CN108920665B - Recommendation scoring method and device based on network structure and comment text - Google Patents

Recommendation scoring method and device based on network structure and comment text Download PDF

Info

Publication number
CN108920665B
CN108920665B CN201810729637.2A CN201810729637A CN108920665B CN 108920665 B CN108920665 B CN 108920665B CN 201810729637 A CN201810729637 A CN 201810729637A CN 108920665 B CN108920665 B CN 108920665B
Authority
CN
China
Prior art keywords
feature matrix
sample
class
target
commodity
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
Application number
CN201810729637.2A
Other languages
Chinese (zh)
Other versions
CN108920665A (en
Inventor
石川
韩霄天
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.)
Beijing University of Posts and Telecommunications
Original Assignee
Beijing University of Posts and Telecommunications
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 Beijing University of Posts and Telecommunications filed Critical Beijing University of Posts and Telecommunications
Priority to CN201810729637.2A priority Critical patent/CN108920665B/en
Publication of CN108920665A publication Critical patent/CN108920665A/en
Application granted granted Critical
Publication of CN108920665B publication Critical patent/CN108920665B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q30/00—Commerce
    • G06Q30/06—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
    • G06Q30/00—Commerce
    • G06Q30/06—Buying, selling or leasing transactions
    • G06Q30/0601—Electronic shopping [e-shopping]
    • G06Q30/0623—Item investigation
    • G06Q30/0625—Directed, with specific intent or strategy
    • G06Q30/0627—Directed, with specific intent or strategy using item specifications

Landscapes

  • Business, Economics & Management (AREA)
  • Accounting & Taxation (AREA)
  • Finance (AREA)
  • Development Economics (AREA)
  • Economics (AREA)
  • Marketing (AREA)
  • Strategic Management (AREA)
  • Physics & Mathematics (AREA)
  • General Business, Economics & Management (AREA)
  • General Physics & Mathematics (AREA)
  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

The embodiment of the invention provides a recommendation scoring method and a recommendation scoring device based on a network structure and a comment text, wherein the method comprises the following steps: determining a target user in the plurality of sample users and determining a target commodity in the plurality of sample commodities; acquiring a first class feature matrix for a target user and a second class feature matrix for a target commodity; acquiring a third type feature matrix for a target user and a fourth type feature matrix for a target commodity; and inputting the first class characteristic matrix and the third class characteristic matrix aiming at the target user and the second class characteristic matrix and the fourth class characteristic matrix aiming at the target commodity into a recommendation network model to obtain the prediction score value of the target user on the target commodity. Therefore, interactive information between the user and the commodity is fully considered, the interactive information comprises comment text information and grading information, and the purchase expectation of the user to the commodity can be more accurately predicted.

Description

Recommendation scoring method and device based on network structure and comment text
Technical Field
The invention relates to the technical field of machine learning, in particular to a recommendation scoring method and device based on a network structure and comment texts.
Background
With the development of electronic commerce, internet companies can provide a large number of commodities for users to select, and users are difficult to select from the large number of commodities. Currently, recommendation systems are mainly used to help users make selections. The recommendation system is obtained by training a neural network model by using the purchase information of the user to the commodity by using a deep learning method.
When a recommending system is used for recommending commodities for a user, for each candidate commodity, the purchasing information of the user on different commodities and the purchased information of the candidate commodity are input into the recommending system, and then the score value of the user on the candidate commodity is obtained, wherein the score value represents the purchasing expectation of the user on the commodity. And the recommending system recommends the candidate commodity with the higher score value to the user.
From the above, when recommending commodities to users, the method plays an important role in scoring the candidate commodities.
However, the current recommendation system is only trained by the user to purchase the merchandise. The purchase information of the user to the commodity cannot completely represent the interaction information of the user and the commodity, the neural network model is trained only by using the purchase information of the user to the commodity, and the obtained recommendation system cannot accurately score the candidate commodity, so that the user cannot be effectively helped to select the commodity.
Disclosure of Invention
The embodiment of the invention aims to provide a recommendation scoring method and device based on a network structure and a comment text, so as to improve the accuracy of predicting the scoring of a user on a commodity. The specific technical scheme is as follows:
in order to improve the accuracy of predicting the scores of the commodities by the users, the embodiment of the invention provides a recommendation scoring method based on a network structure and a comment text, which comprises the following steps:
determining a target user in the plurality of sample users and determining a target commodity in the plurality of sample commodities;
acquiring a first class feature matrix for the target user and a second class feature matrix for the target commodity; the first type feature matrix and the second type feature matrix are determined according to comment text information of the sample users on the sample commodities;
acquiring a third type feature matrix for the target user and a fourth type feature matrix for the target commodity; the third type feature matrix and the fourth type feature matrix are determined according to the purchasing network structure information of the plurality of sample users on the plurality of sample commodities;
inputting the first class feature matrix and the third class feature matrix for the target user and the second class feature matrix and the fourth class feature matrix for the target commodity into a recommendation network model to obtain a prediction score value of the target user on the target commodity;
the recommended network model is a model obtained by training according to a training set, wherein the training set comprises: the real scoring values of the plurality of sample users on the plurality of sample commodities are determined according to the first class feature matrix for each sample user, the second class feature matrix for each sample commodity, the third class feature matrix for each sample user, the fourth class feature matrix for each sample commodity and the real scoring values of the plurality of sample users on the plurality of sample commodities.
Optionally, the first class feature matrix for the target user and the second class feature matrix for the target product are determined by the following steps:
acquiring a plurality of comment text messages of the plurality of sample users on the plurality of sample commodities;
performing word segmentation processing on the plurality of comment text messages to obtain a word vector of each comment text message;
determining a first class feature matrix aiming at the target user according to the word vectors of the comment text information of the target user;
and determining a second class characteristic matrix aiming at the target commodity according to the word vectors of the plurality of comment text messages of the target commodity.
Optionally, the purchasing network structure information is network structure information of a heterogeneous information network based on purchasing information of the plurality of sample users on the plurality of sample commodities, a node of the heterogeneous information network includes the plurality of sample users and the plurality of sample commodities, an edge of the heterogeneous information network is used for connecting the sample users and the sample commodities, and an edge of the heterogeneous information network is used for indicating that the sample users have real scores on the sample commodities;
the third type feature matrix for the target user and the fourth type feature matrix for the target commodity are determined by the following steps:
based on the connection relation between the sample user and the sample commodity in the heterogeneous information network, randomly walking by taking the target user as an initial node to determine a plurality of first random walking sequences, randomly walking by taking the target commodity as the initial node to determine a plurality of second random walking sequences; the first random walk sequence comprises a first preset number of nodes; the second random walk sequence comprises a second preset number of nodes;
determining a third type of feature matrix aiming at the target user according to the plurality of first random walk sequences;
and determining a fourth type feature matrix aiming at the target commodity according to the plurality of second random walk sequences.
Optionally, the recommended network model includes: a first neural network model, a second neural network model, a third neural network model, and a fourth neural network model;
the step of inputting the first class feature matrix and the third class feature matrix for the target user and the second class feature matrix and the fourth class feature matrix for the target commodity into a recommendation network model to obtain the predicted score value of the target user on the target commodity comprises:
inputting a first class feature matrix aiming at the target user into the first neural network model to obtain a first feature vector;
inputting a second type of feature matrix aiming at the target commodity into the second neural network model to obtain a second feature vector;
inputting a third type of feature matrix aiming at the target user into the third neural network model to obtain a third feature vector;
inputting a fourth type of feature matrix aiming at the target commodity into the fourth neural network model to obtain a fourth feature vector;
fusing the first feature vector, the second feature vector, the third feature vector and the fourth feature vector according to a multi-view machine learning algorithm to obtain a fused feature vector;
and processing the fusion feature vector through an activation function to obtain the prediction score value of the target user on the target commodity.
Optionally, the recommended network model is obtained by training through the following steps:
acquiring a preset neural network model and the training set;
inputting a first class feature matrix aiming at a plurality of sample users, a second class feature matrix aiming at a plurality of sample commodities, a third class feature matrix aiming at a plurality of sample users and a fourth class feature matrix aiming at a plurality of sample commodities into the neural network model to obtain the prediction score values of the plurality of sample users on the plurality of sample commodities;
determining a loss value according to the obtained prediction score value and a real score value included in the training set;
determining whether the neural network model converges according to the loss value;
if not, adjusting parameter values in the neural network model, and returning to the step of inputting the first class feature matrix aiming at the plurality of sample users, the second class feature matrix aiming at the plurality of sample commodities, the third class feature matrix aiming at the plurality of sample users and the fourth class feature matrix aiming at the plurality of sample commodities into the neural network model to obtain the predicted score values of the plurality of sample users on the plurality of sample commodities;
and if so, determining the current neural network model as the recommended network model.
In order to achieve the above object, an embodiment of the present invention further provides a recommendation scoring apparatus based on a network structure and a comment text, where the apparatus includes:
the determining module is used for determining a target user in the plurality of sample users and determining a target commodity in the plurality of sample commodities;
the first acquisition module is used for acquiring a first class feature matrix aiming at the target user and a second class feature matrix aiming at the target commodity; the first type feature matrix and the second type feature matrix are determined according to comment text information of the sample users on the sample commodities;
the second acquisition module is used for acquiring a third type feature matrix aiming at the target user and a fourth type feature matrix aiming at the target commodity; the third type feature matrix and the fourth type feature matrix are determined according to the purchasing network structure information of the plurality of sample users on the plurality of sample commodities;
and the prediction module is used for inputting the first class characteristic matrix and the third class characteristic matrix aiming at the target user and the second class characteristic matrix and the fourth class characteristic matrix aiming at the target commodity into a recommendation network model to obtain the prediction score value of the target user on the target commodity.
Optionally, the first obtaining module is specifically configured to:
acquiring a plurality of comment text messages of the plurality of sample users on the plurality of sample commodities;
performing word segmentation processing on the plurality of comment text messages to obtain a word vector of each comment text message;
determining a first class feature matrix aiming at the target user according to the word vectors of the comment text information of the target user;
and determining a second class characteristic matrix aiming at the target commodity according to the word vectors of the plurality of comment text messages of the target commodity.
Optionally, the purchasing network structure information is network structure information of a heterogeneous information network based on purchasing information of the plurality of sample users on the plurality of sample commodities, a node of the heterogeneous information network includes the plurality of sample users and the plurality of sample commodities, an edge of the heterogeneous information network is used for connecting the sample users and the sample commodities, and an edge of the heterogeneous information network is used for indicating that the sample users have real scores on the sample commodities;
the second obtaining module is specifically configured to:
based on the connection relation between the sample user and the sample commodity in the heterogeneous information network, randomly walking by taking the target user as an initial node to determine a plurality of first random walking sequences, randomly walking by taking the target commodity as the initial node to determine a plurality of second random walking sequences; the first random walk sequence comprises a first preset number of nodes; the second random walk sequence comprises a second preset number of nodes;
determining a third type of feature matrix aiming at the target user according to the plurality of first random walk sequences;
and determining a fourth type feature matrix aiming at the target commodity according to the plurality of second random walk sequences.
Optionally, the recommended network model includes: a first neural network model, a second neural network model, a third neural network model, and a fourth neural network model;
the prediction module is specifically configured to:
inputting a first class feature matrix aiming at the target user into the first neural network model to obtain a first feature vector;
inputting a second type of feature matrix aiming at the target commodity into the second neural network model to obtain a second feature vector;
inputting a third type of feature matrix aiming at the target user into the third neural network model to obtain a third feature vector;
inputting a fourth type of feature matrix aiming at the target commodity into the fourth neural network model to obtain a fourth feature vector;
fusing the first feature vector, the second feature vector, the third feature vector and the fourth feature vector according to a multi-view machine learning algorithm to obtain a fused feature vector;
and processing the fusion feature vector through an activation function to obtain the prediction score value of the target user on the target commodity.
Optionally, the apparatus further comprises:
the training module is used for training the recommendation network model;
the training module is specifically configured to:
acquiring a preset neural network model and the training set;
inputting a first class feature matrix aiming at a plurality of sample users, a second class feature matrix aiming at a plurality of sample commodities, a third class feature matrix aiming at a plurality of sample users and a fourth class feature matrix aiming at a plurality of sample commodities into the neural network model to obtain the prediction score values of the plurality of sample users on the plurality of sample commodities;
determining a loss value according to the obtained prediction score value and a real score value included in the training set;
determining whether the neural network model converges according to the loss value;
if not, adjusting parameter values in the neural network model, and returning to the step of inputting the first class feature matrix aiming at the plurality of sample users, the second class feature matrix aiming at the plurality of sample commodities, the third class feature matrix aiming at the plurality of sample users and the fourth class feature matrix aiming at the plurality of sample commodities into the neural network model to obtain the predicted score values of the plurality of sample users on the plurality of sample commodities;
and if so, determining the current neural network model as the recommended network model.
In order to achieve the above object, an embodiment of the present invention further provides an electronic device, including a processor, a communication interface, a memory, and a communication bus; the processor, the communication interface and the memory complete mutual communication through a communication bus;
a memory for storing a computer program;
and the processor is used for realizing any method step when executing the program stored in the memory.
To achieve the above object, an embodiment of the present invention further provides a computer-readable storage medium, in which a computer program is stored, and the computer program, when executed by a processor, implements any of the above method steps.
In the embodiment of the invention, a recommendation network model is obtained by training based on the first class feature matrix and the third class feature matrix for each sample user, the second class feature matrix and the fourth class feature matrix for each sample commodity, and the real score values of a plurality of sample commodities of a plurality of sample users. Inputting the first class characteristic matrix and the third class characteristic matrix aiming at the target user and the second class characteristic matrix and the fourth class characteristic matrix aiming at the target commodity into the trained recommendation network model to obtain the prediction score value of the target user on the target commodity.
The first type feature matrix and the second type feature matrix are determined according to comment text information of a plurality of sample users on a plurality of sample commodities, and the third type feature matrix and the fourth type feature matrix are determined according to purchasing network structure information of the plurality of sample users on the plurality of sample commodities. The method comprises the steps of determining a predicted grade value of a target user to a target commodity according to a first class feature matrix, a second class feature matrix, a third class feature matrix and a fourth class feature matrix, and fully considering interaction information between the user and the commodity, namely fully considering comment text information and grade information between the user and the commodity, so that the purchase expectation of the user to the commodity can be more accurately predicted, and the accuracy of grade of the commodity is improved.
Of course, not all of the advantages described above need to be achieved at the same time in the practice of any one product or method of the invention.
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 a flowchart of a recommendation scoring method based on a network structure and a comment text according to an embodiment of the present invention;
FIG. 2 is a diagram of a heterogeneous information network according to an embodiment of the present invention;
FIG. 3 is an exemplary diagram of a recommendation scoring method based on a network structure and a comment text according to an embodiment of the present invention;
fig. 4 is a schematic structural diagram of a recommendation scoring apparatus based on a network structure and a comment text according to an embodiment of the present invention;
fig. 5 is a schematic structural diagram of an electronic device according to an embodiment of the present invention.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and 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.
In order to solve the problem of inaccurate product scoring, an embodiment of the present invention provides a recommendation scoring method based on a network structure and a comment text, which can be seen in fig. 1, where fig. 1 is a flowchart of the recommendation scoring method based on a network structure and a comment text provided by the embodiment of the present invention, and the method includes the following steps:
step S101: a target user of the plurality of sample users is determined, and a target item of the plurality of sample items is determined.
In the embodiment of the invention, a large number of sample users and a large number of sample commodities can be obtained in advance. The part of sample users have scoring information on the part of sample commodities, and the part of sample users have comment text information on the part of sample commodities. For example, the sample users are users of a certain shopping platform, the sample goods are goods provided by the shopping platform, the sample users purchase and score part of the sample goods of the shopping platform, and some of the sample users comment on the purchased sample goods.
Based on the information data, for a certain sample user and a sample commodity, wherein the sample user has no scoring information and no comment text information for the sample commodity, the embodiment of the invention is to realize that: and predicting the predicted scoring value of the sample user for the sample commodity, wherein the predicted scoring value can also represent the purchase expectation of the sample user for the sample commodity.
That is, in step S101, a target user may be determined from a plurality of sample users, and a target product may be determined from a plurality of sample products, where the target user has no scoring information and comment text information for the target product.
Step S102: and acquiring a first class feature matrix aiming at the target user and a second class feature matrix aiming at the target commodity. The first-class feature matrix and the second-class feature matrix are determined according to comment text information of a plurality of sample users on a plurality of sample commodities.
In the embodiment of the invention, comment text information of a target user on a plurality of sample commodities can be converted into a matrix representation form, and a first-class feature matrix for the target user can be obtained. The comment text information of a plurality of sample users on the target commodity can be converted into a matrix representation form, and a second type feature matrix for the target commodity is obtained.
In one implementation of the present invention, the first class feature matrix for the target user and the second class feature matrix for the target commodity may be determined by the following steps.
Step 11: and acquiring a plurality of comment text messages of a plurality of sample users to a plurality of sample commodities.
In this step, all comment text information made by the plurality of sample users on the plurality of sample commodities can be acquired.
Step 12: and performing word segmentation processing on the plurality of comment text messages to obtain a word vector of each comment text message.
In this step, word segmentation processing may be performed on all the comment text information acquired in step 11, so as to obtain a word vector of each comment text information.
In one implementation, for each comment text message, all punctuations in the comment text message may be deleted during word segmentation, and then the comment text message may be divided into words. And correspondingly converting each word of the comment text information into a digital form to obtain a word vector which comprises a string of numbers and corresponds to the comment text information.
Step 13: and determining a first class characteristic matrix aiming at the target user according to the word vectors of the plurality of comment text messages of the target user.
The text information of the multiple comments of the target user is as follows: made by target users for a plurality of sample goodsThe sum of the text information is reviewed. For example, the target user purchased a sample item I1Sample article I2Sample article I3And sample goods I4And for sample goods I1Sample article I2And sample article I4After commenting, the target user is matched with the sample commodity I1Sample article I2And sample article I4And all the comment text information is determined as the comment text information of the target user.
In one implementation, after a target user is determined, word vectors corresponding to a plurality of comment text messages of the target user can be obtained, and the word vectors are combined to obtain a first-class feature matrix for the target user.
Step 14: and determining a second type of feature matrix aiming at the target commodity according to the word vectors of the plurality of comment text messages of the target commodity.
The text information of the comments of the target commodity is as follows: and summing the text information of comments made by a plurality of different sample users on the target commodity. For example, the target commodity is used by the user U1User U2User U3And user U4Purchased and user U1User U2And user U4After the target commodity is commented, the user U is selected1User U2And user U4And determining the comment text information of the target commodity as the comment text information of the target commodity.
In one implementation, after the target commodity is determined, word vectors of a plurality of comment text messages of the target commodity can be obtained, and the word vectors are combined to obtain a second-class feature matrix for the target commodity.
In the embodiment of the present invention, the execution order of step 13 and step 14 is not limited.
Of course, the text information of the comments of each sample user or each sample commodity can be converted into a matrix form by using the method flows of the steps 11 to 14 provided by the embodiment of the present invention, which is not limited herein.
In an implementation manner of the present invention, when the predicted score value of the target user for the target commodity needs to be determined, the first class feature matrix for the target user and the second class feature matrix for the target commodity may be determined according to the method flows of the above steps 11 to 14, so as to improve the accuracy of the determined first class feature matrix for the target user and the determined second class feature matrix for the target commodity, and further improve the accuracy of the determined predicted score value.
In another implementation manner of the present invention, a first class feature matrix for each sample user and a second class feature matrix for each sample commodity may be pre-calculated, and the plurality of first class feature matrices and the plurality of second class feature matrices obtained by calculation are pre-stored in a database as data in a training set.
When the prediction score value of the target user on the target commodity needs to be determined, the first class feature matrix for the target user and the second class feature matrix for the target commodity are directly obtained from the database, so that the calculation efficiency is improved.
Step S103: and acquiring a third type feature matrix aiming at the target user and a fourth type feature matrix aiming at the target commodity. The third type feature matrix and the fourth type feature matrix are determined according to the purchasing network structure information of a plurality of sample users on a plurality of sample commodities.
In the embodiment of the present invention, the purchase network structure information indicates which sample users have purchased which sample goods and scored the purchased goods, and the purchase network structure information may be represented in the form of a heterogeneous information network, referring to fig. 2, fig. 2 is a schematic diagram of the heterogeneous information network provided in the embodiment of the present invention, where the heterogeneous information network includes nodes and edges, the nodes include sample users and sample goods, and the edges connecting the sample users and the sample goods indicate that the sample users have made scores for the sample goods.
In one implementation of the present invention, the third class feature matrix for the target user and the fourth class feature matrix for the target commodity may be determined by the following steps.
Step 21: based on the connection relation between the sample users and the sample commodities in the heterogeneous information network, the target users are used as starting nodes to carry out random walk, a plurality of first random walk sequences are determined, the target commodities are used as the starting nodes to carry out random walk, and a plurality of second random walk sequences are determined. The first random walk sequence comprises a first preset number of nodes; the second random walk sequence includes a second preset number of nodes.
In practical application, when the first random walk sequence and the second random walk sequence are determined, because there are many sample users and many sample commodities, one sample user can score many sample commodities, and thus there are many first random walk sequences using a target user as a starting node. Similarly, there may be a plurality of second random walk sequences.
Now, referring to fig. 2, it is assumed that the first predetermined number is 5, and the target user is user U3Then with user U3As a starting node, randomly walking along an edge connecting the user and the commodity to obtain 2 first random walk sequences, including: { U3、I2、U2、I1、U1}，{U3、I3、U4、I4、U5}。
Similarly, a plurality of second random walk sequences obtained by performing random walks using the target product as the starting node may also be determined.
In one implementation of the present invention, each sample user and each sample commodity may be numbered in advance, and each of the first random walk sequence and the second random walk sequence may be converted into a vector including a string of numbers.
For example, in the heterogeneous information network shown in FIG. 2, user U may be connected1User U2User U3User U4User U5Numbered 01, 02, 03, 04, 05 respectively, and the commercial products I1Article I2Article I3Article I4Numbered 11, 12, 13, 14, respectively, for U3、I2、U2、I1、U1This first random walk sequence may be represented as a vector containing {03, 12, 02, 11, 01 }.
Step 22: and determining a third type of feature matrix aiming at the target user according to the plurality of first random walk sequences.
And after a plurality of first random walk sequences are obtained, combining the plurality of first random walk sequences to obtain a third type feature matrix aiming at the target user.
In one implementation of the invention, the plurality of first random walk sequences of the target user are each a vector representation comprising a string of numbers. A third class of feature matrices for the target user can be combined from these vectors.
The description is still made in connection with the example in step 21. For user U3The plurality of first random walk sequences of (a), comprising: { U3、I2、U2、I1、U1}，{U3、I3、U4、I4、U5}. Then a plurality of first random walk sequences can be concatenated to obtain a long sequence: { U3、I2、U2、I1、U1、U3、I3、U4、I4、U5}。
In the embodiment of the present invention, both the user and the merchandise can be represented by a vector, and the vector may include a certain number of numerical values. For example, user U may be connected1User U2User U3User U4User U5Respectively expressed as {0.1, 0.2, 0.3}, {0.1, 0.3, 0.1}, {0.2, 0.1, 0.3}, {0.2, 0.3, 0.1}, {0.3, 0.1, 0.2}, and product I1Article I2Article I3Article I4Which can be expressed as {0.7, 0.8, 0.9}, {0.7, 0.9, 0.8}, {0.8, 0.7, 0.9}, {0.8, 0.9, 0.7}, respectively, the above description is for user U3A long sequence of a plurality of first random walk sequences { U }3、I2、U2、I1、U1、U3、I3、U4、I4、U5The conversion can be to the following matrix:
Figure BDA0001720533600000121
of course, the above-described embodiment is merely an example. In practical applications, any method for converting a plurality of sequences into a matrix can be applied to the embodiment of the present invention.
Step 23: and determining a fourth type characteristic matrix aiming at the target commodity according to the plurality of second random walk sequences.
And after a plurality of second random walk sequences are obtained, combining the plurality of second random walk sequences to obtain a fourth type feature matrix for the target user.
In one implementation of the invention, the plurality of second random walk sequences of the target user are each a vector representation comprising a string of numbers. A fourth class feature matrix for the target user can be obtained by combining these vectors. Specifically, reference may be made to the example in step 22, which is not described herein again.
In the embodiment of the present invention, the execution sequence of step 22 and step 23 is not limited.
In an implementation manner of the present invention, when the predicted score value of the target user for the target commodity needs to be determined, the third class feature matrix for the target user and the fourth class feature matrix for the target commodity may be determined according to the method flows of the above steps 21 to 23, so as to improve the accuracy of the determined third class feature matrix for the target user and the determined fourth class feature matrix for the target commodity, and further improve the accuracy of the determined predicted score value.
In another implementation manner of the present invention, a third class feature matrix for each sample user and a fourth class feature matrix for each sample commodity may be pre-calculated, and the calculated third class feature matrices and the fourth class feature matrices are pre-stored in a database as data in a training set.
When the prediction score value of the target user on the target commodity needs to be determined, the third class feature matrix for the target user and the fourth class feature matrix for the target commodity are directly obtained from the database, so that the calculation efficiency is improved.
In the embodiment of the present invention, the execution sequence of step S103 and step S102 is not limited.
Step S104: and inputting the first class characteristic matrix and the third class characteristic matrix aiming at the target user and the second class characteristic matrix and the fourth class characteristic matrix aiming at the target commodity into a recommendation network model to obtain the prediction score value of the target user on the target commodity.
The recommended network model is a model obtained by training according to a training set, wherein the training set comprises: the real scoring values of the plurality of sample users on the plurality of sample commodities are determined according to the first class feature matrix for each sample user, the second class feature matrix for each sample commodity, the third class feature matrix for each sample user, the fourth class feature matrix for each sample commodity and the real scoring values of the plurality of sample users on the plurality of sample commodities.
After the first class feature matrix and the third class feature matrix for the target user and the second class feature matrix and the fourth class feature matrix for the target commodity are determined, the feature matrices can be input into a recommendation network model, and the prediction score value of the target user on the target commodity can be obtained.
Referring to fig. 3, in the embodiment of the present invention, the recommendation network model may include a first neural network model, a second neural network model, a third neural network model, and a fourth neural network model, and in the step of inputting the first class feature matrix and the third class feature matrix for the target user, and the second class feature matrix and the fourth class feature matrix for the target commodity into the recommendation network model, the following refinement steps may be included:
inputting a first class feature matrix aiming at a target user into a first neural network model to obtain a first feature vector; inputting a second type of feature matrix aiming at the target commodity into a second neural network model to obtain a second feature vector; inputting a third type of feature matrix aiming at the target user into a third neural network model to obtain a third feature vector; inputting a fourth type of feature matrix aiming at the target commodity into a fourth neural network model to obtain a fourth feature vector; and fusing the first feature vector, the second feature vector, the third feature vector and the fourth feature vector according to a multi-view machine learning algorithm to obtain a fused feature vector.
The first, second, third and fourth neural network models may be a recurrent neural network model, a convolutional neural network model, a cyclic convolutional neural network model, a deep neural network model, or the like. The embodiment of the present invention is not limited thereto.
Each neural network model can output a feature vector, and in order to comprehensively consider comment text information of a user on a commodity and rating information of the user on the commodity, in the embodiment of the present invention, the first feature vector, the second feature vector, the third feature vector, and the fourth feature vector may be fused.
In one implementation of the invention, the fusion can be performed according to a multi-view machine learning algorithm, and the numerical values in the feature vectors can interact with each other when the algorithm is used for fusion, so that four feature vectors can be better integrated. Fusing a plurality of vectors based on a multi-view machine learning algorithm belongs to the category of the prior art, and is not described herein.
In one implementation manner of the present invention, a linear rectification function (Rectified L initial unit, Re L U) may be used to process the fused feature vector, a Sigmoid activation function may be used to process the fused feature vector, or other activation functions may be used to process the fused feature vector.
Therefore, in the embodiment of the invention, the recommendation network model is obtained by training based on the first class feature matrix and the third class feature matrix for each sample user, the second class feature matrix and the fourth class feature matrix for each sample commodity, and the real score values of a plurality of sample commodities for a plurality of sample users. Inputting the first class characteristic matrix and the third class characteristic matrix aiming at the target user and the second class characteristic matrix and the fourth class characteristic matrix aiming at the target commodity into the trained recommendation network model to obtain the prediction score value of the target user on the target commodity.
The first type feature matrix and the second type feature matrix are determined according to comment text information of a plurality of sample users on a plurality of sample commodities, and the third type feature matrix and the fourth type feature matrix are determined according to purchasing network structure information of the plurality of sample users on the plurality of sample commodities. The method comprises the steps of determining a predicted grade value of a target user to a target commodity according to a first class feature matrix, a second class feature matrix, a third class feature matrix and a fourth class feature matrix, and fully considering interaction information between the user and the commodity, namely fully considering comment text information and grade information between the user and the commodity, so that the purchase expectation of the user to the commodity can be more accurately predicted, and the accuracy of grade of the commodity is improved.
The prediction score value determined in the embodiment of the invention represents the purchase expectation of the target user on the target commodity, a threshold value can be preset, if the prediction score value is greater than the threshold value, the purchase expectation of the target user on the target commodity is higher, and the target commodity can be recommended to the target user.
In the embodiment of the present invention, the recommended network model may be obtained by training through the following steps.
Step 31: and acquiring a preset neural network model and the training set.
The recommendation network model for predicting the score values is obtained by training according to a training set, wherein the training set comprises a first class feature matrix for each sample user, a second class feature matrix for each sample commodity, a third class feature matrix for each sample user, a fourth class feature matrix for each sample commodity and the real score values of a plurality of sample users on the sample commodities.
In the above embodiments, the method for determining the feature matrix included in the training set has been introduced, and is not described again.
Step 32: inputting the first class feature matrix aiming at the plurality of sample users, the second class feature matrix aiming at the plurality of sample commodities, the third class feature matrix aiming at the plurality of sample users and the fourth class feature matrix aiming at the plurality of sample commodities into a neural network model to obtain the prediction score values of the plurality of sample users on the plurality of sample commodities.
The step of inputting the four types of feature matrices into the neural network model to obtain the predicted score value may be referred to as the description of step S104 in the above recommendation scoring method embodiment.
In the embodiment of the present invention, the real score value of a sample user on a sample commodity, the first class feature matrix and the third class feature matrix of the sample user, and the second class feature matrix and the fourth class feature matrix of the sample commodity may be used as a set of training set data.
In an implementation manner of the present invention, a set of training set data may be input each time, and a prediction score value corresponding to the set of training set data is obtained.
In a preferred implementation manner of the present invention, a plurality of sets of training set data may be input each time, so as to obtain a prediction score value corresponding to each set of training set data.
For example, if the training set data includes 10000 real score data, each real score data corresponds to one sample user and sample commodity, 100 real score data of the real score data may be selected for each round of training, that is, the 100 real score data and the feature matrix of the corresponding sample user and sample commodity are input into the neural network model, so as to obtain 100 predicted score values.
Step 33: and determining a loss value according to the obtained prediction score value and the real score value included in the training set.
In the embodiment of the present application, the loss value is obtained by using, but not limited to, Mean Squared Error (MSE) formula as the loss function.
Step 34: and determining whether the neural network model converges according to the obtained loss value. If so, step 35 is performed. If not, step 36.
Step 35: and determining the current neural network model as the recommended network model.
Step 36: the values of the parameters in the neural network model are adjusted and execution returns to step 32.
Based on the same inventive concept, according to the above embodiment of the recommendation scoring method based on the network structure and the comment text, the embodiment of the present invention further provides a recommendation scoring apparatus based on the network structure and the comment text, referring to fig. 4, which may include the following modules:
a determining module 401, configured to determine a target user in the multiple sample users, and determine a target product in the multiple sample products;
a first obtaining module 402, configured to obtain a first class feature matrix for a target user and a second class feature matrix for a target commodity; the first type characteristic matrix and the second type characteristic matrix are determined according to comment text information of a plurality of sample users on a plurality of sample commodities;
a second obtaining module 403, configured to obtain a third class feature matrix for the target user and a fourth class feature matrix for the target product; the third type characteristic matrix and the fourth type characteristic matrix are determined according to the purchasing network structure information of a plurality of sample users on a plurality of sample commodities;
the prediction module 404 is configured to input the first class feature matrix and the third class feature matrix for the target user, and the second class feature matrix and the fourth class feature matrix for the target commodity into the recommendation network model, so as to obtain a prediction score value of the target user on the target commodity;
the recommendation scoring device based on the network structure and the comment text, provided by the embodiment of the invention, inputs a trained recommendation network model for a first class feature matrix and a third class feature matrix of a target user and for a second class feature matrix and a fourth class feature matrix of a target commodity to obtain a prediction scoring value of the target user on the target commodity, wherein the first class feature matrix and the second class feature matrix are determined according to comment text information of a plurality of sample users on a plurality of sample commodities, and the third class feature matrix and the fourth class feature matrix are determined according to purchase network structure information of the plurality of sample users on the plurality of sample commodities. Therefore, interactive information between the user and the commodity is fully considered, the interactive information comprises comment text information and grading information, and the purchase expectation of the user to the commodity can be more accurately predicted.
In an embodiment of the present invention, the first obtaining module 402 may be specifically configured to:
acquiring a plurality of comment text messages of the plurality of sample users on the plurality of sample commodities;
performing word segmentation processing on the plurality of comment text messages to obtain a word vector of each comment text message;
determining a first class feature matrix aiming at the target user according to the word vectors of the comment text information of the target user;
and determining a second class characteristic matrix aiming at the target commodity according to the word vectors of the plurality of comment text messages of the target commodity.
In one embodiment of the invention, the purchase network structure information is network structure information of a heterogeneous information network based on purchase information of a plurality of sample users on a plurality of sample commodities, nodes of the heterogeneous information network comprise the plurality of sample users and the plurality of sample commodities, edges of the heterogeneous information network are used for connecting the sample users and the sample commodities, and edges of the heterogeneous information network are used for indicating that the sample users have real scoring values on the sample commodities.
The second obtaining module 403 may specifically be configured to:
based on the connection relation between a sample user and a sample commodity in a heterogeneous information network, randomly walking by taking a target user as an initial node, determining a plurality of first random walking sequences, randomly walking by taking the target commodity as the initial node, and determining a plurality of second random walking sequences; the first random walk sequence comprises a first preset number of nodes; the second random walk sequence comprises a second preset number of nodes;
determining a third type of feature matrix aiming at the target user according to the plurality of first random walk sequences;
and determining a fourth type characteristic matrix aiming at the target commodity according to the plurality of second random walk sequences.
In one embodiment of the invention, recommending a network model includes: a first neural network model, a second neural network model, a third neural network model, and a fourth neural network model;
the prediction module 404 may be specifically configured to:
inputting a first class feature matrix aiming at a target user into a first neural network model to obtain a first feature vector; inputting a second type of feature matrix aiming at the target commodity into a second neural network model to obtain a second feature vector; inputting a third type of feature matrix aiming at the target user into a third neural network model to obtain a third feature vector; inputting a fourth type of feature matrix aiming at the target commodity into a fourth neural network model to obtain a fourth feature vector; fusing the first feature vector, the second feature vector, the third feature vector and the fourth feature vector according to a multi-view machine learning algorithm to obtain a fused feature vector; and processing the fusion feature vector through an activation function to obtain a prediction score value of the target user on the target commodity.
In an embodiment of the present invention, on the basis of the embodiment of the apparatus shown in fig. 3, the apparatus may further include a training module, configured to train a recommendation network model, specifically, to:
acquiring a preset neural network model and a training set;
inputting a first class feature matrix aiming at a plurality of sample users, a second class feature matrix aiming at a plurality of sample commodities, a third class feature matrix aiming at a plurality of sample users and a fourth class feature matrix aiming at a plurality of sample commodities into a neural network model to obtain the prediction score values of the plurality of sample users on the plurality of sample commodities;
determining a loss value according to the obtained prediction score value and a real score value included in the training set;
determining whether the neural network model converges according to the loss value;
if not, adjusting parameter values in the neural network model, returning to a step of inputting a first class feature matrix aiming at a plurality of sample users, a second class feature matrix aiming at a plurality of sample commodities, a third class feature matrix aiming at a plurality of sample users and a fourth class feature matrix aiming at a plurality of sample commodities into the neural network model, and obtaining the predicted score values of the plurality of sample users on the plurality of sample commodities;
and if so, determining the current neural network model as the recommended network model.
Based on the same inventive concept, according to the above embodiment of the recommendation scoring method based on the network structure and the comment text, the embodiment of the present invention further provides an electronic device, as shown in fig. 5, which includes a processor 501, a communication interface 502, a memory 503, and a communication bus 504, wherein the processor 501, the communication interface 502, and the memory 503 complete mutual communication through the communication bus 504,
a memory 503 for storing a computer program;
the processor 501 is configured to implement the embodiment of the recommendation scoring method based on the network structure and the comment text shown in fig. 1 when executing the program stored in the memory 503. The recommendation scoring method based on the network structure and the comment text comprises the following steps:
determining a target user in the plurality of sample users and determining a target commodity in the plurality of sample commodities;
acquiring a first class feature matrix for the target user and a second class feature matrix for the target commodity; the first type feature matrix and the second type feature matrix are determined according to comment text information of the sample users on the sample commodities;
acquiring a third type feature matrix for the target user and a fourth type feature matrix for the target commodity; the third type feature matrix and the fourth type feature matrix are determined according to the purchasing network structure information of the plurality of sample users on the plurality of sample commodities;
inputting the first class feature matrix and the third class feature matrix for the target user and the second class feature matrix and the fourth class feature matrix for the target commodity into a recommendation network model to obtain a prediction score value of the target user on the target commodity;
the recommended network model is a model obtained by training according to a training set, wherein the training set comprises: the real scoring values of the plurality of sample users on the plurality of sample commodities are determined according to the first class feature matrix for each sample user, the second class feature matrix for each sample commodity, the third class feature matrix for each sample user, the fourth class feature matrix for each sample commodity and the real scoring values of the plurality of sample users on the plurality of sample commodities.
In the embodiment of the invention, a trained recommended network model is input into a first class feature matrix and a third class feature matrix for a target user and a second class feature matrix and a fourth class feature matrix for a target commodity to obtain a predicted score value of the target user on the target commodity, wherein the first class feature matrix and the second class feature matrix are determined according to comment text information of a plurality of sample users on a plurality of sample commodities, and the third class feature matrix and the fourth class feature matrix are determined according to network structure information of the plurality of sample users on the plurality of sample commodities. Therefore, interactive information between the user and the commodity is fully considered, the interactive information comprises comment text information and grading information, and the purchase expectation of the user to the commodity can be more accurately predicted.
The communication bus 504 mentioned above for the electronic device may be a Peripheral Component Interconnect (PCI) bus, an Extended Industry Standard Architecture (EISA) bus, or the like. The communication bus 504 may be divided into an address bus, a data bus, a control bus, etc. For ease of illustration, only one thick line is shown in FIG. 5, but this is not intended to represent only one bus or type of bus.
The communication interface 502 is used for communication between the above-described electronic apparatus and other apparatuses.
The Memory 503 may include a Random Access Memory (RAM) or a Non-Volatile Memory (NVM), such as at least one disk Memory. Optionally, the memory 503 may also be at least one storage device located remotely from the aforementioned processor.
The Processor 501 may be a general-purpose Processor, including a Central Processing Unit (CPU), a Network Processor (NP), and the like; but also Digital Signal Processors (DSPs), Application Specific Integrated Circuits (ASICs), Field Programmable Gate Arrays (FPGAs) or other Programmable logic devices, discrete Gate or transistor logic devices, discrete hardware components.
Based on the same inventive concept, according to the above recommendation scoring method based on the network structure and the comment text, in yet another embodiment provided by the present invention, a computer readable storage medium is further provided, in which a computer program is stored, and the computer program, when being executed by a processor, implements the above recommendation scoring method based on the network structure and the comment text shown in fig. 1.
It is noted that, herein, 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.
All the embodiments in the present specification are described in a related manner, and the same and similar parts among the embodiments may be referred to each other, and each embodiment focuses on the differences from the other embodiments. In particular, as for the device embodiment, the electronic device embodiment and the storage medium embodiment, since they are basically similar to the method embodiment, the description is relatively simple, and relevant points can be referred to the partial description of the recommendation scoring method embodiment based on the network structure and the comment text.
The above description is only for the preferred embodiment of the present invention, and is not intended to limit the scope of the present invention. Any modification, equivalent replacement, or improvement made within the spirit and principle of the present invention shall fall within the protection scope of the present invention.

Claims (6)

1. A recommendation scoring method based on a network structure and comment texts is characterized by comprising the following steps:
determining a target user in the plurality of sample users and determining a target commodity in the plurality of sample commodities;
acquiring a first class feature matrix for the target user and a second class feature matrix for the target commodity; the first type feature matrix and the second type feature matrix are determined according to comment text information of the sample users on the sample commodities;
acquiring a third type feature matrix for the target user and a fourth type feature matrix for the target commodity; the third type feature matrix and the fourth type feature matrix are determined according to the purchasing network structure information of the plurality of sample users on the plurality of sample commodities;
inputting the first class feature matrix and the third class feature matrix for the target user and the second class feature matrix and the fourth class feature matrix for the target commodity into a recommendation network model to obtain a prediction score value of the target user on the target commodity;
the recommended network model is a model obtained by training according to a training set, wherein the training set comprises: the first class feature matrix aiming at each sample user, the second class feature matrix aiming at each sample commodity, the third class feature matrix aiming at each sample user, the fourth class feature matrix aiming at each sample commodity and the real score values of a plurality of sample users on a plurality of sample commodities;
the first class characteristic matrix aiming at the target user and the second class characteristic matrix aiming at the target commodity are determined by the following steps:
acquiring a plurality of comment text messages of the plurality of sample users on the plurality of sample commodities;
performing word segmentation processing on the plurality of comment text messages to obtain a word vector of each comment text message;
determining a first class feature matrix aiming at the target user according to the word vectors of the comment text information of the target user;
determining a second class of feature matrix aiming at the target commodity according to the word vectors of the comment text messages of the target commodity;
the purchasing network structure information is network structure information of a heterogeneous information network based on purchasing information of the plurality of sample users on the plurality of sample commodities, nodes of the heterogeneous information network comprise the plurality of sample users and the plurality of sample commodities, edges of the heterogeneous information network are used for connecting the sample users and the sample commodities, and edges of the heterogeneous information network are used for indicating that the sample users have real scoring values on the sample commodities;
the third type feature matrix for the target user and the fourth type feature matrix for the target commodity are determined by the following steps:
based on the connection relation between the sample user and the sample commodity in the heterogeneous information network, randomly walking by taking the target user as an initial node to determine a plurality of first random walking sequences, randomly walking by taking the target commodity as the initial node to determine a plurality of second random walking sequences; the first random walk sequence comprises a first preset number of nodes; the second random walk sequence comprises a second preset number of nodes;
determining a third type of feature matrix aiming at the target user according to the plurality of first random walk sequences;
determining a fourth type feature matrix aiming at the target commodity according to the plurality of second random walk sequences;
the recommended network model includes: a first neural network model, a second neural network model, a third neural network model, and a fourth neural network model;
the step of inputting the first class feature matrix and the third class feature matrix for the target user and the second class feature matrix and the fourth class feature matrix for the target commodity into a recommendation network model to obtain the predicted score value of the target user on the target commodity comprises:
inputting a first class feature matrix aiming at the target user into the first neural network model to obtain a first feature vector;
inputting a second type of feature matrix aiming at the target commodity into the second neural network model to obtain a second feature vector;
inputting a third type of feature matrix aiming at the target user into the third neural network model to obtain a third feature vector;
inputting a fourth type of feature matrix aiming at the target commodity into the fourth neural network model to obtain a fourth feature vector;
fusing the first feature vector, the second feature vector, the third feature vector and the fourth feature vector according to a multi-view machine learning algorithm to obtain a fused feature vector;
and processing the fusion feature vector through an activation function to obtain the prediction score value of the target user on the target commodity.
2. The method of claim 1, wherein the recommended network model is obtained by training using the following steps:
acquiring a preset neural network model and the training set;
inputting a first class feature matrix aiming at a plurality of sample users, a second class feature matrix aiming at a plurality of sample commodities, a third class feature matrix aiming at a plurality of sample users and a fourth class feature matrix aiming at a plurality of sample commodities into the neural network model to obtain the prediction score values of the plurality of sample users on the plurality of sample commodities;
determining a loss value according to the obtained prediction score value and a real score value included in the training set;
determining whether the neural network model converges according to the loss value;
if not, adjusting parameter values in the neural network model, and returning to the step of inputting the first class feature matrix aiming at the plurality of sample users, the second class feature matrix aiming at the plurality of sample commodities, the third class feature matrix aiming at the plurality of sample users and the fourth class feature matrix aiming at the plurality of sample commodities into the neural network model to obtain the predicted score values of the plurality of sample users on the plurality of sample commodities;
and if so, determining the current neural network model as the recommended network model.
3. The method of claim 1, further comprising:
judging whether the predicted score value of the target user on the target commodity reaches a threshold value; and if so, recommending the target commodity to the target user.
4. A recommendation scoring apparatus based on a network structure and a comment text, the apparatus comprising:
the determining module is used for determining a target user in the plurality of sample users and determining a target commodity in the plurality of sample commodities;
the first acquisition module is used for acquiring a first class feature matrix aiming at the target user and a second class feature matrix aiming at the target commodity; the first type feature matrix and the second type feature matrix are determined according to comment text information of the sample users on the sample commodities;
the second acquisition module is used for acquiring a third type feature matrix aiming at the target user and a fourth type feature matrix aiming at the target commodity; the third type feature matrix and the fourth type feature matrix are determined according to the purchasing network structure information of the plurality of sample users on the plurality of sample commodities;
the prediction module is used for inputting the first class characteristic matrix and the third class characteristic matrix aiming at the target user and the second class characteristic matrix and the fourth class characteristic matrix aiming at the target commodity into a recommendation network model to obtain a prediction score value of the target user on the target commodity;
the first obtaining module is specifically configured to:
acquiring a plurality of comment text messages of the plurality of sample users on the plurality of sample commodities;
performing word segmentation processing on the plurality of comment text messages to obtain a word vector of each comment text message;
determining a first class feature matrix aiming at the target user according to the word vectors of the comment text information of the target user;
determining a second class of feature matrix aiming at the target commodity according to the word vectors of the comment text messages of the target commodity;
the purchasing network structure information is network structure information of a heterogeneous information network based on purchasing information of the plurality of sample users on the plurality of sample commodities, nodes of the heterogeneous information network comprise the plurality of sample users and the plurality of sample commodities, edges of the heterogeneous information network are used for connecting the sample users and the sample commodities, and edges of the heterogeneous information network are used for indicating that the sample users have real scoring values on the sample commodities;
the second obtaining module is specifically configured to:
based on the connection relation between the sample user and the sample commodity in the heterogeneous information network, randomly walking by taking the target user as an initial node to determine a plurality of first random walking sequences, randomly walking by taking the target commodity as the initial node to determine a plurality of second random walking sequences; the first random walk sequence comprises a first preset number of nodes; the second random walk sequence comprises a second preset number of nodes;
determining a third type of feature matrix aiming at the target user according to the plurality of first random walk sequences;
and determining a fourth type feature matrix aiming at the target commodity according to the plurality of second random walk sequences.
5. An electronic device is characterized by comprising a processor, a communication interface, a memory and a communication bus, wherein the processor and the communication interface are used for realizing mutual communication by the memory through the communication bus;
a memory for storing a computer program;
a processor for implementing the method steps of any of claims 1 to 3 when executing a program stored in the memory.
6. A computer-readable storage medium, characterized in that a computer program is stored in the computer-readable storage medium, which computer program, when being executed by a processor, carries out the method steps of any one of the claims 1-3.
CN201810729637.2A 2018-07-05 2018-07-05 Recommendation scoring method and device based on network structure and comment text Active CN108920665B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201810729637.2A CN108920665B (en) 2018-07-05 2018-07-05 Recommendation scoring method and device based on network structure and comment text

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201810729637.2A CN108920665B (en) 2018-07-05 2018-07-05 Recommendation scoring method and device based on network structure and comment text

Publications (2)

Publication Number Publication Date
CN108920665A CN108920665A (en) 2018-11-30
CN108920665B true CN108920665B (en) 2020-08-07

Family

ID=64425289

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201810729637.2A Active CN108920665B (en) 2018-07-05 2018-07-05 Recommendation scoring method and device based on network structure and comment text

Country Status (1)

Country Link
CN (1) CN108920665B (en)

Families Citing this family (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109740728B (en) * 2018-12-10 2019-11-01 杭州世平信息科技有限公司 A kind of measurement of penalty calculation method based on a variety of neural network ensembles
CN111414561B (en) * 2019-01-08 2023-07-28 百度在线网络技术（北京）有限公司 Method and device for presenting information
CN111444334B (en) * 2019-01-16 2023-04-25 阿里巴巴集团控股有限公司 Data processing method, text recognition device and computer equipment
CN110059245B (en) * 2019-02-13 2023-06-30 创新先进技术有限公司 Method and device for predicting user scoring and displaying object
CN110796512B (en) * 2019-09-20 2023-09-05 平安科技（深圳）有限公司 Product information recommendation method and device, storage medium and computer equipment
CN110837602B (en) * 2019-11-05 2022-10-04 重庆邮电大学 User recommendation method based on representation learning and multi-mode convolutional neural network
CN112966173B (en) * 2019-12-13 2024-02-27 北京达佳互联信息技术有限公司 Classification operation method and device for information comments

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105761154A (en) * 2016-04-11 2016-07-13 北京邮电大学 Socialized recommendation method and device
CN106503028A (en) * 2015-09-08 2017-03-15 北京邮电大学 Recommend method and system
CN107292648A (en) * 2016-04-01 2017-10-24 株式会社理光 A kind of user behavior analysis method and device
CN107944629A (en) * 2017-11-30 2018-04-20 北京邮电大学 A kind of recommendation method and device based on heterogeneous information network representation

Family Cites Families (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US8079910B2 (en) * 2007-12-12 2011-12-20 Russell Acree System and method of penalty data compilation, analysis and report generation

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106503028A (en) * 2015-09-08 2017-03-15 北京邮电大学 Recommend method and system
CN107292648A (en) * 2016-04-01 2017-10-24 株式会社理光 A kind of user behavior analysis method and device
CN105761154A (en) * 2016-04-11 2016-07-13 北京邮电大学 Socialized recommendation method and device
CN107944629A (en) * 2017-11-30 2018-04-20 北京邮电大学 A kind of recommendation method and device based on heterogeneous information network representation

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
融合异质网络与主题模型的方面分预测;吉余岗等;《计算机应用》;20171110;第37卷(第11期);第3201-3206页 *

Also Published As

Publication number Publication date
CN108920665A (en) 2018-11-30

Similar Documents

Publication Publication Date Title
CN108920665B (en) Recommendation scoring method and device based on network structure and comment text
CN108985830B (en) Recommendation scoring method and device based on heterogeneous information network
US11551007B2 (en) Determining intent from a historical vector of a to-be-analyzed statement
US10497013B2 (en) Purchasing behavior analysis apparatus and non-transitory computer readable medium
TWI718422B (en) Method, device and equipment for fusing model prediction values
CN111444334A (en) Data processing method, text recognition device and computer equipment
CN109189921B (en) Comment evaluation model training method and device
US11875241B2 (en) Aspect pre-selection using machine learning
CN109584006B (en) Cross-platform commodity matching method based on deep matching model
CN108446297B (en) Recommendation method and device and electronic equipment
CN111291158B (en) Information query method and device, electronic equipment and storage medium
CN113781149A (en) Information recommendation method and device, computer-readable storage medium and electronic equipment
CN111523315B (en) Data processing method, text recognition device and computer equipment
CN108984777B (en) Customer service method, apparatus and computer-readable storage medium
JP5945206B2 (en) Product recommendation device, method and program
CN106933798B (en) Information analysis method and device
CN111680213A (en) Information recommendation method, data processing method and device
CN112199500A (en) Emotional tendency identification method and device for comments and electronic equipment
JP5835802B2 (en) Purchase forecasting apparatus, method, and program
CN110796504B (en) Article recommendation method and device
CN114429384B (en) Intelligent product recommendation method and system based on e-commerce platform
CN112395398A (en) Question and answer processing method, device and equipment
CN116401372A (en) Knowledge graph representation learning method and device, electronic equipment and readable storage medium
CN113297467A (en) Recommendation method, recommendation device and computer storage medium
CN110807646A (en) Data analysis method, device and computer readable storage medium

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