CN111324807A

CN111324807A - Collaborative filtering recommendation method based on trust degree

Info

Publication number: CN111324807A
Application number: CN202010030767.4A
Authority: CN
Inventors: 王凯旋
Original assignee: Beijing University of Technology
Current assignee: Beijing University of Technology
Priority date: 2020-01-13
Filing date: 2020-01-13
Publication date: 2020-06-23

Abstract

The invention discloses a collaborative filtering recommendation method based on credibility, which is expected to recommend items to the most mental apparatus of users in the large environment of Internet information blowout, thereby saving the precious time of the users and improving the information acquisition efficiency. The algorithm is characterized in that: on the basis of a traditional collaborative filtering algorithm, attribute features (including age factors, occupational factors and gender factors), interestingness (a certain category of items preferred by a user), trust (the satisfaction degree of the user on recommendation of a certain part of users) and other dimensions are fused to model the preference of the user, so that the cold start and sparsity problems of a user scoring matrix are relieved as much as possible, and the most valuable, most desirable and most surprised recommendation is made for the user.

Description

Collaborative filtering recommendation method based on trust degree

Technical Field

The invention relates to computer recommendation, in particular to a computer system applying a recommendation system, collaborative filtering, similarity modeling and trust modeling to all scenes needing recommendation.

Background

The 21 st century is an information explosion era, with the development of computer technology, massive information poses a huge challenge to information search, how to find needed contents in huge information data oceans is achieved, search is more efficient, and a user can have better internet experience

Based on the practical market demands, the recommendation system comes along, which is a method for finding new demands according to the previous behavior information of the user, the personalized recommendation algorithm widely used in each e-commerce at present belongs to a collaborative filtering recommendation algorithm, but the scoring matrix of the user is a sparse matrix, so that: the cold start problem, data sparsity and recommendation accuracy of new users still need continuous research improvement.

Disclosure of Invention

The invention aims to optimize the defects of the existing recommendation algorithm and provides methods for solving the problems of data sparsity and cold start of users and improving the recommendation accuracy.

The contribution points of the invention to similarity are as follows:

attribute features are combined. The attribute features related to the invention comprise age, occupation and gender, and the features identify the living state of the user. The recommendation accuracy can be improved by analyzing the similarity of the attribute features, which also influences the preference of the user.

Combining the interestingness. The interestingness is the core recommendation basis of the recommendation system, and the recommendation accuracy can be greatly improved by a good interestingness model.

In combination with the degree of trust. The trust degree is divided into direct trust degree and indirect trust degree, the direct trust degree can improve the recommendation accuracy, the indirect trust degree can fill a scoring matrix, and the accuracy is improved while the data sparsity is improved.

Table 1 user-rating matrix R (m, n) of m x n based on m users n items

	I₁	I₂	…	I_n
					U₁	R₁₁	R₁₂	…	R_1n
U₂	R₂₁	R₂₂	…	R_2n
					…	…	…	…	…
U_m	R_m1	R_m2	…	R_mn

The technical scheme adopted by the invention is a collaborative filtering recommendation method based on the trust degree, and the realization process of the method is as follows:

1) user attribute feature similarity modeling

The user attribute of the user attribute feature similarity model integrates the influence of gender, age and occupation on the interest of the user, and the establishing process is as follows:

a) gender similarity modeling

The interests of different sexes may vary widely, for example, the interests of women who like cosmetics, clothing, jewelryObviously higher than men who are generally more prone to electronic products. Suppose user u has a gender S_uGender of user v is S_vThen, the gender similarity S (u, v) of the user u and the user v is as follows:

b) age similarity modeling

The different ages will be different, so the age model is introduced to optimize the similarity model, let the age of user u be A_uAge of user v is A_vThe age similarity a (u, v) of user u and user v is as follows:

c) occupational similarity modeling

Classifying careers into a tree structure according to categories as shown in the following figure 2, wherein the length of any two nodes in the figure 2 is set as 1, the total length of the parent node nearest to careers a and b is represented by Height, the number of layers in the career tree is called as the Height of the parent node, and H is taken as_a,bThe professions of user u and user v are denoted by profession a and profession b, respectively, and the profession similarity O (u, v) is expressed as follows:

in conclusion, the attribute similarity sim of the user is obtained by comprehensively considering the gender, the age and the occupational factors_a(u, v) are as follows:

sim_a(u,v)＝αS(u,v)+βA(u,v)+γO(u,v) (4)

α, adjusting the parameters of gamma according to the specific system;

2) modeling the similarity of user interest characteristics;

the attribute characteristics discussed in the upper part provide a recommendation scheme under the condition of cold start, but the user specific interest degree can be modeled after the comment footprints of the user are rich, and the interest is selectedThe influence of selection is extremely important, and the interest level discussed by the method is the interest level of the user in a certain type of items, namely the evaluation proportion of the user to the certain type of items is used as the interest level measurement of the user in the certain type of items. Provided with NI_u,iIndicates the total number of evaluations, NI, of user u on the i-class item_uRepresenting the total number of the items evaluated by the user, the interest degree I of the user u in the I-type items_u,iThe following were used:

obtaining the interest degree set I of the user u by using the formula (5)_u＝(I_u,1,I_u,2,...I_u,i) Is shown in the formula I_uvThe number of categories of items that represent user u and user v to score together,

the method is used for identifying the average interest degree of the user u to all the item types, and the interest degree sim between the two users is obtained through calculation according to the Pearson correlation similarity_I(u, v) are as follows:

3) similarity calculation:

the calculation of user similarity relies on a user-item scoring table, with each user interest represented by a scoring vector to map to each row in table 1, so the calculation of user similarity essentially calculates the distance between the user scoring vectors. The calculation methods of the user similarity are numerous, and the constrained Pearson correlation similarity and the jaccard similarity are selected for correlation calculation.

And (3) constraining the Pearson correlation similarity: r_uvRepresenting a common set of scores for user u and user v, R_u,iIdentifying a rating, R, for item i by user u_v,iThe rating of item i by user v is identified,

representing all ratings of user uThe average value of the average value is calculated,

identifying the mean of all scores of user v, the score similarity of user u and user v is expressed in terms of pearson's relative similarity as follows:

jaccard similarity: the cosine similarity and the constrained pearson correlation similarity measure the similarity based on a common scoring item of the users, and under the condition of sparse data, the two similarity calculation methods have a common problem, namely when the data is sparse until the user u and the user v only have one common scoring item, and both people score the same low score for the item, at the moment, the user u and the user v are not better than the item, but the cosine similarity and the constrained pearson correlation similarity consider that the similarity of the two is the maximum value 1, so that the preference of the two users is preferentially recommended to each other. In order to alleviate the situation, the similarity of the Jarcard is introduced, and the similarity calculation method is measured based on the number of common scores among users, so that the problem can be effectively avoided. The calculation formula is as follows:

in the formula | Iu_,vI represents the total number of commonly scored items for user u and user v, I_u|，|I_vI represents the number of items, Jac, evaluated by user u and user v, respectively_(u,v)Has a value range of [0, 1 ]]A larger value indicates a higher correlation between the two.

4) Modeling the user trust:

according to the method, trust factors in social contact are fused into a collaborative filtering recommendation algorithm, and trust is defined as follows: the target user has certain safety, integrity and effectiveness on the recommendation opinions proposed by other users. If the user b accepts the recommendation, the recommendation is not accepted as distrust, and the trust degree of the user b on the user a is the ratio of the accepted number of strokes to n.

4.1) direct confidence calculation

The direct trust degree is that a user actively and directly gives trust scores of other users, but the scheme requires that each user gives trust scores of other users, and the actual operation is difficult, so the method adopts direct trust degree calculation based on a common user score project, and comprises the following steps:

DTrust_u,v＝|sim(u,v)|×Jac_(u,v)(9)

4.2) Indirect confidence computation

The trust relationship between users has the characteristics of subjectivity, weak transitivity, asymmetry and dynamicity. Because the trust degree has transitivity, the problem of sparsity of a scoring matrix is solved, the direct trust degree between two users depends on whether the two users have a common scoring item, and the direct trust degree exists between the users with common scoring, but in life, the direct trust degree often finds that if the user u trusts the user i, and the user i trusts the user v, then the trust relationship of the user u to the user v often exists, which is the transitivity of trust, but in view of the fast attenuation of the trust degree, only the second-order trust degree is taken for improving the accuracy, and the calculation formula is as follows:

wherein, PTrust_u,vRepresents the indirect trust between the user u and the user v, and adj (u, v) represents the satisfaction of the condition DTrust_u,b≥λ,DTrust_b,vAnd the lambda parameter is an adjustable credit threshold value, which indicates that only the intermediate users with direct credit larger than lambda can be included in the indirect trust calculation.

The Trust level between user u and user v is denoted Trust_u,vThen:

5) in summary, the following steps:

integrating user attributes and user interest degrees, score similarity and trust degree_z(u, v) is expressed by the following formula: the delta, epsilon, mu and rho can be dynamically adjusted according to different dependence degrees of a specific system on various factors

sim_z(u,v)＝δsim(u,v)+εsim_I(u,v)+μsim_a(u,v)+ρTrust_u,v(12)

6) Generating a recommendation result:

according to the comprehensive similarity sim of the user u and the user v_z(u,v)，

Let user v score item i as R_v,iThe average of the scores of the user v for all the items is

P_u,iDenotes the rating, N, of item i by user u based on the rating inference of v_uThe top-N sets with the highest scores in all the prediction scores are the final recommended item combination

7) Evaluation criteria for the method:

the recommendation quality of the algorithm is evaluated by taking a standard mean error (MAE) as a rating standard, the MAE is a generally accepted evaluation index in the research of the recommendation algorithm, the average absolute deviation of the predicted score of the item by a computing system and the actual score of the item by a user is calculated, the smaller the MAE value is, the higher the recommendation precision of the recommendation system is, and the calculation formula is as follows:

wherein: r_u,iRepresents the actual rating, P, of user u for item i_u,iAnd (4) the predicted scores of the user u on the item i are obtained according to the method.

Drawings

FIG. 1 is a flow chart of the method implementation.

FIG. 2 is a career tree diagram.

Detailed Description

The present invention will be described in detail below with reference to the accompanying drawings and examples.

The collaborative filtering recommendation method based on the trust degree is characterized by comprising the following steps: on the basis of a traditional collaborative filtering algorithm, multiple dimensions such as attribute characteristics, interestingness, trust and the like are fused to model the hobbies of the user, the problems of cold start and sparsity of a user scoring matrix are relieved as much as possible, and the most valuable, most desirable and most surprised recommendation is made for the user.

Based on the traditional collaborative filtering recommendation algorithm, the attribute similarity of the user is fused, and the attributes of age, occupation and gender which possibly influence the user choice are fused into the recommendation process of the algorithm

On the basis of the traditional collaborative filtering algorithm, the similarity of the user attribute characteristics is combined, the preference of a user for a certain type of items is fully considered into the algorithm,

on the basis of the traditional collaborative filtering algorithm, the modeling of the user trust adopts the calculation of two dimensions of the direct trust and the indirect trust, so that the problem of the sparsity of a user scoring matrix is solved

And finally, the calculation mode of the similarity can dynamically adjust the proportion parameters of different influence factors according to different scenes, so as to obtain the project which best meets the requirements of the user to the maximum extent.

Claims

1. The collaborative filtering recommendation method based on the trust degree is characterized by comprising the following steps: the method comprises the following implementation processes:

1) user attribute feature similarity modeling

a) gender similarity modeling

Suppose user u has a gender S_uGender of user v is S_vThen, the gender similarity S (u, v) of user u and user v) As follows:

b) age similarity modeling

Let the age of user u be A_uAge of user v is A_vThe age similarity a (u, v) of user u and user v is as follows:

c) occupational similarity modeling

Classifying careers into a tree structure according to categories, setting the length of any two nodes to be 1, expressing the total length of the layer number of the nearest parent node of career a and career b in the career tree by Height to be called the Height thereof, and recording H_a,bThe professions of user u and user v are denoted by profession a and profession b, respectively, and the profession similarity O (u, v) is expressed as follows:

sim_a(u,v)＝αS(u,v)+βA(u,v)+γO(u,v) (4)

α, adjusting the parameters of gamma according to the specific system;

2) modeling the similarity of user interest characteristics;

provided with NI_u,iIndicates the total number of evaluations, NI, of user u on the i-class item_uRepresenting the total number of the items evaluated by the user, the interest degree I of the user u in the I-type items_u,iThe following were used:

3) similarity calculation:

the calculation of the user similarity relies on the user-item score table, each user interest is represented by a score vector to map with each row in table 1, so the calculation of the user similarity essentially calculates the distance between the user score vectors; the method for calculating the user similarity is numerous, and restricted Pearson correlation similarity and jaccard similarity are selected for correlation calculation;

represents the mean of all the scores of the user u,

jaccard similarity: the calculation formula is as follows:

in the formula I_u,vI represents the total number of commonly scored items for user u and user v, I_u|，|I_vI represents the number of items, Jac, evaluated by user u and user v, respectively_(u,v)Has a value range of [0, 1 ]]A larger value indicates a higher correlation between the two;

4) modeling the user trust:

fusing trust factors in social contact into a collaborative filtering recommendation algorithm, wherein if the user a receives a recommendation, the recommendation is not accepted as untrustworthy, and the trust degree of the user b on the user a is the ratio of the accepted number of strokes to n, and all recommended items n made by the user a on the user b are recommended;

4.1) direct confidence calculation

DTrust_u,v＝|sim(u,v)|×Jac_(u,v)(9)

4.2) Indirect confidence computation

The direct trust between two users depends on whether the two users have a common scoring item, the direct trust exists between the users with common scoring, if the user u trusts the user i, and the user i trusts the user v, the user u often has a trust relationship to the user v, which is the transitivity of trust, and only the second-order trust is taken, and the calculation formula is as follows:

DTrust_u,b≥λ,DTrust_b,v≥λ (10)

wherein, PTrust_u,vRepresents the indirect trust between the user u and the user v, and adj (u, v) represents the satisfaction of the condition DTrust_u,b≥λ,DTrust_b,vIn ≧ λThe lambda parameter is an adjustable credit threshold value, which indicates that only the intermediate users with direct credit larger than lambda can be brought into indirect trust calculation;

the Trust level between user u and user v is denoted Trust_u,vThen:

5) in summary, the following steps:

sim_z(u,v)＝δsim(u,v)+εsim_I(u,v)+μsim_a(u,v)+ρTrust_u,v(12)

6) Generating a recommendation result:

according to the comprehensive similarity sim of the user u and the user v_z(u, v) let user v score item i be R_v,iThe average of the scores of the user v for all the items is

7) Evaluation criteria for the method:

the recommendation quality of the algorithm is evaluated by taking the standard average error MAE as a rating standard, the average absolute deviation of the prediction score of the computing system to the item and the actual score of the user to the item is calculated, the smaller the MAE value is, the higher the recommendation precision of the recommendation system is, and the calculation formula is as follows: