A method, device, computer equipment and storage medium for enriching user tags

Abstract

Embodiments of the present invention provide a method, apparatus, computer device, and storage medium for enriching user tags. In the technical solution provided by the embodiments of the present invention, a state value of a user terminal is obtained; a state matrix is ​​generated based on the state value; a similarity matrix is ​​generated by calculating the state matrix of the user terminal; a set of related user vectors for the current user is generated based on a preset similarity threshold, a stored content preference association matrix, and the similarity matrix. The set of related user vectors for the current user includes the tag association of the current user and the tag associations of related users; and the tag association of the current user is traversed and updated based on the tag associations of related users. This ensures the integrity of user information and improves the accuracy and richness of user tags.

Description

Technical field

[0001] The present invention relates to the field of network applications, and in particular to a method, device, computer equipment and storage medium for enriching user tags. [Background Technology]

[0002] With the rapid development of the Internet, more and more netizens are using mobile terminals. Mobile terminals generate a large amount of mobile data every day based on network behavior. Enterprises can conduct big data analysis on mobile data through the Internet, thereby achieving precise marketing based on user preferences.

[0003] Related technologies primarily analyze user behavior through big data and assign different tags to users based on the analysis results, recommending products based on these tags. Due to the large number of products, user tags often contain a large number of null values. Related technologies use single value decomposition or clustering methods to enrich user tags. However, these methods often lose user information when extracting key information, resulting in low accuracy in the generated user tags. [Summary of the invention]

[0004] In view of this, embodiments of the present invention provide a method, apparatus, computer device, and storage medium for enriching user tags, which can ensure the integrity of user information, thereby improving the accuracy and richness of user tags.

[0005] In one aspect, an embodiment of the present invention provides a method for enriching user tags, the method comprising:

[0006] Get the status value of the user terminal;

[0007] Generate a state matrix based on the state value;

[0008] Calculate the state matrix of the user terminal to generate a similarity matrix;

[0009] Generate a set of related user vectors for the current user based on a preset similarity threshold, a stored content preference association matrix, and a similarity matrix. The set of related user vectors for the current user includes the tag association of the current user and the tag association of related users.

[0010] According to the tag relevance of related users, the tag relevance of the current user is traversed and updated.

[0011] Optionally, the status value of the user terminal includes an average processor usage, an average memory usage, and an average screen lighting time;

[0012] According to the state value, generate the state matrix, including:

[0013] Normalizing the average processor usage to generate a first matrix;

[0014] Normalizing the average memory usage to generate a second matrix;

[0015] Normalize the average screen lighting time to generate a third matrix;

[0016] The first matrix, the second matrix, and the third matrix are concatenated to generate a state matrix.

[0017] Optionally, the content preference association matrix includes at least one user and a tag association degree of each user at each time interval;

[0018] Based on the preset similarity threshold, the stored content preference association matrix and the similarity matrix, a set of relevant user vectors for the current user is generated, including:

[0019] Select related users whose similarity with the current user is greater than a similarity threshold from the similarity matrix;

[0020] Match the relevant users with the content preference association matrix to obtain the tag association degree corresponding to the relevant users;

[0021] The tag association degrees of related users and the tag association degree of the current user are determined as a user vector set.

[0022] Optionally, based on the tag relevance of related users, the tag relevance of the current user is traversed and updated, including:

[0023] Traverse the tag relevance of the current user and obtain the time interval corresponding to the null value of the tag relevance of the current user;

[0024] Calculate the tag relevance of relevant users corresponding to the time interval to generate the updated relevance of the current user;

[0025] The updated relevance is determined as the tag relevance of the current user.

[0026] Optionally, calculating the tag relevance of relevant users corresponding to the time interval to generate the updated relevance of the current user includes:

[0027] Add the tag relevance of the relevant users corresponding to the time interval to generate an addition result;

[0028] Divide the sum by the total number of users to generate the update relevance of the current user.

[0029] Optionally, obtaining a status value of a user terminal includes:

[0030] During the set time period, the status value of the user terminal is obtained from the terminal log at preset time intervals.

[0031] Optionally, calculating the state matrix of the user terminal to generate a similarity matrix includes:

[0032] The state matrix of the user terminal is calculated using the Pearson coefficient to generate a similarity matrix.

[0033] On the other hand, an embodiment of the present invention provides a device for determining an auxiliary device, including:

[0034] An acquiring unit, configured to acquire a status value of a user terminal;

[0035] A first generating unit, configured to generate a state matrix according to the state value;

[0036] A second generating unit, configured to calculate a state matrix of the user terminal to generate a similarity matrix;

[0037] A third generating unit is configured to generate a set of related user vectors of the current user based on a preset similarity threshold, the stored content preference association matrix, and the similarity matrix, wherein the set of related user vectors of the current user includes a tag association degree of the current user and a tag association degree of related users;

[0038] The updating unit is used to traverse and update the tag relevance of the current user according to the tag relevance of related users.

[0039] On the other hand, an embodiment of the present invention provides a storage medium, which includes a stored program, wherein when the program is running, the device where the storage medium is located is controlled to execute the above-mentioned method for enriching user tags.

[0040] On the other hand, an embodiment of the present invention provides a computer device comprising a memory and a processor, wherein the memory is used to store information including program instructions, and the processor is used to control the execution of the program instructions, and is characterized in that the program instructions implement the above-mentioned user tag enrichment method when loaded and executed by the processor.

[0041] In the solution of the embodiment of the present invention, a state value of a user terminal is obtained; a state matrix is ​​generated based on the state value; the state matrix of the user terminal is calculated to generate a similarity matrix; a set of related user vectors of the current user is generated based on a preset similarity threshold, a stored content preference association matrix, and the similarity matrix, the set of related user vectors of the current user including the tag association of the current user and the tag association of related users; the tag association of the current user is traversed and updated based on the tag association of related users, thereby ensuring the integrity of user information and improving the accuracy and richness of user tags.

Brief Description of the Drawings

[0042] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0043] Figure 1 A flowchart of a method for enriching user tags provided by an embodiment of the present invention;

[0044] Figure 2 A flowchart of another method for enriching user tags provided by an embodiment of the present invention;

[0045] Figure 3 A schematic diagram of the structure of a user tag enrichment device provided by an embodiment of the present invention;

[0046] Figure 4 A schematic diagram of a computer device provided in an embodiment of the present invention. [Specific implementation method]

[0047] In order to better understand the technical solution of the present invention, the embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0048] It should be understood that the embodiments described are only a portion of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by persons of ordinary skill in the art without creative work are within the scope of protection of the present invention.

[0049] The terms used in the embodiments of the present invention are only for the purpose of describing specific embodiments and are not intended to limit the present invention. The singular forms "a", "an", "the" and "the" used in the embodiments of the present invention and the appended claims are also intended to include plural forms unless the context clearly indicates otherwise.

[0050] It should be understood that the term "and / or" as used herein is merely a description of the relationship between associated objects, indicating that three possible relationships exist. For example, "A and / or B" can represent: A exists alone, A and B exist simultaneously, or B exists alone. Furthermore, the character " / " in this document generally indicates that the associated objects are in an "or" relationship.

[0051] It should be understood that although the terms "first," "second," etc. may be used to describe the set thresholds in embodiments of the present invention, these set thresholds should not be limited to these terms. These terms are merely used to distinguish one set threshold from another. For example, without departing from the scope of embodiments of the present invention, the first set threshold may also be referred to as the second set threshold, and similarly, the second set threshold may also be referred to as the first set threshold.

[0052] Currently, in big data systems, the richness and accuracy of user tags directly determine the personalized relevance of product recommendations. The richer and more accurate the user tags, the more relevant the product recommendations. Related technologies add user tags by collecting basic user information and analyzing user usage behavior based on terminal applications. User tags primarily include descriptive tags, rule tags, behavior tags, interest tags, and group tags. User usage behavior based on terminal applications includes operations such as content search, content browsing, and purchases.

[0053] The data source of the embodiment of the present invention not only focuses on the user's terminal usage behavior, but also focuses on the terminal logs generated by the terminal hardware. It profiles the user through three factors: processor (CPU), memory, and screen lighting time per unit time, and proposes a method for analyzing the similarity between users from a new perspective.

[0054] Figure 1 A flowchart of a method for enriching user tags provided by an embodiment of the present invention is shown in FIG. Figure 1 As shown, the method includes:

[0055] Step 101: Obtain the status value of the user terminal.

[0056] Step 102: Generate a state matrix according to the state value.

[0057] Step 103: Calculate the state matrix of the user terminal to generate a similarity matrix.

[0058] Step 104: Generate a set of related user vectors for the current user based on a preset similarity threshold, the stored content preference association matrix, and the similarity matrix. The set of related user vectors for the current user includes the tag association of the current user and the tag association of related users.

[0059] Step 105: Based on the tag relevance of related users, the tag relevance of the current user is traversed and updated.

[0060] In the technical solution provided by an embodiment of the present invention, a state value of a user terminal is obtained; a state matrix is ​​generated based on the state value; the state matrix of the user terminal is calculated to generate a similarity matrix; a set of related user vectors of the current user is generated based on a preset similarity threshold, a stored content preference association matrix, and the similarity matrix, the set of related user vectors of the current user including the tag association of the current user and the tag association of related users; and the tag association of the current user is traversed and updated based on the tag association of related users, thereby ensuring the integrity of user information and improving the accuracy and richness of user tags.

[0061] Figure 2 A flowchart of another method for enriching user tags provided by an embodiment of the present invention is shown in FIG. Figure 2 As shown, the method includes:

[0062] Step 201: Obtain the status value of the user terminal.

[0063] In the embodiment of the present invention, each step is executed by a server.

[0064] In the embodiment of the present invention, the user terminal includes but is not limited to a mobile phone, a tablet computer, a wearable device, a speaker or a personal computer.

[0065] In the embodiment of the present invention, the status value of the user terminal includes the average usage rate of the processor (CPU), the average usage rate of the memory (RAM), and the average screen lighting time.

[0066] Specifically, within the set time period, the status value of the user terminal is obtained from the terminal log at a preset time interval. During the operation of the user terminal, a large number of terminal logs will be generated, and the terminal logs will record information about the user terminal during operation, such as: CPU usage, RAM usage, and screen lighting time. Among them, the time period and time interval can be set according to actual needs. As an optional solution, the time period is set to 24 hours and the time interval is set to 1 hour. The server calculates the average CPU usage within 1 hour based on the obtained CPU usage; calculates the average RAM usage within 1 hour based on the obtained RAM usage; and calculates the average screen lighting time within 1 hour based on the obtained screen lighting time.

[0067] Step 202: Generate a state matrix according to the state value.

[0068] In the embodiment of the present invention, step 202 specifically includes:

[0069] Step 2021: Normalize the average CPU usage to generate a first matrix.

[0070] In the embodiment of the present invention, the average CPU usage is normalized by using a Min-max normalization method to generate a first matrix. The first matrix includes multiple users and the normalized CPU usage of each user in each time interval.

[0071] Specifically, the average CPU usage is calculated using the formula Cij = (Cxj - Cmin) / (Cmax - Cmin) to generate a first matrix. Here, Cij is the normalized CPU usage of user i in the jth hour, Cxj is the average CPU usage of user x in the jth hour, Cmin is the minimum average CPU usage of user x across all time intervals in the time period, and Cmax is the maximum average CPU usage of user x across all time intervals in the time period. For example, the first matrix is ​​shown in Table 1.

[0072] Table 1

[0073] T1 T2 T3 … T24 U1 <![CDATA[C 1,1 ]]> <![CDATA[C 1,2 ]]> <![CDATA[C 1,3 ]]> … <![CDATA[C 1,24 ]]> U2 <![CDATA[C 2,1 ]]> <![CDATA[C 2,2 ]]> <![CDATA[C 2,3 ]]> … <![CDATA[C 2,24 ]]> … … … … … …

[0074] Among them, T1 is the first time interval, T2 is the second time interval, T3 is the third time interval, T24 is the 24th time interval, U1 is user 1, U2 is user 2, C 1,1 is the normalized CPU usage of user 1 in the first hour, C 1,2 is the normalized CPU usage of user 1 in the second hour, C 2,1 The normalized CPU usage of user 2 in the first hour, and so on.

[0075] Step 2022: Normalize the average RAM usage to generate a second matrix.

[0076] In the embodiment of the present invention, the average RAM usage is normalized by using a Min-max normalization method to generate a second matrix. The second matrix includes multiple users and the normalized RAM usage of each user in each time interval.

[0077] Specifically, the average RAM usage is calculated using the formula Rij = (Rxj - Rmin) / (Rmax - Rmin) to generate a second matrix. Rij is the normalized RAM usage of user i in the jth hour, Rxj is the average RAM usage of user x in the jth hour, Rmin is the minimum average RAM usage of user x across all time intervals in the time period, and Rmax is the maximum average RAM usage of user x across all time intervals in the time period. For example, the second matrix is ​​shown in Table 2.

[0078] Table 2

[0079] T1 T2 T3 … T24 U1 <![CDATA[R 1,1 ]]> <![CDATA[R 1,2 ]]> <![CDATA[R 1,3 ]]> … <![CDATA[R 1,24 ]]> U2 <![CDATA[R 2,1 ]]> <![CDATA[R 2,2 ]]> <![CDATA[R 2,3 ]]> … <![CDATA[R 2,24 ]]> … … … … … …

[0080] Where, T1 is the first time interval, T2 is the second time interval, T3 is the third time interval, T24 is the 24th time interval, U1 is user 1, U2 is user 2, R 1,1 is the normalized RAM usage of user 1 in the first hour, R 1,2 is the normalized RAM usage of user 1 in the second hour, R 2,1 is the normalized RAM usage of user 2 in the first hour, and so on.

[0081] Step 2023: Normalize the average screen lighting time to generate a third matrix.

[0082] In an embodiment of the present invention, the average screen lighting time is normalized by a Min-max normalization method to generate a third matrix. The third matrix includes multiple users and the normalized screen lighting time of each user in each time interval.

[0083] Specifically, the average screen on time is calculated using the formula Dij = (Dxj - Rmin) / (Dmax - Dmin) to generate a third matrix. Dij is the normalized screen on time of user i in the jth hour, Dxj is the average screen on time of user x in the jth hour, Dmin is the minimum average screen on time of user x across all time intervals in the time period, and Dmax is the maximum average screen on time of user x across all time intervals in the time period. For example, the third matrix is ​​shown in Table 3.

[0084] Table 3

[0085] T1 T2 T3 … T24 U1 <![CDATA[D 1,1 ]]> <![CDATA[D 1,2 ]]> <![CDATA[D 1,3 ]]> … <![CDATA[D 1,24 ]]> U2 <![CDATA[D 2,1 ]]> <![CDATA[D 2,2 ]]> <![CDATA[D 2,3 ]]> … <![CDATA[D 2,24 ]]> … … … … … …

[0086] Among them, T1 is the first time interval, T2 is the second time interval, T3 is the third time interval, T24 is the 24th time interval, U1 is user 1, U2 is user 2, D 1,1 D is the normalized screen on time of user 1 in the first hour. 1,2 D is the normalized screen on time of user 1 in the second hour. 2,1 The normalized screen on time for user 2 in the first hour, and so on.

[0087] Step 2024: Concatenate the first matrix, the second matrix, and the third matrix to generate a state matrix.

[0088] Specifically, the first matrix, the second matrix, and the third matrix are vector-concatenated to generate a state matrix, which includes multiple eigenvalues. The state matrix can represent the usage status of the user terminal in a time period. Taking the time period as 24 hours as an example, the usage status of the user terminal of the i-th user in 24 hours is Ui=(C i,1 , C i,2 ,…,C i,24 , R i,1 , R i,2 ,…,R i,24 , D i,1 , D i,2 ,…,D i,24 ), where C i,1 , C i,2 ,…,C i,24 , R i,1 , R i,2 ,…,R i,24 , D i,1 , D i,2 ,…,D i,24 are the 72 eigenvalues ​​of user i. For example, the state matrix is ​​shown in Table 4.

[0089] Table 4

[0090]

[0091] Step 203: Calculate the state matrix of the user terminal using the Pearson coefficient to generate a similarity matrix, where the similarity matrix includes similarities between users.

[0092] Specifically, through the Pearson coefficient The state matrix of the user terminal is calculated to generate a similarity matrix, which includes the similarities between multiple users. ij is the similarity between user i and user j, U i,k is the value of user i at the kth eigenvalue, U j,k is the value of user j at the kth eigenvalue, is the average value of user i, is the average value of user j, and G is the total number of eigenvalues ​​of the user. For example, the similarity matrix is ​​shown in Table 5.

[0093] Table 5

[0094] User similarity U1 U2 U3 U4 U5 … U1 <![CDATA[w 1,1 ]]> <![CDATA[w 1,2 ]]> <![CDATA[w 1,3 ]]> <![CDATA[w 1,4 ]]> <![CDATA[w 1,5 ]]> … U2 <![CDATA[w 2,1 ]]> <![CDATA[w 2,2 ]]> <![CDATA[w 2,3 ]]> <![CDATA[w 2,4 ]]> <![CDATA[w 2,5 ]]> … U3 <![CDATA[w 3,1 ]]> <![CDATA[w 3,2 ]]> <![CDATA[w 3,3 ]]> <![CDATA[w 3,4 ]]> <![CDATA[w 3,5 ]]> … U4 <![CDATA[w 4,1 ]]> <![CDATA[w 4,2 ]]> <![CDATA[w 4,3 ]]> <![CDATA[w 4,4 ]]> <![CDATA[w 4,5 ]]> … U5 <![CDATA[w 5,1 ]]> <![CDATA[w 5,2 ]]> <![CDATA[w 5,3 ]]> <![CDATA[w 5,4 ]]> <![CDATA[w 5,5 ]]> … … … … … … … …

[0095] Among them, w 1,1 is the user similarity between user 1 and user 1, w 1,2 is the user similarity between user 1 and user 2, w 1,3is the user similarity between user 1 and user 3, and so on.

[0096] Step 204: Select related users whose similarity to the current user is greater than a similarity threshold from the similarity matrix.

[0097] In the embodiment of the present invention, the similarity threshold has a similarity range, for example, the similarity range includes (0.5, 1), and the specific value of the similarity threshold can be set according to actual test results. As an optional solution, the similarity threshold is 0.75.

[0098] Specifically, a current user is pre-selected. Based on the current user, users whose similarity to the current user exceeds a similarity threshold are selected from the similarity matrix. These selected users are then identified as related users. For example, if the current user is user 1 (U1), the related users identified from the similarity matrix are user 3 (U3), user 4 (U4), and user 7 (U7).

[0099] Step 205: Match the relevant users with the content preference association matrix to obtain the tag association degrees corresponding to the relevant users.

[0100] In this embodiment of the present invention, the content preference association matrix is ​​a user-tag association matrix based on content preferences pre-stored in a storage medium. The content preference association matrix includes at least one user and the tag association degree of each user at each time interval. For example, the content preference association matrix is ​​shown in Table 6.

[0101] Table 6

[0102]

[0103] Among them, NULL means that the tag similarity of user 1 in the first time interval is null value, UsrTag 1,2 represents the tag similarity value of user 1 in the second time interval, and so on. As can be seen from Table 6, there are many null values ​​in the content preference matrix, that is, the sparsity of the content preference matrix is ​​high.

[0104] For example: the current user is selected as user 1 (U1), and the related users are user 3 (U3), user 4 (U4) and user 7 (U7); the related users are matched with the content preference association matrix to match the tag association degree corresponding to the related users. The tag association degree of the related user U3 is UsrTag 3,1 ,NULL,NULL,UsrTag 3,4 、UsrTag 3,5 ; The tag association degree of the relevant user U4 is NULL, UsrTag 4,2 、UsrTag 4,3 、UsrTag4,4 、UsrTag 4,5 ; The tag relevance of the relevant user U7 is UsrTag 7,1 ,NULL,UsrTag 7,3 ,NULL,NULL.

[0105] Step 206: Determine the tag association degrees of the related users and the tag association degree of the current user as a user vector set.

[0106] For example: the tag association of the current user U1 is NULL, UsrTag 1,2 ,NULL,UsrTag 1,4 、UsrTag 1,5 ; The tag relevance of the relevant user U3 is UsrTag 3,1 ,NULL,NULL,UsrTag 3,4 、UsrTag 3,5 ; The tag association degree of the relevant user U4 is NULL, UsrTag 4,2 、UsrTag 4,3 、UsrTag 4,4 、UsrTag 4,5 ; The tag relevance of the relevant user U7 is UsrTag 7,1 ,NULL,UsrTag 7,3 , NULL, NULL. The user vector set Us is shown in Table 7.

[0107] Table 7

[0108] Us T1 T2 T3 T4 T5 … U1 NULL <![CDATA[UsrTag 1,2 ]]> NULL <![CDATA[UsrTag 1,4 ]]> <![CDATA[UsrTag 1,5 ]]> … U3 <![CDATA[UsrTag 3,1 ]]> NULL NULL <![CDATA[UsrTag 3,4 ]]> <![CDATA[UsrTag 3,5 ]]> … U4 NULL <![CDATA[UsrTag 4,2 ]]> <![CDATA[UsrTag 4,3 ]]> <![CDATA[UsrTag 4,4 ]]> <![CDATA[UsrTag 4,5 ]]> … U7 <![CDATA[UsrTag 7,1 ]]> NULL <![CDATA[UsrTag 7,3 ]]> NULL NULL …

[0109] Step 207: Traverse the tag relevance of the current user to obtain the time interval corresponding to the null value of the tag relevance of the current user.

[0110] In this embodiment of the present invention, the tag association degree of the current user U1 is NULL, UsrTag 1,2 ,NULL,UsrTag 1,4 、UsrTag 1,5 , the time intervals corresponding to the null values ​​are the first time interval T1 and the third time interval T3.

[0111] Step 208: Calculate the tag relevance of the relevant users corresponding to the time interval to generate the updated relevance of the current user.

[0112] Specifically, through the formula Calculate the tag relevance of the relevant users corresponding to the time interval to generate the updated relevance of the current user. h,lis the tag association of user h, S is all related users in the user vector set, and M is the total number of users.

[0113] In the embodiment of the present invention, step 208 specifically includes:

[0114] Step 2081: Add the tag associations of the relevant users corresponding to the time interval to generate an addition result.

[0115] In the embodiment of the present invention, the tag association degree of the relevant user U3 corresponding to the first time interval is UsrTag 3,1 , the tag association degree corresponding to the relevant user U4 is NULL, and the tag association degree corresponding to the relevant user U7 is UsrTag 7,1 . Add up the tag associations of the relevant users corresponding to the time interval, i.e.: UsrTag 3,1 +UsrTag 7,1 .

[0116] Step 2082: Divide the sum by the total number of users to generate the update relevance of the current user.

[0117] In the embodiment of the present invention, the addition result is UsrTag 3,1 +UsrTag 7,1 , the total number of users is M, then the update relevance of the current user U1 is (UsrTag 3,1 +UsrTag 7,1 ) / M.

[0118] Step 209: Determine the updated relevance as the tag relevance of the current user.

[0119] In the embodiment of the present invention, the updated user vector machine is shown in Table 8.

[0120] Table 8

[0121]

[0122] In the technical solution of the user tag enrichment method provided by an embodiment of the present invention, a state value of a user terminal is obtained; a state matrix is ​​generated based on the state value; the state matrix of the user terminal is calculated to generate a similarity matrix, wherein the similarity matrix includes similarities between users; a set of related user vectors of the current user is generated based on a preset similarity threshold, a stored content preference association matrix, and the similarity matrix, wherein the set of related user vectors of the current user includes the tag association of the current user and the tag association of related users; and the tag association of the current user is traversed and updated based on the tag association of related users, thereby ensuring the integrity of user information and improving the accuracy and enrichment of user tags.

[0123] Figure 3 This is a schematic diagram of the structure of a user tag enrichment device provided by an embodiment of the present invention, which is used to execute the above-mentioned user tag enrichment method, such as Figure 3 As shown, the device includes: an acquisition unit 11, a first generation unit 12, a second generation unit 13, a third generation unit 14 and an update unit 15.

[0124] The acquiring unit 11 is configured to acquire a status value of a user terminal.

[0125] The first generating unit 12 is used to generate a state matrix according to the state value.

[0126] The second generating unit 13 is used to calculate the state matrix of the user terminal to generate a similarity matrix.

[0127] The third generating unit 14 is configured to generate a set of related user vectors of the current user according to a preset similarity threshold, the stored content preference association matrix and the similarity matrix. The set of related user vectors of the current user includes the tag association of the current user and the tag association of related users.

[0128] The updating unit 15 is configured to traverse and update the tag relevance of the current user according to the tag relevance of related users.

[0129] In an embodiment of the present invention, the first generation unit 12 is specifically used to normalize the average processor usage to generate a first matrix; normalize the average memory usage to generate a second matrix; normalize the average screen lighting time to generate a third matrix; and splice the first matrix, the second matrix, and the third matrix to generate a state matrix.

[0130] In an embodiment of the present invention, the third generation unit 14 is specifically used to select related users whose similarity with the current user is greater than a similarity threshold from the similarity matrix; match the related users with the content preference association matrix to match the tag associations corresponding to the related users; and determine the tag associations of the related users and the tag associations of the current user as a user vector set.

[0131] In an embodiment of the present invention, the updating unit 15 is specifically configured to traverse the tag associations of the current user, obtain the time interval corresponding to the null value of the tag association of the current user; calculate the tag associations of the relevant users corresponding to the time interval to generate the updated association of the current user; and determine the updated association as the tag association of the current user.

[0132] In the embodiment of the present invention, the updating unit 15 is specifically configured to add the tag associations of the relevant users corresponding to the time interval to generate an addition result; and divide the addition result by the total number of users to generate the updated association of the current user.

[0133] In the embodiment of the present invention, the acquiring unit 11 is specifically configured to acquire the status value of the user terminal from the terminal log at preset time intervals within a set time period.

[0134] In the embodiment of the present invention, the second generating unit 13 is specifically configured to calculate the state matrix of the user terminal by using the Pearson coefficient to generate a similarity matrix.

[0135] In the solution of the embodiment of the present invention, a state value of a user terminal is obtained; a state matrix is ​​generated based on the state value; the state matrix of the user terminal is calculated to generate a similarity matrix; a set of related user vectors of the current user is generated based on a preset similarity threshold, a stored content preference association matrix, and the similarity matrix, the set of related user vectors of the current user including the tag association of the current user and the tag association of related users; the tag association of the current user is traversed and updated based on the tag association of related users, thereby ensuring the integrity of user information and improving the accuracy and richness of user tags.

[0136] An embodiment of the present invention provides a storage medium, which includes a stored program, wherein when the program is running, the device where the storage medium is located is controlled to execute the steps of the embodiment of the above-mentioned user tag enrichment method. For a specific description, please refer to the embodiment of the above-mentioned user tag enrichment method.

[0137] An embodiment of the present invention provides a computer device including a memory and a processor, wherein the memory is used to store information including program instructions, and the processor is used to control the execution of the program instructions. When the program instructions are loaded and executed by the processor, the steps of the embodiment of the above-mentioned method for enriching user tags are implemented. For a specific description, please refer to the embodiment of the above-mentioned method for enriching user tags.

[0138] Figure 4 Schematic diagram of a computer device provided by an embodiment of the present invention. Figure 4 As shown, the computer device 30 of this embodiment includes: a processor 31, a memory 32, and a computer program 33 stored in the memory 32 and executable on the processor 31. When executed by the processor 31, the computer program 33 implements the method for enriching user tags in the embodiment. To avoid repetition, a detailed description is not given here. Alternatively, when executed by the processor 31, the computer program implements the functions of each model / unit in the device for enriching user tags in the embodiment. To avoid repetition, a detailed description is not given here.

[0139] The computer device 30 includes, but is not limited to, a processor 31 and a memory 32. Those skilled in the art will understand that Figure 4This is merely an example of the computer device 30 and does not constitute a limitation of the computer device 30 . The computer device 30 may include more or fewer components than shown in the figure, or a combination of certain components, or different components. For example, the computer device may also include input and output devices, network access devices, buses, etc.

[0140] The processor 31 may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or any conventional processor.

[0141] The memory 32 can be an internal storage unit of the computer device 30, such as a hard drive or memory of the computer device 30. The memory 32 can also be an external storage device of the computer device 30, such as a plug-in hard drive, a Smart Media (SM) card, a Secure Digital (SD) card, a Flash Card, etc. equipped on the computer device 30. Furthermore, the memory 32 can include both an internal storage unit of the computer device 30 and an external storage device. The memory 32 is used to store computer programs and other programs and data required by the computer device. The memory 32 can also be used to temporarily store data that has been output or is about to be output.

[0142] The various devices and products described in the above embodiments include modules / units, which may be software modules / units, hardware modules / units, or partially software modules / units and partially hardware modules / units. For example, for various devices and products that apply to or integrate chips, the various modules / units contained therein may all be implemented in the form of hardware such as circuits, or at least some of the modules / units may be implemented in the form of software programs, which run on the integrated processor inside the chip, and the remaining (if any) modules / units may be implemented in the form of hardware such as circuits; for various devices and products that apply to or integrate chip modules, the various modules / units contained therein may all be implemented in the form of hardware such as circuits, and different modules / units may be located in the same part of the chip module (such as a chip, circuit module, etc.) or in different components, and at least some / units may be implemented in the form of software programs. The software program runs on the processor integrated in the chip module, and the remaining modules / units (if any) can be implemented in hardware such as circuits. For various devices and products applied to or integrated in the terminal, the modules / units contained therein can all be implemented in hardware such as circuits. Different modules / units can be located in the same component (for example, chip, circuit module, etc.) or different components in the terminal, or at least some modules / units can be implemented in software program, which runs on the processor integrated in the terminal, and the remaining modules / units (if any) can be implemented in hardware such as circuits.

[0143] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A method for enriching user tags, characterized in that: The method comprises: Get the status value of the user terminal; generating a state matrix according to the state value; Calculating the state matrix of the user terminal to generate a similarity matrix; Generate a set of related user vectors for the current user based on a preset similarity threshold, a stored content preference association matrix, and the similarity matrix, wherein the set of related user vectors for the current user includes a tag association degree of the current user and a tag association degree of related users; According to the tag relevance of the related users, the tag relevance of the current user is traversed and updated; The status values ​​of the user terminal include average processor usage, average memory usage, and average screen lighting time; Generating a state matrix according to the state value includes: Normalizing the average processor usage to generate a first matrix; Normalizing the average memory usage to generate a second matrix; Normalizing the average screen lighting time to generate a third matrix; concatenating the first matrix, the second matrix, and the third matrix to generate the state matrix; The content preference association matrix includes at least one user and the tag association degree of each user in each time interval; The generating of a set of relevant user vectors of the current user according to a preset similarity threshold, a stored content preference association matrix and the similarity matrix includes: Selecting the related users whose similarity to the current user is greater than the similarity threshold from the similarity matrix; Matching the relevant users with the content preference association matrix to obtain the tag association degrees corresponding to the relevant users; The tag association degrees of the related users and the tag association degree of the current user are determined as the user vector set.

2. The method according to claim 1, characterized in that The step of traversing and updating the tag relevance of the current user according to the tag relevance of the related users includes: Traversing the tag relevance of the current user to obtain the time interval corresponding to the null value of the tag relevance of the current user; Calculating the tag association degree of the relevant users corresponding to the time interval to generate an updated association degree of the current user; The updated relevance is determined as the tag relevance of the current user.

3. The method according to claim 2, characterized in that The calculating the tag association degree of the relevant users corresponding to the time interval to generate the updated association degree of the current user includes: Adding the tag association degrees of the relevant users corresponding to the time interval to generate an addition result; The added result is divided by the total number of users to generate the update relevance of the current user.

4. The method according to claim 1, wherein The obtaining of the status value of the user terminal includes: During the set time period, the status value of the user terminal is obtained from the terminal log at preset time intervals.

5. The method according to claim 1, characterized in that The calculating the state matrix of the user terminal to generate a similarity matrix includes: The state matrix of the user terminal is calculated using the Pearson coefficient to generate a similarity matrix.

6. A user tag enrichment device, characterized in that: The device comprises: An acquiring unit, configured to acquire a status value of a user terminal; A first generating unit, configured to generate a state matrix according to the state value; A second generating unit, configured to calculate the state matrix of the user terminal to generate a similarity matrix; a third generating unit, configured to generate a set of related user vectors of the current user based on a preset similarity threshold, a stored content preference association matrix, and the similarity matrix, wherein the set of related user vectors of the current user includes a tag association degree of the current user and a tag association degree of related users; An updating unit, configured to traverse and update the tag relevance of the current user according to the tag relevance of the related users; The status values ​​of the user terminal include average processor usage, average memory usage, and average screen lighting time; The first generating unit is specifically configured to normalize the average processor usage to generate a first matrix; normalize the average memory usage to generate a second matrix; normalize the average screen lighting time to generate a third matrix; and concatenate the first matrix, the second matrix, and the third matrix to generate the state matrix; The content preference association matrix includes at least one user and the tag association degree of each user in each time interval; The third generation unit is specifically configured to select, from the similarity matrix, the related users whose similarity with the current user is greater than the similarity threshold; match the related users with the content preference association matrix to obtain the tag associations corresponding to the related users; and determine the tag associations of the related users and the tag associations of the current user as the user vector set.

7. A storage medium, characterized in that: The storage medium includes a stored program, wherein when the program is running, the device where the storage medium is located is controlled to execute the user tag enrichment method according to any one of claims 1 to 5.

8. A computer device comprising a memory and a processor, wherein the memory is used to store information including program instructions, and the processor is used to control the execution of the program instructions, characterized in that: When the program instructions are loaded and executed by a processor, the method for enriching user tags according to any one of claims 1 to 5 is implemented.

Images