CN115098797A - Friend recommendation system and method based on ternary closure - Google Patents
Friend recommendation system and method based on ternary closure Download PDFInfo
- Publication number
- CN115098797A CN115098797A CN202210932275.3A CN202210932275A CN115098797A CN 115098797 A CN115098797 A CN 115098797A CN 202210932275 A CN202210932275 A CN 202210932275A CN 115098797 A CN115098797 A CN 115098797A
- Authority
- CN
- China
- Prior art keywords
- user
- users
- data
- weight
- friend
- 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 26
- 230000000694 effects Effects 0.000 claims abstract description 68
- 230000007246 mechanism Effects 0.000 claims abstract description 10
- 238000013480 data collection Methods 0.000 claims abstract description 9
- 238000012545 processing Methods 0.000 claims abstract description 9
- 230000008447 perception Effects 0.000 claims abstract description 8
- 238000007781 pre-processing Methods 0.000 claims abstract description 6
- 230000036651 mood Effects 0.000 claims description 7
- 230000008569 process Effects 0.000 claims description 6
- 230000006996 mental state Effects 0.000 claims description 5
- 230000006872 improvement Effects 0.000 description 4
- 238000004891 communication Methods 0.000 description 3
- 206010012374 Depressed mood Diseases 0.000 description 2
- 238000011161 development Methods 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 230000006870 function Effects 0.000 description 2
- 230000003993 interaction Effects 0.000 description 2
- 230000008450 motivation Effects 0.000 description 2
- 206010044565 Tremor Diseases 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000015572 biosynthetic process Effects 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 230000002708 enhancing effect Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 239000002699 waste material Substances 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/95—Retrieval from the web
- G06F16/953—Querying, e.g. by the use of web search engines
- G06F16/9536—Search customisation based on social or collaborative filtering
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/95—Retrieval from the web
- G06F16/953—Querying, e.g. by the use of web search engines
- G06F16/9537—Spatial or temporal dependent retrieval, e.g. spatiotemporal queries
-
- 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
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/01—Social networking
Landscapes
- Engineering & Computer Science (AREA)
- Databases & Information Systems (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Business, Economics & Management (AREA)
- Data Mining & Analysis (AREA)
- General Engineering & Computer Science (AREA)
- Economics (AREA)
- Health & Medical Sciences (AREA)
- Computing Systems (AREA)
- General Health & Medical Sciences (AREA)
- Human Resources & Organizations (AREA)
- Marketing (AREA)
- Primary Health Care (AREA)
- Strategic Management (AREA)
- Tourism & Hospitality (AREA)
- General Business, Economics & Management (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention belongs to the technical field of social software development, and particularly relates to a friend recommendation system and method based on a ternary closure, which comprises a data collection module, a data processing module, a weight calculation module, a state matching module and a friend recommendation module; the data collection module judges whether to update the data according to the geographic position and time of the user; the data processing module is used for preprocessing the activity data of the user and clustering the activity data of the user; the weight calculation module calculates the weight of a user to be recommended which may form a side with other users, adds the calculated weight into a queue to be recommended, introduces an opening coefficient, updates the weight, and then adds the queue to be recommended; the state matching module introduces a state perception mechanism before recommendation, and matches the state of the user with the states of other users; and the friend recommending module selects m users for recommending, sorts the weights of other users and the user after state matching according to the magnitude sequence for each user, and selects the first m users for recommending.
Description
Technical Field
The invention belongs to the technical field of social software development, and particularly relates to a friend recommendation system and method based on a ternary closure.
Background
With the research of social networks and the rapid development of online social networks, online social activities become an indispensable part of human lives, and people share their lives and find like-minded friends through social platforms such as QQ, microblog, tremble, twitter and the like. In order to meet the demand of people for making friends, the platforms provide friend recommendation functions, and the friend recommendation functions analyze the preference of users according to personal information filled by the users and the number of common friends and recommend the users to be matched with the interests of the users.
However, most of current friend recommendations are concentrated on communication devices such as mobile phones and computers, friend recommendations on a car machine are few, and with the popularization of new energy cars, more and more people are believed to use the car machine for communication, so that people also need to cover the friend recommendations on the car machine. Current friend recommendations mainly focus on the social network of users, personal information of users and activities of users on a platform, but these methods usually focus on only the external factors (homogeneity) of the formation of the ternary closure, and ignore the influence (motivation) of the internal factors, thereby causing inaccuracy of recommendations and waste of resources. For example, friend recommendation based on a user social network mostly concerns the number of common friends, and does not consider the motivation of the user whether to add strange friends or not; for the personal information friend recommendation of the user, the completeness and timeliness of the information are ignored, and inaccurate recommendation can be caused; analyzing interest and hobbies based on the online activities of the users, and ignoring the influence of the social network relation on the ternary closure. Therefore, the friend recommendation systems are too simple, and various influence factors are not fully considered, so that accurate recommendation is difficult to achieve, and the requirements of people who really have friend making ideas cannot be met.
Disclosure of Invention
Aiming at the defects of the existing friend recommendation system, the invention provides a friend recommendation system and method based on a ternary closure, and aims to solve the technical problems that the recommendation success rate in the friend recommendation system is low, the users are wrongly matched, the considered factors are too extensive, and humanized recommendation is not achieved.
A friend recommendation system and method based on ternary closure includes the following steps:
step one, judging whether to update data according to the geographical position and time of the user, if the geographical position of the user changes and the time difference between the geographical position of the user and the last data acquisition time is one week, updating the information of the user, and if not, ending the recommendation process to avoid repeated recommendation.
And secondly, preprocessing the data set of the user, wherein the extracted activities of the user have certain repeatability, and the data redundancy is reduced by clustering the activities, so that the data is more representative.
And step three, calculating the weight of the side possibly formed by the user to be recommended and other users, adding the calculated weight value into a queue to be recommended, introducing an opening coefficient, and updating the weight value.
And step four, introducing a state perception mechanism, analyzing the current state of the user, and performing user state matching according to the mood, the mental state and the busy/idle state analysis of the user.
And fifthly, sorting the users subjected to state matching according to the weight value, and then selecting the users for recommendation.
In a further development of the invention, the first step comprises: judging whether to acquire data or not by acquiring the geographical position of a user and the time of requesting updating, and setting two conditions: the first condition is whether the geographic position obtained during the recommendation is consistent with the geographic position recommended for the last time, the second condition is whether the time for the recommendation is different from the time for the recommendation for the last time by one week, if the time for the recommendation is different from the time for the last time and the time is different from the time for the last time by one week, the data updating is carried out, otherwise, the whole recommendation process is ended, the data updating is not carried out, and the repeated recommendation is reduced.
In a further improvement of the present invention, the second step includes: the data obtained are very redundant, some activities are related to each other and can be classified into a class, so that the activities are carried outClustering, for how to measure the correlation between different activities, a set of characteristics such as (challenging, team, etc.) is selected, if a certain activity has a certain characteristic, the weight of the activity pointing to the edge is 1, otherwise, the weight is 0, and for the correlation between activity a and activity b, ρ can be used ab Where cosine similarity can be chosen, denoted by Correlation (a, b):
and performing active clustering to reduce data redundancy.
In a further improvement of the present invention, the third step includes: and calculating the weight of the edge which can be formed by the user to be recommended and other users according to the principle of the ternary closure, and calculating the weight from two aspects including user-user and user-activity-user. The calculated weight value is added into the queue to be recommended, considering that the user may not add strange users, an openness coefficient gamma is introduced, the gamma can be 0 or 1, when the gamma is 0, the user does not add strange friends, when the gamma is 1, the user experiences the strange friends, and for updating the weight value of the queue to be recommended, the calculation formula is as follows: w ═ γ × w ab +γ
Where w represents the weight of the user forming an edge with other users, γ is the openness coefficient, w ab The initial weight obtained in the two ternary closure modes is shown. And sorting the calculated weights in a descending order, and adding the weights into a queue to be recommended.
In a further improvement of the present invention, the fourth step includes: a state sensing mechanism is introduced before recommendation, and the mood, the mental state, the busy-idle state and the like of a user are observed and analyzed by using sensors and cameras on a vehicle machine and a mobile phone, so that the state of the user can be matched with the states of other users.
In a further improvement of the present invention, the fifth step includes: and selecting m users for the user to recommend, and for each user, sorting the weights of the other users after state matching with the user according to the order of magnitude, and selecting the first m users from the weights to recommend.
The invention also discloses a friend recommendation system based on the ternary closure, which comprises a data collection module, a data processing module, a weight calculation module, a state matching module and a friend recommendation module.
And the data collection module judges whether to update the data according to the geographic position and the time of the user.
And the data processing module is used for preprocessing the activity data of the user and clustering the activity data of the user.
And the weight calculation module is used for calculating the weight of the side possibly formed by the user to be recommended and other users, introducing an opening coefficient, updating the weight value and then adding the weight value into the queue to be recommended.
And the state matching module introduces a state perception mechanism before recommendation, and can match the state of the user at the moment with the states of other users.
And the friend recommending module selects m users for recommending, sorts the weights of other users and the user after state matching according to the magnitude sequence for each user, and selects the first m users for recommending.
The invention has the beneficial effects that: the existing friend recommendation method only considers a single factor influencing the ternary closure, often neglects the influence of a user in many aspects when the user becomes a friend, and the existing friend recommendation is mostly concentrated on communication equipment such as a mobile phone, a computer and the like, and the coverage of friend recommendation on a vehicle machine is insufficient, so that the interaction of the mobile phone and the vehicle machine is utilized to obtain user data, the correlation among data in the data set of the user is considered, so that a cluster of user activity data is provided, the correlation among the data is reduced, then two aspects influencing the ternary closure are utilized, a user-user and a user-activity-user work out initial weights which are possibly formed by the user to be recommended and other users, then the condition that possible users do not add strange friends is also considered, and an openness coefficient is introduced, the method comprises the steps of updating the weight, introducing a human state perception mechanism before recommending the user, performing state matching on the user to be recommended and other users, and recommending the user with the state matching completed and the weight being front to the user.
Drawings
FIG. 1 is a flow chart of a method embodying the present invention.
FIG. 2 is a schematic diagram of clustering user activities.
Fig. 3 is a schematic diagram of a recommendation system implemented by the present invention.
Detailed Description
For the purpose of enhancing the understanding of the present invention, the present invention will be described in further detail with reference to the accompanying drawings and examples, which are provided for the purpose of illustration only and are not intended to limit the scope of the present invention.
Fig. 1 is a flowchart of a method implemented by the present invention, and a friend recommendation system and method based on a ternary closure, including the following steps:
step one, judging whether to update data according to the geographical position and time of the user, if the geographical position of the user changes and the time difference between the geographical position of the user and the last data acquisition time is one week, updating the information of the user, and if not, ending the recommendation process to avoid repeated recommendation.
In the data acquisition stage, data acquisition is carried out through interaction of a vehicle machine and a mobile phone, online social software acquires open-source data through collecting geographic position information, online social activity information and personal data of a user, in the data acquisition stage, whether data updating is carried out or not is judged, repeated recommendation of the user can be caused if the geographic position of the user is not changed during recommendation, interests and hobbies of people are timeliness, some activities are favored within a period of time, other activities are favored possibly within a period of time, whether the two times of data acquisition time are different by one week is also judged, and the whole friend recommendation process is finished if the two conditions are not met simultaneously. Otherwise, updating data is carried out, and the obtained data comprises activities and friend relations of all users in the geographical position in one year.
And secondly, preprocessing the data set of the user, wherein the extracted activities of the user have certain repeatability, and the data redundancy is reduced by clustering the activities, so that the data is more representative.
The obtained data has great redundancy, some activities are related to each other and can be classified into one type, so as to perform clustering of the activities, for how to measure the correlation among different activities, a group of characteristics such as (challenging, team and the like) is selected, if a certain activity has a certain characteristic, the weight of the certain activity pointing to the edge is 1, otherwise, the certain activity is 0. As shown in FIG. 2, for activities a and b, it can be characterized
a＝(w a1 ，w a2 ，...w an-1 ，w an )，b＝(w b1 ，w b2 ，...w bn-1 ，w bn ) Wherein w is ai ，w bi The correlation between activity a and activity b can be represented by ρ, where 0 is taken to indicate that an activity does not have the property, and 1 is taken to indicate that an activity has the property ab As denoted by Correlation (a, b), the cosine similarity may be chosen here:
the similarity between different activities is calculated, several thresholds can be set by utilizing a control variable method for selecting the threshold, the threshold with the best classification effect is selected, and finally, the activity data is updated.
And step three, calculating the weight of the side possibly formed by the user to be recommended and other users, adding the calculated weight value into the queue to be recommended, introducing an opening coefficient, updating the weight value, and then updating the queue to be recommended.
And calculating the weight of the edge which can be formed by the user to be recommended and other users according to the principle of the ternary closure, and considering from two aspects, including user-user and user-activity-user, and calculating the weight from the two aspects.
For user-user patterns, a familiarity is defined:
where u1, u2 represents the user, friend (ui) represents the number of friends of the user ui,
friend (u1) andgatefriend (u2) indicates the number of common friends of u1 and u2, and friend (u1) Ufriend (u2) indicates the union of the numbers of friends of u1 and u 2.
For user-activity-user patterns, a similarity is defined:
wherein u1, the user denoted u2, activity (ui) denotes the number of activities in which the user ui participates,
activity (u1) anduactivity (u2) indicates the number of activities u1 and u2 that participate in together,
the activity (u1) U activity (u2) indicates that u1 and u2 are the union of activities.
By combining the two triangle closure modes of user-user and user-activity-user, the final weight calculation formula can be obtained:
w ab ＝α×similarity(u1，u2)+β×familarity(u1，u2)
wherein α is a similarity influence factor, β is a familiarity influence factor, α, β may be (0,1), and according to the specific situation of each user, α, β values may also be dynamically adjusted, ensuring α + β is 1, if the user is enthusiastic to participate in the activity, the α value is kept larger than the β value, and if the user is enthusiastic to make friends, the β value is kept larger than the α value.
Adding weights calculated by a user to be recommended and other users into a queue to be recommended, introducing an openness coefficient gamma considering that the user may not add strange users, wherein gamma can be 0 or 1, when gamma is 0, the user does not add strange friends, when gamma is 1, the user experiences the strange friends, and for updating the weights of the queue to be recommended, the calculation formula is as follows:
w＝γ×w ab +γ
wherein w ab Are the weights calculated in two ternary closure modes. And updating the weight between the user to be treated and the recommendation queue user through the formula.
Step four: a state perception mechanism is introduced to analyze the state of the user at the moment, and the user state can be matched according to the mood, the mental state and the busy-idle state analysis of the user.
A state perception mechanism is introduced before recommendation, the mood, the mental state and the busy state of a user are observed and analyzed by utilizing sensors and cameras on a vehicle and a mobile phone, if the user is in a state of low mood, the user can be recommended to the user with pleasure mood, the user is guided, the user can walk away the mood of hurting the mind early, a humanized recommendation system can be realized, correspondingly, if the user is in a happy state, the user can be matched with the user with low mood, the recommendation system meets the requirements of the real society, and the humanized design is full of.
Step five: and sorting the users subjected to state matching according to the weight value, and then selecting the users for recommendation.
And selecting m users for the user to recommend, and for each user, sorting the weights of other users after state matching with the user according to the magnitude sequence, and selecting the first m users from the weights to recommend.
As shown in fig. 3, the friend recommendation system based on the ternary closure includes a data collection module, a data processing module, a weight calculation module, a state matching module, and a friend recommendation module. And the data collection module judges whether to update the data according to the geographic position and the time of the user. And the data processing module is used for preprocessing the activity data of the user and clustering the activity data of the user. And the weight calculation module is used for calculating the weight of a user to be recommended which possibly forms a side with other users, adding the calculated weight into a queue to be recommended, introducing an opening coefficient, updating the weight and then adding into the queue to be recommended. And the state matching module introduces a state perception mechanism before recommendation, and can match the state of the user at the moment with the states of other users. And the friend recommending module selects m users for recommending for the user, sorts the weights of other users and the user after state matching according to the magnitude sequence for each user, and selects the first m users for recommending.
The foregoing illustrates and describes the principles, general features, and advantages of the present invention. It will be understood by those skilled in the art that the present invention is not limited to the embodiments described above, which are described in the specification and illustrated only to illustrate the principle of the present invention, but that various changes and modifications may be made therein without departing from the spirit and scope of the present invention, which fall within the scope of the invention as claimed. The scope of the invention is defined by the appended claims and equivalents thereof.
Claims (7)
1. A friend recommendation system based on a ternary closure is characterized by comprising a data collection module, a data processing module, a weight calculation module, a state matching module and a friend recommendation module; the data collection module judges whether to update data according to the geographic position and time of the user; the data processing module is used for preprocessing the activity data of the user and clustering the activity data of the user; the weight calculation module calculates the weight of a user to be recommended and other users which may form a side, adds the calculated weight into a queue to be recommended, introduces an opening coefficient, updates the weight, and then adds the weight into the queue to be recommended; the state matching module introduces a state perception mechanism before recommendation to match the state of the user with the states of other users; the friend recommending module selects m users for recommending, and for each user, the weights of other users and the user after state matching are sorted according to the magnitude sequence, and the first m users are selected for recommending, wherein m is a dynamic numerical value selected according to the user condition.
2. A friend recommendation method based on a ternary closure, characterized in that, using the friend recommendation system based on a ternary closure as claimed in claim 1, specifically comprises the following steps:
step one, judging whether to update data or not by using a data collection module according to the geographical position and time of a user, if the geographical position of the user changes and the time difference is one week from the acquisition time of the last data, updating the information of the user, and if not, ending the recommendation process to avoid repeated recommendation;
secondly, the data processing module preprocesses the data set of the user, the extracted activities of the user have certain repeatability, and the data redundancy is reduced by clustering the activities, so that the data is more representative;
step three, a weight calculation module calculates the weight of a user to be recommended and other users which may form a side, adds the calculated weight into a queue to be recommended, introduces an opening coefficient and updates the weight;
a state sensing mechanism is introduced into the state matching module, the state of the user at the moment is analyzed, and the user state matching is carried out according to the mood, the mental state and the busy-idle state analysis of the user;
and fifthly, sorting the users subjected to state matching according to the weight value by using a friend recommending module, and then selecting the users for recommending.
3. The friend recommendation method based on ternary closure according to claim 2, wherein in the first step, it is determined whether to acquire data by acquiring a geographical location of a user and a time for requesting update, it is determined whether the geographical location acquired during the current recommendation is consistent with a latest recommended geographical location, it is determined whether a week difference between the current recommended time and the latest recommended time exists, it is set that the data is updated when the current geographical location is different from the previous time and the time difference is one week, otherwise, the whole recommendation process is ended, the data is updated, the acquired data includes activities and friend relationships of all users in one year at the geographical location, and if the two conditions are not satisfied, the data is not updated, and repeated recommendation is reduced.
4. The friend recommendation method based on ternary closure as claimed in claim 3, wherein in said second step, a set of characteristics is selected, a weight value pointing to the edge is set to 1, otherwise, the weight value is set to 0, and the correlation p for the activity a and the activity b is set to ab Using rho ab Expressed as Correlation (a, b), cosine similarity is selected:
and performing active clustering to reduce data redundancy.
5. The friend recommendation method based on ternary closure according to claim 4, wherein said activities a and b are represented by the characteristics:
a＝(w a1 ，w a2 ，...w an-1 ，w an )，b＝(w b1 ，w b2 ，...w bn-1 ，w bn )，
for the correlation of Activity a and Activity b, use ρ ab Expressed as Correlation (a, b), cosine similarity is selected:
the similarity among different activities is calculated, the selection of the threshold value can utilize a control variable method to set several threshold values, the threshold value with the best classification effect is selected, and finally the updated activity data i is taken as [1, n ].
6. The friend recommendation method based on ternary closure as claimed in claim 5, wherein said step three is considered from two aspects, including user-user and user-activity-user, to calculate the weight: the calculated weight value is added into a queue to be recommended, considering that a user may not add strange users, an openness coefficient gamma is introduced, wherein gamma can be 0 or 1, when gamma is 0, the user does not add strange friends, when gamma is 1, the user experiences the strange friends, and the calculation formula is as follows for updating the weight value of the queue to be recommended:
w＝γ×w ab +γ
where w represents the weight of the user forming an edge with other users, γ is the openness coefficient, w ab The initial weights obtained under the two ternary closure modes are integrated, the calculated weights are sequenced from large to small, and then the weights are added into a queue to be recommended.
7. The friend recommendation method based on ternary closure as recited in claim 6, wherein for user-user schema, a familiarity is defined:
where u1, u2 represent users, friend (ui) represents the number of friends of user ui, i takes 1,2 represents the number of common friends of u1 and u2 of two different users u1, u2friend (u1) and u friend (u2), and friend (u1) and u friend (u2) represent the union of the numbers of u1 and u2 friends;
for user-activity-user patterns, a similarity is defined:
wherein u1, u2 denote users, activity (ui) denotes the number of activities in which the user ui participates, activity (u1) anduactivity (u2) denotes the number of activities in which u1 and u2 participate together, and activity (u1) denotes the union of u1 and u2 participation in activities;
and integrating two triangle closure modes of user-user and user-activity-user to obtain a final weight calculation formula:
w ab ＝α×similarity(u1，u2)+β×familarity(u1，u2)
wherein alpha is a similarity influence factor, beta is a familiarity influence factor, alpha, beta can be (0,1), according to the specific situation of each user, the values of alpha, beta are dynamically adjusted, alpha + beta is ensured to be 1, if the user is enthusiastic to participate in the activity, the value of alpha is kept larger than the value of beta, and if the user is enthusiastic to make friends, the value of beta is kept larger than the value of alpha.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210932275.3A CN115098797A (en) | 2022-08-04 | 2022-08-04 | Friend recommendation system and method based on ternary closure |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210932275.3A CN115098797A (en) | 2022-08-04 | 2022-08-04 | Friend recommendation system and method based on ternary closure |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115098797A true CN115098797A (en) | 2022-09-23 |
Family
ID=83300195
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210932275.3A Pending CN115098797A (en) | 2022-08-04 | 2022-08-04 | Friend recommendation system and method based on ternary closure |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115098797A (en) |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106294846A (en) * | 2016-08-19 | 2017-01-04 | 维沃移动通信有限公司 | A kind of recommendation method of information and mobile terminal |
| CN107679249A (en) * | 2017-10-27 | 2018-02-09 | 上海掌门科技有限公司 | Friend recommendation method and apparatus |
| CN107861967A (en) * | 2017-09-02 | 2018-03-30 | 长沙军鸽软件有限公司 | A kind of methods, devices and systems of intelligent Matching good friend |
| CN111581532A (en) * | 2020-05-08 | 2020-08-25 | 深圳市万佳安人工智能数据技术有限公司 | Social network friend-making recommendation method and system based on random block |
-
2022
- 2022-08-04 CN CN202210932275.3A patent/CN115098797A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106294846A (en) * | 2016-08-19 | 2017-01-04 | 维沃移动通信有限公司 | A kind of recommendation method of information and mobile terminal |
| CN107861967A (en) * | 2017-09-02 | 2018-03-30 | 长沙军鸽软件有限公司 | A kind of methods, devices and systems of intelligent Matching good friend |
| CN107679249A (en) * | 2017-10-27 | 2018-02-09 | 上海掌门科技有限公司 | Friend recommendation method and apparatus |
| CN111581532A (en) * | 2020-05-08 | 2020-08-25 | 深圳市万佳安人工智能数据技术有限公司 | Social network friend-making recommendation method and system based on random block |
Non-Patent Citations (5)
| Title |
|---|
| CC\'S WORLD: "协同过滤推荐算法综述&传统推荐算法综述", HTTPS://BLOG.CSDN.NET/QQ_36643449/ARTICLE/DETAILS/122125850, 26 December 2021 (2021-12-26), pages 2 - 8 * |
| 学到老: "社交网络SNS的好友推荐算法", HTTPS://CLOUD.TENCENT.COM/DEVELOPER/ARTICLE/1390456, 13 February 2019 (2019-02-13), pages 1 - 5 * |
| 杨秀璋等: "Python网络数据爬取及分析从入门到精通（分析篇）", 30 June 2018, pages: 182 - 184 * |
| 葛东旭: "数据挖掘原理与应用", 31 March 2020, pages: 201 - 203 * |
| 贺超波等: "应用非负矩阵分解的社交网络好友推荐", 华南师范大学学报（自然科学版）, 31 December 2016 (2016-12-31), pages 1 - 3 * |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN105468598B (en) | Friend recommendation method and device | |
| CN110162717B (en) | Method and device for recommending friends | |
| CN111324789B (en) | Method for calculating network information data heat | |
| CN105281925B (en) | The method and apparatus that network service groups of users divides | |
| Wang et al. | Exploiting social influence for context-aware event recommendation in event-based social networks | |
| CN102141986A (en) | Individualized information providing method and system based on user behaviors | |
| CN107111651A (en) | A kind of matching degree computational methods, device and user equipment | |
| CN110996116A (en) | Anchor information pushing method and device, computer equipment and storage medium | |
| Wachs et al. | Why do men get more attention? Exploring factors behind success in an online design community | |
| CN103198072A (en) | Method and device for mining and recommendation of popular search word | |
| CN104077723A (en) | Social network recommending system and social network recommending method | |
| CN111079009A (en) | User interest detection method and system for government map service | |
| CN109558384A (en) | Log classification method, device, electronic equipment and storage medium | |
| CN110019996A (en) | A kind of family relationship recognition methods and system | |
| CN111581532A (en) | Social network friend-making recommendation method and system based on random block | |
| CN109617999A (en) | Content delivery method, device, computer equipment and storage medium | |
| CN108764369A (en) | Character recognition method, device based on data fusion and computer storage media | |
| CN107368499A (en) | A kind of client's tag modeling and recommendation method and device | |
| JP2013257668A (en) | Interest analysis method, interest analyzer and program of the same | |
| Wang et al. | Top-k socially constrained spatial keyword search in large siot networks | |
| CN115098797A (en) | Friend recommendation system and method based on ternary closure | |
| CN110209704B (en) | User matching method and device | |
| CN111159576A (en) | User classification method, device and system | |
| CN108133296B (en) | Event attendance prediction method combining environmental data under social network based on events | |
| Li et al. | When user interest meets data quality: A novel user filter scheme for mobile crowd sensing |
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 |