CN110209946B - Social and community-based product recommendation method, system and storage medium - Google Patents
Social and community-based product recommendation method, system and storage medium Download PDFInfo
- Publication number
- CN110209946B CN110209946B CN201910496478.0A CN201910496478A CN110209946B CN 110209946 B CN110209946 B CN 110209946B CN 201910496478 A CN201910496478 A CN 201910496478A CN 110209946 B CN110209946 B CN 110209946B
- Authority
- CN
- China
- Prior art keywords
- user
- product
- preference
- feedback
- relationship
- 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.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 60
- 230000003993 interaction Effects 0.000 claims abstract description 41
- 230000002452 interceptive effect Effects 0.000 claims abstract description 23
- 238000012163 sequencing technique Methods 0.000 claims abstract description 11
- 238000012545 processing Methods 0.000 claims abstract description 8
- 230000006870 function Effects 0.000 claims description 46
- 239000011159 matrix material Substances 0.000 claims description 18
- 238000000354 decomposition reaction Methods 0.000 claims description 10
- 230000008569 process Effects 0.000 claims description 6
- 230000010354 integration Effects 0.000 claims description 3
- PDEDQSAFHNADLV-UHFFFAOYSA-M potassium;disodium;dinitrate;nitrite Chemical compound [Na+].[Na+].[K+].[O-]N=O.[O-][N+]([O-])=O.[O-][N+]([O-])=O PDEDQSAFHNADLV-UHFFFAOYSA-M 0.000 claims description 3
- 239000002245 particle Substances 0.000 claims description 2
- 230000009471 action Effects 0.000 description 3
- 230000009286 beneficial effect Effects 0.000 description 3
- 238000012360 testing method Methods 0.000 description 3
- 238000012549 training Methods 0.000 description 3
- 230000007547 defect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/95—Retrieval from the web
- G06F16/953—Querying, e.g. by the use of web search engines
- G06F16/9535—Search customisation based on user profiles and personalisation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q30/00—Commerce
- G06Q30/06—Buying, selling or leasing transactions
- G06Q30/0601—Electronic shopping [e-shopping]
- G06Q30/0631—Item recommendations
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/01—Social networking
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Theoretical Computer Science (AREA)
- Databases & Information Systems (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Marketing (AREA)
- General Business, Economics & Management (AREA)
- Strategic Management (AREA)
- Accounting & Taxation (AREA)
- Finance (AREA)
- Economics (AREA)
- Development Economics (AREA)
- General Engineering & Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Computing Systems (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Human Resources & Organizations (AREA)
- Primary Health Care (AREA)
- Tourism & Hospitality (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention provides a social contact and community-based product recommendation method, a social contact and community-based product recommendation system and a storage medium, and relates to the field of data processing. The method comprises the following steps: acquiring interaction data, social relationship data and community relationship data of a user and a product, and forming historical data; determining a feedback relationship between the user and the product based on the historical data and a preset partial order relationship; determining a preference relationship of the user to the product based on the historical data and the feedback relationship; determining an objective function of the preference relationship; acquiring a preference score of a user for a product based on the target function; and sequencing the non-interactive products of the user based on the preference scores to obtain a product recommendation result. The invention can accurately recommend the product to the user.
Description
Technical Field
The invention relates to the field of data processing, in particular to a product recommendation method, a product recommendation system and a storage medium based on social contact and community.
Background
With the development of internet technology, research on personalized recommendation based on social networks is more active. And the social recommendation method can be divided into a scoring prediction algorithm based on explicit feedback data and a personalized ranking algorithm based on implicit feedback data. Wherein, the explicit feedback data generally refers to the grade, i.e. the grade of 1-5, can obviously reflect the preference degree of the user for the product. The implicit feedback data refers to the interaction relationship between the user and the product, such as browsing, purchasing, collecting and the like, and the preference degree of the user for the product cannot be obviously reflected. For those products where the user does not interact with the product, it cannot be shown that the user does not like the product, and it may be that the user has not yet discovered the product. Because implicit feedback data widely exists, the acquisition cost is low, and the method is close to reality, the sorting algorithm based on the implicit feedback data gets more and more attention.
The social recommendation method provided by the prior art is mainly based on a Bayesian personalized ranking algorithm to optimize ranking. Products interacted with the user are set as positive feedback, products not interacted with the user are set as negative feedback, the preference of the user on the interacted products is assumed to be larger than that of the non-interacted products, and the social relationship of friends of the user is considered, so that the product recommendation is optimized by means of a Bayesian personalized ranking algorithm.
However, the inventor of the present application finds that, when the recommendation method is actually applied, the social relationships of users in an online social network are relatively sparse, and therefore, the recommendation result of a product cannot be accurately obtained only by considering the social relationships. The prior art recommendation methods are therefore not accurate enough.
Disclosure of Invention
Technical problem to be solved
Aiming at the defects of the prior art, the invention provides a product recommendation method, a product recommendation system and a storage medium based on social contact and community, and solves the technical problem that the prior art cannot accurately recommend products.
(II) technical scheme
In order to achieve the purpose, the invention is realized by the following technical scheme:
the invention provides a product recommendation method based on social contact and community, which solves the technical problem, the recommendation method is executed by a computer and comprises the following steps:
acquiring interaction data, social relationship data and community relationship data of a user and a product, and forming historical data;
determining a feedback relationship between the user and the product based on the historical data and a preset partial order relationship;
determining a preference relationship of the user to the product based on the historical data and the feedback relationship;
determining an objective function of the preference relationship;
acquiring a preference score of a user for a product based on the target function;
and sequencing the non-interactive products of the user based on the preference scores to obtain a product recommendation result.
Preferably, the historical data includes:
the system comprises a user set U, a product set P, a set D in which users and products have interaction, a social relationship set S and a community relationship set G.
Preferably, the preset partial order relationship is as follows:
positive feedback > social feedback > community feedback > negative feedback;
the feedback relationship between the user and the product comprises:
and (3) positive feedback assembly: PF (particle Filter)u={<u,pi>},
Wherein: < u, piIs e.g. D, represents user u and product piInteraction exists;
social feedback set: SFu={<u,ps>},
Wherein: < u, psThe social friends of the user u have interaction, but the user u does not have interactive products;
community feedback set: GFu={<u,pg>},
Wherein: < u, pgProducts that show that social friends of the user u have interaction, but the user u does not interact with the social friends of the user u and does not interact with the social friends of the user u;
and (3) negative feedback integration: NFu={<u,pj>},
Wherein: < u, pjThe expression > represents that the user u does not interact with the social friends and the social group friends of the user u, and the social group friends of the user u also do not have interactive products.
Preferably, the preference relationship of the user to the product is represented by a matrix decomposition model based on the user set U to the product set P:
wherein:
w represents a feature matrix of the user set U;
h represents a feature matrix of the product set P;
b represents a deviation term of the product set P.
Preferably, the objective function of the preference relationship is as follows:
wherein:
σ (-) represents a logistic function;
Θ denotes a set of parameters in the matrix decomposition model, i.e., Θ ═ { W, H, b };
λΘis a regularization parameter;
CSrepresenting user u and product psThe number of friends who have interacted with the product in the friends of the user u is not interacted;
Cgrepresenting user u and product pgAnd no interaction exists, and the social friends of the user u have no interaction with the product, but the social friends of the user u have the number of the interactive social friends with the product.
Preferably, the obtaining of the preference score of the user for the product based on the objective function specifically includes:
s501, setting the iteration number as iter, wherein iter is 1;
s502, initializing a parameter set at random based on normal distribution:
Θiter={Witer,Hiter,biter}
s503, setting a circulation variable to be eta, wherein eta is 1;
s504, traversing an interaction relation set D of the user and the product under the iter iteration;
s505, in the eta cycle process, one user u is randomly selected, and meanwhile, a corresponding positive feedback product p is taken out from a positive feedback set, a social feedback set, a community feedback set and a negative feedback set corresponding to the user uiSocial feedback product psCommunity feedback product pgAnd a negative feedback product pj; thereby obtaining a group of user partial order product combinations of the eta traversal under the iter iteration
S506, combining the products in the user partial orderSubstituting the target function into the target function to obtain the target function of the eta cycle under the iter iteration
S507, updating the objective function based on the random gradient rise methodMiddle parameterAnda gradient of (a);
s508, assigning eta +1 to eta, judging whether eta > | D | is true, and if yes, executing the step S509; otherwise, returning to step S505;
s509, judging parametersWhether convergence is achieved, if convergence is achieved, an optimal parameter set is obtainedOtherwise, iter +1 is assigned to iter, and the procedure returns to step S503.
Preferably, the obtaining of the product recommendation result specifically includes:
and randomly selecting a user v in the product set U, and sequencing all non-interactive products of the user v in the product set P in a descending order according to the preference score to obtain a product recommendation result.
Preferably, the method further comprises the following steps:
evaluating the product recommendation, the evaluating comprising: precision @ K assessment, Recall @ K assessment, MRR @ K assessment, MAP @ K assessment.
The invention provides a product recommendation system based on social contact and community, which solves the technical problem, and comprises a computer, wherein the computer comprises:
at least one memory cell;
at least one processing unit;
wherein the at least one memory unit has stored therein at least one instruction that is loaded and executed by the at least one processing unit to perform the steps of:
acquiring interaction data, social relationship data and community relationship data of a user and a product, and forming historical data;
determining a feedback relationship between the user and the product based on the historical data and a preset partial order relationship;
determining a preference relationship of the user to the product based on the historical data and the feedback relationship;
determining an objective function of the preference relationship;
acquiring a preference score of a user for a product based on the target function;
and sequencing the non-interactive products of the user based on the preference scores to obtain a product recommendation result.
The present invention provides a computer readable storage medium, having at least one instruction stored thereon, at least the one instruction being loaded and executed by a processor to implement the method as described above
(III) advantageous effects
The invention provides a social and community-based product recommendation method, a system and a storage medium. Compared with the prior art, the method has the following beneficial effects:
according to the method, historical data is formed by acquiring interaction data, social relationship data and community relationship data of a user and a product; determining a feedback relationship between a user and a product based on historical data and a preset partial order relationship; determining a preference relationship of the user to the product based on the historical data and the feedback relationship; determining an objective function of the preference relationship; acquiring a preference score of a user for a product based on a target function; and sequencing the non-interactive products of the user based on the preference scores to obtain a product recommendation result. According to the method and the system, the interaction between the user and the product is analyzed by combining the social relationship and the community relationship of the user, and the preference of the user is divided in a finer granularity, so that the recommendation result is more accurate.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, it is obvious that the drawings in the following description are only some embodiments of the present invention, and for those skilled in the art, other drawings can be obtained according to the drawings without creative efforts.
Fig. 1 is an overall flowchart of a social and community-based product recommendation method according to an embodiment of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention are clearly and completely described, and it is obvious that the described embodiments are a part of the embodiments of the present invention, but not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
The embodiment of the application solves the problem that products cannot be accurately recommended in the prior art by providing a product recommendation method, a product recommendation system and a storage medium based on social contact and community, and realizes accurate recommendation of products.
In order to solve the technical problems, the general idea of the embodiment of the application is as follows:
according to the method and the device, historical data are formed by acquiring the interaction data, social relation data and community relation data of the user and the product; determining a feedback relationship between a user and a product based on historical data and a preset partial order relationship; determining a preference relationship of the user to the product based on the historical data and the feedback relationship; determining an objective function of the preference relationship; acquiring a preference score of a user for a product based on a target function; and sequencing the non-interactive products of the user based on the preference scores to obtain a product recommendation result. According to the embodiment of the invention, the interaction between the user and the product is analyzed by combining the social relationship and the community relationship of the user, and the preference of the user is divided in a finer granularity, so that the recommendation result is more accurate.
In order to better understand the technical solution, the technical solution will be described in detail with reference to the drawings and the specific embodiments.
An embodiment of the present invention provides a social and community-based product recommendation method, as shown in fig. 1, where the method is executed by a computer and includes the following steps:
s1, acquiring interaction data, social relationship data and community relationship data of the user and the product, and forming historical data;
s2, determining the feedback relationship between the user and the product based on the historical data and the preset partial order relationship;
s3, determining the preference relationship of the user to the product based on the historical data and the feedback relationship;
s4, determining an objective function of the preference relationship;
s5, obtaining the preference score of the user for the product based on the target function;
and S6, sequencing the non-interactive products of the user based on the preference scores to obtain a product recommendation result.
According to the method and the device, historical data are formed by acquiring the interaction data, social relation data and community relation data of the user and the product; determining a feedback relationship between a user and a product based on historical data and a preset partial order relationship; determining a preference relationship of the user to the product based on the historical data and the feedback relationship; determining an objective function of the preference relationship; acquiring a preference score of a user for a product based on a target function; and sequencing the non-interactive products of the user based on the preference scores to obtain a product recommendation result. According to the embodiment of the invention, the interaction between the user and the product is analyzed by combining the social relationship and the community relationship of the user, and the preference of the user is divided in a finer granularity, so that the recommendation result is more accurate.
The following steps are described in detail:
in step S1, interaction data, social relationship data, and community relationship data of the user and the product are acquired, and history data is formed.
Specifically, the embodiment of the invention can obtain the following data information based on an internet platform:
the system comprises a user set U, a product set P, a set D in which users and products have interaction, a social relationship set S and a community relationship set G.
Setting a binary group < u, pi"E D denotes user u and product piThere is over-interaction. Wherein U is E.U, piAnd e.g. P, the total users | U | ═ M, and the products | P | ═ N.
Binary < ui,uj"E S means user uiAnd user ujAre friends; binary < ui,uj"E G denotes user uiAnd user ujJoin the same community.
In step S2, a feedback relationship between the user and the product is determined based on the history data and a pre-set partial order relationship.
Specifically, the feedback relationship is represented by a positive feedback set, a social feedback set, a community feedback set, and a negative feedback set.
In practical application, the social relationship is a two-way connection between friends, and belongs to a strong connection. The community relation is added by the user according to the interest of the user and belongs to one-way connection. One user may join multiple communities, with multiple users in a community. And the users have a plurality of interests, so that the connection between every two users belonging to the same community belongs to weak connection. Therefore, the preset feedback set in the embodiment of the present invention has the following partial order relationship:
positive feedback > social feedback > community feedback > negative feedback.
Based on the above setting, the feedback relationship can be obtained as follows:
the positive feedback is integrated into PFu={<u,pi>}。
Wherein: < u, pi>∈D。
Social feedback set is SFu={<u,ps>}。
Wherein: < u, psIs a product that user u's social friends have interacted with, but user u itself has not interacted with.
Community feedback set is GFu={<u,pg>}。
Wherein: < u, pgIs a product that the social friends of user u have interacted, but user u itself has no interaction and the social friends of user u also have no interaction.
Negative feedback integration into NFu={<u,pj>}。
Wherein: < u, pjIs a product that user u itself has no interaction and that user u's social and social friends have no interaction.
The above set simultaneously satisfies the following conditions:
PFu∪SFu∪GFu∪NFu=P
that is, the four feedback sets include all of the product data, and no product data exists in any two feedback sets at the same time.
In step S3, a preference relationship of the user for the product is determined based on the history data and the feedback relationship.
Specifically, a preference relationship is represented based on a matrix decomposition model of the user set U to the product set P, and the preference relationship specifically includes:
w represents a feature matrix of the user set U;
h represents a feature matrix of the product set P;
b represents a deviation term of the product set P.
The dimension of W is M x K dimensions and the dimension of H is N x K dimensions.
Based on the feedback set corresponding to the preference relationship, the following results are obtained:
wherein:
In step S4, an objective function of the preference relationship is determined.
Specifically, the objective function is constructed based on a Bayesian personalized sorting method. And taking the matrix decomposition model as a measure of the partial order relation. And performing optimization solution by using a gradient ascent algorithm to obtain each parameter value in the matrix decomposition model.
The objective function is specifically:
wherein:
σ (-) represents a logistic function;
Θ denotes a set of parameters in the matrix decomposition model, i.e., Θ ═ { W, H, b };
λΘfor regularizingCounting;
CSrepresenting user u and product psThe number of friends who have interacted with the product in the friends of the user u is not interacted;
Cgrepresenting user u and product pgAnd no interaction exists, and the social friends of the user u have no interaction with the product, but the social friends of the user u have the number of the interactive social friends with the product.
Wherein, CSThe larger the value, the user u is represented to the product piWith product psThe closer the preference of (a); cgThe larger the value, the more the user is indicated to product psWith product pgThe closer the preference difference of (a) is.
The objective function is to train the preference score of the user on the product to meet the order bias relationship proposed by the invention.
In step S5, a preference score of the user for the product is obtained based on the objective function.
Specifically, the method comprises the following steps:
s501, setting the iteration number as iter, wherein iter is 1;
s502, initializing a parameter set at random based on normal distribution:
Θiter={Witer,Hiter,biter}
s503, setting a circulation variable to be eta, wherein eta is 1;
s504, traversing an interaction relation set D of the user and the product under iter iteration, and randomly sampling the partial sequence relation for | D | times;
s505, in the eta cycle process, randomly selecting a user u, and meanwhile, selecting a positive feedback set PF corresponding to the user uuRandomly selecting a positive feedback product piFrom the social feedback set SF corresponding to the user uuRandomly selecting a social feedback product psFrom the community feedback set GF corresponding to the user uuRandomly selecting a community feedback product pgNF in negative feedback set corresponding to the user uuRandomly selecting a negative feedback product pj(ii) a Thereby obtainingObtaining a group of user partial order product combinations of the eta traversal under the iter iteration
S506, combining the products in the user partial orderSubstituting the target function into the target function to obtain the target function of the eta cycle under the iter iteration
S507, updating the objective function based on the random gradient rise methodMiddle parameterAnda gradient of (a);
s508, assigning eta +1 to eta, judging whether eta > | D | is true, and if yes, executing the step S509; otherwise, returning to step S505;
s509, judging parametersWhether convergence is achieved, if convergence is achieved, an optimal parameter set is obtainedOtherwise, iter +1 is assigned to iter, and the procedure returns to step S503.
In step S6, the non-interactive products of the user are sorted based on the preference scores, and a product recommendation result is obtained.
Specifically, one user v in the product set U is randomly selected, all non-interactive products of the user v in the product set P are sorted in a descending order according to the preference score, and the method can be preset to extract the first n products to form a product recommendation list, so as to obtain a product recommendation result and recommend the product recommendation result to the user.
According to the method and the device, on the basis of interaction between the user and the product, the social information and the community information are considered, the preference sequence of the user to the non-interacted product is constructed, and therefore the user preference is divided in a finer granularity mode.
The embodiment of the invention also comprises the following steps:
and S7, evaluating the product recommendation result.
Specifically, the embodiment of the invention evaluates the recommendation result by using Precision @ K and Recall @ K, MRR @ K, MAP @ K.
First, the data set is divided into a training set and a test set. The invention uses a training set to train a model to obtain the preference score x of a user on a product. And obtaining a recommendation list of each user according to the preference scores, and comparing and evaluating the recommendation list obtained by the training set and the test set to obtain the quality of the recommendation result.
Wherein precision @ K represents the proportion of items that the user likes in the recommendation list. The recommendation accuracy for a single user u is:
wherein:
hit represents how many of the user u's recommendation list are actually purchased by the user u;
and K is the length of the recommended result.
The overall accuracy is:
i.e. the prediction accuracy of the population sample is the average of the individual accuracies.
Recall @ K indicates how many items the user likes in the test set appear in the recommendation list.
The recommendation recall rate for a single user u is:
wherein:
hit represents how many of the user u's recommendation list are actually purchased by the user u.
The overall recall ratio was:
i.e. the recall of the population sample is the average of the individual recall.
MRR @ K is specifically:
wherein:
rankuindicating the location of the first predicted accurate product in the recommendation list of user u.
MAP @ K is specifically:
wherein:
hit represents how many of the recommendation list of user u hit the product that user u actually purchased.
An embodiment of the present invention further provides a product recommendation system based on social contact and community, where the system includes a computer, and the computer includes:
at least one memory cell;
at least one processing unit;
wherein the at least one memory unit has stored therein at least one instruction that is loaded and executed by the at least one processing unit to perform the steps of:
acquiring interaction data, social relationship data and community relationship data of a user and a product, and forming historical data;
determining a feedback relationship between the user and the product based on the historical data and a preset partial order relationship;
determining a preference relationship of the user to the product based on the historical data and the feedback relationship;
determining an objective function of the preference relationship;
acquiring a preference score of a user for a product based on the target function;
and sequencing the non-interactive products of the user based on the preference scores to obtain a product recommendation result.
It can be understood that the recommendation system provided by the embodiment of the present invention corresponds to the recommendation method, and the explanation, examples, and beneficial effects of the relevant contents may refer to the corresponding contents in the product recommendation method based on social contact and community, which are not described herein again.
An embodiment of the present invention further provides a computer-readable storage medium, where at least one instruction is stored in the storage medium, and the at least one instruction is loaded and executed by a processor to implement the method.
In summary, compared with the prior art, the method has the following beneficial effects:
according to the method and the device, historical data are formed by acquiring the interaction data, social relation data and community relation data of the user and the product; determining a feedback relationship between a user and a product based on historical data and a preset partial order relationship; determining a preference relationship of the user to the product based on the historical data and the feedback relationship; determining an objective function of the preference relationship; acquiring a preference score of a user for a product based on a target function; and sequencing the non-interactive products of the user based on the preference scores to obtain a product recommendation result. According to the embodiment of the invention, the interaction between the user and the product is analyzed by combining the social relationship and the community relationship of the user, and the preference of the user is divided in a finer granularity, so that the recommendation result is more accurate.
It should be noted that, through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by software plus a necessary general hardware platform. With this understanding in mind, the above-described technical solutions may be embodied in the form of a software product, which can be stored in a computer-readable storage medium such as ROM/RAM, magnetic disk, optical disk, etc., and includes instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute the methods described in the embodiments or some parts of the embodiments.
In the description provided herein, numerous specific details are set forth. It is understood, however, that embodiments of the invention may be practiced without these specific details. In some instances, well-known methods, structures and techniques have not been shown in detail in order not to obscure an understanding of this description.
In this document, relational terms such as first and second, and the like may be used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions. Also, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising an … …" does not exclude the presence of other identical elements in a process, method, article, or apparatus that comprises the element.
The above examples are only intended to illustrate the technical solution of the present invention, but not to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some technical features may be equivalently replaced; and such modifications or substitutions do not depart from the spirit and scope of the corresponding technical solutions of the embodiments of the present invention.
Claims (8)
1. A social and community based product recommendation method, the recommendation method being executed by a computer and comprising the steps of:
acquiring interaction data, social relationship data and community relationship data of a user and a product, and forming historical data;
determining a feedback relationship between the user and the product based on the historical data and a preset partial order relationship;
determining a preference relationship of the user to the product based on the historical data and the feedback relationship;
determining an objective function of the preference relationship;
acquiring a preference score of a user for a product based on the target function;
ranking the non-interactive products of the user based on the preference scores to obtain a product recommendation result;
the preference relation of the user to the product is represented by a matrix decomposition model based on the user set U to the product set P:
wherein:
w represents a feature matrix of the user set U;
h represents a feature matrix of the product set P;
b represents a deviation item of the product set P;
the objective function of the preference relationship is:
wherein:
σ (-) represents a logistic function;
Θ denotes a set of parameters in the matrix decomposition model, i.e., Θ ═ { W, H, b };
λΘis a regularization parameter;
CSrepresenting user u and product psWithout interaction, but of user uThe number of friends who have interacted with the product in the friends;
Cgrepresenting user u and product pgAnd no interaction exists, and the social friends of the user u have no interaction with the product, but the social friends of the user u have the number of the interactive social friends with the product.
2. The recommendation method of claim 1, wherein the historical data comprises:
the system comprises a user set U, a product set P, a set D in which users and products have interaction, a social relationship set S and a community relationship set G.
3. The recommendation method according to claim 2, wherein the preset partial order relationship is:
positive feedback > social feedback > community feedback > negative feedback;
the feedback relationship between the user and the product comprises:
and (3) positive feedback assembly: PF (particle Filter)u={<u,pi>},
Wherein: < u, piIs e.g. D, represents user u and product piInteraction exists;
social feedback set: SFu={<u,ps>},
Wherein: < u, psThe social friends of the user u have interaction, but the user u does not have interactive products;
community feedback set: GFu={<u,pg>},
Wherein: < u, pgProducts that show that social friends of the user u have interaction, but the user u does not interact with the social friends of the user u and does not interact with the social friends of the user u;
and (3) negative feedback integration: NFu={<u,pj>},
Wherein: < u, pjThe expression > represents that the user u does not interact with the social friends and the social group friends of the user u, and the social group friends of the user u also do not have interactive products.
4. The recommendation method according to claim 1, wherein the obtaining of the preference score of the user for the product based on the objective function specifically comprises:
s501, setting the iteration number as iter, wherein iter is 1;
s502, initializing a parameter set at random based on normal distribution:
Θiter={Witer,Hiter,biter}
s503, setting a circulation variable to be eta, wherein eta is 1;
s504, traversing an interaction relation set D of the user and the product under the iter iteration;
s505, in the eta cycle process, one user u is randomly selected, and meanwhile, a corresponding positive feedback product p is taken out from a positive feedback set, a social feedback set, a community feedback set and a negative feedback set corresponding to the user uiSocial feedback product psCommunity feedback product pgAnd negative feedback product pj(ii) a Thereby obtaining a group of user partial order product combinations of the eta traversal under the iter iteration
S506, combining the products in the user partial orderSubstituting the target function into the target function to obtain the target function of the eta cycle under the iter iteration
S507, updating the objective function based on the random gradient rise methodMiddle parameterAnda gradient of (a);
s508, assigning eta +1 to eta, judging whether eta > | D | is true, and if yes, executing the step S509; otherwise, returning to step S505;
5. The recommendation method of claim 2, wherein the obtaining of the product recommendation result specifically comprises:
and randomly selecting a user v in the product set U, and sequencing all non-interactive products of the user v in the product set P in a descending order according to the preference score to obtain a product recommendation result.
6. The recommendation method of claim 1, further comprising:
evaluating the product recommendation, the evaluating comprising: precision @ K assessment, Recall @ K assessment, MRR @ K assessment, MAP @ K assessment.
7. A social and community based product recommendation system, the recommendation system comprising a computer, the computer comprising:
at least one memory cell;
at least one processing unit;
wherein the at least one memory unit has stored therein at least one instruction that is loaded and executed by the at least one processing unit to perform the steps of:
acquiring interaction data, social relationship data and community relationship data of a user and a product, and forming historical data;
determining a feedback relationship between the user and the product based on the historical data and a preset partial order relationship;
determining a preference relationship of the user to the product based on the historical data and the feedback relationship;
determining an objective function of the preference relationship;
acquiring a preference score of a user for a product based on the target function;
ranking the non-interactive products of the user based on the preference scores to obtain a product recommendation result;
the preference relation of the user to the product is represented by a matrix decomposition model based on the user set U to the product set P:
wherein:
w represents a feature matrix of the user set U;
h represents a feature matrix of the product set P;
b represents a deviation item of the product set P;
the objective function of the preference relationship is:
wherein:
σ (-) represents a logistic function;
Θ denotes a set of parameters in the matrix decomposition model, i.e., Θ ═ { W, H, b };
λΘis a regularization parameter;
CSrepresenting user u and product psThe number of friends who have interacted with the product in the friends of the user u is not interacted;
Cgrepresenting user u and product pgAnd no interaction exists, and the social friends of the user u have no interaction with the product, but the social friends of the user u have the number of the interactive social friends with the product.
8. A computer readable storage medium having stored thereon at least one instruction, at least the one instruction being loaded and executed by a processor to implement the method of claim 1.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910496478.0A CN110209946B (en) | 2019-06-10 | 2019-06-10 | Social and community-based product recommendation method, system and storage medium |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910496478.0A CN110209946B (en) | 2019-06-10 | 2019-06-10 | Social and community-based product recommendation method, system and storage medium |
Publications (2)
Publication Number | Publication Date |
---|---|
CN110209946A CN110209946A (en) | 2019-09-06 |
CN110209946B true CN110209946B (en) | 2021-03-09 |
Family
ID=67791600
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201910496478.0A Active CN110209946B (en) | 2019-06-10 | 2019-06-10 | Social and community-based product recommendation method, system and storage medium |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN110209946B (en) |
Families Citing this family (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110837598B (en) * | 2019-11-11 | 2021-03-19 | 腾讯科技(深圳)有限公司 | Information recommendation method, device, equipment and storage medium |
CN112231579B (en) * | 2019-12-30 | 2022-10-28 | 北京邮电大学 | Social video recommendation system and method based on implicit community discovery |
CN111667298B (en) * | 2020-04-15 | 2022-09-30 | 合肥工业大学 | User personalized demand prediction method and system |
CN112149003B (en) * | 2020-10-28 | 2023-06-20 | 浙江集享电子商务有限公司 | Commodity community recommendation method and device and computer equipment |
CN116594712B (en) * | 2023-05-29 | 2024-02-23 | 上海炫稷网络科技有限公司 | User center display method and system based on data analysis |
Family Cites Families (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20120123837A1 (en) * | 2010-11-17 | 2012-05-17 | Wellstar Gmbh & Co Kg | Social network shopping system and method |
WO2014070293A1 (en) * | 2012-11-05 | 2014-05-08 | Nara Logics, Inc. | Systems and methods for providing enhanced neural network genesis and recommendations to one or more users |
CN106777123B (en) * | 2016-12-16 | 2019-07-19 | 合肥工业大学 | A kind of group's recommended method based on two-way tensor resolution model |
CN107391637B (en) * | 2017-07-10 | 2021-01-26 | 江苏省现代企业信息化应用支撑软件工程技术研发中心 | Group recommendation method aiming at geographic social information |
CN107833117B (en) * | 2017-12-13 | 2020-03-17 | 合肥工业大学 | Bayesian personalized sorting recommendation method considering tag information |
CN108763314B (en) * | 2018-04-26 | 2021-01-19 | 深圳市腾讯计算机系统有限公司 | Interest recommendation method, device, server and storage medium |
-
2019
- 2019-06-10 CN CN201910496478.0A patent/CN110209946B/en active Active
Also Published As
Publication number | Publication date |
---|---|
CN110209946A (en) | 2019-09-06 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN110209946B (en) | Social and community-based product recommendation method, system and storage medium | |
CN111797321B (en) | Personalized knowledge recommendation method and system for different scenes | |
CN107451894B (en) | Data processing method, device and computer readable storage medium | |
CN108920503A (en) | A kind of micro- video personalized recommendation algorithm based on social networks degree of belief | |
CN104199818B (en) | Method is recommended in a kind of socialization based on classification | |
Zhang et al. | Prediction uncertainty in collaborative filtering: Enhancing personalized online product ranking | |
CN109840833B (en) | Bayesian collaborative filtering recommendation method | |
CN113379494B (en) | Commodity recommendation method and device based on heterogeneous social relationship and electronic equipment | |
WO2020220757A1 (en) | Method and device for pushing object to user based on reinforcement learning model | |
CN112115358A (en) | Personalized recommendation method using multi-hop path features in knowledge graph | |
CN110727872A (en) | Method and device for mining ambiguous selection behavior based on implicit feedback | |
CN113420221A (en) | Interpretable recommendation method integrating implicit article preference and explicit feature preference of user | |
CN109190040B (en) | Collaborative evolution-based personalized recommendation method and device | |
CN110059257B (en) | Project recommendation method based on score correction | |
CN105260458A (en) | Video recommendation method for display apparatus and display apparatus | |
CN108491477B (en) | Neural network recommendation method based on multi-dimensional cloud and user dynamic interest | |
CN108763515B (en) | Time-sensitive personalized recommendation method based on probability matrix decomposition | |
CN117216376A (en) | Fair perception recommendation system and recommendation method based on depth map neural network | |
KR20200142871A (en) | Method and apparatus for recommending items using explicit and implicit feedback | |
CN114418701A (en) | Method and device for generating recommendation list, electronic equipment and storage medium | |
CN110956528B (en) | Recommendation method and system for e-commerce platform | |
CN110717103B (en) | Improved collaborative filtering method based on stack noise reduction encoder | |
CN110825967A (en) | Recommendation list re-ranking method for improving diversity of recommendation system | |
Sridhar et al. | Extending Deep Neural Categorisation Models for Recommendations by Applying Gradient Based Learning | |
KR20200122652A (en) | Nutrient Profiling-based Pet Food Recommendation System |
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 |