CN111768230A - Label recommendation method and device for client portrait system and computer equipment - Google Patents
Label recommendation method and device for client portrait system and computer equipment Download PDFInfo
- Publication number
- CN111768230A CN111768230A CN202010584585.1A CN202010584585A CN111768230A CN 111768230 A CN111768230 A CN 111768230A CN 202010584585 A CN202010584585 A CN 202010584585A CN 111768230 A CN111768230 A CN 111768230A
- Authority
- CN
- China
- Prior art keywords
- label
- target user
- user
- tag
- target
- 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
- 238000000034 method Methods 0.000 title claims abstract description 45
- 230000003993 interaction Effects 0.000 claims abstract description 55
- 238000012216 screening Methods 0.000 claims abstract description 23
- 238000012545 processing Methods 0.000 claims abstract description 12
- 238000010606 normalization Methods 0.000 claims description 18
- 238000004364 calculation method Methods 0.000 claims description 15
- 238000004590 computer program Methods 0.000 claims description 6
- 238000003384 imaging method Methods 0.000 abstract description 2
- 238000004458 analytical method Methods 0.000 description 12
- 239000013598 vector Substances 0.000 description 7
- 238000003032 molecular docking Methods 0.000 description 6
- 238000007405 data analysis Methods 0.000 description 5
- 238000004891 communication Methods 0.000 description 4
- 238000004422 calculation algorithm Methods 0.000 description 3
- 238000010586 diagram Methods 0.000 description 3
- 230000000694 effects Effects 0.000 description 3
- 238000001914 filtration Methods 0.000 description 3
- 238000007726 management method Methods 0.000 description 3
- 230000008901 benefit Effects 0.000 description 2
- 238000011161 development Methods 0.000 description 2
- 230000004069 differentiation Effects 0.000 description 2
- 238000000605 extraction Methods 0.000 description 2
- 230000005611 electricity Effects 0.000 description 1
- 230000010365 information processing Effects 0.000 description 1
- 230000002452 interceptive effect Effects 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 238000005065 mining Methods 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 230000008569 process Effects 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000011524 similarity measure Methods 0.000 description 1
- 238000012546 transfer Methods 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
- 239000002699 waste material Substances 0.000 description 1
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
- G06Q30/02—Marketing; Price estimation or determination; Fundraising
- G06Q30/0201—Market modelling; Market analysis; Collecting market data
-
- 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
- G06Q30/02—Marketing; Price estimation or determination; Fundraising
- G06Q30/0201—Market modelling; Market analysis; Collecting market data
- G06Q30/0202—Market predictions or forecasting for commercial activities
Landscapes
- Business, Economics & Management (AREA)
- Strategic Management (AREA)
- Engineering & Computer Science (AREA)
- Accounting & Taxation (AREA)
- Development Economics (AREA)
- Finance (AREA)
- Entrepreneurship & Innovation (AREA)
- Game Theory and Decision Science (AREA)
- Data Mining & Analysis (AREA)
- Economics (AREA)
- Marketing (AREA)
- Physics & Mathematics (AREA)
- General Business, Economics & Management (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The application discloses a label recommendation method and device for a customer portrait system and computer equipment, relates to the technical field of computers, and can solve the problems that labels in the customer portrait system cannot be accurately screened, so that label resources are seriously wasted, and the individual label selection cannot meet the individual requirements of users. The method comprises the following steps: acquiring label attribute information of a target user, wherein the label attribute information comprises a historical label interaction state of the target user and attribute characteristics influencing a target user to pay attention to a label; screening out similar users of the target user according to a preset rule corresponding to the label attribute information; and determining a target recommendation label of the target user according to the label interaction state of the similar user, and sending the target recommendation label to the target user. The method and the device are suitable for recommendation processing of the label in the customer imaging system.
Description
Technical Field
The application relates to the technical field of computers, in particular to a label recommendation method and device for a client portrait system and computer equipment.
Background
The client representation system has wide application in the aspects of enterprise client management, marketing activities, client services and the like. However, there are thousands of client tags in enterprise-level client figures, so that it is very important to guide the user to select the tags meeting the client requirements and to utilize different tags to perform more dimensional client exploration and client operation analysis in the application of the client figure system.
The inventor of the application finds that users of the portrait system often do not know which tags the portrait system has, and the users are always limited to a few tags which are frequently used, and do not spend a lot of time to mine different attribute tags according to the needs of the customers, so that the values of a plurality of tags are not fully displayed, the serious waste of tag resources in the portrait system of the customers is caused, and the single tag selection cannot meet the personalized needs of the customers.
Disclosure of Invention
In view of this, the application provides a method and an apparatus for recommending a tag in a client representation system, and a computer device, which mainly solve the problems that the tag in the client representation system cannot be accurately screened, so that the tag resource is seriously wasted, and the single tag selection cannot meet the personalized requirement of a user.
According to one aspect of the application, a tag recommendation method for a client representation system is provided, the method comprising:
acquiring label attribute information of a target user, wherein the label attribute information comprises a historical label interaction state of the target user and attribute characteristics influencing a target user to pay attention to a label;
screening out similar users of the target user according to a preset rule corresponding to the label attribute information;
and determining a target recommendation label of the target user according to the label interaction state of the similar user, and sending the target recommendation label to the target user.
According to another aspect of the present application, there is provided a tag recommendation apparatus for a client representation system, the apparatus comprising:
the system comprises an acquisition module, a processing module and a display module, wherein the acquisition module is used for acquiring label attribute information of a target user, and the label attribute information comprises a historical label interaction state of the target user and attribute characteristics influencing a target user to pay attention to a label;
the screening module is used for screening out similar users of the target user according to a preset rule corresponding to the label attribute information;
and the determining module is used for determining the target recommendation label of the target user according to the label interaction state of the similar user and sending the target recommendation label to the target user.
According to yet another aspect of the present application, there is provided a non-transitory readable storage medium having stored thereon a computer program which, when executed by a processor, implements a tag recommendation method for a client representation system as described above.
According to yet another aspect of the present application, there is provided a computer apparatus comprising a non-volatile readable storage medium, a processor, and a computer program stored on the non-volatile readable storage medium and executable on the processor, the processor implementing the tag recommendation method of the client representation system when executing the program.
Compared with the label application of the current client portrait system, the label recommendation method, the label recommendation device and the computer equipment of the portrait system customize the label recommendation model of the portrait system for the user based on the collaborative filtering recommendation algorithm of the internet, namely similar users similar to the target user preference can be screened out according to the label attribute information of the target user and the corresponding preset rule, the label recommendation of the target user is generated by using the labels selected by the similar users, so that the user can accurately and efficiently select the target label suitable for the user according to the recommended label, the label in the portrait system can be maximally applied, and the personalized requirement of the user can be met.
Drawings
The accompanying drawings, which are included to provide a further understanding of the application and are incorporated in and constitute a part of this application, illustrate embodiment(s) of the application and together with the description serve to explain the application and not to limit the application to the disclosed embodiment. In the drawings:
FIG. 1 is a flow chart illustrating a tag recommendation method for a client representation system according to an embodiment of the present application;
FIG. 2 is a flow chart illustrating a tag recommendation method for a client representation system according to an embodiment of the present application;
FIG. 3 is a schematic structural diagram illustrating a tag recommendation apparatus of a client representation system according to an embodiment of the present application;
FIG. 4 is a schematic structural diagram of a tag recommendation device of another client representation system according to an embodiment of the present application.
Detailed Description
The present application will be described in detail below with reference to the accompanying drawings in conjunction with embodiments. It should be noted that the embodiments and features of the embodiments in the present application may be combined with each other without conflict.
Aiming at the problems that when the label of the existing customer portrait system is applied, the label in the customer portrait system cannot be accurately screened, so that the label resource is seriously wasted, and the single label selection cannot meet the personalized requirement of a user, the embodiment of the application provides a label recommendation method of the customer portrait system, and as shown in fig. 1, the method comprises the following steps:
101. and acquiring label attribute information of the target user, wherein the label attribute information comprises historical label interaction state of the target user and attribute characteristics influencing the target user to pay attention to the label.
The tag attribute information can include historical tag interaction states of the target user, such as interaction states of the target user to each tag, wherein the interaction states include click rate, collection states and tag use states, and the tag use states can correspond to whether the tags are applied to a client grouping docking marketing platform or not and whether the tags are applied to a multidimensional analysis platform or not; the tag attribute information may further include a work attribute feature and a person attribute feature that affect the target user to pay attention to the tag, wherein the work attribute feature may include: department responsibilities (off-line marketing, electric marketing, underwriting, claims settlement, planning, repayment actuarial and finance), post properties (management post, development post, product post, data analysis post, operation post and sale post), working years (1-2 years, 3-5 years, 6-8 years, 9-10 years, 11-15 years and more than 15 years), affiliated institution grades (five-grade, four-grade, three-grade, two-grade and headquarter) and the like; the person attribute features may include: sex of the employee (male, female), student's calendar (university, this family, graduate, doctor), etc.
102. And screening out similar users of the target user according to a preset rule corresponding to the label attribute information.
For this embodiment, in a specific application scenario, since different types of tag attribute information all affect the selection of tags, different scoring rules need to be configured for different tag attributes respectively in combination with service requirements, so that similar users of a target user can be accurately screened out.
103. And determining a target recommendation label of the target user according to the label interaction state of the similar user, and sending the target recommendation label to the target user.
For this embodiment, after the target user is screened out, the tag interaction state of the similar user can be used as a tag screening basis of the target user, and the target recommended tag is determined by further analyzing the selected tag of the similar user, and then recommended to the target user for selection.
By the label recommendation method of the client portrait system in the embodiment, a label recommendation model of the portrait system for the user can be customized based on a collaborative filtering recommendation algorithm of the internet, namely, similar users similar to the preference of the target user are screened out according to the label attribute information of the target user and corresponding preset rules, and label recommendation of the target user is generated by using the labels selected by the similar users, so that the user can accurately and efficiently select the target label suitable for the user according to the recommended label, the label in the portrait system can be maximally applied, and the personalized requirements of the user can be met.
Further, as a refinement and an extension of the embodiments of the foregoing embodiments, for fully explaining the implementation process in this embodiment, another tag recommendation method for a client representation system is provided, as shown in fig. 2, the method includes:
201. and acquiring label attribute information of the target user, wherein the label attribute information comprises historical label interaction state of the target user and attribute characteristics influencing the target user to pay attention to the label.
For this embodiment, in a specific application scenario, the tag attribute information of the target user needs to be obtained from multiple dimensions in advance, so that the tag selection tendency of the target user can be accurately determined based on the tag attribute information, and further effective recommendations are generated.
202a, if the label attribute information is the historical label interaction state of the target user, determining the interaction state of the target user to each label based on the historical browsing record.
The interactive state comprises click quantity, a collection state and a tag using state, and if the tag is applied to the customer grouping and docking marketing platform, the tag is applied to the multidimensional analysis platform and the like.
For the embodiment, the tag attribute information may be an interaction state of the target user with respect to each tag, and specifically, the interaction information of the target user with respect to the tag in the image system in a preset historical time period may be obtained by embedding the point, that is, the click and collection actions of the user with respect to the tag may be obtained, and behavior data such as marketing activity, business analysis or prediction may be performed on the tag reference. The preset historical time period can be set according to actual requirements, and if the preset historical time period is the previous month of the current time, the preset historical time period can be set.
203a, calculating a first intimacy degree of the target user to each label according to a first scoring rule.
When the affinity is calculated, firstly, the label interaction state data contained in the first scoring rule is screened out according to the historical data record of the user, then the sum of products of each label interaction state data and the corresponding score is calculated, and the sum of the products is determined as the first affinity of the target user to the label.
For example, a first scoring rule is set as in table 1:
table 1:
if it is determined that the browsing volume of the target user for the tag 1 is 12, the tag 1 is collected, and the tag 1 is respectively applied to the client group docking marketing platform and the multidimensional analysis platform, the intimacy degree of the target user for the tag 1 can be calculated according to the customized first scoring rule as follows: 1 × 12+2+3+3 ═ 20. Correspondingly, if it is determined that the browsing volume of the target user for the tag 2 is 0, the tag 2 is not collected, and the tag 2 is not applied to the customer grouping docking marketing platform and the multidimensional analysis platform, the intimacy degree of the target user for the tag 2 can be calculated according to the customized first scoring rule as follows: 1 × 0+0+0+0 ═ 0.
204a, calculating a second intimacy degree of the existing users to each label in the database by using the first scoring rule.
For this embodiment, based on the example in the embodiment step 203a, a plurality of existing users may be screened from the database, and the second affinity of the existing users to each tag is calculated by using the determined first scoring rule, for example, if it is determined that the browsing volume of the user a to the tag 1 is 14, and the tag 1 is collected, and the tag 1 is applied to the customer grouping docking marketing platform and the multidimensional analysis platform, the affinity of the user B to the tag 1 may be calculated as: 1 × 14+2+3+3 ═ 22. If it is determined that the browsing volume of the user B for the tag 1 is 0, but the tag 1 is collected, but the tag 1 is not applied to the customer clustering docking marketing platform and the multidimensional analysis platform, according to the first scoring rule, the intimacy of the user B for the tag 1 can be calculated as follows: 1 × 0+2+0+0 ═ 2. By the above method, as shown in table 2, the intimacy of any existing user (e.g., target user and existing user A, B, C, D, E) with respect to any tag (e.g., tags 1, 2, 3, 4, 5, 6) can be calculated.
Table 2:
205a, screening out the first similar users of the target user according to the first intimacy degree and the second intimacy degree.
In a specific application scenario, in order to filter out similar users of the target user, the embodiment step 205a may specifically include: normalizing the first intimacy degree and the second intimacy degree according to a preset first calculation formula; calculating a first similarity between the target user and each existing user according to the normalization processing result; and determining the existing users with the first similarity higher than a first preset threshold as the first similar users.
Wherein the first calculation formula is characterized by: in the formula, X1 is a normalization result of the intimacy degree of the current tag corresponding to the target user or the existing user, X is the intimacy degree of the tag corresponding to the target user or the existing user, min is a minimum value of the intimacy degree of the corresponding tag in the target user and the existing user, and max is a maximum value of the intimacy degree of the corresponding tag in the target user and the existing user.
In a specific application scenario, in order to facilitate the judgment of the first similar user, noise reduction and normalization processing are performed on the label intimacy degree by using a first calculation formula, so that the result falls in a [0, 1] interval. For this embodiment, based on the example in the embodiment step 204a, the normalization processing result shown in table 3 can be obtained by performing the normalization processing on the label intimacy degree in table 2 by using the first calculation formula.
Table 3:
accordingly, the similarity between the target user and the existing user can be calculated by using the cosine similarity. The cosine similarity measures the difference between two individuals by using the cosine value of the included angle between two vectors in the vector space. The cosine value is closer to 1, which indicates that the included angle is closer to 0 degree, namely the two vectors are more similar, which is called cosine similarity. For two vectors, the two vectors are considered to be more similar if the angle between them is smaller. Cosine similarity takes advantage of this theoretical idea. The similarity value between the vectors is measured by calculating the cosine value of the included angle of the two vectors.
The calculation formula may be:
wherein, XiAnd obtaining the intimacy normalization result of each label corresponding to the target user, wherein Yi is the intimacy normalization result of each label corresponding to the existing user, and n is the label serial number.
For example, based on the intimacy degree normalization result in table 3, the cosine similarity calculation formula is used to calculate the first similarity between the target user and the current user a as follows:
accordingly, the first similarity between the target user and the current user, which can be calculated by the above method, is shown in table 4.
Table 4:
| target user | |
| Target user | 1 |
| User A | 0.759869 |
| User B | 0.445346 |
| User C | 0.273041 |
| User D | 0.425422 |
| User E | 0.639317 |
In a specific application scenario, according to the tag similarity analysis result between the target user and each existing user calculated in table 4, the similar user with higher similarity to the target user can be determined, and the N users with the highest similarity can be selected as the similar users of the target user; a threshold value may also be set, for example, a similar user as the target user with a cosine value higher than 0.6. From table 4, if the filtering threshold of the similar user is set to 0.6, it can be determined that the user a and the user E have higher similarity to the target user. Tags of interest to user a and user E should be recommended to the target user. At this time, the tag data corresponding to the target user, the existing user a and the user E extracted from table 1 may be as shown in table 5.
Table 5:
| label 1 | Label 2 | Label 3 | Label 4 | Label 5 | Label 6 | |
| Target user | 20 | 0 | 25 | 15 | 0 | 12 |
| User A | 22 | 0 | 18 | 3 | 2 | 15 |
| User E | 18 | 8 | 20 | 0 | 2 | 10 |
206a, extracting a first label with the first intimacy degree of 0, and screening out a first to-be-recommended label with the second intimacy degree of more than 0 from the first label.
For this embodiment, based on table 5 in the embodiment step 205a, determining, by using the tag data corresponding to the target user, the existing user a, and the user E, that the first tag having the first affinity of 0 corresponding to the target user includes: the second intimacy degree between the tag 2 and the user E is greater than 0, and the second intimacy degree between the tag 5 and the user a and the user E is greater than 0, so that the tag 2 and the tag 5 can be determined as the first to-be-recommended tag.
207a, calculating a first recommendation value of each first to-be-recommended label.
In a specific application scenario, based on the example in the embodiment step 206a, after determining that the tags 2 and 5 are tags that may be needed by the selected target user, the tags may be weighted, and then recommendation scores of the tags are respectively calculated, where the specific calculation formula is: and recommending the score, which is the intimacy degree of the similar user corresponding to the label to be recommended, between the target user and the similar user. For example, the label 2 only has the recommendation of the user E, the 'intimacy degree' between the user E and the label 2 is 8 (see table 5), and the similarity degree between the target user and the user E is 0.639317 (see table 4), so the recommendation score of the label 2 is calculated as: 8 × 0.639317 ═ 5.114536. Similarly, the tag 5 has recommendations of the user a and the user E, the 'intimacy degree' between the user a and the tag 5 is 2, the 'intimacy degree' between the user E and the tag 5 is 2, the similarity between the target user and the user a is 0.759869, and the similarity between the target user and the user E is 0.639317, so that the recommendation score of the tag 5 is calculated as: 2 × 0.759869+2 × 0.639317 ═ 2.798372. Thus, tag 2 is considered more valuable than the recommendation.
208a, determining a first to-be-recommended label with the highest first recommendation value as a first target recommendation label, and sending the first target recommendation label to the target user.
For this embodiment, if it is determined through step 207a that the tag 2 is the first to-be-recommended tag with the highest first recommendation value, the tag 2 may be sent to the target user as the first target recommendation tag.
By the label recommendation method of the client portrait system, the interaction state of the target user on each label can be determined based on the historical browsing record, the first intimacy degree of the target user on each label and the second intimacy degree of the existing user on each label in the database are calculated according to the first scoring rule of the interaction state, the first similar user of the target user is screened out according to the first intimacy degree and the second intimacy degree, the first to-be-recommended label with the highest first recommendation value is selected from the selected labels of the first similar user and is sent to the target user as the first target recommendation label. The method has the advantages that the user can accurately and efficiently select the target label suitable for the user, the selected target label can meet the individual requirements of the user, the label utilization rate can be improved, and the label in the image system can be maximally applied.
In an embodiment step 202b, which is parallel to the embodiment step 202a, if the tag attribute information is an attribute feature that affects the target user to pay attention to the tag, the attribute feature corresponding to the target user is acquired.
Wherein the attribute features comprise work attribute features and personnel attribute features. The work attribute features may include a first broad class: such as department responsibilities, position properties, working years, affiliated institution levels, etc., and the different first major classes respectively comprise corresponding second minor classes, such as department responsibilities may include: offline marketing, electricity marketing, underwriting, claim settlement, planning, repayment actuarial and finance; the station properties may include: a management post, a development post, a product post, a data analysis post, an operation post and a sales post; the operating age may include: 1-2 years, 3-5 years, 6-8 years, 9-10 years, 11 years-15 years, more than 15 years; the mechanism grades can include five grades, four grades, three grades, two grades, headquarters and the like; person attribute features may include a first broad category of: the gender of the employee, the academic calendar of the employee, etc., and the different first categories respectively include the corresponding second categories, for example, the gender of the employee may include: male and female; the employee's calendar may include: dachun, Ben Ke, research student, doctor, etc.
For the embodiment, as a parallel scheme of the embodiment steps 202a to 208a, since new employees and some users with low liveness lack a historical tag interaction state when using the representation system, and cannot effectively obtain the "intimacy" between the new employees and the tags, the similar users who cannot be determined by the previous embodiment are a preferred mode of the present application, and the following embodiment is a supplementary optimization based on the basic recommendation model for solving these problems.
203b, calculating the difference value between the target user and each existing user according to the attribute characteristics and the second grading rule.
For this embodiment, in a specific application scenario, according to the service suggestions, a differentiation rule may be defined for each first broad class attribute feature according to the second broad class in each first broad class attribute feature, taking the first broad class of department responsibilities as an example, if, inside a company, it is considered that the difference between the marketing class position and the position related to data analysis such as planning, accounting, and finance is relatively large, a differentiation rule related to department responsibilities may be defined as the following table 6.
Table 6:
it should be noted that, in table 6, only the difference degree scores of the attribute feature of the department responsibilities are listed simply, and the score settings of other attribute features such as the post property, the working age, the employee gender, the employee academic history, the affiliated institution level and the like can refer to the listing form of the department responsibilities, and specific scores can be set according to the actual application requirements. After the discrimination rules are configured for the attribute features of each first large category, similar users of each target user can be determined according to the rule analysis of each divided attribute feature, and a label to be recommended is determined.
First, the work attribute features and the person attribute features of the target user and each existing user need to be obtained from the representation system database, and for example, as shown in table 7, the attribute feature information of the target user and corresponding three existing users is obtained.
Table 7:
second, it is necessary to find the user most similar to the target work content among A, B, C three existing users. From table 7, it can be determined that the target job belongs to the telemarketing department as user a, but the post properties are different, the product post is more concerned with the operation of the product, and the data analysis post is more concerned with the achievement of the telemarketing department and the mining of the customer. B, C and target work are all data analysis positions, but in different departments, the analysis labels of interest should be different. At this time, the difference between the target user and the rest of the existing employee attributes can be calculated according to the difference values of the attribute features shown in table 6, for example, the difference values between the target job and the first category attribute features corresponding to A, B, C of the three existing users are calculated and obtained as shown in table 8.
Table 8:
further, the importance of the attributes can be unified through normalization. Wherein, the same attribute feature can be taken as a group for normalization. The normalization algorithm is a linear transformation of the raw data, dropping the result to [0, 1]]Interval, the transfer function is as follows: x2(X-min)/(max-min), where X2The normalization result of the difference value of each attribute feature between the target user and the existing user is obtained, x is the difference value of the current attribute feature corresponding to the target user and the existing user, min is the minimum difference value of the current attribute feature corresponding to the target user and each existing user, and max is the maximum difference value of the current attribute feature corresponding to the target user and each existing user. For example, by performing normalization processing on the difference values in table 8, the normalization processing result indicated by the difference one in table 9 can be obtained.
Table 9:
in a specific application scenario, the influence weight of different attribute characteristics on the judgment result can be defined according to the influence degree of the different attribute characteristics on the user difference degree by combining with an experience coefficient. For example, table 10 shows the influence weights corresponding to the attribute features.
Table 10:
the attribute features in table 9 are 'weighted' by the above-mentioned impact weights, highlighting the degree of impact of the important attribute. The difference value between users after weighting, i.e. the difference value of each attribute feature in table 9, is multiplied by the weighting coefficient in the corresponding table 10, and is calculated as follows:
table 11:
finally, the difference between the individuals is obtained by using euclidean distance calculation, for example, the difference value calculation formula between the target user and the user a may be:
further, the difference between the target user and A, B, C can be calculated as shown in Table 12.
Table 12:
204b, determining the existing user with the difference value smaller than the second preset threshold value as a second similar user.
For this embodiment, a second similar user of the target user may be determined according to the difference value between the target user and the A, B, C in step 203 b. For example, if the second preset threshold is set to 0.25, it is determined that the user a and the user B are the second similar users.
205b, extracting a second label with an interaction state of a second similar user.
For the embodiment, in a specific application scenario, the screened second similar user should default to the staff with the tag interaction data, and after the second similar user is obtained, the interaction information of the second similar user on the tag in the imaging system in a preset historical time period can be obtained through the embedded point, that is, the clicking and collecting actions of the second similar user on the tag can be obtained, and the behavior data such as marketing activity, operation analysis or prediction can be performed on the tag reference. The preset historical time period can be set according to actual requirements, and if the preset historical time period is the previous month of the current time, the preset historical time period can be set. And further, a second label with an interaction state of a second similar user can be obtained, and the second label can be directly determined as a second label to be recommended to be generated and recommended to the target user due to the target user label interaction data.
206b, determining the second tag as a second target recommended tag, and sending the second target recommended tag to the target user.
According to the tag recommendation method of the client portrait system, tag recommendation can be generated for a target user based on two modes, one mode is that the interaction state of the target user to each tag is determined based on a historical browsing record, the first intimacy degree of the target user to each tag and the second intimacy degree of the existing user to each tag in a database are calculated according to the first grading rule of the interaction state, the first similar user of the target user is screened out by utilizing the first intimacy degree and the second intimacy degree, the first to-be-recommended tag with the highest first recommendation value is selected from the selected tags of the first similar user and is sent to the target user as the first target recommendation tag; and the other method is to obtain the attribute characteristics of the target user and the existing users, calculate the difference value between the target user and each existing user by using a second grading rule and the attribute characteristics, determine a second similar user of the target user by using the difference value, screen out a second target recommendation label with an interaction state based on the label interaction state of the second similar user, and send the second target recommendation label as a recommendation to the target user. Through the two methods for determining the recommended tags, a user can accurately and efficiently select the target tags suitable for the user, so that the problems of some zombie tags are solved, the tags in the portrait system can be maximally applied, the utilization rate of the tags is improved, the recommended tags are determined from multiple dimensions, and the selected tags can better meet the personalized requirements of the user.
Further, as a specific implementation of the method shown in fig. 1 and fig. 2, an embodiment of the present application provides a tag recommendation apparatus of a client representation system, as shown in fig. 3, the apparatus includes: an acquisition module 31, a screening module 32 and a determination module 33;
the obtaining module 31 is configured to obtain tag attribute information of a target user, where the tag attribute information includes a historical tag interaction state of the target user and an attribute characteristic that affects a tag concerned by the target user;
the screening module 32 may be configured to screen out similar users of the target user according to a preset rule corresponding to the tag attribute information;
the determining module 33 may be configured to determine a target recommended label of the target user according to the label interaction states of the similar users, and send the target recommended label to the target user.
In a specific application scenario, when the tag attribute information is a historical tag interaction state of the target user, and the preset rule is a first scoring rule corresponding to the historical tag interaction state, as shown in fig. 4, the screening module 32 may specifically include: a determination unit 321, a calculation unit 322, a screening unit 323;
the determining unit 321 is configured to determine, based on the historical browsing record, an interaction state of the target user for each tag, where the interaction state includes a click amount, a collection state, and a tag use state;
a calculating unit 322, configured to calculate a first affinity of the target user for each tag according to a first scoring rule;
the calculating unit 322 is further configured to calculate a second affinity of the existing user to each tag in the database by using the first scoring rule;
the screening unit 323 may be configured to screen out a first similar user of the target user according to the first affinity and the second affinity.
Correspondingly, in order to screen out a first similar user of the target user according to the first intimacy degree and the second intimacy degree, the screening unit 323 can be specifically configured to perform normalization processing on the first intimacy degree and the second intimacy degree according to a preset first calculation formula; calculating a first similarity between the target user and each existing user according to the normalization processing result; and determining the existing users with the first similarity higher than a first preset threshold as the first similar users.
In a specific application scenario, the first calculation formula is characterized by: x1 is (X-min)/(max-min), where X1 is a normalization result of the intimacy degree of the corresponding tag of the target user or the existing user, X is the intimacy degree of the corresponding tag of the target user or the existing user, min is a minimum intimacy degree of the corresponding tag of the target user and the existing user, and max is a maximum intimacy degree of the corresponding tag of the target user and the existing user.
Correspondingly, in order to determine the target recommended label of the target user according to the label interaction state of the similar user, as shown in fig. 4, the determining module 33 may further include: an extraction unit 331, a calculation unit 332, a determination unit 333;
the extracting unit 331 is configured to extract a first label with a first affinity of 0, and screen out a first to-be-recommended label with a second affinity greater than 0 from the first label;
a calculating unit 332, configured to calculate a first recommendation value of each first to-be-recommended label;
the determining unit 333 is configured to determine the first to-be-recommended tag with the highest first recommendation value as the first target recommended tag, and send the first target recommended tag to the target user.
In a specific application scenario, when the tag attribute information is an attribute feature that affects a target user to pay attention to a tag, and the preset rule is a second scoring rule of each attribute feature, as shown in fig. 4, the screening module 32 may further include: an acquisition unit 324;
an obtaining unit 324, configured to obtain attribute features corresponding to a target user, where the attribute features include a work attribute feature and a person attribute feature;
the calculating unit 322 is further configured to calculate a difference value between the target user and each existing user according to the attribute feature and the second scoring rule;
the determining unit 321 may be further configured to determine an existing user with a difference value smaller than a second preset threshold as a second similar user.
Correspondingly, in order to determine the target recommended tag of the target user according to the tag interaction state of the similar user, the extraction unit 331 may be further configured to extract a second tag of a second similar user having an interaction state;
the determining unit 333 is further configured to determine the second tag as a second target recommended tag, and send the second target recommended tag to the target user.
It should be noted that other corresponding descriptions of the functional units related to the tag recommendation device of the client representation system provided in this embodiment may refer to the corresponding descriptions in fig. 1 to fig. 2, and are not repeated herein.
Based on the method shown in fig. 1 to 2, correspondingly, the embodiment further provides a non-volatile storage medium, on which computer readable instructions are stored, and the readable instructions, when executed by a processor, implement the tag recommendation method of the client representation system shown in fig. 1 to 2.
Based on such understanding, the technical solution of the present application may be embodied in the form of a software product, which may be stored in a non-volatile storage medium (which may be a CD-ROM, a usb disk, a removable hard disk, etc.), and includes several instructions for enabling a computer device (which may be a personal computer, a server, or a network device, etc.) to execute the method according to the implementation scenarios of the present application.
Based on the method shown in fig. 1 to fig. 2 and the virtual device embodiments shown in fig. 3 and fig. 4, in order to achieve the above object, the present embodiment further provides a computer device, where the computer device includes a storage medium and a processor; a nonvolatile storage medium for storing a computer program; a processor for executing a computer program to implement the tag recommendation method of the client representation system as described above with reference to fig. 1-2.
Optionally, the computer device may further include a user interface, a network interface, a camera, Radio Frequency (RF) circuitry, sensors, audio circuitry, a WI-FI module, and so forth. The user interface may include a Display screen (Display), an input unit such as a keypad (Keyboard), etc., and the optional user interface may also include a USB interface, a card reader interface, etc. The network interface may optionally include a standard wired interface, a wireless interface (e.g., WI-FI interface), etc.
It will be understood by those skilled in the art that the present embodiment provides a computer device structure that is not limited to the physical device, and may include more or less components, or some components in combination, or a different arrangement of components.
The nonvolatile storage medium can also comprise an operating system and a network communication module. The operating system is a program that manages the hardware and software resources of the computer device described above, supporting the operation of information handling programs and other software and/or programs. The network communication module is used for realizing communication among components in the nonvolatile storage medium and communication with other hardware and software in the information processing entity device.
Through the above description of the embodiments, those skilled in the art will clearly understand that the present application can be implemented by software plus a necessary general hardware platform, and can also be implemented by hardware.
By applying the technical scheme, compared with the prior art, the method and the device can generate the label recommendation for the target user based on two modes, wherein the first mode is to determine the interaction state of the target user to each label based on the historical browsing record, calculate the first intimacy degree of the target user to each label and the second intimacy degree of the existing user to each label in the database according to the first grading rule of the interaction state, screen out the first similar user of the target user by utilizing the first intimacy degree and the second intimacy degree, select the first label to be recommended with the highest first recommendation value from the selected labels of the first similar user, and send the first label to be recommended to the target user as the first target recommendation label; and the other method is to obtain the attribute characteristics of the target user and the existing users, calculate the difference value between the target user and each existing user by using a second grading rule and the attribute characteristics, determine a second similar user of the target user by using the difference value, screen out a second target recommendation label with an interaction state based on the label interaction state of the second similar user, and send the second target recommendation label as a recommendation to the target user. Through the two methods for determining the recommended tags, a user can accurately and efficiently select the target tags suitable for the user, so that the problems of some zombie tags are solved, the tags in the portrait system can be maximally applied, the utilization rate of the tags is improved, the recommended tags are determined from multiple dimensions, and the selected tags can better meet the personalized requirements of the user.
Those skilled in the art will appreciate that the figures are merely schematic representations of one preferred implementation scenario and that the blocks or flow diagrams in the figures are not necessarily required to practice the present application. Those skilled in the art will appreciate that the modules in the devices in the implementation scenario may be distributed in the devices in the implementation scenario according to the description of the implementation scenario, or may be located in one or more devices different from the present implementation scenario with corresponding changes. The modules of the implementation scenario may be combined into one module, or may be further split into a plurality of sub-modules.
The above application serial numbers are for description purposes only and do not represent the superiority or inferiority of the implementation scenarios. The above disclosure is only a few specific implementation scenarios of the present application, but the present application is not limited thereto, and any variations that can be made by those skilled in the art are intended to fall within the scope of the present application.
Claims (10)
1. A tag recommendation method for a client representation system, comprising:
acquiring label attribute information of a target user, wherein the label attribute information comprises a historical label interaction state of the target user and attribute characteristics influencing a target user to pay attention to a label;
screening out similar users of the target user according to a preset rule corresponding to the label attribute information;
and determining a target recommendation label of the target user according to the label interaction state of the similar user, and sending the target recommendation label to the target user.
2. The method according to claim 1, wherein if the tag attribute information is a historical tag interaction state of a target user, the preset rule is a first scoring rule corresponding to the historical tag interaction state, and the screening out similar users of the target user according to the preset rule corresponding to the tag attribute information specifically includes:
determining the interaction state of the target user for each label based on historical browsing records, wherein the interaction state comprises click rate, collection state and label use state;
calculating first intimacy of the target user to each label according to the first scoring rule;
calculating a second intimacy degree of the existing user to each label in the database by using the first scoring rule;
and screening out a first similar user of the target user according to the first intimacy degree and the second intimacy degree.
3. The method according to claim 2, wherein the screening out the first similar user of the target user according to the first affinity and the second affinity specifically comprises:
normalizing the first intimacy degree and the second intimacy degree according to a preset first calculation formula;
calculating a first similarity between the target user and each existing user according to the normalization processing result;
and determining the existing users with the first similarity higher than a first preset threshold as first similar users.
4. The method of claim 3, wherein the first calculation formula is characterized by:
x1 ═ X-min)/(max-min), where X1 is a normalized result of the affinity of the current tag for the target user or the existing user, X is the affinity of the tag for the target user or the existing user, min is a minimum affinity of the corresponding tag between the target user and the existing user, and max is a maximum affinity of the corresponding tag between the target user and the existing user.
5. The method according to claim 4, wherein the determining the target recommended label of the target user according to the label interaction state of the similar user and sending the target recommended label to the target user specifically includes:
extracting the first label with the first intimacy degree of 0, and screening out a first to-be-recommended label with the second intimacy degree of more than 0 from the first label;
calculating a first recommendation value of each first to-be-recommended label;
and determining the first to-be-recommended label with the highest first recommendation value as a first target recommended label, and sending the first target recommended label to the target user.
6. The method according to claim 1, wherein if the tag attribute information is an attribute feature that affects a tag concerned by the target user, the preset rule is a second scoring rule for each attribute feature, and the screening out the similar users of the target user according to the preset rule corresponding to the tag attribute information specifically includes:
acquiring attribute characteristics corresponding to the target user, wherein the attribute characteristics comprise work attribute characteristics and personnel attribute characteristics;
calculating difference values between the target user and each existing user according to the attribute characteristics and the second grading rule;
and determining the existing user with the difference value smaller than a second preset threshold value as a second similar user.
7. The method according to claim 6, wherein the determining the target recommended label of the target user according to the label interaction state of the similar user and sending the target recommended label to the target user specifically includes:
extracting a second label of which the second similar user has an interaction state;
and determining the second tag as a second target recommended tag, and sending the second target recommended tag to the target user.
8. A tag recommendation device for a client representation system, comprising:
the system comprises an acquisition module, a processing module and a display module, wherein the acquisition module is used for acquiring label attribute information of a target user, and the label attribute information comprises a historical label interaction state of the target user and attribute characteristics influencing a target user to pay attention to a label;
the screening module is used for screening out similar users of the target user according to a preset rule corresponding to the label attribute information;
and the determining module is used for determining the target recommendation label of the target user according to the label interaction state of the similar user and sending the target recommendation label to the target user.
9. A non-transitory readable storage medium having stored thereon a computer program, wherein the program when executed by a processor implements the tag recommendation method of a client representation system as claimed in any one of claims 1 to 7.
10. A computer device comprising a non-volatile readable storage medium, a processor and a computer program stored on the non-volatile readable storage medium and executable on the processor, wherein the processor when executing the program implements a tag recommendation method for a customer representation system as claimed in any one of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010584585.1A CN111768230A (en) | 2020-06-24 | 2020-06-24 | Label recommendation method and device for client portrait system and computer equipment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010584585.1A CN111768230A (en) | 2020-06-24 | 2020-06-24 | Label recommendation method and device for client portrait system and computer equipment |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN111768230A true CN111768230A (en) | 2020-10-13 |
Family
ID=72722320
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010584585.1A Pending CN111768230A (en) | 2020-06-24 | 2020-06-24 | Label recommendation method and device for client portrait system and computer equipment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111768230A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114240548A (en) * | 2021-12-10 | 2022-03-25 | 杭州米络星科技(集团)有限公司 | Data processing method, device, equipment and storage medium of social customer relationship management system |
| CN114969519A (en) * | 2022-05-13 | 2022-08-30 | 海知见信息技术(北京)有限公司 | Attention-degree-based customer recommendation method and device |
| CN117648497A (en) * | 2024-01-29 | 2024-03-05 | 贵州大学 | Method and system for realizing intelligent acquisition of user information based on big data |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106997358A (en) * | 2016-01-22 | 2017-08-01 | 中移（杭州）信息技术有限公司 | Information recommendation method and device |
| CN109165975A (en) * | 2018-08-09 | 2019-01-08 | 平安科技（深圳）有限公司 | Label recommendation method, device, computer equipment 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 |
| CN109543111A (en) * | 2018-11-28 | 2019-03-29 | 广州虎牙信息科技有限公司 | Recommendation information screening technique, device, storage medium and server |
| CN109978608A (en) * | 2019-03-05 | 2019-07-05 | 广州海晟科技有限公司 | The marketing label analysis extracting method and system of target user's portrait |
| CN110291544A (en) * | 2017-02-13 | 2019-09-27 | 北京嘀嘀无限科技发展有限公司 | System and method for determining cohesion between user |
| CN111190939A (en) * | 2019-12-27 | 2020-05-22 | 深圳市优必选科技股份有限公司 | User portrait construction method and device |
-
2020
- 2020-06-24 CN CN202010584585.1A patent/CN111768230A/en active Pending
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106997358A (en) * | 2016-01-22 | 2017-08-01 | 中移（杭州）信息技术有限公司 | Information recommendation method and device |
| CN110291544A (en) * | 2017-02-13 | 2019-09-27 | 北京嘀嘀无限科技发展有限公司 | System and method for determining cohesion between user |
| CN109165975A (en) * | 2018-08-09 | 2019-01-08 | 平安科技（深圳）有限公司 | Label recommendation method, device, computer equipment 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 |
| CN109543111A (en) * | 2018-11-28 | 2019-03-29 | 广州虎牙信息科技有限公司 | Recommendation information screening technique, device, storage medium and server |
| CN109978608A (en) * | 2019-03-05 | 2019-07-05 | 广州海晟科技有限公司 | The marketing label analysis extracting method and system of target user's portrait |
| CN111190939A (en) * | 2019-12-27 | 2020-05-22 | 深圳市优必选科技股份有限公司 | User portrait construction method and device |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114240548A (en) * | 2021-12-10 | 2022-03-25 | 杭州米络星科技(集团)有限公司 | Data processing method, device, equipment and storage medium of social customer relationship management system |
| CN114969519A (en) * | 2022-05-13 | 2022-08-30 | 海知见信息技术(北京)有限公司 | Attention-degree-based customer recommendation method and device |
| CN114969519B (en) * | 2022-05-13 | 2023-05-12 | 海知见信息技术(北京)有限公司 | Attention-based client recommendation method and device |
| CN117648497A (en) * | 2024-01-29 | 2024-03-05 | 贵州大学 | Method and system for realizing intelligent acquisition of user information based on big data |
| CN117648497B (en) * | 2024-01-29 | 2024-04-30 | 贵州大学 | Method and system for realizing intelligent acquisition of user information based on big data |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111768230A (en) | Label recommendation method and device for client portrait system and computer equipment | |
| CN111340246A (en) | Processing method and device for enterprise intelligent decision analysis and computer equipment | |
| US20150242793A1 (en) | Systems and methods for auto-optimization of gamification mechanics for workforce motivation | |
| US7761797B2 (en) | Interactive user interface for displaying correlation | |
| JP2016206878A (en) | Salesperson raising support system and salesperson raising support method | |
| Micheni | Diffusion of big data and analytics in developing countries | |
| Theys et al. | The evolution of inter-regional spatial mismatch in the USA: The role of skills and spatial structure | |
| Qatawneh | The influence of data mining on accounting information system performance: a mediating role of information technology infrastructure | |
| Bhambri | Data mining as a tool to predict churn behavior of customers | |
| Nagaraj et al. | E-Commerce Customer Churn Prediction Scheme Based on Customer Behaviour Using Machine Learning | |
| Dmowska et al. | Improving assessment of urban racial segregation by partitioning a region into racial enclaves | |
| Sharma et al. | Data mining research for customer relationship management systems: a framework and analysis | |
| CN107729424B (en) | Data visualization method and equipment | |
| US11295325B2 (en) | Benefit surrender prediction | |
| Stavrou et al. | Flexible working arrangements in context: An empirical investigation through self-organizing maps | |
| US10853763B1 (en) | Onboarding new service providers utilizing portable service provider attributes | |
| CN111400663B (en) | Model training method, device, equipment and computer readable storage medium | |
| JP6031165B1 (en) | Promising customer prediction apparatus, promising customer prediction method, and promising customer prediction program | |
| CN112131477A (en) | Library book recommendation system and method based on user portrait | |
| KR102035141B1 (en) | Planning system and method of financial contents | |
| WO2015127333A1 (en) | Systems and methods for auto-optimization of gamification mechanics for workforce motivation | |
| CN113762675A (en) | Information generation method, device, server, system and storage medium | |
| CN110852392A (en) | User grouping method, device, equipment and medium | |
| Ezeh et al. | Factors influencing the choice of commercial bank among transport workers in Anambra state Nigeria | |
| US20180189696A1 (en) | System and method for measuring and monitoring innovation intelligence |
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 |