CN113034241B - Product information recommendation method and computer equipment - Google Patents
Product information recommendation method and computer equipment Download PDFInfo
- Publication number
- CN113034241B CN113034241B CN202110440594.8A CN202110440594A CN113034241B CN 113034241 B CN113034241 B CN 113034241B CN 202110440594 A CN202110440594 A CN 202110440594A CN 113034241 B CN113034241 B CN 113034241B
- Authority
- CN
- China
- Prior art keywords
- target account
- data
- account
- similarity
- behavior data
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 47
- 230000006399 behavior Effects 0.000 claims description 177
- 239000013598 vector Substances 0.000 claims description 58
- 238000006243 chemical reaction Methods 0.000 claims description 15
- 238000004590 computer program Methods 0.000 claims description 15
- 238000001816 cooling Methods 0.000 claims description 8
- 238000004458 analytical method Methods 0.000 claims description 6
- 230000003542 behavioural effect Effects 0.000 claims description 5
- 230000008569 process Effects 0.000 abstract description 11
- 238000004364 calculation method Methods 0.000 abstract description 6
- 230000004927 fusion Effects 0.000 abstract description 2
- 238000005065 mining Methods 0.000 description 7
- 238000005259 measurement Methods 0.000 description 6
- 238000010586 diagram Methods 0.000 description 5
- 230000006870 function Effects 0.000 description 4
- 238000013480 data collection Methods 0.000 description 2
- 230000001419 dependent effect Effects 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 230000005484 gravity Effects 0.000 description 2
- 238000012545 processing Methods 0.000 description 2
- 238000005057 refrigeration Methods 0.000 description 2
- 230000002238 attenuated effect Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 238000002156 mixing Methods 0.000 description 1
- 239000000203 mixture Substances 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012216 screening Methods 0.000 description 1
- 238000012163 sequencing technique Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
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
- G06Q30/0629—Directed, with specific intent or strategy for generating comparisons
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 application relates to the technical field of big data, and provides a product information recommending method, a product information recommending device, computer equipment and a computer readable storage medium. According to the product information recommendation method, when the behavior data of the target account does not meet the preset condition, the portrait data of the target account is obtained, N similarity values between the target account and N sample accounts are calculated based on the behavior data and the portrait data of the target account, wherein N is an integer greater than 1, the similarity calculation between the target account and the N sample accounts is achieved through fusion of the portrait data when the behavior data of the target account is small, and then the product information is pushed to the target account based on the sorting result of the N similarity values, so that the phenomenon that the behavior data of the target account is excessively relied on in the process of pushing the product information to the target account is avoided, and the application range of a product information recommendation scheme is expanded.
Description
Technical Field
The present invention relates to the field of big data technologies, and in particular, to a product information recommendation method, a product information recommendation device, a computer device, and a computer readable storage medium.
Background
At present, with the continuous development of internet technology, users can utilize different application programs corresponding to different shopping platforms to make online shopping by installing the application programs of different shopping platforms on mobile phones. When the user selects and purchases the product by using the application program, the browsing data and the consumption data of the user are stored in the corresponding account in an electronic form, so that the browsing data and the consumption data of the account can be collected conveniently, and an implementation basis is provided for product demand mining and product information recommendation of the user corresponding to the account.
When the existing product information recommendation is carried out for the user, the requirement mining is mainly carried out according to the existing data, and then the mined potential product requirement is recommended to the account in a product information mode. However, in the existing product information recommendation scheme, specifically, after a user logs in on an application program by using a target account, behavior data of a browsing product and a consumer product is based, corresponding candidate product information is predicted and matched, and then the candidate product information is pushed to the target account for display. However, when the number of times that the user logs in the application program using the target account is small, or the time interval between two uses of the application program is long, the collected behavior data corresponding to the target account will show large discreteness, so that the requirement mining cannot be performed based on the behavior data of the target account, and the product information recommendation cannot be performed to the target account. Therefore, the existing product information recommendation scheme has the problem of smaller application range.
Disclosure of Invention
In view of the above, the embodiments of the present application provide a product information recommendation method, a product information recommendation device, a computer device, and a computer readable storage medium, so as to solve the problem that the application range of the existing product information recommendation scheme is smaller.
A first aspect of an embodiment of the present application provides a product information recommendation method, including:
If the behavior data of the target account does not meet the preset condition, obtaining the portrait data of the target account;
based on the behavior data and the portrait data, measuring and calculating the similarity between the target account and N sample accounts respectively to obtain N similarity values; wherein N is an integer greater than 1;
And pushing product information to the target account based on the sorting result of the N similarity values.
A second aspect of an embodiment of the present application provides a product information recommendation apparatus, including:
the system comprises an acquisition unit, a storage unit and a storage unit, wherein the acquisition unit is used for acquiring portrait data of a target account if the behavior data of the target account does not meet preset conditions;
The first execution unit is used for measuring and calculating the similarity between the target account and N sample accounts respectively based on the behavior data and the portrait data to obtain N similarity values; wherein N is an integer greater than 1;
and the second execution unit is used for pushing the product information to the target account based on the sorting result of the N similarity values.
A third aspect of the embodiments of the present application provides a computer device comprising a memory, a processor and a computer program stored in the memory and executable on the processor, the processor implementing the steps of the first aspect when executing the computer program.
A fourth aspect of the embodiments of the present application provides a computer readable storage medium storing a computer program which, when executed by a processor, implements the steps of the first aspect.
The product information recommending method, the product information recommending device, the computer equipment and the computer readable storage medium provided by the embodiment of the application have the following beneficial effects:
According to the embodiment of the application, when the behavior data of the target account does not meet the preset condition, the portrait data of the target account is obtained, and then N similarity values between the target account and N sample accounts are calculated based on the behavior data and the portrait data of the target account, wherein N is an integer greater than 1, so that when the behavior data of the target account is smaller, the similarity calculation between the target account and the N sample accounts is realized by fusing the portrait data, and then the product information is pushed to the target account based on the sorting result of the N similarity values, the phenomenon that the behavior data of the target account is excessively dependent in the process of pushing the product information to the target account is avoided, the flexibility of recommending a product information scheme is improved, and the application range of the product information recommending scheme is enlarged.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings that are needed in the embodiments or the description of the prior art will be briefly described below, it being obvious that the drawings in the following description are only some embodiments of the present application, and that other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.
Fig. 1 is a flowchart of an implementation of a product information recommendation method according to an embodiment of the present application;
FIG. 2 is a flow chart of one implementation of step S11 in an embodiment of the present application;
FIG. 3 is a flowchart of another implementation of step S11 in an embodiment of the present application;
FIG. 4 is a block diagram of a product information recommendation device according to an embodiment of the present application;
Fig. 5 is a block diagram of a computer device according to an embodiment of the present application.
Detailed Description
The present application will be described in further detail with reference to the drawings and examples, in order to make the objects, technical solutions and advantages of the present application more apparent. It should be understood that the specific embodiments described herein are for purposes of illustration only and are not intended to limit the scope of the application.
The execution main body of the product information recommendation method provided in the embodiment is a server, and specifically may be a server configured with the function of the method, or any server in a server cluster. Here, the server cluster may be a server cluster composed of a plurality of servers, and a distributed system is constructed based on the server cluster, so that data sharing or data synchronization between the plurality of servers in the server cluster can be achieved. On the basis, a target script file is configured to any server in the server cluster, and the target script file describes the product information recommending method provided by the embodiment, so that the server configured with the target script file can execute the target script file and further execute each step in the product information recommending method.
When the method is realized, the server or any server in the server cluster receives a data access request sent by the target account through the terminal, returns a corresponding page file according to the data access request, loads the page file by the terminal, and further displays product information recommended by the server for the target account. The server can determine a database corresponding to the target account according to the target account identifier carried in the data access request, and acquire behavior data of the target account from the database. If the behavior data of the target account does not meet the preset condition, the portrait data of the target account is obtained, then the similarity between the target account and N sample accounts is calculated based on the behavior data and the portrait data, and the product information is pushed to the target account based on the sorting result of the N similarity values. When the behavior data of the target account is small, the similarity value measurement and calculation between the target account and the N sample accounts are realized through fusion of the portrait data, and the product information is pushed to the target account based on the sorting results of the N similarity values obtained through the similarity value measurement and calculation, so that the situation that the behavior data of the target account is excessively depended in the process of pushing the product information to the target account is avoided, the flexibility of a recommended product information scheme is improved, and the application range of the product information recommendation scheme is enlarged.
For example, taking the server as the server of the shopping platform application program as an example, enabling the shopping platform application program on a terminal by a user, logging in a target account on the shopping platform application program, sending a data access request to the server by the terminal through the shopping platform application program logging in the target account, determining a corresponding database of the target account by the server according to a target account identifier carried in the data access request, acquiring behavior data of the target account from the database, if the behavior data of the target account does not meet a preset condition, acquiring portrait data of the target account by the server, measuring and calculating similarity values between the target account and N sample accounts based on the behavior data and the portrait data, and pushing product information to the target account based on the sequencing result of the N similarity values. The method and the device have the advantages that when the behavior data of the target account are small, the similarity value measurement and calculation between the target account and the sample account can be realized by fusing the behavior data and the portrait data, the phenomenon that the behavior data of the target account is excessively depended on in the process of pushing the product information to the target account is avoided, the flexibility of recommending the product information scheme is improved, and the application range of the product information recommendation scheme is widened.
The following describes in detail a product information recommendation method provided in this embodiment through a specific implementation manner.
Fig. 1 shows a flowchart of an implementation of a product information recommendation method according to an embodiment of the present application, which is described in detail below:
S11: and if the behavior data of the target account does not meet the preset condition, acquiring the portrait data of the target account.
In step S11, the target account is used as a unique identifier for logging in a related application program or data presentation page on the terminal. The user can access the data resources in the corresponding server through the application program or the data presentation page by utilizing the terminal. Here, the application is an application installed on the terminal that can be used to present product information; when a user accesses the server through the data display page by using the terminal, the page file returned by the server can be loaded and displayed through the data display page, so that the display operation of the product information is realized.
Taking the example that a user uses a target account to perform login operation on an application program of a terminal, when the user starts the application program by using the terminal, the user can send a data access request to a server through the application program by using the identity of a tourist, and can also send the data access request to the server by using the identity of a member or a common user by registering the account and logging in the account on the application program. The server is used for returning the page file to the terminal according to the data access request, and the terminal loads the page file through the application program so as to display the corresponding operable page. The user uses the terminal to operate on the operable page, the operation behavior of the user is collected by the embedded point configured in the display page, corresponding behavior data is obtained, and then the behavior data is sent to the server through the application program.
In this embodiment, the behavior data includes product browsing data and/or product consumption data of the user corresponding to the target account on the application program. Here, the product browsing data may include, but is not limited to, clicking on a picture of the product, browsing detailed information of the product, saving a picture of the product, or a screenshot operation on a product detail page, etc.; product consumption data may include, but is not limited to, attributes of products purchased, number or times of products purchased, and the like. The preset condition is used for characterizing the characteristic index of the behavior data capable of carrying out target account preference analysis, and can include, but is not limited to, a quantity threshold value of the behavior data, a category number threshold value of the behavior data and the like. Correspondingly, if the number of the behavior data of the target account or the number of the types of the behavior data is equal to or greater than a preset threshold, user preference analysis can be directly carried out on the target account according to the behavior data of the target account, and product information recommendation is further achieved according to an analysis result.
When the method is realized, the behavior data of the user can be collected and configured by conducting behavior data collection and configuration on the application program, for example, configuring corresponding page buried points or configuring an execution environment script. In the process of browsing or consuming the product by the user in actual application, the application program can correspondingly collect and store the behavior data and the operation time of the user according to the actual operation condition of the user, so that the obtained behavior data can be used for describing the behavior content of the user corresponding to the target account when the application program is used.
In practical application, because the user is likely to have lower practicability due to misoperation or less operation when actually using the application program, in order to avoid the problem of lower recommending operation efficiency caused by directly recommending product information according to the behavior data of the target account, the scheme of the embodiment also blends portrait data of the target account, namely, further performs demand mining or preference mining on the user corresponding to the target account in a manner of blending portrait data.
In this embodiment, by configuring a preset condition to characterize a feature index of behavior data capable of performing account preference analysis, and taking the preset condition as a determination mechanism for obtaining portrait data of a target account, the portrait data of the target account can be obtained under the condition that the behavior data of the target account does not meet the feature index, so that a realization basis is provided by taking portrait data of the target account as a consideration factor in the process of realizing product information recommendation.
Fig. 2 shows a flowchart of an implementation of step S11 in this embodiment. As shown in fig. 2, as one embodiment, the preset condition includes a quantity threshold; the step S11 specifically includes: s111 to S112.
S111: and if the number of the behavior data of the target account does not meet the number threshold, carrying out the associated server query according to the target account.
S112: and if the association server is queried, acquiring the portrait data of the target account from the association server.
In this embodiment, the association server refers to a server that a user can log in an association application program using a target account and access data to the association application program through the association application program. The portrait data is used for describing user attribute information corresponding to the target account, such as the gender of the user, the age of the user, the income range of the user, the investment category of the user, the classification group label of the user and the like.
It should be noted that, when providing product related services such as product browsing or product consumption for the user, multiple applications may be developed according to different demands of the user, so that there is a business relationship correlation or association between the multiple applications.
In this embodiment, if the number of the behavior data of the target account does not meet the number threshold, it indicates that the operations such as product information browsing or product consumption performed by the user using the current application program are too rare, and the user's needs or preferences cannot be mined based on the behavior data. Therefore, the related server query is performed according to the target account to acquire more user data from the related server, and an implementation basis is provided for realizing the requirement mining or the preference mining of the user.
In the specific implementation process, the query of the association server is carried out according to the target account, all the association application programs of the current application program logged in with the target account can be determined first, then the server set is determined according to all the association application programs, and finally the association server is determined from the server set according to the target account. That is, the server determines candidate servers with which data interaction is possible, uses all candidate servers as a server set, and determines an associated server from the server set according to the target account. Here, the associated server is determined from the server set according to the target account, specifically, whether the candidate server in the server set stores the data access record and the portrait data of the target account is determined, if the candidate server in the server set stores the data access record of the target account, the candidate server is used by the user to access the candidate server, and the user attribute information is stored in the candidate server, so the candidate server is used as the associated server, and the associated server is determined to be queried.
It should be understood that the portrait data of the target account is obtained from the association server, where the portrait data of the target account may be portrait information that is filled in by a user corresponding to the target account and uploaded to the association server, or portrait data that is generated by the association server according to habit data of the user using the association application program when the user uses the association application program, using an existing user portrait generation tool, such as a portrait generation model or a portrait generation network, based on habit data of the user using the association application program. In some prior arts, in order to realize classification of user groups, portrait data is generated by using a portrait data generation model according to portrait information of users or habit data of users using application programs, which belongs to common technical means, so that a description of how to obtain portrait data in an associated server is not repeated here.
Fig. 3 shows another implementation flowchart of step S11 in the present embodiment. As shown in fig. 3, in addition to the embodiment shown in fig. 2, as another embodiment, step S111 is followed by step S113, which is parallel to step S112. Specifically:
S113: if no associated server is queried, displaying a data acquisition page; the data acquisition page is used for acquiring portrait data of the target account.
In this embodiment, the content in the data collection page includes a form for collecting user information, such as a user behavior questionnaire, a user consumption data questionnaire, and the like. Here, the data acquisition page for acquiring the portrait data is displayed so that the user fills out the form in the data acquisition page by himself, and the portrait data is acquired based on the filling situation of the user.
When the method is realized, a portrait data generation strategy can be pre-configured in a file of the data acquisition page, when no relevant server is queried, a user can complete filling of form information in the page by displaying the data acquisition page, and then the portrait data is generated by the data acquisition page based on the pre-configured portrait data generation strategy according to the information on the form. Here, a preconfigured image data generation policy describes a method of how image data is generated from form contents. For example, after the user fills out the form in the data acquisition page, based on the content in the form information, the corresponding user portrait data can be matched from the local database, that is, the user labels are matched from the local database according to the information on the form, and then portrait data is obtained by summarizing all the user labels.
It can be understood that the purpose of querying the associated server is to obtain the portrait data, and in order to obtain the portrait data even if the associated server is not queried, in this embodiment, the user fills out relevant information in the data acquisition page by displaying the data acquisition page for obtaining the portrait data by the user, and further generates the portrait data based on the relevant information filled out by the user through the data acquisition page, so that the use scenario and application range of the scheme of this embodiment are widened.
S12: based on the behavior data and the portrait data, measuring and calculating the similarity between the target account and N sample accounts respectively to obtain N similarity values; wherein N is an integer greater than 1.
In step S12, the sample account may be a real account or a virtual account. In the implementation, an account with higher activity can be selected from the existing account set according to the number of the existing accounts to be used as a sample account, and/or virtual accounts with different categories can be constructed to be used as sample accounts.
It should be noted that, the sample account refers to a pre-designated reference account, in practical application, N reference accounts may be designated as a plurality of sample accounts according to the needs of the account classification, that is, when N classifications of accounts are required, N reference accounts may be designated. By measuring and calculating the similarity between the target account and the accounts with different samples, the type of the target account which tends to be more can be determined, and a data basis is conveniently provided for product information recommendation of the target account.
In this embodiment, when the similarity between the target account and each sample account is measured, not only the behavior data between the target account and each sample account is considered, but also the portrait data between the target account and each sample account is considered, so that under the condition that the behavior data of the target account is less, the portrait data is used as a compensation part for similarity comparison, and is fused into the process of similarity comparison between the target account and each sample account, and the accuracy of the similarity measurement result between the target account and each sample account is further improved.
When the method is realized, the similarity value between the target account and the N sample accounts is calculated based on the behavior data and the portrait data, and the similarity value between the target account and each sample account can be calculated in the same mode. If the behavior similarity is calculated according to the behavior data of the target account and the behavior data of the sample account, the portrait similarity is calculated according to the portrait data of the target account and the portrait data of the sample account, and the behavior similarity and the portrait similarity are fused to obtain a similarity value between the target account and the sample account. Here, the similarity value between the target account and the sample account mentioned in this embodiment is used to describe the degree of similarity between the target account and the sample account, that is, when the similarity value is larger, the target account is more similar to the sample account, and when the similarity value is smaller, the target account is less similar to the sample account.
In all embodiments of the present application, the similarity value between the target account and the sample account is calculated to determine whether the user consumption habit corresponding to the sample account is suitable for being used as a reference for product information recommendation based on the similarity. Because the larger the similarity value is, the more similar the target account and the sample account are, so the sample account corresponds to the consumption habit of the user, can be used as a reference for pushing product information to the target account, and the product consumption data of the sample account can be used as a reference for recommending products by further determining the sample account with larger similarity to the target account, thereby providing a technical means and an implementation basis with higher rationalization degree for screening out the target product information and pushing the target account.
As an embodiment, step S12 may include:
correcting the behavior data by utilizing a Newton cooling method to obtain new behavior data;
And calculating the similarity between the target account and each sample account based on the new behavior data and the portrait data, so as to obtain N similarity values.
In the present embodiment, newton's cooling method is used to describe the correspondence between the heat decay and time of the behavior data.
It should be noted that, because the behavior data includes product browsing data and/or product consumption data of the user corresponding to the target account on the application program, the behavior data may also be used to characterize the attention heat of the user corresponding to the target account to some products. When the potential product demands of the users or the products favored by the users are mined according to the user behavior data, the product demands of the users or the products favored by the users can change according to time changes, for example, the demands or the favorites of the users on the refrigeration household appliances such as fans, air conditioners and the like can reach peak values in summer, and the demands or the favorites of the users on the refrigeration household appliances can be greatly attenuated after the time reaches autumn or even winter from summer. In consideration of the phenomenon that the attention heat of a user to certain products decays with the passage of time, the Newton cooling method is utilized to correct the behavior data, so that the obtained new behavior data has a reference value, and the scientificalness of similarity comparison is improved.
When the method is realized, the Newton cooling method is adopted to correct the behavior data, and the time attenuation is considered, so the Newton cooling method is adopted to correct the behavior data, namely, the attenuation degree of each group of behavior data or behavior samples in the behavior data is calculated, and the behavior data or behavior samples with serious attenuation degree in the behavior data are removed. For example, the heat attenuation measurement is performed on all the behavior data to the same degree, then the behavior data with serious heat attenuation degree is removed, and the behavior data with small heat attenuation degree is left as new behavior data.
As one possible implementation of this embodiment, newton's cooling method may be expressed by the following formula: Wherein H (t) is a heat value corresponding to the current time of the behavior data; h (t 0) is a heat value corresponding to the initial time of the behavior data; /(I) Is a heat decay coefficient, wherein e is a natural constant base, k is a decay rate constant, t is a current time, and t0 is an initial time. Here, the initial time is the behavior occurrence time corresponding to the behavior data, and the current time is the time for measuring and calculating the heat decay time, so that the current time is necessarily after the initial time, that is, the value of the current time is necessarily greater than the initial time. Assuming that after 30 days the score decays to half of the original, the calculated decay coefficients are as follows:
Wherein/>
In this embodiment, by correcting the behavior data, the accuracy of the similarity between the target account and the sample account can be improved, and it should be understood that, in practical application, if the current time is less apart from the initial time, the situation that the behavior data decays with time may not need to be considered. The embodiment provides a scheme for correcting the behavior data, because considering that the release time of the product directly affects the demand or preference heat of the user for the product in the actual product sales process, when recommending the product information to the target account, the latest released product information is more recommended instead of the former product information, thereby bringing more convenient experience to the user.
As an embodiment, in the above scheme, the steps are as follows: based on the new behavior data and the portrait data, calculating the similarity between the target account and each sample account respectively to obtain N similarity values, wherein the method comprises the following steps:
performing vector conversion on the new behavior data and the portrait data respectively to obtain a first behavior vector and a first portrait vector;
Calculating the similarity between the first behavior vector and the second behavior vector of each sample account to obtain N behavior similarities, and calculating the similarity between the first image vector and the second image vector of each sample account to obtain N image similarities;
And obtaining N similarity values based on the N behavior similarities and the N image similarities.
In this embodiment, the similarity measurement between the target account and the sample account takes into account both behavioral data and portrait data. That is, the similarity between the target account and the sample account in the behavior dimension and the similarity between the target account and the sample account in the portrait dimension are fused as the similarity value. Before the two dimension similarity is fused, vector conversion is needed to be carried out on the behavior data and the portrait data respectively, namely, the behavior data is converted into a behavior vector, the portrait data is converted into a portrait vector, the behavior similarity of the target account and the sample account in the behavior dimension is measured in a vector mode, and the portrait similarity of the target account and the sample account in the portrait dimension is measured.
In the implementation, the first behavior vector and the first image vector obtained by performing vector conversion on the behavior data and the image data respectively may be Spark dense vectors. Based on this, when a person skilled in the art performs vector conversion on new behavior data and performs vector conversion on image data, the person skilled in the art can learn and use the existing Spark dense vector conversion method to implement, so that detailed description of how to perform vector conversion on new behavior data and the image data respectively to obtain a first behavior vector and a first image vector is not repeated here.
It should be noted that, in this embodiment, the second behavior vector is obtained by performing a vector conversion according to the behavior data of the sample account, the second behavior vector is used for characterizing the behavior habit of the sample account, the second image vector is obtained by performing a vector conversion according to the portrait data of the sample account, and the second portrait vector is used for characterizing the portrait data of the sample account. Here, the specific conversion process of the second behavior vector and the second portrait vector may be the same as the conversion method of the first behavior vector and the first portrait vector. That is, an existing Spark dense vector conversion method may be used to perform vector conversion based on the behavior data of the sample account to obtain a second behavior vector, and perform vector conversion based on the portrait data of the sample account to obtain a second portrait vector, which is not described herein.
In this embodiment, the behavior similarity between the first behavior vector and the second behavior vector of the sample account is calculated, which may be a distance between the first behavior vector and the second behavior vector is calculated, and then the behavior similarity is calculated according to the distance between the two vectors; or directly measuring and calculating cosine similarity between the first behavior vector and the second behavior vector as the behavior similarity. Similar to the behavior similarity, the similarity between the first portrait vector and the second portrait vector of the sample account may be calculated by measuring the distance between the first portrait vector and the second portrait vector, and then calculating the similarity between the two portrait vectors according to the distance between the two vectors; or directly measuring and calculating cosine similarity between the first image vector and the second image vector as the image similarity.
As a possible implementation manner of this embodiment, in the foregoing solution, the steps include: based on the N behavior similarities and the N image similarities, N similarity values are obtained, including:
Based on the N behavior similarities and the N image similarities, calculating a similarity value between the target account and each sample account through the following formula;
Si＝K1×Pi+K2×Ai
wherein i is [2, N ]; s i is a similarity value between the target account and the ith sample account; p i is the portrait similarity between the target account and the ith sample account; a i is the behavioral similarity between the target account and the ith sample account; k 1 is a first preset weight parameter, K 2 is a second preset weight parameter, and K 1 and K 2 are both greater than 0.
In this embodiment, the behavior similarity is used to indicate the similarity degree between the target account and the sample account in the behavior dimension, and the portrait similarity is used to indicate the similarity degree between the target account and the sample account in the portrait dimension, that is, the similarity degree between the target account and the sample account is described by two different dimensions, but the behavior similarity and the portrait similarity are fused, and the fused numerical value is used as the similarity between the target account and the sample account, so that the consideration factors for measuring and calculating the similarity not only include the behavior data but also include the portrait data, the consideration factors of the similarity are richer, and the rationalization degree of the similarity is improved.
It should be understood that, since K 1 is a first preset weight parameter, K 2 is a second preset weight parameter, and K 1 and K 2 are both greater than 0, in this embodiment, since the behavior data cannot satisfy the preset condition, and thus the image data is introduced, K 2 is a second preset weight parameter, and represents a weight constant for adjusting the behavior similarity, the specific gravity of which may be relatively smaller, and K 1 is a first preset weight parameter, and represents a weight constant for adjusting the image similarity, and the specific gravity of which may be relatively larger, that is, K 1＞K2. In practical application, specific values of K 1 and K 2 can be further configured according to the side reconfiguration between the behavior data and the portrait data.
S13: and pushing product information to the target account based on the sorting result of the N similarity values.
In step S13, by sorting the N similarity values, the sample account most similar to the target account can be determined.
When the method is implemented, the sorting of the N similarity values can be from big to small, the product information is pushed to the target account based on the sorting result of the N similarity values, the sample account corresponding to a plurality of similarity values in the sorting result can be directly selected, and the product information is pushed to the target account based on the behavior data of the sample account. For example, for N similarity values arranged from large to small, if no parallel nouns exist, N sorting ranks are obtained, N-X sample accounts ranked in the N sorting ranks are selected, a behavior data sample set of the N-X sample accounts is obtained, and product information is pushed to a target account based on the behavior data sample set, wherein X is an integer smaller than N. For another example, the N similarity values are arranged from large to small, a sample account with the largest similarity value in the sorting result is selected, a behavior data sample of the sample account is obtained, and then product information is pushed to a target account based on the behavior data sample. Or the sorting of the N similarity values can be from big to small, the product information is pushed to the target account based on the sorting result of the N similarity values, or whether the parallel ranking exists in the sorting result or not can be judged first, when the parallel ranking exists in the sorting result and is higher, the parallel ranking is used as a cut-off point, at least one sample account corresponding to a plurality of similarity values before the parallel ranking is determined, and the product information is pushed to the target account based on the behavior data of the at least one sample account.
As an embodiment, in the above scheme, the steps are as follows: based on the sorting result of the N similarity values, pushing product information to the target account comprises the following steps:
Determining a target sample account from the N sample accounts based on the sorting results of the N similarity values;
and pushing product information to the target account according to the behavior data sample of the target sample account and the behavior data of the target account.
In this embodiment, the target sample account is a sample account with higher similarity to the target account. Here, the N similarity value sorting result may be a result of sorting N similarity values from large to small, and further determining a sample account corresponding to a plurality of similarity values ranked first in the sorting result as the target sample account.
It should be noted that, the behavior data sample of the target sample account includes product browsing data and/or product consumption data of the user corresponding to the sample account on the application program. When the method is realized, the product information is pushed to the target account according to the behavior data sample of the target sample account, which is equivalent to the product information which is screened from the product browsing data and/or the product consumption data corresponding to the target sample account and is pushed to the target account. Here, product information is screened out from product browsing data and/or product consumption data corresponding to the target sample account to push the product information to the target account, specifically, product information which is not paid attention to by the screened target account is pushed. For example, by comparing the difference between the behavior data sample and the behavior data, a target behavior which is included in the behavior data sample but not included in the behavior data is determined, and then product information corresponding to the target behavior is recommended to the target account.
It should be understood that the pushing of the product information to the target account may be that the product information is configured into a page file to obtain a new page file, when the user sends a data access request to the server by using the terminal that has logged in to the target account, the server returns the new page file to the terminal according to the data access request, and the terminal loads the new page file to further display the corresponding product information, so as to complete the pushing operation of the product information to the target account.
According to the scheme, when the behavior data of the target account does not meet the preset condition, the portrait data of the target account is obtained, and then N similarity values between the target account and N sample accounts are calculated based on the behavior data and the portrait data of the target account, wherein N is an integer greater than 1, so that when the behavior data of the target account is small, the similarity calculation between the target account and the N sample accounts is realized by fusing the portrait data, and then the product information is pushed to the target account based on the sorting result of the N similarity values, the phenomenon that the behavior data of the target account is excessively dependent in the process of pushing the product information to the target account is avoided, the flexibility of recommending the product information scheme is improved, and the application range of the product information recommending scheme is enlarged.
Referring to fig. 4, fig. 4 is a block diagram illustrating a product information recommendation apparatus according to an embodiment of the present application. The mobile terminal in this embodiment includes units for performing the steps in the embodiments corresponding to fig. 1 and 3. Refer specifically to fig. 1 and 3 and the related descriptions in the embodiments corresponding to fig. 1 and 3. For convenience of explanation, only the portions related to the present embodiment are shown. Referring to fig. 4, the product information recommending apparatus 40 includes: an acquisition unit 41, a first execution unit 42 and a second execution unit 43. Specifically:
An obtaining unit 41, configured to obtain portrait data of a target account if behavioral data of the target account does not meet a preset condition;
the first execution unit 42 is configured to calculate, based on the behavior data and the portrait data, similarities between the target account and N sample accounts, respectively, to obtain N similarity values; wherein N is an integer greater than 1;
The second execution unit 43 is configured to push product information to the target account based on the sorting result of the N similarity values.
It should be understood that, in the block diagram of the product information recommendation device shown in fig. 4, each unit is configured to perform each step in the embodiment corresponding to fig. 1 and 3, and each step in the embodiment corresponding to fig. 1 and 3 has been explained in detail in the foregoing embodiment, and specific reference is made to fig. 1 and 3 and the related description in the embodiment corresponding to fig. 1 and 3, which are not repeated herein.
Fig. 5 is a block diagram of a computer device according to an embodiment of the present application. As shown in fig. 5, the computer device 50 of this embodiment includes: a processor 51, a memory 52 and a computer program 53 stored in said memory 52 and executable on said processor 51, for example a program of a product information recommendation method. The steps in each embodiment of the above-mentioned product information recommendation method are implemented by the processor 51 when the computer program 53 is executed, for example, S11 to S13 shown in fig. 1, and the functions of each unit in the embodiment corresponding to fig. 4, for example, the functions of units 41 to 43 shown in fig. 4, are implemented by the processor 51 when the computer program 53 is executed, and are specifically please refer to the related descriptions in the embodiment corresponding to fig. 4, which are not repeated herein.
By way of example, the computer program 53 may be divided into one or more units, which are stored in the memory 52 and executed by the processor 51 to complete the present application. The one or more elements may be a series of computer program instruction segments capable of performing the specified functions, which instruction segments are used to describe the execution of the computer program 53 in the computer device 50. For example, the computer program 53 may be divided into an acquisition unit, a first execution unit and a second execution unit, each unit functioning specifically as described above.
The turntable device may include, but is not limited to, a processor 51, a memory 52. It will be appreciated by those skilled in the art that fig. 5 is merely an example of the computer device 50 and is not meant to be limiting of the computer device 50, and may include more or fewer components than shown, or may combine certain components, or different components, e.g., the turntable device may also include an input-output device, a network access device, a bus, etc.
The Processor 51 may be a central processing unit (Central Processing Unit, CPU), other general purpose Processor, digital signal Processor (DIGITAL SIGNAL Processor, DSP), application SPECIFIC INTEGRATED Circuit (ASIC), off-the-shelf Programmable gate array (Field-Programmable GATE ARRAY, FPGA) or other Programmable logic device, discrete gate or transistor logic device, discrete hardware components, or the like. A general purpose processor may be a microprocessor or the processor may be any conventional processor or the like.
The memory 52 may be an internal storage unit of the computer device 50, such as a hard disk or a memory of the computer device 50. The memory 52 may also be an external storage device of the computer device 50, such as a plug-in hard disk, a smart memory card (SMART MEDIA CARD, SMC), a Secure Digital (SD) card, a flash memory card (FLASH CARD), etc. that are provided on the computer device 50. Further, the memory 52 may also include both internal and external storage units of the computer device 50. The memory 52 is used for storing the computer program as well as other programs and data required by the turntable device. The memory 52 may also be used to temporarily store data that has been output or is to be output.
The above embodiments are only for illustrating the technical solution of the present application, and not for limiting the same; although the application has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical scheme described in the foregoing embodiments can be modified or some technical features thereof can be replaced by equivalents; such modifications and substitutions do not depart from the spirit and scope of the technical solutions of the embodiments of the present application, and are intended to be included in the scope of the present application.
Claims (7)
1. A product information recommendation method, comprising:
if the behavior data of the target account does not meet the preset condition, obtaining the portrait data of the target account; the preset conditions are used for representing characteristic indexes of behavior data capable of carrying out target account preference analysis;
based on the behavior data and the portrait data, measuring and calculating the similarity between the target account and N sample accounts respectively to obtain N similarity values; wherein N is an integer greater than 1;
pushing product information to the target account based on the sorting result of the N similarity values;
The preset condition comprises a quantity threshold; and if the behavior data of the target account does not meet the preset condition, acquiring the portrait data of the target account, wherein the acquiring comprises the following steps:
If the number of the behavior data of the target account does not meet the number threshold, carrying out associated server query according to the target account;
if the associated server is queried, obtaining portrait data of the target account from the associated server;
And if the number of the behavior data of the target account does not meet the number threshold, performing the associated server query according to the target account, including:
If no associated server is queried, displaying a data acquisition page; the data acquisition page is used for acquiring portrait data of the target account;
Based on the behavior data and the portrait data, measuring and calculating the similarity between the target account and N sample accounts respectively to obtain N similarity values, wherein the measuring and calculating comprises the following steps:
correcting the behavior data by utilizing a Newton cooling method to obtain new behavior data;
And calculating the similarity between the target account and each sample account based on the new behavior data and the portrait data, so as to obtain N similarity values.
2. The product information recommendation method according to claim 1, wherein the calculating the similarity between the target account and each sample account based on the new behavior data and the portrait data to obtain N similarity values includes:
performing vector conversion on the new behavior data and the portrait data respectively to obtain a first behavior vector and a first portrait vector;
Calculating the similarity between the first behavior vector and the second behavior vector of each sample account to obtain N behavior similarities, and calculating the similarity between the first image vector and the second image vector of each sample account to obtain N image similarities;
And obtaining N similarity values based on the N behavior similarities and the N image similarities.
3. The product information recommendation method according to claim 2, wherein the obtaining N similarity values based on the N behavioral similarities and the N portrait similarities includes:
Based on the N behavior similarities and the N image similarities, calculating a similarity value between the target account and each sample account through the following formula;
Si=K1×Pi+K2×Ai
Wherein, ; S i is a similarity value between the target account and the ith sample account; p i is the portrait similarity between the target account and the ith sample account; a i is the behavioral similarity between the target account and the ith sample account; k 1 is a first preset weight parameter, K 2 is a first preset weight parameter, and K 1 and K 2 are both greater than 0.
4. The product information recommendation method according to claim 1, wherein pushing product information to the target account based on the ranking results of the N similarity values includes:
Determining a target sample account from the N sample accounts based on the sorting results of the N similarity values;
and pushing product information to the target account according to the behavior data sample of the target sample account and the behavior data of the target account.
5. A product information recommendation device, characterized by comprising:
The system comprises an acquisition unit, a storage unit and a storage unit, wherein the acquisition unit is used for acquiring portrait data of a target account if the behavior data of the target account does not meet preset conditions; the preset conditions are used for representing characteristic indexes of behavior data capable of carrying out target account preference analysis;
The first execution unit is used for measuring and calculating the similarity between the target account and N sample accounts respectively based on the behavior data and the portrait data to obtain N similarity values; wherein N is an integer greater than 1;
The second execution unit is used for pushing product information to the target account based on the sorting results of the N similarity values;
wherein the preset condition includes a quantity threshold; and if the behavior data of the target account does not meet the preset condition, acquiring the portrait data of the target account, wherein the acquiring comprises the following steps:
If the number of the behavior data of the target account does not meet the number threshold, carrying out associated server query according to the target account;
if the associated server is queried, obtaining portrait data of the target account from the associated server;
And if the number of the behavior data of the target account does not meet the number threshold, performing the associated server query according to the target account, including:
If no associated server is queried, displaying a data acquisition page; the data acquisition page is used for acquiring portrait data of the target account;
Based on the behavior data and the portrait data, measuring and calculating the similarity between the target account and N sample accounts respectively to obtain N similarity values, wherein the measuring and calculating comprises the following steps:
correcting the behavior data by utilizing a Newton cooling method to obtain new behavior data;
And calculating the similarity between the target account and each sample account based on the new behavior data and the portrait data, so as to obtain N similarity values.
6. A computer device comprising a memory, a processor and a computer program stored in the memory and executable on the processor, the processor executing the computer program steps of the method according to any of claims 1 to 4.
7. A computer readable storage medium storing a computer program, characterized in that the computer program when executed by a processor implements the steps of the method according to any one of claims 1 to 4.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110440594.8A CN113034241B (en) | 2021-04-23 | 2021-04-23 | Product information recommendation method and computer equipment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110440594.8A CN113034241B (en) | 2021-04-23 | 2021-04-23 | Product information recommendation method and computer equipment |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113034241A CN113034241A (en) | 2021-06-25 |
| CN113034241B true CN113034241B (en) | 2024-06-11 |
Family
ID=76457723
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110440594.8A Active CN113034241B (en) | 2021-04-23 | 2021-04-23 | Product information recommendation method and computer equipment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113034241B (en) |
Families Citing this family (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114416198A (en) * | 2021-12-31 | 2022-04-29 | 成都易达数安科技有限公司 | APP column intelligent sequencing method based on user image |
| CN115348353B (en) * | 2022-08-10 | 2024-06-18 | 青岛海信智慧生活科技股份有限公司 | Content pushing method and server |
| CN115544994B (en) * | 2022-12-01 | 2023-05-05 | 爱集微咨询(厦门)有限公司 | Data pushing method, device, electronic equipment and readable storage medium |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2010134733A (en) * | 2008-12-05 | 2010-06-17 | Dainippon Printing Co Ltd | Information recommendation device, information recommendation method, and program |
| CN108460629A (en) * | 2018-02-10 | 2018-08-28 | 深圳壹账通智能科技有限公司 | User, which markets, recommends method, apparatus, terminal device and storage medium |
| CN109345348A (en) * | 2018-09-30 | 2019-02-15 | 重庆誉存大数据科技有限公司 | The recommended method of multidimensional information portrait based on travel agency user |
| CN109582857A (en) * | 2018-10-15 | 2019-04-05 | 深圳壹账通智能科技有限公司 | Based on big data information-pushing method, device, computer equipment and storage medium |
| KR20190114703A (en) * | 2018-03-29 | 2019-10-10 | 네이버 주식회사 | Method for recommending items and server using the same |
| CN111159561A (en) * | 2019-12-31 | 2020-05-15 | 青梧桐有限责任公司 | Method for constructing recommendation engine according to user behaviors and user portrait |
| CN112685639A (en) * | 2020-12-31 | 2021-04-20 | 苏宁金融科技(南京)有限公司 | Activity recommendation method and device, computer equipment and storage medium |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106228386B (en) * | 2016-07-12 | 2018-09-25 | 腾讯科技（深圳）有限公司 | A kind of information-pushing method and device |
-
2021
- 2021-04-23 CN CN202110440594.8A patent/CN113034241B/en active Active
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2010134733A (en) * | 2008-12-05 | 2010-06-17 | Dainippon Printing Co Ltd | Information recommendation device, information recommendation method, and program |
| CN108460629A (en) * | 2018-02-10 | 2018-08-28 | 深圳壹账通智能科技有限公司 | User, which markets, recommends method, apparatus, terminal device and storage medium |
| KR20190114703A (en) * | 2018-03-29 | 2019-10-10 | 네이버 주식회사 | Method for recommending items and server using the same |
| CN109345348A (en) * | 2018-09-30 | 2019-02-15 | 重庆誉存大数据科技有限公司 | The recommended method of multidimensional information portrait based on travel agency user |
| CN109582857A (en) * | 2018-10-15 | 2019-04-05 | 深圳壹账通智能科技有限公司 | Based on big data information-pushing method, device, computer equipment and storage medium |
| CN111159561A (en) * | 2019-12-31 | 2020-05-15 | 青梧桐有限责任公司 | Method for constructing recommendation engine according to user behaviors and user portrait |
| CN112685639A (en) * | 2020-12-31 | 2021-04-20 | 苏宁金融科技(南京)有限公司 | Activity recommendation method and device, computer equipment and storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113034241A (en) | 2021-06-25 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110825957B (en) | Deep learning-based information recommendation method, device, equipment and storage medium | |
| CN113034241B (en) | Product information recommendation method and computer equipment | |
| CN110941740B (en) | Video recommendation method and computer-readable storage medium | |
| WO2018041168A1 (en) | Information pushing method, storage medium and server | |
| JP5962926B2 (en) | Recommender system, recommendation method, and program | |
| CN110413870B (en) | Commodity recommendation method and device and server | |
| US20160285672A1 (en) | Method and system for processing network media information | |
| CN102934113B (en) | Information providing system, information providing method, information providing apparatus | |
| JP2011529224A (en) | Recommendation of correlated information | |
| CN108319646B (en) | Vehicle source searching method and device based on user historical behaviors | |
| CN105678335B (en) | It estimates the method, apparatus of clicking rate and calculates equipment | |
| CN113220657B (en) | Data processing method and device and computer equipment | |
| CN111310032B (en) | Resource recommendation method, device, computer equipment and readable storage medium | |
| CN112613938B (en) | Model training method and device and computer equipment | |
| CN106354867A (en) | Multimedia resource recommendation method and device | |
| Shani et al. | Tutorial on application-oriented evaluation of recommendation systems | |
| JP2006053616A (en) | Server device, web site recommendation method and program | |
| CN111680213A (en) | Information recommendation method, data processing method and device | |
| CN111177564B (en) | Product recommendation method and device | |
| CN110708361B (en) | System, method and device for determining grade of digital content publishing user and server | |
| CN114625973B (en) | Anonymous information cross-domain recommendation method and device, electronic equipment and storage medium | |
| CN110851708A (en) | Negative sample extraction method and device, computer equipment and storage medium | |
| KR101663359B1 (en) | Method and apparatus for providing updated news contents | |
| CN114090891A (en) | Personalized content recommendation method, device, equipment and storage medium | |
| CN115345635A (en) | Processing method and device for recommended content, computer 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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |