CN110910215A - Product recommendation method, device, equipment and computer-readable storage medium - Google Patents
Product recommendation method, device, equipment and computer-readable storage medium Download PDFInfo
- Publication number
- CN110910215A CN110910215A CN201911145185.4A CN201911145185A CN110910215A CN 110910215 A CN110910215 A CN 110910215A CN 201911145185 A CN201911145185 A CN 201911145185A CN 110910215 A CN110910215 A CN 110910215A
- Authority
- CN
- China
- Prior art keywords
- score
- product
- target user
- user
- preset
- 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.)
- Pending
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
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)
Abstract
The invention relates to the technical field of financial technology (Fintech), and discloses a product recommendation method, which comprises the following steps: when a product recommendation request is received, determining a target user corresponding to the product recommendation request; adjusting a first actual score of the target user in a preset score sample set to obtain an adjusted score of the target user; bringing the adjusted scores into a preset similarity formula, and calculating the similarity between the target user and a reference user through the similarity formula, wherein the reference user is a user in the score sample set except the target user; and recommending products for the target user according to the similarity. The invention also discloses a product recommendation device, equipment and a computer readable storage medium. The invention improves the accuracy of product recommendation by effectively eliminating noise data.
Description
Technical Field
The invention relates to the technical field of financial technology (Fintech), in particular to a product recommendation method, a device, equipment and a computer readable storage medium.
Background
With the development of computer technology, more and more technologies are applied in the financial field, the traditional financial industry is gradually changing to financial technology (Fintech), and the financial platform interaction technology is no exception, but due to the requirements of the financial industry on safety and real-time performance, higher requirements are also put forward on the technologies.
In order to accurately recommend products to users, the similarity of the users is determined by cosine similarity, Pearson similarity, Jaccard similarity and other methods according to the preference degrees of the users to the products, a neighbor cluster is divided for the users, then the evaluation of the users is predicted by the evaluation of the neighbor cluster to a certain product, and finally whether the products are recommended to the users is determined according to evaluation feedback, so that the personalized difference of the users is ignored in the product recommendation; for example, two users have many items that are commonly rated, but neither user scores the items high, which to some extent can only indicate that neither user likes the product, which is noisy data for the recommendation algorithm, and traditional algorithms do not distinguish effectively.
Disclosure of Invention
The invention mainly aims to provide a product recommendation method, a product recommendation device, equipment and a computer readable storage medium, and aims to solve the technical problem that the product recommendation is inaccurate because the used recommendation algorithm cannot exclude noise data when the current product recommendation is carried out.
In order to achieve the above object, the present invention provides a product recommendation method, including the steps of:
when a product recommendation request is received, determining a target user corresponding to the product recommendation request;
adjusting a first actual score of the target user in a preset score sample set to obtain an adjusted score of the target user;
bringing the adjusted scores into a preset similarity formula, and calculating the similarity between the target user and a reference user through the similarity formula, wherein the reference user is a user in the score sample set except the target user;
and recommending products for the target user according to the similarity.
In an embodiment, the step of adjusting the first actual score of the target user in a preset score sample set to obtain the adjusted score of the target user includes:
extracting a first actual score and a second actual score of the target user and the reference user for the target product in a preset score sample set;
counting the first actual scores to calculate a first average score of the target user on all products participating in scoring, and counting the second actual scores to calculate a second average score of the reference user on all products participating in scoring;
obtaining a maximum product score of the target product, and taking the first average score, the second average score and the maximum product score as independent variables of a preset adjusting formula to obtain an adjusting score of the target user, wherein the preset adjusting formula is as follows:
the above-mentioned
An adjusted score representing the target user, RmaxRepresenting a maximum product score for the target product, the
Representing a target user u1For a first actual score of the target product, the
Representing a target user u1A first average score of all products involved in the scoring, said
Representing reference users u2For all ginsengAnd a second average score of the scored product.
In an embodiment, the preset similarity formula is:
the Sim (u)1,u2) ' representing target user u1And reference user u2The similarity of (A) to (B), the
Representing a target user u1For a first actual score of the target product, the
Representing reference users u2For a second actual score of the target product, the
Representing a target user u1A first average score, said
Representing reference users u2A second average score of (a);
the above-mentioned
Representing a target user u1And reference user u2The preference degrees of the two are consistent;
the adjustment score is
The above-mentioned
Representing a target user u1And reference user u2The variance of the scores.
In an embodiment, the step of recommending the product to the target user according to the similarity includes:
sorting the similarity from big to small, and obtaining a preset number of reference users with the similarity sorted in the front to form an adjacent cluster;
and determining a recommended product according to the reference user in the adjacent cluster, and pushing the recommended product to the target user.
In an embodiment, the step of determining a recommended product according to the reference user in the adjacent cluster and pushing the recommended product to the target user includes:
acquiring a reference user in the adjacent cluster, inputting the similarity between the reference user and the target user into a preset scoring formula, and obtaining a product prediction score of the target user on a product;
when the prediction score is higher than a preset threshold value, the product is used as a recommended product and pushed to the target user;
wherein, the preset scoring formula is as follows:
In an embodiment, the step of recommending the product to the target user according to the similarity includes:
outputting a product evaluation prompt when a purchase request based on a recommended product is detected;
and acquiring a product actual score input based on the product evaluation prompt, and adding the product actual score as a first product score of the target user to a preset score sample set.
In an embodiment, after the step of obtaining the actual product score input based on the product evaluation prompt, and adding the actual product score as the first product score of the target user to a preset score sample set, the method includes:
when the absolute value of the difference value between the actual product score and the predicted product score is larger than a preset difference value, marking the preset adjusting formula;
and when the marking frequency of the preset adjusting formula is greater than the preset frequency, outputting a formula adjusting prompt.
In addition, to achieve the above object, the present invention also provides a product recommendation apparatus, including:
the request receiving module is used for determining a target user corresponding to a product recommendation request when the product recommendation request is received;
the score adjusting module is used for adjusting a first actual score of the target user in a preset score sample set to obtain an adjusted score of the target user;
the similarity calculation module is used for substituting the adjusted scores into a preset similarity formula and calculating the similarity between the target user and a reference user according to the similarity formula, wherein the reference user is a user except the target user in the score sample set;
and the product recommendation module is used for recommending products for the target user according to the similarity.
In addition, to achieve the above object, the present invention also provides a product recommendation apparatus, including: a memory, a processor and a product recommendation program stored on the memory and executable on the processor, the product recommendation program when executed by the processor implementing the steps of the product recommendation method as described above.
In addition, to achieve the above object, the present invention also provides a computer-readable storage medium having a product recommendation program stored thereon, which when executed by a processor implements the steps of the product recommendation method as described above.
The invention provides a product recommendation method, a device, equipment and a computer readable storage medium, wherein when a product recommendation request is received by product recommendation equipment in the embodiment of the invention, a target user corresponding to the product recommendation request is determined; adjusting a first actual score of the target user in a preset score sample set to obtain an adjusted score of the target user so as to eliminate noise data in the preset score sample set; the product recommendation equipment brings the adjusted scores into a preset similarity formula, and the similarity between the target user and a reference user is calculated through the similarity formula, wherein the reference user is a user in the score sample set except the target user; according to the similarity, product recommendation is carried out on the target user, and the accuracy of product recommendation is improved by carrying out noise reduction processing on the data.
Drawings
FIG. 1 is a schematic diagram of an apparatus architecture of a hardware operating environment according to an embodiment of the present invention;
FIG. 2 is a flowchart illustrating a first embodiment of a product recommendation method according to the present invention;
fig. 3 is a functional block diagram of an embodiment of a product recommendation device according to the present invention.
The implementation, functional features and advantages of the objects of the present invention will be further explained with reference to the accompanying drawings.
Detailed Description
It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.
As shown in fig. 1, fig. 1 is a schematic device structure diagram of a hardware operating environment according to an embodiment of the present invention.
The product recommendation device according to the embodiment of the present invention may be a terminal or a server, and as shown in fig. 1, the product recommendation device may include: a processor 1001, such as a CPU, a network interface 1004, a user interface 1003, a memory 1005, a communication bus 1002. Wherein a communication bus 1002 is used to enable connective communication between these components. The user interface 1003 may include a Display screen (Display), an input unit such as a Keyboard (Keyboard), and the optional user interface 1003 may also include a standard wired interface, a wireless interface. The network interface 1004 may optionally include a standard wired interface, a wireless interface (e.g., WI-FI interface). The memory 1005 may be a high-speed RAM memory or a non-volatile memory (e.g., a magnetic disk memory). The memory 1005 may alternatively be a storage device separate from the processor 1001.
Those skilled in the art will appreciate that the configuration of the apparatus shown in fig. 1 is not intended to be limiting of the apparatus and may include more or fewer components than those shown, or some components may be combined, or a different arrangement of components.
As shown in fig. 1, a memory 1005, which is a kind of computer storage medium, may include therein an operating system, a network communication module, a user interface module, and a product recommendation program.
In the device shown in fig. 1, the network interface 1004 is mainly used for connecting to a backend server and performing data communication with the backend server; the user interface 1003 is mainly used for connecting a client (user side) and performing data communication with the client; and the processor 1001 may be configured to call the product recommendation program stored in the memory 1005 and perform the operations of the product recommendation method described below.
In the prior art, the preference degree of a user on a product is determined through a collaborative filtering algorithm, the similarity of the user and a neighbor cluster of the user are determined through algorithms such as the similarity, the evaluation of the user is predicted through the evaluation of the neighbor cluster on a certain product, and finally whether the user is recommended to the user is determined according to evaluation feedback. The user-based collaborative filtering algorithm ignores the influence of the user on the grade of the common evaluation item and the grade differentiation. Assuming that two users have many items that are commonly rated, but neither user scores the items high, this can only to some extent indicate that neither user likes the item, which is noisy data for the recommendation algorithm; in addition, conventional algorithms do not distinguish effectively if there is a large variance in the user's scores for a common assessment item.
For example, in practical applications, there are generally three situations:
1. the user A scores 2 points for a certain type of movies, and the user B scores 5 points for the same type of movies;
2. the user A scores 2 points for a certain type of movies, and the user B scores 2 points for the same type of movies;
3. user a scores 5 for a certain category of movies and user B scores 5 for the same category of movies.
In the conventional algorithm, the similarity obtained by the user a and the user B in the above three cases is the same, which is obviously not logical, and the actual situation should be that, for the case 1, the preference of a and B is obviously different, for the case 2, it can only be judged that both a and B do not like such a movie, and for the case 3, it can be predicted that both a and B like such a movie. The traditional collaborative filtering recommendation algorithm based on the users cannot perform noise reduction processing on data samples, and noise data can influence the similarity between the users and have great influence on the final recommendation effect.
The invention provides an improved collaborative filtering algorithm based on user similarity, which combines the difference between scores of the same products by a user and the preference degree of the same products by the user to perform noise reduction processing on data to form a more reliable user neighbor cluster, thereby finally improving the scoring precision of the user and promoting the recommendation quality. Specifically, the method comprises the following steps:
the embodiment of the product recommendation method is provided based on the hardware structure.
Referring to fig. 2, fig. 2 is a schematic flowchart of a first embodiment of a product recommendation method of the present invention, where the method includes:
step S10, when a product recommendation request is received, determining a target user corresponding to the product recommendation request;
the product recommendation method in this embodiment is applied to a product recommendation device, the product recommendation device receives a product recommendation request, and a triggering manner of the product recommendation request is not specifically limited, that is, the product recommendation request may be actively triggered, for example, a client clicks a "service query" on a terminal to actively trigger the product recommendation request; in addition, the product recommendation request may also be automatically triggered, for example, when a preset product in the product recommendation device is updated, the product recommendation request is automatically triggered, and when the product recommendation device detects the product update, the product recommendation request is automatically triggered.
When the product recommendation device receives a product recommendation request, the product recommendation device obtains a target user corresponding to the product recommendation request, where the target user refers to a target user identifier for recommending a product, the number of the target users is not specifically limited, and in this embodiment, a target user is taken as an example for explanation.
Step S20, adjusting the first actual score of the target user in a preset score sample set, and obtaining an adjusted score of the target user.
In this embodiment, a preset scoring sample set is provided in the product recommendation device, where the preset scoring sample set is preset historical scoring data for product recommendation, and the preset scoring sample set is a user product scoring matrix as shown in table 1:
| user/product | Product i | Product j | Product n |
| User 1 | R11 | R1j | R1n |
| User 2 | Ri1 | Rij | Rin |
| User m | Rm1 | Rmj | Rmn |
TABLE 1
The product recommendation device adjusts the first actual score of the target user in the preset score sample set to obtain the adjusted score of the target user, and specifically includes:
step a1, extracting a first actual score and a second actual score of the target user and the reference user for the target product in a preset score sample set;
step a2, counting the first actual scores to calculate the first average scores of all the products participating in the scoring by the target user, counting the second actual scores to calculate the second average scores of all the products participating in the scoring by the reference user;
step a3, obtaining a maximum product score of the target product, and taking the first average score, the second average score and the maximum product score as independent variables of a preset adjustment formula to obtain an adjustment score of the target user, wherein the preset adjustment formula is as follows:
In the embodiment, the product recommendation equipment extracts a first actual score of a target user for a target product in a preset score sample set and a second actual score of a reference user for the target product; the product recommending device counts each first actual score, calculates a first average score of the target user on all products participating in the scoring, and counts each second actual score to calculate a second average score of the reference user on all products participating in the scoring; the product recommendation device obtains a maximum product score of a target product, and the product recommendation device takes the first average score, the second average score and the maximum product score as independent variables of a preset adjustment formula to obtain an adjustment score of a target user, wherein the preset adjustment formula is as follows:
That is, since the scoring criteria of each user are different, the scoring criteria of the user can be reflected to some extent by the average scoring of all products by the user, in this embodiment, by
To embody the target user u1And reference user u2Average score of (3). Then the preference of each user is different, and the differentiation degree of the common scoring products can be passed
To be embodied.
In addition, in the embodiment, the scoring of the user on the product is also related to the scoring standard of the user, and the user u is a target user according to the scoring standard1And reference user u2Average score of (3)
While taking into account the maximum score RmaxIn comparing the target users u1And reference user u2When the degree of the difference of the preference of the target product i is reached, the target user u can be selected1First actual score for target product i
Is optimized to obtainTo adjust the score
And calculating the similarity according to the adjusted scores after the optimization processing. Specifically, the method comprises the following steps:
and step S30, substituting the adjusted scores into a preset similarity formula, and calculating the similarity between the target user and a reference user through the similarity formula, wherein the reference user is a user in the score sample set except the target user.
The similarity formula preset in the product recommendation device in this embodiment is as follows:
the Sim (u)1,u2) ' representing target user u1And reference user u2The similarity of (A) to (B), the
Representing a target user u1For a first actual score of product i, the
Representing reference users u2For a second actual score of product i, the
Representing a target user u1A first average score, said
Representing reference users u2A second average score of (a);
the above-mentioned
Representing a target user u1And reference user u2The preference degrees of the two are consistent;
the above-mentioned
Representing a target user u1And reference user u2The variance of the scores.
In the present embodiment, the covariance
The significance of (1): covariance is used to measure the degree of deviation of each dimension from the mean, and if the value of covariance is positive, it means that the two are positively correlated, if it is negative, it means that the two are negatively correlated, and if it is 0, it means that they are independent of each other.
Representing a target user u1Deviation of score from user u for target product i1Average score of (3)
To the extent that (a) is present,
representing user u2Scoring product i off user u2Average score of (3)
To the extent of (c).
Can be used for measuring user u1And user u2And (4) consistency of the change trend of the preference degree of the target product i.
Representation reflects for target user u1And reference user u2All common scoring items
Target user u1And reference user u2The degree of preference of (a) is consistent with the trend of change.
The optimization is carried out on the basis of the covariance of the product recommendation equipment,
can represent a target user u1And reference user u2The preference degree variation trend of the user score is consistent, but only the covariance is used to ignore the influence caused by the user score differentiation, so that the adjustment score is used in the embodiment, and the target product i which is jointly scored in the preset scoring sample set can be used
Represents the optimized covariance. Then pass through
Representing a target user u1And reference user u2Variance of scores, the purpose of dividing variance is to score user u1And user u2The variation trend of the preference degree is normalized.
And step S40, recommending the product for the target user according to the similarity.
The similarity of the product recommendation equipment is used for recommending products for target users, and specifically, the implementation mode is as follows: the product recommending device selects a reference user with the highest similarity as a standard user, acquires a product with the highest standard user score, which is not purchased by a target user, as a recommended product, and sends the recommended product to the target user; the implementation mode two is as follows: the product recommendation device sorts all reference users according to the similarity and the size, selects a plurality of reference users ranked in the front to form a neighbor cluster, and carries out product recommendation on target users according to products evaluated by the reference users in the neighbor cluster.
In this embodiment, when a product recommendation device receives a product recommendation request, a target user corresponding to the product recommendation request is determined; adjusting a first actual score of the target user in a preset score sample set to obtain an adjusted score of the target user so as to eliminate noise data in the preset score sample set; the product recommendation equipment brings the adjusted scores into a preset similarity formula, and the similarity between the target user and a reference user is calculated through the similarity formula, wherein the reference user is a user in the score sample set except the target user; according to the similarity, product recommendation is carried out on the target user, and the accuracy of product recommendation is improved by carrying out noise reduction processing on the data.
The technical scheme in this embodiment may be applied in a financial scenario, for example, when a user a applies for a loan, a bank B1 needs to perform credit rating on the user a, since the user a has not had any loan record in the bank in the past, the bank needs to analyze and process credit assessment information of the user a at other financial institutions in the past, the bank finds that there are many same borrowers for other financial institutions B2, B3 …, Bn and the bank B1 when collecting data, and the bank may calculate similarities between the institutions based on the credit ratings of the same borrowers by the institutions, predict the credit rating of the user a, and issue a product for the user.
Further, based on the first embodiment of the product recommendation method of the present invention, a second embodiment of the product recommendation method of the present invention is provided.
This embodiment is a refinement of step S40 in the first embodiment, and in this embodiment, includes:
and step S41, sorting the similarity from big to small, and acquiring a preset number of reference users with the similarity sorted in the front to form an adjacent cluster.
The product recommendation device sorts the similarity from large to small, and the product recommendation device acquires a preset number of reference users with the similarity ranked in the front to form an adjacent cluster.
And step S42, determining a recommended product according to the reference user in the adjacent cluster, and pushing the recommended product to the target user.
The product recommendation device determines a recommended product according to the user scores in the adjacent clusters, and pushes the recommended product to the target user, and specifically includes:
b1, acquiring a reference user in the adjacent cluster, inputting the similarity between the reference user and the target user into a preset scoring formula, and acquiring the product prediction score of the target user on the product;
step b2, when the prediction score is higher than a preset threshold value, the product is used as a recommended product and pushed to the target user;
wherein, the preset scoring formula is as follows:
That is, the product recommendation device determines Top-n neighbor cluster U of target user U { U ═ U } according to the similarity ranking from large to small1,u2.....unTherein of
Here, n may be 1 to len (U)u)(len(Uu) Representing the size of a set of all users having a common score product with the target user u), the final range of n needs to be determined by comparing the prediction error over multiple training samples.
The score of a Top-n neighbor cluster generated by the product recommendation equipment on a certain product j to be recommended is predicted for a user u1For the productThe score of (2):
In the embodiment, the neighbor cluster is accurately constructed, so that the target user is recommended according to the grade of the reference user in the neighbor cluster for the product, and the target user is recommended more accurately.
Further, based on the above embodiment of the product recommendation method of the present invention, a third embodiment of the product recommendation method of the present invention is provided.
This embodiment may be combined with any other embodiment, where this embodiment is a step after step S40 in the first embodiment, and the difference between this embodiment and the foregoing embodiment is that:
outputting a product evaluation prompt when a purchase request based on a recommended product is detected;
and acquiring a product actual score input based on the product evaluation prompt, and adding the product actual score as a first product score of the target user to a preset score sample set.
In this embodiment, when the product recommendation device detects a purchase request based on a recommended product, the product recommendation device outputs a product evaluation prompt; the product recommending device obtains a product actual score input based on a product evaluation prompt, and the product recommending device adds the product actual score as a first product score of a target user to a preset score sample set.
In this embodiment, the product recommendation device updates the preset scoring sample set in real time, so that the information in the preset scoring sample set is more accurate, and the similarity calculated according to the preset scoring sample set is more accurate, thereby further improving the product recommendation precision.
Further, based on the above embodiment of the product recommendation method of the present invention, a fourth embodiment of the product recommendation method of the present invention is provided.
This embodiment is a step after the third embodiment, and the present embodiment is different from the above embodiments in that:
when the absolute value of the difference value between the actual product score and the predicted product score is larger than a preset difference value, marking the preset adjusting formula;
and when the marking frequency of the preset adjusting formula is greater than the preset frequency, outputting a formula adjusting prompt.
The product recommending equipment compares the actual product score with the predicted product score to determine the output values of the actual product score and the predicted product score; judging whether the absolute value of the difference between the actual product score and the predicted product score is greater than a preset difference, wherein the preset difference can be flexibly set according to a specific scene, for example, the preset difference is set to be 2 points, and if the absolute value of the difference between the actual product score and the predicted product score is less than or equal to the preset difference, determining that the product score is accurately estimated and recommending the product for a target user; and if the absolute value of the difference between the actual product score and the predicted product score is greater than a preset difference, marking a preset adjustment formula.
The product recommendation device counts the marking frequency of a preset adjustment formula in real time, and outputs a formula adjustment prompt when the marking frequency of the preset adjustment formula is greater than the preset frequency. In this embodiment, when the product recommendation is inaccurate, a formula adjustment prompt is output, so that the user can reversely adjust the preset adjustment formula, thereby further improving the product recommendation accuracy.
Referring to fig. 3, an embodiment of the present invention further provides a product recommendation device, where the product recommendation device includes:
the request receiving module 10 is configured to determine, when a product recommendation request is received, a target user corresponding to the product recommendation request;
the score adjusting module 20 is configured to adjust a first actual score of the target user in a preset score sample set, so as to obtain an adjusted score of the target user;
a similarity calculation module 30, configured to bring the adjusted score into a preset similarity formula, and calculate a similarity between the target user and a reference user according to the similarity formula, where the reference user is a user in the score sample set except the target user;
and the product recommending module 40 is used for recommending products for the target user according to the similarity.
In one embodiment, the score adjusting module 20 includes:
the score extraction unit is used for extracting a first actual score and a second actual score of the target user and the reference user for the target product in a preset score sample set;
the average technology unit is used for counting the first actual scores, calculating first average scores of all products participating in scoring by the target user, counting the second actual scores, and calculating second average scores of all products participating in scoring by the reference user;
the adjustment scoring unit is configured to obtain a maximum product score of the target product, and obtain an adjustment score of the target user by using the first average score, the second average score, and the maximum product score as independent variables of a preset adjustment formula, where the preset adjustment formula is:
the above-mentioned
An adjusted score representing the target user, RmaxRepresenting for the target productMaximum product score of, said
Representing a target user u1For a first actual score of the target product, the
Representing a target user u1A first average score of all products involved in the scoring, said
Representing reference users u2A second average score is given to all products participating in the scoring.
In an embodiment, the preset similarity formula in the similarity calculation module is:
the Sim (u)1,u2) ' representing target user u1And reference user u2The similarity of (A) to (B), the
Representing a target user u1For a first actual score of the target product, the
Representing reference users u2For a second actual score of the target product, the
Representing a target user u1A first average score, said
Representing reference users u2A second average score of (a);
the above-mentioned
Representing a target user u1And reference user u2The preference degrees of the two are consistent;
the adjustment score is
The above-mentioned
Representing a target user u1And reference user u2The variance of the scores.
In one embodiment, the product recommendation module 40 includes:
the group determination submodule is used for sequencing the similarity from large to small and acquiring a preset number of reference users with the similarity sequenced in the front to form an adjacent cluster;
and the product pushing submodule is used for determining a recommended product according to the reference user in the adjacent cluster and pushing the recommended product to the target user.
In one embodiment, the product push sub-module includes:
the information input unit is used for acquiring a reference user in the adjacent cluster, and inputting the similarity between the reference user and the target user into a preset scoring formula to obtain a product prediction score of the target user for a product;
the product pushing unit is used for taking the product as a recommended product and pushing the recommended product to the target user when the prediction score is higher than a preset threshold value;
wherein, the preset scoring formula is as follows:
In one embodiment, the product recommendation device includes:
the evaluation prompt module is used for outputting a product evaluation prompt when a purchase request based on a recommended product is detected;
and the score updating module is used for acquiring the actual product score input based on the product evaluation prompt, and adding the actual product score as the first product score of the target user to a preset score sample set.
In one embodiment, the product recommendation device includes:
the formula marking module is used for marking the preset adjusting formula when the absolute value of the difference value between the actual product score and the predicted product score is larger than a preset difference value;
and the adjustment prompting module is used for outputting a formula adjustment prompt when the marking frequency of the preset adjustment formula is greater than the preset frequency.
The method executed by each program module can refer to each embodiment of the product recommendation method of the present invention, and is not described herein again.
The invention also provides a computer readable storage medium.
The computer-readable storage medium of the present invention has stored thereon a product recommendation program which, when executed by a processor, implements the steps of the product recommendation method as described above.
The method implemented when the product recommendation program running on the processor is executed may refer to each embodiment of the product recommendation method of the present invention, and details are not described herein.
It should be noted that, in this document, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or system 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 system. Without further limitation, an element defined by the phrase "comprising an … …" does not exclude the presence of other like elements in a process, method, article, or system that comprises the element.
The above-mentioned serial numbers of the embodiments of the present invention are merely for description and do not represent the merits of the embodiments.
Through the above description of the embodiments, those skilled in the art will clearly understand that the method of the above embodiments can be implemented by software plus a necessary general hardware platform, and certainly can also be implemented by hardware, but in many cases, the former is a better implementation manner. Based on such understanding, the technical solution of the present invention may be embodied in the form of a software product, which is stored in a storage medium (e.g., ROM/RAM, magnetic disk, optical disk) as described above and includes instructions for enabling a terminal device (e.g., a mobile phone, a computer, a server, an air conditioner, or a network device) to execute the method according to the embodiments of the present invention.
The above description is only a preferred embodiment of the present invention, and not intended to limit the scope of the present invention, and all modifications of equivalent structures and equivalent processes, which are made by using the contents of the present specification and the accompanying drawings, or directly or indirectly applied to other related technical fields, are included in the scope of the present invention.
Claims (10)
1. A product recommendation method, characterized in that the product recommendation method comprises the steps of:
when a product recommendation request is received, determining a target user corresponding to the product recommendation request;
adjusting a first actual score of the target user in a preset score sample set to obtain an adjusted score of the target user;
bringing the adjusted scores into a preset similarity formula, and calculating the similarity between the target user and a reference user through the similarity formula, wherein the reference user is a user in the score sample set except the target user;
and recommending products for the target user according to the similarity.
2. The product recommendation method of claim 1, wherein the step of adjusting the first actual score of the target user in a preset score sample set to obtain the adjusted score of the target user comprises:
extracting a first actual score and a second actual score of the target user and the reference user for the target product in a preset score sample set;
counting the first actual scores to calculate a first average score of the target user on all products participating in scoring, and counting the second actual scores to calculate a second average score of the reference user on all products participating in scoring;
obtaining a maximum product score of the target product, and taking the first average score, the second average score and the maximum product score as independent variables of a preset adjusting formula to obtain an adjusting score of the target user, wherein the preset adjusting formula is as follows:
the above-mentioned
An adjusted score representing the target user, RmaxRepresenting a maximum product score for the target product, the
Representing a target user u1For a first actual score of the target product, the
Representing a target user u1A first average score of all products involved in the scoring, said
Representing reference users u2A second average score is given to all products participating in the scoring.
3. The product recommendation method of claim 1, wherein the preset similarity formula is:
the Sim (u)1,u2) ' representing target user u1And reference user u2The similarity of (A) to (B), the
Representing a target user u1For a first actual score of the target product, the
Representing reference users u2For a second actual score of the target product, the
Representing a target user u1A first average score, said
Representing reference users u2A second average score of (a);
the above-mentioned
Representing a target user u1And reference user u2The preference degrees of the two are consistent;
the adjustment score is
The above-mentioned
Representing a target user u1And reference user u2The variance of the scores.
4. The product recommendation method of claim 1, wherein the step of recommending products for the target user based on the similarity comprises:
sorting the similarity from big to small, and obtaining a preset number of reference users with the similarity sorted in the front to form an adjacent cluster;
and determining a recommended product according to the reference user in the adjacent cluster, and pushing the recommended product to the target user.
5. The product recommendation method of claim 4, wherein said step of determining recommended products according to reference users in said neighboring cluster, and pushing said recommended products to said target users comprises:
acquiring a reference user in the adjacent cluster, inputting the similarity between the reference user and the target user into a preset scoring formula, and obtaining a product prediction score of the target user on a product;
when the prediction score is higher than a preset threshold value, the product is used as a recommended product and pushed to the target user;
wherein, the preset scoring formula is as follows:
6. The product recommendation method of claim 5, wherein the step of recommending products for the target user based on the similarity comprises:
outputting a product evaluation prompt when a purchase request based on a recommended product is detected;
and acquiring a product actual score input based on the product evaluation prompt, and adding the product actual score as a first product score of the target user to a preset score sample set.
7. The product recommendation method of claim 6, wherein the step of obtaining the actual product score based on the product rating prompt input and adding the actual product score as the first product score of the target user to a preset score sample set comprises:
when the absolute value of the difference value between the actual product score and the predicted product score is larger than a preset difference value, marking the preset adjusting formula;
and when the marking frequency of the preset adjusting formula is greater than the preset frequency, outputting a formula adjusting prompt.
8. A product recommendation device, characterized in that the product recommendation device comprises:
the request receiving module is used for determining a target user corresponding to a product recommendation request when the product recommendation request is received;
the score adjusting module is used for adjusting a first actual score of the target user in a preset score sample set to obtain an adjusted score of the target user;
the similarity calculation module is used for substituting the adjusted scores into a preset similarity formula and calculating the similarity between the target user and a reference user according to the similarity formula, wherein the reference user is a user except the target user in the score sample set;
and the product recommendation module is used for recommending products for the target user according to the similarity.
9. A product recommendation device, characterized in that the product recommendation device comprises: memory, processor and a product recommendation program stored on the memory and executable on the processor, the product recommendation program when executed by the processor implementing the steps of the product recommendation method of any of claims 1 to 7.
10. A computer-readable storage medium, characterized in that a product recommendation program is stored thereon, which when executed by a processor implements the steps of the product recommendation method according to any one of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911145185.4A CN110910215A (en) | 2019-11-20 | 2019-11-20 | Product recommendation method, device, equipment and computer-readable storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911145185.4A CN110910215A (en) | 2019-11-20 | 2019-11-20 | Product recommendation method, device, equipment and computer-readable storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN110910215A true CN110910215A (en) | 2020-03-24 |
Family
ID=69818455
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201911145185.4A Pending CN110910215A (en) | 2019-11-20 | 2019-11-20 | Product recommendation method, device, equipment and computer-readable storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110910215A (en) |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111414548A (en) * | 2020-05-09 | 2020-07-14 | 中国工商银行股份有限公司 | Object recommendation method and device, electronic equipment and medium |
| CN112100489A (en) * | 2020-08-27 | 2020-12-18 | 北京百度网讯科技有限公司 | Object recommendation method, device and computer storage medium |
| CN112348532A (en) * | 2020-10-19 | 2021-02-09 | 前海飞算科技(深圳)有限公司 | Optimal recommendation information processing method and device, storage medium and electronic equipment |
| CN113744030A (en) * | 2021-09-08 | 2021-12-03 | 未鲲(上海)科技服务有限公司 | Recommendation method, device, server and medium based on AI user portrait |
| CN117763492A (en) * | 2023-12-04 | 2024-03-26 | 淮阴工学院 | Network security tool intelligent recommendation method and device based on time sequence spatial characteristics and preference fluctuation |
-
2019
- 2019-11-20 CN CN201911145185.4A patent/CN110910215A/en active Pending
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111414548A (en) * | 2020-05-09 | 2020-07-14 | 中国工商银行股份有限公司 | Object recommendation method and device, electronic equipment and medium |
| CN111414548B (en) * | 2020-05-09 | 2023-08-01 | 中国工商银行股份有限公司 | Object recommendation method, device, electronic equipment and medium |
| CN112100489A (en) * | 2020-08-27 | 2020-12-18 | 北京百度网讯科技有限公司 | Object recommendation method, device and computer storage medium |
| CN112100489B (en) * | 2020-08-27 | 2022-07-15 | 北京百度网讯科技有限公司 | Object recommendation method, device and computer storage medium |
| CN112348532A (en) * | 2020-10-19 | 2021-02-09 | 前海飞算科技(深圳)有限公司 | Optimal recommendation information processing method and device, storage medium and electronic equipment |
| CN113744030A (en) * | 2021-09-08 | 2021-12-03 | 未鲲(上海)科技服务有限公司 | Recommendation method, device, server and medium based on AI user portrait |
| CN117763492A (en) * | 2023-12-04 | 2024-03-26 | 淮阴工学院 | Network security tool intelligent recommendation method and device based on time sequence spatial characteristics and preference fluctuation |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110910215A (en) | Product recommendation method, device, equipment and computer-readable storage medium | |
| CN105022761B (en) | Group searching method and device | |
| WO2020048084A1 (en) | Resource recommendation method and apparatus, computer device, and computer-readable storage medium | |
| CN107622072B (en) | Identification method for webpage operation behavior, server and terminal | |
| US20160239865A1 (en) | Method and device for advertisement classification | |
| WO2015120798A1 (en) | Method for processing network media information and related system | |
| US10387805B2 (en) | System and method for ranking news feeds | |
| US20200210459A1 (en) | Method and apparatus for classifying samples | |
| CN109922379B (en) | Advertisement video optimization method, device and equipment and computer readable storage medium | |
| CN109271592B (en) | Article recommendation method, device, storage medium and program product | |
| WO2015062359A1 (en) | Method and device for advertisement classification, server and storage medium | |
| US11734312B2 (en) | Feature transformation and missing values | |
| CN108647985B (en) | Article recommendation method and device | |
| CN111444930B (en) | Method and device for determining prediction effect of two-classification model | |
| CN110008396B (en) | Object information pushing method, device, equipment and computer readable storage medium | |
| CN111325228B (en) | Model training method and device | |
| CN108305174B (en) | Resource processing method, device, storage medium and computer equipment | |
| KR20170030401A (en) | User Group Clustering Method On the Basis of Analysis for Customer Experience Data | |
| US11847142B2 (en) | Document analysis system | |
| CN113886628A (en) | Image retrieval method, device and storage medium | |
| CN114201687A (en) | Information recommendation method and device, electronic equipment and storage medium | |
| CN114912031A (en) | Mixed recommendation method and system based on clustering and collaborative filtering | |
| CN110825971B (en) | Article cold start recommendation algorithm integrating relationship mining and collaborative filtering | |
| CN112085526A (en) | Information matching method and device based on user group, computer equipment and storage medium | |
| CN113220741A (en) | Internet advertisement false flow identification method, system, equipment and 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 |