of (2) is as follows:
Figure SMS_112
distribution approximation loss
Figure SMS_113
The calculation formula of (2) is as follows:
Figure SMS_114
dynamic mainstream eigenvector approximation loss
Figure SMS_115
Is calculated byThe formula is as follows:
Figure SMS_116
wherein,
Figure SMS_124
a hidden vector representing the user interaction is shown,
Figure SMS_119
representing the data of the user interaction(s),
Figure SMS_125
representing the posterior distribution of each user data sample,
Figure SMS_121
representing variation distribution
Figure SMS_127
And posterior distribution
Figure SMS_133
In the approximation that the difference between the first and second values,
Figure SMS_134
to represent
Figure SMS_123
To pair
Figure SMS_129
In the expectation that the position of the target is not changed,
Figure SMS_117
in order to be a hyper-parameter,
Figure SMS_126
is a user interaction implicit vector
Figure SMS_122
KL represents the KL divergence,
Figure SMS_130
a distribution is represented a priori, which is,
Figure SMS_120
which represents a vector of the variance (m) of the signal,
Figure SMS_128
which represents the square of the mean vector and,
Figure SMS_118
in order to be a hyper-parameter,
Figure SMS_132
representing the original dynamic mainstream feature vector,
Figure SMS_131
representing the reconstructed dynamic mainstream feature vector.
A system for correcting deviations in a mainstream of a recommendation system, comprising:
the data collection and processing module is used for acquiring the user information, the article information and the user article interaction information in the recommendation system and respectively constructing the user co-occurrence vectors
Figure SMS_135
Co-occurrence vector of article
Figure SMS_136
；
A mainstream score calculation module for calculating a mainstream score based on the co-occurrence vector of the article
Figure SMS_138
Calculating the total interaction times of the articles
Figure SMS_142
(ii) a According to user co-occurrence vectors
Figure SMS_144
Calculating the total number of interactions
Figure SMS_139
(ii) a According to the total number of interaction times of the articles
Figure SMS_143
Total number of user interactions
Figure SMS_145
Item categories, computing users
Figure SMS_148
Dynamic mainstream level score of
Figure SMS_137
(ii) a According to the user
Figure SMS_141
Dynamic mainstream degree score of
Figure SMS_146
Calculating the average value of the dynamic mainstream degree scores of all the users to obtain the global dynamic mainstream degree score
Figure SMS_147
And forming the global dynamic mainstream degree scores of all the article categories into a global dynamic mainstream degree vector
Figure SMS_140
；
The dynamic mainstream degree characteristic model building module is used for building a dynamic mainstream degree characteristic model based on a three-layer perceptron MLP model, the ReLU function is used as an activation function in the first two layers of the dynamic mainstream degree characteristic model, and the softmax activation function is used in the last layer of the dynamic mainstream degree characteristic model; with user information vectors
Figure SMS_149
And the global dynamic mainstream degree vector output by the mainstream fraction calculation module
Figure SMS_150
Spliced as input of dynamic mainstream degree characteristic model
Figure SMS_151
Outputting a dynamic mainstream characteristic hidden vector by the dynamic mainstream degree characteristic model;
the collaborative filtering module construction module is used for constructing a collaborative filtering module comprising an encoder and a decoder;
the encoder is constructed by adopting a three-layer perceptron MLP model; user interaction data
Figure SMS_152
Input encoder, encoder calculating user interaction data
Figure SMS_153
And each dimension of (1), and separately generate
Figure SMS_154
Mean value of
Figure SMS_155
Variance, forming a mean vector of users
Figure SMS_156
Sum variance vector
Figure SMS_157
Wherein the two vectors are both in t dimension, and then h dimension user interaction hidden vectors are generated by random sampling
Figure SMS_158
；
The decoder is constructed by adopting a four-layer perceptron MLP model, the first three layers of activation functions of the decoder are tanh functions, and the last layer of activation function of the decoder is softmax functions; dynamic mainstream feature hidden vectors output by a dynamic mainstream degree feature model building module and user interaction hidden vectors output by an encoder
Figure SMS_159
As input to the decoder, the decoder outputs reconstructed user interaction data
Figure SMS_160
And reconstructing dynamic mainstream feature vectors
Figure SMS_161
Reconstructing dynamic mainstream feature vectors
Figure SMS_162
For completing the reconstruction of the decoder;
a recommendation result generation module for inputting the user interaction data currently observed by the user to be predicted according to the collaborative filtering module obtained by the collaborative filtering module construction module after training
Figure SMS_163
The encoder outputs a user interaction hidden vector
Figure SMS_164
(ii) a Then the user interaction is hidden
Figure SMS_165
And inputting the dynamic mainstream characteristic hidden vector output by the dynamic mainstream degree characteristic model building module into a decoder, and outputting reconstructed user interaction data by the decoder.
A computer device comprising a memory and a processor, the memory storing a computer program which, when executed by the processor, causes the processor to carry out the steps of the above method.
A computer-readable storage medium, storing a computer program which, when executed by a processor, causes the processor to carry out the steps of the above-mentioned method.
The invention has the following beneficial effects:
1. in the invention, in the process of dynamically correcting the mainstream deviation, the mainstream score of the user is calculated and weighted, and then the weighted user interaction data is used as the training data of model training.
2. In the invention, a collaborative filtering module based on an asymmetric variational self-encoder is constructed, and the capability of capturing and utilizing dynamic mainstream characteristics by a model is enhanced through asymmetric structural design and introduction of dynamic mainstream characteristic vectors.
3. In the invention, two scenes of sensitive interaction times and insensitive interaction times are fully considered, a method for dynamically calculating the mainstream degree score of the user is provided, the mainstream degree score can be used as input data of a variational self-encoder through a weighting normalization process, the influence of mainstream deviation is fully considered, the overall recommendation accuracy of a recommendation system is improved, and the recommendation fairness of the recommendation system is higher.
Drawings
FIG. 1 is a schematic flow diagram of the present invention;
FIG. 2 is a detailed flow diagram of the present invention;
FIG. 3 is a schematic diagram of the structure of the collaborative filtering module according to the present invention.
Detailed Description
Example 1
The embodiment provides a method for correcting a mainstream deviation of a recommendation system, a flow diagram of which is shown in fig. 1, and the method comprises the steps of data collection and processing, mainstream score calculation, dynamic mainstream degree feature model construction, collaborative filtering module construction and recommendation result generation. The detailed flow diagram of the method is shown in fig. 2, and specifically includes:
step S1, data collection and processing
Obtaining user information, article information and user article interaction information in a recommendation system, and respectively constructing user co-occurrence vectors
Figure SMS_166
Co-occurrence vector of article
Figure SMS_167
。
Constructing any user according to user article interaction information
Figure SMS_168
User co-occurrence vectors with all items
Figure SMS_173
，
Figure SMS_175
Representing a user
Figure SMS_169
And articles
Figure SMS_174
The interaction scenario of (2). In the scene with insensitive interaction times (only paying attention to whether there is interaction, but not paying attention to the interaction times, such as movie recommendation, book recommendation and the like), if explicit interaction is generated
Figure SMS_178
Otherwise
Figure SMS_179
(ii) a In the case of sensitive interaction times (concerning whether there is an interaction, and also concerning the number of interactions, such as music recommendation, short video recommendation, etc.), if an explicit interaction is generated
Figure SMS_171
Wherein
Figure SMS_172
Representing a user
Figure SMS_176
And articles
Figure SMS_177
The number of interactions; otherwise
Figure SMS_170
。
Constructing any article according to user article interaction information
Figure SMS_183
Co-occurrence vector with all articles
Figure SMS_185
，
Figure SMS_188
Representing an article
Figure SMS_182
And the user
Figure SMS_184
The interaction scenario of (2). In the scene with sensitive interaction times (only paying attention to whether there is interaction, but not paying attention to the interaction times, such as movie recommendation, book recommendation and the like), if explicit interaction is generated
Figure SMS_187
Otherwise, otherwise
Figure SMS_190
. In case of sensitive interaction times (concerning whether there is any interaction, and also concerning the number of interactions, such as music recommendations, short video recommendations, etc.), if an explicit interaction is generated
Figure SMS_180
In which
Figure SMS_186
Representing an article
Figure SMS_189
And the user
Figure SMS_191
The number of interactions; otherwise
Figure SMS_181
。
Wherein,
Figure SMS_192
which represents the total number of users,
Figure SMS_193
the total number of items is indicated and,
Figure SMS_194
to represent an item of material that is,
Figure SMS_195
denotes the first
Figure SMS_196
The number of the articles is increased, and the articles,
Figure SMS_197
representing the user.
The data cleaning is mainly to filter out part of users and articles according to a threshold (for example, to filter out articles with interaction times smaller than a certain threshold), and the purpose of the data cleaning is to remove abnormal data to ensure normal operation of a recommendation process.
Step S2, calculating the mainstream score
The step mainly quantizes the user and the global mainstream degree, so that a dynamic mainstream degree feature vector can be generated in the step S3 conveniently. Since the concept of mainstream level involves both individual users and overall users, the mainstream level scores of individual users and global can be calculated separately here.
According to the article
Figure SMS_198
Co-occurrence vector with article
Figure SMS_199
The total number of interactions of the article can be calculated
Figure SMS_200
. Since the degree of mainstream is a dynamically changing concept, the number of item interactions can be filtered based on time, e.g. only considering the occurrence
Figure SMS_201
Interaction in the time period only takes the release date into consideration
Figure SMS_202
Previous item with total number of item interactions of
Figure SMS_203
Number of interactions of any article therein
Figure SMS_204
Only data within this time period is considered.
According to the user
Figure SMS_205
Co-occurrence vector with user
Figure SMS_206
The total number of interactions of the user can be calculated
Figure SMS_207
. Since the degree of mainstream is a dynamically changing concept, the number of user interactions can be filtered according to time, e.g. only considering the occurrence
Figure SMS_208
And (3) the total interaction times of the user in the time period are as follows:
Figure SMS_209
wherein is arbitrary
Figure SMS_210
Only the interactive data in the time period is considered, and the release date is
Figure SMS_211
The subsequent item interaction data are all set to 0.
According to the idea of collaborative filtering, the degree of user's mainstream depends on whether their interactive items are interacted with by other users. Meanwhile, the method considers that the preference degrees of users to different categories of articles are inconsistent, so the influence of the article category factor is also considered in the calculation process of the mainstream degree score of the user. According to the total number of interaction times of the articles
Figure SMS_214
Total number of interactions of user
Figure SMS_217
Item category, computing user
Figure SMS_220
Dynamic mainstream degree score of
Figure SMS_213
. In the case of insensitive number of interactions, for the category
Figure SMS_216
Article of, user
Figure SMS_219
Dynamic mainstream degree score of
Figure SMS_221
Calculating according to the formula (1); in the case of sensitive number of interactions, for the category
Figure SMS_212
Article of, user
Figure SMS_215
Dynamic mainstream degree score of
Figure SMS_218
Calculating according to the formula (2);
Figure SMS_222
（1）
Figure SMS_223
（2）
wherein,
Figure SMS_226
will be provided with
Figure SMS_231
Within a time period, the user
Figure SMS_233
All items that have interacted add up to the number of interactions of all users. The phenomenon of power law distribution (i.e. small part) due to the interactive recording of real world objectsItems occupy most of the interactions, and most items have little), so the total number of interactions with other users for each item using a logarithmic function
Figure SMS_225
Inhibition, base number
Figure SMS_228
Is a hyper-parameter. Dynamic mainstream score of user in case of insensitive interaction times
Figure SMS_230
Each interaction record is given the same weight. Dynamic mainstream score of user in case of sensitive interaction times
Figure SMS_232
Each interaction record is given a different weight,
Figure SMS_224
representing a user
Figure SMS_227
And articles
Figure SMS_229
This means that the greater the number of user interactions, the greater the weight of the item in the user mainstream level score evaluation.
According to the user
Figure SMS_234
Dynamic mainstream level score of
Figure SMS_235
For categories of
Figure SMS_236
Article of (1), global dynamic mainstream level score
Figure SMS_237
Is the average of the dynamic mainstream level scores of all users, so the global dynamic masterFractional degree of flow
Figure SMS_238
The calculation formula of (c) is:
Figure SMS_239
then, the global dynamic mainstream degree scores of all the article categories are divided
Figure SMS_240
Make up a dimension of
Figure SMS_241
Global dynamic mainstream level vector of
Figure SMS_242
Expressed as:
Figure SMS_243
wherein,
Figure SMS_245
、
Figure SMS_247
both of which represent the time that it takes,
Figure SMS_250
representing the hyper-parameter (for controlling the logarithmic curve),
Figure SMS_246
representing an article
Figure SMS_248
Belong to the category
Figure SMS_251
，
Figure SMS_252
Which represents the total number of users,
Figure SMS_244
a set of all the users is represented,
Figure SMS_249
representing the total number of categories of items.
And then weighting the co-occurrence vectors of the users according to the obtained mainstream degree scores of the users, namely introducing the mainstream information of the users into the co-occurrence vectors. The weighting process is described as: user will be connected
Figure SMS_253
To belong to the category
Figure SMS_254
Article of
Figure SMS_255
Of the interaction situation
Figure SMS_256
Multiplication by the user
Figure SMS_257
In the category of
Figure SMS_258
Dynamic mainstream degree score of
Figure SMS_259
；
For user co-occurrence vectors
Figure SMS_260
Each item in the list is weighted, and the co-occurrence vector of the whole user is obtained after the weighting is finished
Figure SMS_261
Normalizing by using softmax function to obtain user interaction data for inputting into the collaborative filtering module
Figure SMS_262
：
Figure SMS_263
Wherein any one of the user interaction data
Figure SMS_264
Has a value range of [0,1 ]]。
User interaction data derived therefrom
Figure SMS_265
As well as the input of the encoder followed by the collaborative filtering module.
S3, constructing a dynamic mainstream degree characteristic model
Can be calculated according to the step (2) in
Figure SMS_267
Time slot arbitrary user
Figure SMS_269
With respect to any category of item collections
Figure SMS_271
Main stream degree score of
Figure SMS_268
. For any user
Figure SMS_270
Calculating its mainstream level scores for all the categories of the item set, these mainstream level scores may constitute one
Figure SMS_272
Vector of dimensions, denoted as
Figure SMS_273
The vector may characterize the user
Figure SMS_266
The mainstream degree of all the item categories, the larger the score of the mainstream degree of an item category is, the more the user prefers popular items in the category, and the smaller the score is, the more the user prefers popular items in the category.
In order to enable the collaborative filtering model in step S4 to make full use of the user information and the global information, a dynamic mainstream feature expression based on the user information and the global information needs to be obtained. A dynamic mainstream level feature model is thus constructed here. And constructing a dynamic mainstream degree feature model based on the three-layer perceptron MLP model, wherein the dynamic mainstream degree feature model is used for extracting key features in user information and global information and reducing the original input into a hidden vector with a lower dimensionality. The ReLU function is used as an activation function in the first two layers of the dynamic mainstream degree characteristic model, and the softmax activation function is used in the last layer of the dynamic mainstream degree characteristic model. With user information vectors
Figure SMS_274
And the global dynamic mainstream degree vector output by the step S2
Figure SMS_275
Spliced as input of dynamic mainstream degree characteristic model
Figure SMS_276
。
The user information vector
Figure SMS_277
Expressed as:
Figure SMS_278
wherein,
Figure SMS_279
representing a user
Figure SMS_280
The age information of the person to be treated is quantified,
Figure SMS_281
representing a user
Figure SMS_282
The binary gender information of (1);
inputting the dynamic mainstream degree characteristic model
Figure SMS_283
Expressed as:
Figure SMS_284
；
wherein,
Figure SMS_285
representing vector stitching operations
Input the method
Figure SMS_286
And outputting the dynamic mainstream characteristic hidden vector after dimensionality reduction of the three-layer perceptron of the dynamic mainstream degree characteristic model.
S4, constructing a collaborative filtering module
An asymmetric variational self-encoder (VAE) is used as a main structure of the collaborative filtering model, and the structure is shown in fig. 3. The method adopts an asymmetric variational self-encoder to carry out collaborative filtering, and aims to add extra dynamic mainstream degree information into a hidden layer between an encoder and a decoder, so that the decoder can directly utilize personal information of a user and global mainstream degree during decoding.
The collaborative filtering module comprises an encoder and a decoder;
the encoder is constructed by adopting a three-layer perceptron MLP model. User interaction data
Figure SMS_288
Input encoder, encoder calculating user interaction data
Figure SMS_291
And each dimension of (1), and separately generate
Figure SMS_293
Mean value of
Figure SMS_289
Variance, forming a mean vector of the user
Figure SMS_290
Sum variance vector
Figure SMS_292
Wherein the two vectors are both in t dimension, and then h dimension user interaction hidden vector is generated by random sampling
Figure SMS_294
，
Figure SMS_287
. Since the network cannot perform back propagation due to the random sampling method, the re-parameterization method is adopted to complete the sampling process.
The decoder is constructed by adopting a four-layer perceptron MLP model, the first three layers of activation functions of the decoder are tanh functions, and the last layer of activation functions of the decoder are softmax functions and are used for generating probability distribution
Figure SMS_295
. The dynamic mainstream characteristic hidden vector output by the step S3 and the user interaction hidden vector output by the encoder
Figure SMS_296
As input to the decoder, the decoder outputs reconstructed user interaction data
Figure SMS_297
And reconstructing dynamic mainstream feature vectors
Figure SMS_298
Reconstructing dynamic mainstream feature vectors
Figure SMS_299
For completing the reconstruction of the decoder.
The variational autoencoder reasoning process is as follows, assuming the user
Figure SMS_300
Corresponding user interaction hidden vector
Figure SMS_301
Compliance
Figure SMS_302
Normal distribution of (c). Assuming user interaction data input to the encoder according to the recommendation system interaction data characteristics
Figure SMS_303
Obey probability of
Figure SMS_304
The likelihood function of the polynomial distribution of (1) is as follows:
Figure SMS_305
wherein,
Figure SMS_306
representing hidden vectors interacted by user
Figure SMS_307
Is determined and is
Figure SMS_308
Probability of individual item interaction
In order to enable the network to learn the parameters by back-propagation, a posteriori distributions for each user data sample must be found
Figure SMS_311
Since this posterior distribution is not easy to find, the variation distribution is used here by means of variation estimation
Figure SMS_313
To approximate
Figure SMS_316
. Suppose that
Figure SMS_310
Satisfy a Gaussian distribution
Figure SMS_314
In which
Figure SMS_317
Is a variance vector
Figure SMS_319
Diagonal covariance matrix of. Then the optimization goal of the network at this time is to optimize the parameter generation mean vector
Figure SMS_309
Sum variance vector
Figure SMS_312
Make variation distribute
Figure SMS_315
And posterior distribution
Figure SMS_318
As similar as possible.
Wherein the reparameterization method operates as follows, assuming noise
Figure SMS_320
Obey normal distribution
Figure SMS_321
User interaction implicit vector
Figure SMS_322
May be represented by a variance vector
Figure SMS_323
Mean vector of
Figure SMS_324
And noise are linearly combined, so that the network can learn. The reparameterization formula is as follows:
Figure SMS_325
wherein,
Figure SMS_326
is the standard deviation;
unlike the standard variational self-encoder network, an asymmetric structure is used to obtain the user interaction implicit vector
Figure SMS_327
Thereafter, the input generated in step S3 is inputted
Figure SMS_328
Is spliced at
Figure SMS_329
And then fed into the decoder. The generating part of the decoder may be divided into reconstructing user interaction data
Figure SMS_330
And reconstructing dynamic mainstream feature vectors
Figure SMS_331
。
In conclusion, loss function of collaborative filtering model based on asymmetric variational self-encoder
Figure SMS_332
Dividing into reconstructed target losses
Figure SMS_333
Distribution approximation loss
Figure SMS_334
And dynamic mainstream eigenvector approximation loss
Figure SMS_335
Three parts, loss function
Figure SMS_336
The calculation formula of (2) is as follows:
Figure SMS_337
reconstructing object loss
Figure SMS_338
The purpose of this is to make the reconstructed user interaction data output by the decoder as identical as possible to the user interaction data input to the encoder, the calculation formula being:
Figure SMS_339
distribution approximation loss
Figure SMS_340
Is to make the variation distributed
Figure SMS_341
As close to a posterior distribution as possible
Figure SMS_342
For measuring the approximation degree of two distributions, the calculation formula is:
Figure SMS_343
in order to convert the original dynamic mainstream feature vector
Figure SMS_344
And reconstructing dynamic mainstream feature vectors
Figure SMS_345
With its negative number as a loss term, with the aim of enabling the reconstruction of dynamic mainstream feature vectors
Figure SMS_346
In the direction of original dynamic main flow feature vector
Figure SMS_347
Approximate losses by dynamic mainstream feature vectors as close as possible
Figure SMS_348
The method can enable a decoder to complete the reconstruction process by using the dynamic mainstream characteristics as much as possible. Dynamic mainstream feature vectorApproximate loss
Figure SMS_349
The calculation formula of (c) is:
Figure SMS_350
wherein,
Figure SMS_356
a hidden vector representing the user interaction is shown,
Figure SMS_355
representing the data of the user interaction(s),
Figure SMS_367
representing the posterior distribution of each user data sample,
Figure SMS_358
representing variation distribution
Figure SMS_363
And posterior distribution
Figure SMS_357
In the approximation that the difference between the first and second values,
Figure SMS_361
to represent
Figure SMS_364
To pair
Figure SMS_368
In the expectation of the above-mentioned method,
Figure SMS_351
is a hyper-parameter, is used for controlling the punishment of the distribution similarity degree to the whole objective function,
Figure SMS_359
is a user interaction implicit vector
Figure SMS_353
Of (2)The degree, KL, indicates the KL divergence,
Figure SMS_365
which represents a distribution a priori, and,
Figure SMS_354
the variance vector is represented by a vector of variances,
Figure SMS_360
which represents the square of the mean vector and,
Figure SMS_352
in order to be a super-parameter,
Figure SMS_362
representing the original dynamic mainstream feature vector,
Figure SMS_366
representing the reconstructed dynamic mainstream feature vector.
Step S5, recommendation result generation
Inputting the user to be predicted according to the collaborative filtering module obtained by the training completion of the step S4
Figure SMS_369
Currently observed user interaction data
Figure SMS_370
Encoder output mean vector
Figure SMS_371
Sum variance vector
Figure SMS_372
Then is represented by the formula
Figure SMS_373
Calculating to obtain a user interaction hidden vector
Figure SMS_374
(ii) a Then the user interaction is hidden
Figure SMS_375
And step S3 outputThe decoder outputs reconstructed user interaction data, the dimensionality of the decoder is n-dimensional and is the same as the quantity of all articles, and the value of each dimensionality is 0,1]In between.
Aiming at the obtained reconstructed user interaction data, firstly, eliminating articles which do not meet the time requirement, and only considering
Figure SMS_376
And the previous time period, then any release time is
Figure SMS_377
And setting the value of the dimension where the subsequent article serial number is positioned as 0. Secondly, removing the objects which have appeared in the historical data and the currently observed interaction situation
Figure SMS_378
And recording the dimension serial numbers with the values not being 0, and setting the dimensions of the reconstructed user interaction data as 0. And finally, sequencing the reconstructed user interaction data from large to small, wherein the dimension serial number of top-N is the user to be predicted
Figure SMS_379
The top N item recommendation lists.
Example 2
The embodiment provides a system for correcting mainstream deviation of a recommendation system, which comprises a data collection and processing module, a mainstream score calculation module, a dynamic mainstream degree feature model construction module, a collaborative filtering module construction module and a recommendation result generation module, wherein the specific content of each module is as follows:
the data collection and processing module is used for acquiring the user information, the article information and the user article interaction information in the recommendation system and respectively constructing the user co-occurrence vectors
Figure SMS_380
Co-occurrence vector of article
Figure SMS_381
。
Constructing any user according to user article interaction information
Figure SMS_383
User co-occurrence vectors with all items
Figure SMS_387
，
Figure SMS_389
Representing a user
Figure SMS_384
And articles
Figure SMS_386
The interaction scenario of (2). In the scene with insensitive interaction times (only paying attention to whether there is excessive interaction, but not paying attention to the interaction times, such as movie recommendation, book recommendation and the like), if the explicit interaction is generated
Figure SMS_390
Otherwise, otherwise
Figure SMS_392
(ii) a In the case of sensitive interaction times (concerning whether there is an interaction, and also concerning the number of interactions, such as music recommendation, short video recommendation, etc.), if an explicit interaction is generated
Figure SMS_382
Wherein
Figure SMS_388
Representing a user
Figure SMS_391
And articles
Figure SMS_393
The number of interactions; otherwise
Figure SMS_385
。
Constructing any article according to user article interaction information
Figure SMS_397
Co-occurrence vector with all articles
Figure SMS_399
，
Figure SMS_402
Representing an article
Figure SMS_396
And the user
Figure SMS_400
The interaction scenario of (2). In the scene with sensitive interaction times (only paying attention to whether there is interaction, but not paying attention to the interaction times, such as movie recommendation, book recommendation and the like), if explicit interaction is generated
Figure SMS_403
Otherwise
Figure SMS_405
. In case of sensitive interaction times (concerning whether there is any interaction, and also concerning the number of interactions, such as music recommendations, short video recommendations, etc.), if an explicit interaction is generated
Figure SMS_394
In which
Figure SMS_398
Representing an article
Figure SMS_401
And the user
Figure SMS_404
The number of interactions; otherwise
Figure SMS_395
。
Wherein,
Figure SMS_406
which represents the total number of users,
Figure SMS_407
the total number of items is indicated and,
Figure SMS_408
the items are shown as being in the form of objects,
Figure SMS_409
is shown as
Figure SMS_410
The number of the articles is increased, and the articles,
Figure SMS_411
representing the user.
Data cleaning is mainly to filter out part of users and articles according to a threshold (for example, to filter out articles with interaction times smaller than a certain threshold), and the purpose is to clear abnormal data to ensure normal operation of a recommendation process.
And the mainstream score calculating module is mainly used for quantizing the user and the global mainstream degree and facilitating the generation of the dynamic mainstream degree characteristic vector in the dynamic mainstream degree characteristic model building module. Since the concept of mainstream level involves both individual users and overall users, the mainstream level scores of individual users and global can be calculated separately here.
According to the article
Figure SMS_412
Co-occurrence vector with article
Figure SMS_413
The total number of interactions of the article can be calculated
Figure SMS_414
. Since the mainstream degree is a dynamically changing concept, the article interaction times can be filtered according to time, such as only considering occurrence
Figure SMS_415
Interaction in the time period only takes the release date into consideration
Figure SMS_416
Previous article with total number of article interactions of
Figure SMS_417
Number of interactions of any article therein
Figure SMS_418
Only data within this time period is considered.
According to the user
Figure SMS_419
Co-occurrence vector with user
Figure SMS_420
The total number of interactions of the user can be calculated
Figure SMS_421
. Since the degree of main flow is a dynamically changing concept, the number of user interactions can be filtered according to time, e.g. only considering the occurrence
Figure SMS_422
And (3) the total interaction times of the user in the time period are as follows:
Figure SMS_423
wherein is arbitrary
Figure SMS_424
Only the interactive data in the time period are considered, and the release date is
Figure SMS_425
The subsequent item interaction data are all set to 0.
According to the idea of collaborative filtering, the degree of user's mainstream depends on whether their interactive items are interacted with by other users. Meanwhile, the method considers that the preference degrees of users to different categories of articles are inconsistent, so the influence of the article category factor is also considered in the calculation process of the mainstream degree score of the user. According to the total number of interactions of the article
Figure SMS_428
Total number of interactions of user
Figure SMS_429
Item category, computing user
Figure SMS_433
Dynamic mainstream degree score of
Figure SMS_427
. In the case of insensitive number of interactions, for the category
Figure SMS_431
Article of, user
Figure SMS_432
Dynamic mainstream degree score of
Figure SMS_435
Calculating according to the formula (1); in the case of sensitive number of interactions, for the category
Figure SMS_426
Article of, user
Figure SMS_430
Dynamic mainstream level score of
Figure SMS_434
Calculating according to the formula (2);
Figure SMS_436
（1）
Figure SMS_437
（2）
wherein,
Figure SMS_439
will be provided with
Figure SMS_442
During the period of time, the user can select the time period,user' s
Figure SMS_444
All items that have interacted add up to the number of interactions of all users. Because the interaction records of real-world objects have the phenomenon of power law distribution (namely a small part of the objects occupy most of the interactions, and most of the objects have no interactions), the total number of interactions between each object and other users is determined by using a logarithmic function
Figure SMS_440
Inhibition, base number
Figure SMS_443
Is a hyper-parameter. Dynamic mainstream score of user in case of insensitive interaction times
Figure SMS_446
Each interaction record is given the same weight. Dynamic mainstream score of user in case of sensitive interaction times
Figure SMS_447
Each interaction record is given a different weight,
Figure SMS_438
representing a user
Figure SMS_441
And articles
Figure SMS_445
This means that the greater the number of user interactions, the greater the weight in the evaluation of the degree of item occupancy in the score of the degree of mainstream of the user.
According to the user
Figure SMS_448
Dynamic mainstream degree score of
Figure SMS_449
For the category of
Figure SMS_450
Global dynamic mainstream level score
Figure SMS_451
Is the average of the dynamic mainstream level scores of all users, so the global dynamic mainstream level score
Figure SMS_452
The calculation formula of (c) is:
Figure SMS_453
then, the global dynamic mainstream degree scores of all the article categories are divided
Figure SMS_454
Make up a dimension of
Figure SMS_455
Global dynamic mainstream level vector of
Figure SMS_456
Expressed as:
Figure SMS_457
wherein,
Figure SMS_458
、
Figure SMS_462
both of which represent the time that it takes,
Figure SMS_464
representing the hyper-parameter (for controlling the logarithmic curve),
Figure SMS_459
representing an article
Figure SMS_463
Belong to the category
Figure SMS_465
，
Figure SMS_466
Which represents the total number of users,
Figure SMS_460
a set of all the users is represented,
Figure SMS_461
representing the total number of item categories.
And then weighting the co-occurrence vectors of the users according to the obtained mainstream degree scores of the users, namely introducing the mainstream information of the users into the co-occurrence vectors. The weighting process is described as: user will be
Figure SMS_467
To belong to the category
Figure SMS_468
Article of
Figure SMS_469
Of the interaction situation
Figure SMS_470
Multiplication by the user
Figure SMS_471
In the category of
Figure SMS_472
Dynamic mainstream degree score of
Figure SMS_473
；
For user co-occurrence vectors
Figure SMS_474
Each item in the list is weighted, and the co-occurrence vector of the whole user is obtained after the weighting is finished
Figure SMS_475
Normalization using softmax function, resulting in input collaborative filteringUser interaction data for modules
Figure SMS_476
：
Figure SMS_477
Wherein any of the user interaction data
Figure SMS_478
Has a value range of [0,1 ]]。
User interaction data obtained here
Figure SMS_479
As well as the input of the encoder followed by the collaborative filtering module.
And the dynamic mainstream degree characteristic model building module is used for building the dynamic mainstream degree characteristic model.
Can be calculated according to the mainstream score calculation module when
Figure SMS_481
Time slot arbitrary user
Figure SMS_484
With respect to any category of item collections
Figure SMS_486
Score of degree of mainstream
Figure SMS_482
. For any user
Figure SMS_483
Calculating its mainstream level scores for the set of all categories of items, which may constitute one mainstream level score
Figure SMS_485
Vector of dimensions, denoted as
Figure SMS_487
The vector may characterize the user
Figure SMS_480
The mainstream degree of all the item categories, the larger the score of the mainstream degree of an item category is, the more the user prefers popular items in the category, and the smaller the score is, the more the user prefers popular items in the category.
In order to enable the collaborative filtering model of the collaborative filtering module building module to fully utilize the user information and the global information, a dynamic mainstream feature expression based on the user information and the global information needs to be obtained. A dynamic mainstream level feature model is thus constructed here. And constructing a dynamic mainstream degree feature model based on the three-layer perceptron MLP model, wherein the dynamic mainstream degree feature model is used for extracting key features in user information and global information and reducing the original input into a hidden vector with a lower dimensionality. The first two layers of the dynamic mainstream degree characteristic model use a ReLU function as an activation function, and the last layer of the dynamic mainstream degree characteristic model uses a softmax activation function. With user information vectors
Figure SMS_488
And the global dynamic mainstream degree vector output by the mainstream fraction calculation module
Figure SMS_489
Spliced and used as input of dynamic mainstream degree characteristic model
Figure SMS_490
。
The user information vector
Figure SMS_491
Expressed as:
Figure SMS_492
wherein,
Figure SMS_493
representing a user
Figure SMS_494
The age information of the patient is quantified by the age-information-measuring device,
Figure SMS_495
representing a user
Figure SMS_496
The binary gender information of (1);
inputting the dynamic mainstream degree characteristic model
Figure SMS_497
Expressed as:
Figure SMS_498
；
wherein,
Figure SMS_499
representing vector stitching operations
Input device
Figure SMS_500
And outputting the dynamic mainstream characteristic hidden vector after dimensionality reduction of the three-layer perceptron of the dynamic mainstream degree characteristic model.
The collaborative filtering module construction module is used for constructing a collaborative filtering module, and adopts an asymmetric variational self-encoder (VAE) as a main structure of a collaborative filtering model, and the structure of the collaborative filtering module is shown in FIG. 3. The method adopts an asymmetric variational self-encoder to carry out collaborative filtering, and aims to add extra dynamic mainstream degree information into a hidden layer between an encoder and a decoder, so that the decoder can directly utilize personal information of a user and global mainstream degree during decoding.
The collaborative filtering module comprises an encoder and a decoder;
the encoder is constructed by adopting a three-layer perceptron MLP model. User interaction data
Figure SMS_502
Input encoder, encoder calculating user interaction data
Figure SMS_504
And each dimension of (1), and separately generate
Figure SMS_506
Mean value of
Figure SMS_503
Variance, forming a mean vector of users
Figure SMS_505
Sum variance vector
Figure SMS_507
Wherein the two vectors are both in t dimension, and then h dimension user interaction hidden vectors are generated by random sampling
Figure SMS_508
，
Figure SMS_501
. Since the network cannot perform back propagation due to the random sampling method, the re-parameterization method is adopted to complete the sampling process.
The decoder is constructed by adopting a four-layer perceptron MLP model, the first three layers of activation functions of the decoder are tanh functions, and the last layer of activation functions of the decoder are softmax functions and are used for generating probability distribution
Figure SMS_509
. Dynamic mainstream characteristic hidden vectors output by the dynamic mainstream degree characteristic model building module and user interaction hidden vectors output by the encoder
Figure SMS_510
As input to a decoder, the decoder output reconstructs user interaction data
Figure SMS_511
And reconstructing dynamic mainstream feature vectors
Figure SMS_512
Reconstructing dynamic mainstream feature vectors
Figure SMS_513
For completing the reconstruction of the decoder.
Variational autocoder inference procedures are as follows, assuming a user
Figure SMS_514
Corresponding user interaction hidden vector
Figure SMS_515
Compliance
Figure SMS_516
Is normally distributed. Assuming user interaction data input to the encoder according to the recommendation system interaction data characteristics
Figure SMS_517
Obey probability of
Figure SMS_518
The likelihood function of the polynomial distribution of (1) is as follows:
Figure SMS_519
wherein,
Figure SMS_520
representing hidden vectors interacted by user
Figure SMS_521
Is determined and is
Figure SMS_522
Probability of individual item interaction;
in order to enable the network to learn the parameters by back-propagation, a posteriori distributions for each user data sample must be found
Figure SMS_524
Since this posterior distribution is not easy to find, the variation distribution is used here by means of variation estimation
Figure SMS_527
To approximate
Figure SMS_529
. Suppose that
Figure SMS_525
Satisfy the Gaussian distribution
Figure SMS_526
In which
Figure SMS_531
Is a variance vector
Figure SMS_533
Diagonal covariance matrix of. Then the optimization goal of the network at this time is to optimize the parameter generation mean vector
Figure SMS_523
Sum variance vector
Figure SMS_528
Make variation distribution
Figure SMS_530
And posterior distribution
Figure SMS_532
As similar as possible.
Wherein the reparameterization method operates as follows, assuming noise
Figure SMS_534
Obey normal distribution
Figure SMS_535
Hidden vector of user interaction
Figure SMS_536
Can be represented by a variance vector
Figure SMS_537
Mean vector of
Figure SMS_538
Linearly combined with noise to obtainThereby enabling the network to learn. The reparameterization formula is as follows:
Figure SMS_539
wherein,
Figure SMS_540
is the standard deviation;
unlike the standard variational self-encoder network, an asymmetric structure is used to obtain the user interaction implicit vector
Figure SMS_541
Then, the input generated in the dynamic mainstream degree characteristic model building module is input
Figure SMS_542
Is spliced at
Figure SMS_543
And then fed into the decoder. The generating part of the decoder may be divided into reconstructing user interaction data
Figure SMS_544
And reconstructing dynamic mainstream feature vectors
Figure SMS_545
。
In conclusion, loss function of collaborative filtering model based on asymmetric variational self-encoder
Figure SMS_546
Dividing into reconstructed target losses
Figure SMS_547
Distribution approximation loss
Figure SMS_548
And dynamic mainstream eigenvector approximation loss
Figure SMS_549
Three parts, loss function
Figure SMS_550
The calculation formula of (c) is:
Figure SMS_551
reconstructing object loss
Figure SMS_552
The purpose of this is to make the reconstructed user interaction data output by the decoder as identical as possible to the user interaction data input to the encoder, the calculation formula being:
Figure SMS_553
distribution approximation loss
Figure SMS_554
Is to make the variation distributed
Figure SMS_555
As close to a posterior distribution as possible
Figure SMS_556
For measuring the approximation degree of the two distributions, the calculation formula is:
Figure SMS_557
to convert the original dynamic mainstream feature vector
Figure SMS_558
And reconstructing dynamic mainstream feature vectors
Figure SMS_559
With its negative number as a loss term, with the aim of enabling the reconstruction of dynamic mainstream feature vectors
Figure SMS_560
In the direction of and withOriginal dynamic mainstream feature vector
Figure SMS_561
Approximate losses by dynamic mainstream feature vectors as close as possible
Figure SMS_562
The method can enable a decoder to complete the reconstruction process by using the dynamic mainstream characteristics as much as possible. Dynamic mainstream eigenvector approximation loss
Figure SMS_563
The calculation formula of (2) is as follows:
Figure SMS_564
wherein,
Figure SMS_569
a hidden vector representing the user interaction is shown,
Figure SMS_567
representing the data of the user interaction(s),
Figure SMS_578
representing the posterior distribution of each user data sample,
Figure SMS_571
representing variation distribution
Figure SMS_579
And posterior distribution
Figure SMS_572
In the approximation that the difference between the first and second values,
Figure SMS_580
represent
Figure SMS_575
To pair
Figure SMS_581
In the expectation of the above-mentioned method,
Figure SMS_565
is a hyper-parameter, is used for controlling the punishment of the distribution similarity degree to the whole objective function,
Figure SMS_577
is a user interaction implicit vector
Figure SMS_566
KL represents the KL divergence,
Figure SMS_573
which represents a distribution a priori, and,
Figure SMS_576
which represents a vector of the variance (m) of the signal,
Figure SMS_582
which represents the square of the mean vector and,
Figure SMS_568
in order to be a hyper-parameter,
Figure SMS_574
representing the original dynamic mainstream feature vector(s),
Figure SMS_570
representing the reconstructed dynamic mainstream feature vector.
A recommendation result generation module for inputting the user to be predicted according to the collaborative filtering module obtained by the collaborative filtering module construction module after training
Figure SMS_583
Currently observed user interaction data
Figure SMS_584
Encoder output mean vector
Figure SMS_585
Sum variance vector
Figure SMS_586
Then by the formula
Figure SMS_587
Calculating to obtain a user interaction hidden vector
Figure SMS_588
(ii) a Then the user interaction is hidden
Figure SMS_589
Inputting the dynamic mainstream characteristic hidden vector output by the dynamic mainstream degree characteristic model building module into a decoder, outputting the reconstructed user interaction data by the decoder, wherein the dimensionality of the user interaction data is n-dimensional and is the same as the quantity of all articles, and the value of each dimensionality is 0,1]In between.
Aiming at the obtained reconstructed user interaction data, firstly, eliminating articles which do not meet the time requirement, and only considering
Figure SMS_590
And the previous time period, then any release time is
Figure SMS_591
And setting the value of the dimension where the subsequent article serial number is positioned as 0. Secondly, removing the objects which have appeared in the historical data and the currently observed interaction situation
Figure SMS_592
And recording the dimension serial numbers with the values not being 0, and setting the dimensions of the reconstructed user interaction data as 0. And finally, sequencing the reconstructed user interaction data from large to small, wherein the dimension serial number with the large top-N is the user to be predicted
Figure SMS_593
The top N item recommendation lists.
Example 3
The present embodiment provides a computer device, which includes a memory and a processor, wherein the memory stores a computer program, and the computer program, when executed by the processor, causes the processor to execute the steps of the above method for correcting a mainstream deviation of a recommendation system.
The computer device may be a desktop computer, a notebook, a palm computer, a cloud server, or other computing devices. The computer equipment can carry out man-machine interaction with a user through a keyboard, a mouse, a remote controller, a touch panel or voice control equipment and the like.
The memory includes at least one type of readable storage medium including a flash memory, a hard disk, a multimedia card, a card-type memory (e.g., SD or D interface display memory, etc.), a Random Access Memory (RAM), a Static Random Access Memory (SRAM), a read-only memory (ROM), an electrically erasable programmable read-only memory (EEPROM), a programmable read-only memory (PROM), a magnetic memory, a magnetic disk, an optical disk, etc. In some embodiments, the storage may be an internal storage unit of the computer device, such as a hard disk or a memory of the computer device. In other embodiments, the memory may also be an external storage device of the computer device, such as a plug-in hard disk, a Smart Media Card (SMC), a Secure Digital (SD) Card, a Flash memory Card (Flash Card), or the like provided on the computer device. Of course, the memory may also include both internal and external storage units of the computer device. In this embodiment, the memory is usually used for storing an operating system and various types of application software installed on the computer device, for example, program codes of the method for correcting the mainstream deviation of the recommendation system, and the like. In addition, the memory may also be used to temporarily store various types of data that have been output or are to be output.
The processor may be a Central Processing Unit (CPU), controller, microcontroller, microprocessor, or other data Processing chip in some embodiments. The processor is typically used to control the overall operation of the computer device. In this embodiment, the processor is configured to execute the program code stored in the memory or process data, for example, execute the program code of the method for correcting the deviation of the main stream of the recommendation system.
Example 4
The present embodiment provides a computer-readable storage medium, in which a computer program is stored, and the computer program, when executed by a processor, causes the processor to execute the steps of the above method for correcting a deviation of a mainstream of a recommendation system.
Wherein the computer readable storage medium stores an interface display program executable by at least one processor to cause the at least one processor to perform the steps of the method for correcting a recommended system mainstream deviation as described above.
Through the above description of the embodiments, those skilled in the art will clearly understand that the method of the above embodiments can be implemented by software plus a necessary general hardware platform, and certainly can also be implemented by hardware, but in many cases, the former is a better implementation manner. Based on such understanding, the technical solutions of the present application or portions contributing to the prior art may be embodied in the form of a software product, where the computer software product is stored in a storage medium (e.g., ROM/RAM, magnetic disk, optical disk), and includes several instructions for enabling a terminal device (which may be a mobile phone, a computer, a server, or a network device, etc.) to execute the method for correcting mainstream deviations of the recommendation system according to the embodiments of the present application.

Claims (9)

1. A method for correcting deviations in a recommendation system mainstream comprising the steps of:
step S1, data collection and processing
Obtaining user information, article information and user article interaction information in a recommendation system, and respectively constructing user co-occurrence vectors
Figure QLYQS_1
Co-occurrence vector of article
Figure QLYQS_2
；
Step S2, calculating the mainstream score
According to co-occurrence vector of articles
Figure QLYQS_4
Calculating the total interaction times of the articles
Figure QLYQS_7
(ii) a According to user co-occurrence vectors
Figure QLYQS_10
Calculating the total number of interactions
Figure QLYQS_6
(ii) a According to the total number of interaction times of the articles
Figure QLYQS_8
Total number of interactions of user
Figure QLYQS_11
Item categories, computing users
Figure QLYQS_13
Dynamic mainstream degree score of
Figure QLYQS_5
(ii) a According to the user
Figure QLYQS_9
Dynamic mainstream level score of
Figure QLYQS_12
Calculating the average value of the dynamic mainstream degree scores of all the users to obtain the global dynamic mainstream degree score
Figure QLYQS_14
And forming global dynamic mainstream degree scores of all article categories into a global dynamic mainstream degree vector
Figure QLYQS_3
；
S3, constructing a dynamic mainstream degree characteristic model
MLP model structure based on three-layer perceptronBuilding a dynamic mainstream degree characteristic model, wherein the first two layers of the dynamic mainstream degree characteristic model use a ReLU function as an activation function, and the last layer of the dynamic mainstream degree characteristic model uses a softmax activation function; with user information vectors
Figure QLYQS_15
And the global dynamic mainstream degree vector output by the step S2
Figure QLYQS_16
Spliced as input of dynamic mainstream degree characteristic model
Figure QLYQS_17
Outputting a dynamic mainstream characteristic hidden vector by the dynamic mainstream degree characteristic model;
s4, constructing a collaborative filtering module
Constructing a collaborative filtering module comprising an encoder and a decoder;
the encoder is constructed by adopting a three-layer perceptron MLP model; user interaction data
Figure QLYQS_18
Input encoder, encoder calculating user interaction data
Figure QLYQS_19
And each dimension of (a) and generate separately
Figure QLYQS_20
Mean value of
Figure QLYQS_21
Variance, forming a mean vector of users
Figure QLYQS_22
Sum variance vector
Figure QLYQS_23
Wherein the two vectors are both in t dimension and are randomly sampled to generate h dimensionUser interaction hidden vector of
Figure QLYQS_24
；
The decoder is constructed by adopting a four-layer perceptron MLP model, the first three layers of activation functions of the decoder are tanh functions, and the last layer of activation functions of the decoder are softmax functions; the dynamic mainstream characteristic hidden vector output by the step S3 and the user interaction hidden vector output by the encoder
Figure QLYQS_25
As input to a decoder, the decoder output reconstructs user interaction data
Figure QLYQS_26
And reconstructing dynamic mainstream feature vectors
Figure QLYQS_27
Reconstructing dynamic mainstream feature vectors
Figure QLYQS_28
For completing the reconstruction of the decoder;
step S5, recommendation result generation
Inputting user interaction data currently observed by the user to be predicted according to the collaborative filtering module obtained by the training completion in the step S4
Figure QLYQS_29
The encoder outputs a user interaction hidden vector
Figure QLYQS_30
(ii) a Then the user interaction is hidden
Figure QLYQS_31
And inputting the dynamic mainstream characteristic hidden vector output in the step S3 into a decoder, and outputting reconstructed user interaction data by the decoder.
2. The method of claim 1, wherein the deviation of the mainstream of the recommendation system is correctedThe method is characterized in that: in step S1, any user is constructed according to the user article interaction information
Figure QLYQS_32
User co-occurrence vector with all items
Figure QLYQS_33
Constructing any article according to user article interaction information
Figure QLYQS_34
Co-occurrence vector with all articles
Figure QLYQS_35
；
Wherein,
Figure QLYQS_36
which represents the total number of users,
Figure QLYQS_39
the total number of the items is represented,
Figure QLYQS_42
the items are shown as being in the form of objects,
Figure QLYQS_38
denotes the first
Figure QLYQS_41
The number of the articles is one,
Figure QLYQS_45
representing a user
Figure QLYQS_46
And articles
Figure QLYQS_37
The interaction situation of (a) is,
Figure QLYQS_40
representing an article
Figure QLYQS_43
And the user
Figure QLYQS_44
The interaction scenario of (2).
3. A method for correcting deviations in a recommendation system mainstream according to claim 1, wherein: in step S2, according to the article
Figure QLYQS_47
Co-occurrence vector with article
Figure QLYQS_48
Calculating the total number of article interactions
Figure QLYQS_49
；
According to the user
Figure QLYQS_50
Co-occurrence vector with user
Figure QLYQS_51
And the total number of times of interaction of the user is calculated,
Figure QLYQS_52
；
according to the total number of interactions of the article
Figure QLYQS_55
Total number of interactions of user
Figure QLYQS_58
Item category, computing user
Figure QLYQS_60
Dynamic mainstream degree score of
Figure QLYQS_54
(ii) a In the case of insensitive number of interactions, for the category
Figure QLYQS_57
Article of, user
Figure QLYQS_59
Dynamic mainstream degree score of
Figure QLYQS_62
Calculating according to the formula (1); in the case of sensitive number of interactions, for the category
Figure QLYQS_53
Article of, user
Figure QLYQS_56
Dynamic mainstream degree score of
Figure QLYQS_61
Calculating according to the formula (2);
Figure QLYQS_63
（1）
Figure QLYQS_64
（2）
according to the user
Figure QLYQS_65
Dynamic mainstream degree score of
Figure QLYQS_66
Calculating the average value of the dynamic mainstream degree scores of all the users to obtain the global dynamic mainstream degree score
Figure QLYQS_67
The calculation formula is as follows:
Figure QLYQS_68
then, the global dynamic mainstream degree scores of all the article categories are divided
Figure QLYQS_69
Make up a dimension of
Figure QLYQS_70
Global dynamic mainstream level vector of
Figure QLYQS_71
Expressed as:
Figure QLYQS_72
wherein,
Figure QLYQS_74
、
Figure QLYQS_80
all of which represent the time of day,
Figure QLYQS_81
the representation of the hyper-parameter is,
Figure QLYQS_75
representing an article
Figure QLYQS_76
Belong to the category
Figure QLYQS_77
，
Figure QLYQS_78
Which represents the total number of users,
Figure QLYQS_73
a set of all the users is represented,
Figure QLYQS_79
representing the total number of item categories.
4. A method of correcting deviations in a recommendation system mainstream according to claim 3, wherein: and carrying out co-occurrence vector weighting processing, wherein the weighting processing is described as follows: user will be
Figure QLYQS_82
To belong to the category
Figure QLYQS_83
Article of
Figure QLYQS_84
Of the interaction situation
Figure QLYQS_85
Multiplication by the user
Figure QLYQS_86
In the category of
Figure QLYQS_87
Dynamic mainstream degree score of
Figure QLYQS_88
；
For user co-occurrence vectors
Figure QLYQS_89
Each item in the list is weighted, and after the weighting is finished, the co-occurrence vector of the whole user is obtained
Figure QLYQS_90
Normalizing by using a softmax function to obtain the user interaction number for inputting the collaborative filtering moduleAccording to
Figure QLYQS_91
：
Figure QLYQS_92
Wherein any of the user interaction data
Figure QLYQS_93
Has a value range of [0,1 ]]。
5. A method for correcting deviations in a recommendation system mainstream according to claim 1, wherein: in step S3, the user information vector
Figure QLYQS_94
Expressed as:
Figure QLYQS_95
wherein,
Figure QLYQS_96
representing a user
Figure QLYQS_97
The age information of the person to be treated is quantified,
Figure QLYQS_98
representing a user
Figure QLYQS_99
The binary gender information of (1);
input of dynamic mainstream degree characteristic model
Figure QLYQS_100
Expressed as:
Figure QLYQS_101
；
wherein,
Figure QLYQS_102
a vector stitching operation is represented.
6. A method for correcting deviations in a recommendation system mainstream according to claim 1, wherein: in step S4, the loss function of the collaborative filtering module
Figure QLYQS_103
Divided into reconstructed target losses
Figure QLYQS_104
Distribution approximation loss
Figure QLYQS_105
And dynamic mainstream eigenvector approximation loss
Figure QLYQS_106
Three parts, loss function
Figure QLYQS_107
The calculation formula of (2) is as follows:
Figure QLYQS_108
reconstructing object losses
Figure QLYQS_109
The calculation formula of (2) is as follows:
Figure QLYQS_110
distribution approximation loss
Figure QLYQS_111
The calculation formula of (2) is as follows:
Figure QLYQS_112
dynamic mainstream eigenvector approximation loss
Figure QLYQS_113
The calculation formula of (2) is as follows:
Figure QLYQS_114
wherein,
Figure QLYQS_123
a hidden vector representing the user interaction is shown,
Figure QLYQS_117
representing the data of the user interaction(s),
Figure QLYQS_119
representing the posterior distribution of each user data sample,
Figure QLYQS_125
representing variation distribution
Figure QLYQS_128
And posterior distribution
Figure QLYQS_129
In the approximation that the difference between the first and second values,
Figure QLYQS_132
to represent
Figure QLYQS_124
To pair
Figure QLYQS_130
In the expectation that the position of the target is not changed,
Figure QLYQS_118
in order to be a hyper-parameter,
Figure QLYQS_121
is a user interaction hidden vector
Figure QLYQS_120
KL represents the KL divergence,
Figure QLYQS_126
a distribution is represented a priori, which is,
Figure QLYQS_127
the variance vector is represented by a vector of variances,
Figure QLYQS_131
which represents the square of the mean vector and,
Figure QLYQS_116
in order to be a hyper-parameter,
Figure QLYQS_122
representing the original dynamic mainstream feature vector,
Figure QLYQS_115
representing the reconstructed dynamic mainstream feature vector.
7. A system for correcting deviations in a recommendation system mainstream comprising:
a data collecting and processing module for obtaining user information, article information and user article interaction information in the recommendation system and respectively constructing user co-occurrence vectors
Figure QLYQS_133
Co-occurrence vector of article
Figure QLYQS_134
；
A mainstream score calculation module for calculating a mainstream score based on the co-occurrence vector of the article
Figure QLYQS_137
Calculating the total interaction times of the articles
Figure QLYQS_141
(ii) a According to user co-occurrence vectors
Figure QLYQS_143
Calculating the total number of interactions of the user
Figure QLYQS_138
(ii) a According to the total number of interactions of the article
Figure QLYQS_140
Total number of interactions of user
Figure QLYQS_142
Item category, computing user
Figure QLYQS_145
Dynamic mainstream level score of
Figure QLYQS_136
(ii) a According to the user
Figure QLYQS_139
Dynamic mainstream degree score of
Figure QLYQS_144
Calculating the average value of the dynamic mainstream degree scores of all the users to obtain the global dynamic mainstream degree score
Figure QLYQS_146
And forming global dynamic mainstream degree scores of all article categories into a global dynamic mainstream degree vector
Figure QLYQS_135
；
The dynamic mainstream degree characteristic model building module is used for building a dynamic mainstream degree characteristic model based on a three-layer perceptron MLP model, the ReLU function is used as an activation function in the first two layers of the dynamic mainstream degree characteristic model, and the softmax activation function is used in the last layer of the dynamic mainstream degree characteristic model; with user information vectors
Figure QLYQS_147
And the global dynamic mainstream degree vector output by the mainstream fraction calculation module
Figure QLYQS_148
Spliced as input of dynamic mainstream degree characteristic model
Figure QLYQS_149
Outputting a dynamic mainstream characteristic hidden vector by the dynamic mainstream degree characteristic model;
the collaborative filtering module construction module is used for constructing a collaborative filtering module comprising an encoder and a decoder;
the encoder is constructed by adopting a three-layer perceptron MLP model; user interaction data
Figure QLYQS_150
Input encoder, encoder calculating user interaction data
Figure QLYQS_151
And each dimension of (1), and separately generate
Figure QLYQS_152
Mean value of
Figure QLYQS_153
Variance, forming a mean vector of users
Figure QLYQS_154
Sum variance vector
Figure QLYQS_155
Wherein the two vectors are both in t dimension, and then h dimension user interaction hidden vector is generated by random sampling
Figure QLYQS_156
；
The decoder is constructed by adopting a four-layer perceptron MLP model, the first three layers of activation functions of the decoder are tanh functions, and the last layer of activation functions of the decoder are softmax functions; dynamic mainstream characteristic hidden vectors output by the dynamic mainstream degree characteristic model building module and user interaction hidden vectors output by the encoder
Figure QLYQS_157
As input to the decoder, the decoder outputs reconstructed user interaction data
Figure QLYQS_158
And reconstructing dynamic mainstream feature vectors
Figure QLYQS_159
Reconstructing dynamic mainstream feature vectors
Figure QLYQS_160
For completing the reconstruction of the decoder;
a recommendation result generation module for inputting the user interaction data currently observed by the user to be predicted according to the collaborative filtering module obtained by the collaborative filtering module construction module after training
Figure QLYQS_161
The encoder outputs a user interaction hidden vector
Figure QLYQS_162
(ii) a Then the user interaction is hidden
Figure QLYQS_163
And dynamic mainstream level characteristicsAnd inputting the dynamic mainstream characteristic hidden vector output by the model building module into a decoder, and outputting reconstructed user interaction data by the decoder.
8. A computer device, characterized by: comprising a memory and a processor, the memory storing a computer program which, when executed by the processor, causes the processor to perform the steps of the method according to any one of claims 1 to 6.
9. A computer-readable storage medium, characterized in that: stored with a computer program which, when executed by a processor, causes the processor to carry out the steps of the method according to any one of claims 1 to 6.
CN202310104256.6A 2023-02-13 2023-02-13 Method, system, device and storage medium for correcting mainstream deviation of recommendation system Active CN115809374B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202310104256.6A CN115809374B (en) 2023-02-13 2023-02-13 Method, system, device and storage medium for correcting mainstream deviation of recommendation system

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202310104256.6A CN115809374B (en) 2023-02-13 2023-02-13 Method, system, device and storage medium for correcting mainstream deviation of recommendation system

Publications (2)

Publication Number Publication Date
CN115809374A true CN115809374A (en) 2023-03-17
CN115809374B CN115809374B (en) 2023-04-18

Family

ID=85487843

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202310104256.6A Active CN115809374B (en) 2023-02-13 2023-02-13 Method, system, device and storage medium for correcting mainstream deviation of recommendation system

Country Status (1)

Country Link
CN (1) CN115809374B (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116186421A (en) * 2023-05-04 2023-05-30 中国科学技术大学 Recommendation method, system, equipment and storage medium for eliminating popularity deviation

Citations (13)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20080097821A1 (en) * 2006-10-24 2008-04-24 Microsoft Corporation Recommendations utilizing meta-data based pair-wise lift predictions
US20160179950A1 (en) * 2014-12-22 2016-06-23 Rovi Guides, Inc. Systems and methods for filtering techniques using metadata and usage data analysis
CN108647226A (en) * 2018-03-26 2018-10-12 浙江大学 A kind of mixing recommendation method based on variation autocoder
CN112184391A (en) * 2020-10-16 2021-01-05 中国科学院计算技术研究所 Recommendation model training method, medium, electronic device and recommendation model
CN113158024A (en) * 2021-02-26 2021-07-23 中国科学技术大学 Causal reasoning method for correcting popularity deviation of recommendation system
CN113987358A (en) * 2021-11-15 2022-01-28 中国科学技术大学 Training method, recommendation method and recommendation system of recommendation model
CN114428910A (en) * 2022-01-28 2022-05-03 腾讯科技（深圳）有限公司 Resource recommendation method and device, electronic equipment, product and medium
US20220188899A1 (en) * 2019-08-30 2022-06-16 Tata Consultancy Services Limited System and method for handling popularity bias in item recommendations
CN114861783A (en) * 2022-04-26 2022-08-05 北京三快在线科技有限公司 Recommendation model training method and device, electronic equipment and storage medium
CN114912033A (en) * 2022-05-16 2022-08-16 重庆大学 Knowledge graph-based recommendation popularity deviation adaptive buffering method
CN115129945A (en) * 2022-06-23 2022-09-30 阿里巴巴新加坡控股有限公司 Graph structure contrast learning method, equipment and computer storage medium
CN115147192A (en) * 2022-07-29 2022-10-04 华东师范大学 Recommendation method and recommendation system based on double-view-angle deviation correction
CN115438871A (en) * 2022-09-23 2022-12-06 哈尔滨商业大学 Ice and snow scenic spot recommendation method and system integrating preference and eliminating popularity deviation

Patent Citations (13)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20080097821A1 (en) * 2006-10-24 2008-04-24 Microsoft Corporation Recommendations utilizing meta-data based pair-wise lift predictions
US20160179950A1 (en) * 2014-12-22 2016-06-23 Rovi Guides, Inc. Systems and methods for filtering techniques using metadata and usage data analysis
CN108647226A (en) * 2018-03-26 2018-10-12 浙江大学 A kind of mixing recommendation method based on variation autocoder
US20220188899A1 (en) * 2019-08-30 2022-06-16 Tata Consultancy Services Limited System and method for handling popularity bias in item recommendations
CN112184391A (en) * 2020-10-16 2021-01-05 中国科学院计算技术研究所 Recommendation model training method, medium, electronic device and recommendation model
CN113158024A (en) * 2021-02-26 2021-07-23 中国科学技术大学 Causal reasoning method for correcting popularity deviation of recommendation system
CN113987358A (en) * 2021-11-15 2022-01-28 中国科学技术大学 Training method, recommendation method and recommendation system of recommendation model
CN114428910A (en) * 2022-01-28 2022-05-03 腾讯科技（深圳）有限公司 Resource recommendation method and device, electronic equipment, product and medium
CN114861783A (en) * 2022-04-26 2022-08-05 北京三快在线科技有限公司 Recommendation model training method and device, electronic equipment and storage medium
CN114912033A (en) * 2022-05-16 2022-08-16 重庆大学 Knowledge graph-based recommendation popularity deviation adaptive buffering method
CN115129945A (en) * 2022-06-23 2022-09-30 阿里巴巴新加坡控股有限公司 Graph structure contrast learning method, equipment and computer storage medium
CN115147192A (en) * 2022-07-29 2022-10-04 华东师范大学 Recommendation method and recommendation system based on double-view-angle deviation correction
CN115438871A (en) * 2022-09-23 2022-12-06 哈尔滨商业大学 Ice and snow scenic spot recommendation method and system integrating preference and eliminating popularity deviation

Non-Patent Citations (3)

* Cited by examiner, † Cited by third party
Title
P. BEDI 等: "Using novelty score of unseen items to handle popularity bias in recommender systems" *
ZHIHONG CHEN 等: "Co-training Disentangled Domain Adaptation Network for Leveraging Popularity Bias in Recommenders" *
机器之心PRO: "如何斩获KDD Cup两冠一季？美团广告团队公开解决方案" *

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116186421A (en) * 2023-05-04 2023-05-30 中国科学技术大学 Recommendation method, system, equipment and storage medium for eliminating popularity deviation

Also Published As

Publication number Publication date
CN115809374B (en) 2023-04-18

Similar Documents

Publication Publication Date Title
CN110162703B (en) Content recommendation method, training device, content recommendation equipment and storage medium
Catherine et al. Transnets: Learning to transform for recommendation
Kim et al. TWILITE: A recommendation system for Twitter using a probabilistic model based on latent Dirichlet allocation
EP4181026A1 (en) Recommendation model training method and apparatus, recommendation method and apparatus, and computer-readable medium
CN110378731A (en) Obtain method, apparatus, server and the storage medium of user's portrait
CN109903103B (en) Method and device for recommending articles
US11216518B2 (en) Systems and methods of providing recommendations of content items
CN113158024B (en) Causal reasoning method for correcting popularity deviation of recommendation system
CN106610970A (en) Collaborative filtering-based content recommendation system and method
CN114519145A (en) Sequence recommendation method for mining long-term and short-term interests of users based on graph neural network
CN114202061A (en) Article recommendation method, electronic device and medium based on generation of confrontation network model and deep reinforcement learning
CN110598120A (en) Behavior data based financing recommendation method, device and equipment
CN111768239A (en) Property recommendation method, device, system, server and storage medium
CN110781401A (en) Top-n project recommendation method based on collaborative autoregressive flow
CN115809374B (en) Method, system, device and storage medium for correcting mainstream deviation of recommendation system
CN112699310A (en) Cold start cross-domain hybrid recommendation method and system based on deep neural network
US20240037133A1 (en) Method and apparatus for recommending cold start object, computer device, and storage medium
CN113362139A (en) Data processing method and device based on double-tower structure model
JP2017059193A (en) Time series image compensation device, time series image generation method, and program for time series image compensation device
CN116975427A (en) Method, device and computing equipment for determining interest degree of object in resource information
He et al. Interest HD: An interest frame model for recommendation based on HD image generation
CN115544379A (en) Quaternion map convolutional neural network-based recommendation method and device
CN112749332A (en) Data processing method, device and computer readable medium
Hien et al. A Deep Learning Model for Context Understanding in Recommendation Systems
Sun et al. Personalized recommendation for Weibo comic users

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