CN105589966A - Friend recommendation method based on composite scores - Google Patents

Friend recommendation method based on composite scores Download PDF

Info

Publication number
CN105589966A
CN105589966A CN201510980588.6A CN201510980588A CN105589966A CN 105589966 A CN105589966 A CN 105589966A CN 201510980588 A CN201510980588 A CN 201510980588A CN 105589966 A CN105589966 A CN 105589966A
Authority
CN
China
Prior art keywords
user
information
registering
score
time
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
Application number
CN201510980588.6A
Other languages
Chinese (zh)
Inventor
刘海亮
许祥
苏航
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Shenzhen Research Institute of Sun Yat Sen University
Original Assignee
Shenzhen Research Institute of Sun Yat Sen University
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Shenzhen Research Institute of Sun Yat Sen University filed Critical Shenzhen Research Institute of Sun Yat Sen University
Priority to CN201510980588.6A priority Critical patent/CN105589966A/en
Publication of CN105589966A publication Critical patent/CN105589966A/en
Pending legal-status Critical Current

Links

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
    • G06F16/9537—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)
  • Business, Economics & Management (AREA)
  • General Physics & Mathematics (AREA)
  • General Health & Medical Sciences (AREA)
  • Economics (AREA)
  • Primary Health Care (AREA)
  • Strategic Management (AREA)
  • Tourism & Hospitality (AREA)
  • Human Resources & Organizations (AREA)
  • General Business, Economics & Management (AREA)
  • Marketing (AREA)
  • Health & Medical Sciences (AREA)
  • Computing Systems (AREA)
  • Data Mining & Analysis (AREA)
  • General Engineering & Computer Science (AREA)
  • Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

The invention discloses a friend recommendation method based on composite scores. According to the method, the influence of personal information filled by a user on friend recommendation is taken into account, thus a corresponding information score is worked out via the factor of the personal information; the influence of the sign location in a mobile device on friend recommendation is also taken into account, thus a corresponding geographical score is worked out via the signing information of the user; furthermore, the factor of push information is also added, the same point in the push information of the users is also taken into account, thus the friend recommendation accuracy of the method is improved. When computing the geographical score, time slicing and spatial clustering modes are used, thus the computation quantity is reduced. The method has the advantages that the various influence factors are synthesized, and the influences of the personal information, signing information and push information of the user on the friend recommendation are ingeniously combined together in an ingenious manner. The method can be applied to application scenarios such as social network, marriage dating and friend making and recommendation system of mobile terminals.

Description

A kind of friend recommendation method based on integrate score
Technical field
The present invention relates to the friend recommendation field in social networks, particularly aspect solving and optimizing the validity of friend recommendation,Solved emphatically under various factors, how rationally comprehensive many factors obtains the problem of optimum commending friends.
Background technology
Along with the arriving of large data age and the development of mobile intelligent terminal, social networks obtains rapidly on mobile InternetDevelopment, excavating some hiding information becomes the direction of increasing people's research for user provides good friend more accurately. AsModern general research is mainly to consider the impact of single factor on good friend, and the good friend who obtains good friend by good friend's network topology comesRecommend, recommend potential good friend by the similitude of subscriber data, the recommendation by geographical position etc. is all comparatively simply rightSingle influence factor is processed and is obtained corresponding good friend.
Nowadays the data research based on geographical position is valued by the people gradually, but only not only counts by consideration geographical position factorCalculation amount is large, even and if process and alleviate amount of calculation but decline to some extent again aspect validity recommending. Simultaneously along with information pushingSocial demand continue to increase, the information of propelling movement also becomes us must not an irrespective important factor, how from pushingThe information of obtaining good friend in information also becomes very necessary factor.
Patented method, the matching degree of consideration personal information, by dividing the age, compose power, calculating obtaining information score,The information of registering by individual to the processing of register position and time after, alleviate amount of calculation, calculate corresponding geographical score, thenCorresponding hobby score is being obtained in the impact of consideration pushed information by pushed information, finally three scores are weighted to summation and obtainGet final integrate score, judge best commending friends by integrate score.
Summary of the invention
The present invention will solve under various factors, and how rationally comprehensive many factors obtains asking of optimum commending friendsTopic, provides a kind of friend recommendation method, is the method for calculating integrate score based on personal information, the information of registering, pushed informationChoose best commending friends.
The technical scheme that technical solution problem of the present invention is taked is as follows:
1) personal information of filling according to user, determines the information score that participates in recommending screening user, comprises the steps:
A) age is divided, between the age location in definite user's who needs commending friends personal information, if neededWant the user of commending friends not fill in the age, not in the operation that continues computing information score, and all participations recommendations are screenedUser's information must be divided into 0, if filled in the age, proceeds.
B) according to above-mentioned gained interval, different age brackets is composed to weight.
C), by the hobby in userspersonal information, calculate the information weights score of the user i in all ages and classes intervali。
D) the information weights score of the user to all participations recommendation screenings that calculateiCarry out stipulations processing, obtainWhole user i information score infi。
2) according to user's the information of registering, determine the geographical score that participates in the user who recommends screening, comprise the steps:
A) obtain all information of registering that need in the user of commending friends certain hour, wherein the information spinner of registering of the i time is wantedComprise the time T of registeringi, position Addr registersi。
B) adopt DBSCAN clustering algorithm, by needing all positions of registering of registering in information of user of commending friendsIn Addr, through, latitude information, cluster is carried out in the position of registering, the i that obtains the user who needs commending friends registers in informationThe affiliated cluster position addr in position registersi。
C) time of one day is divided, in all times of registering of definite user who needs commending friends, register each timeOrder in the time period of all divisions, the time period at the place of wherein registering for the i time is ti, will need the user of commending friendsThe information table of registering for the i time is shown (ti,addri)。
D) identical processing is carried out to the time of registering in the position of registering of registering in information that participates in the user who recommends screening, obtainGet all information of registering that participate in the user who recommends screening, wherein the k time attendance sheet is shown (tk,addrk)。
E) by recommending the user's of screening the information of registering to compare, calculate and obtain to the user who needs commending friends and participationGet the geographical weights that participate in the user who recommends screening.
F) recommend the user's of screening score to carry out stipulations processing to all participations that calculate, obtain final participation and push awayRecommend the user's of screening geographical score.
3) according to pushed information, determine the hobby score that participates in the user who recommends screening, step is as follows:
A) the initial propelling movement amount of the type to pushed information initializes, and gives the concrete of initial each pushed information typePropelling movement value.
B) record is pushed to user's information type, user is accepted or refuses to push to carry out record, and in the time that propelling movement is rejected,The propelling movement amount of the class record of this pushed information reduces by half and new record more, if pushed information is accepted, and the type of this pushed informationPropelling movement amount increase by 1 and new record more.
C) user's who needs commending friends the propelling movement number of times of all pushed information types is sorted, obtain first three kind and push awayRecommend the type of information.
D) calculate respectively each and participate in recommending user's the hobby weights of screening, obtain the user that participates in recommending screening withNeed commending friends user first three plant the propelling movement amount of corresponding pushed information type, the wherein propelling movement amount of pushed information type in iFor mi, the then initial value divided by this pushed information type by propelling movement amount that is existing each recommendation information of three kinds of recommendation information types,Then hobby weights three kinds of pushed information type result of calculations and recommend the user of screening as participation.
E) recommend the user's of screening hobby weights to carry out stipulations processing to all participations that calculate, obtain final ginsengWith the user preferences score of recommending screening.
4) calculate integrate score:
The information score that each is participated in recommending the user who screens, geographical score, hobby score is weighted summation and obtainsThis user's integrate score eventually, recommends the user of screening to sort by integrate score to all participations and obtains best recommendation wellFriend.
Brief description of the drawings
Accompanying drawing is herein merged in description and forms the part of this description, has explained our ratio juris.
Fig. 1 is the overview flow chart of the inventive method.
Fig. 2 is the flow chart of computing information score.
Fig. 3 is the flow chart that calculates geographical score.
Fig. 4 is the flow chart that calculates hobby score.
Detailed description of the invention
Concrete steps of the invention process are as follows:
Fig. 1 has listed a kind of flow chart of friend recommendation method, and as shown in Figure 1, this friend recommendation method comprises followingStep S101-S103:
Step S101 is computing information score, geographical score, hobby score, the use that participates in recommending screening by calculatingMust assign to provide foundation for the calculating that participates in the integrate score of recommending the user who screens for three kinds of family.
Three kinds of scores that step S102 calculates S101 are weighted summation, obtain each and participate in recommending screeningUser's integrate score.
Step S103 uses the integrate score that S102 calculates to recommend the user of screening to sort to all participations, obtainsGet first 30, the best friend recommendation obtaining is given to the user who needs commending friends
Information score in step S101 is calculated as shown in Figure 2, comprises the steps A1-A6:
Steps A 1 is to obtain the personal information that need to recommend user's user to fill in, must bag in personal information is filled inDraw together the selection of age and consumer taste, personal preference can be shopping, film, books etc.
Steps A 2 was divided the age, is divided into below 13, and 14-22,23-28,28-35,36-45, more than 46,Obtain the age in the user's who needs commending friends personal information, if need to recommend user's age of user not fill out, stopComputing information score, recommends the user's of screening information score unification to be designated as 0 all participations, if filled in definite being somebody's turn to do of ageThe interval at age place, proceeds remaining operation, if age of user is 25 years old, is 23-28. between its location
Steps A 3 is composed weights to the each age group of dividing, the user's who needs commending friends who obtains in steps A 2The interval right weight at age place is the highest, then reduces gradually to age two ends, and as determined, between location be 23-28, can be by itWeight is set to 6, and the weight of all age group is followed successively by: 4,5,6,5,4,3
The hobby that steps A 4 is obtained in recommended user's personal information is selected, by the need that obtain in result and steps A 2Want the hobby in user's the personal information of commending friends to contrast, obtain the number of both identical hobbies, and record. As twoPerson has all selected film, the number of both identical hobbies of doing shopping is 2
This age of Age estimation in the user of steps A 5 to the participation recommendation screening of obtaining in steps A 4 personal informationThe interval that place steps A 2 is divided, obtains the definite weights in steps A 3 in this interval, identical by what obtain in steps A 4Hobby number and definite weights multiply each other, and obtain each recommended user's personal information weights, as recommended userAge is 36, and the interval at its place is 36-45, and the weights of this age bracket are 4, and both identical hobbies are 2, calculateInformation weights are 8
Steps A 6 calculates all participations at A5 and recommends after user's the personal information weights of screening, by formula oneCarry out weights stipulations, [N, M]=[0,10], stipulations interval, obtains the information score that the user of screening is recommended in each participation
y = ( M - N ) * x - x m i n x max - x m i n + N (formula one)
The calculating of the geographical score in step S101 as shown in Figure 3, comprises the steps B1-B7:
Step B1 obtains the user's information of registering of first 6 months that need to recommend user, registering in information each timeThis user's the time of registering and the position of registering be must comprise, corresponding warp, latitude in position, comprised, and record.
In step B2, need the place of registering of registering in information that B1 is obtained to carry out cluster, adopt DBSCAN cluster(a kind of clustering method based on density, this algorithm will have enough highdensity region and be divided into bunch algorithm, and have noiseSpatial database in find arbitrary shape bunch), obtain class bunch set by counting in given radius and minimum field, andWherein the distance between each point is to calculate by longitude and latitude. Obtain the center position in bunch set of each class,The central point of this set institute compositing area. The position of registering that each is registered in information changes the central point of cluster set under it intoPosition, and amendment record, a single point that does not wherein meet cluster requirement is not processed it.
Step B3 carries out segmentation to the time of one day, is divided into 00:01-6:29,6:30-9:00, and 9:01-12:00,12:01-13:30,13:31-18:00,18:01-24:00, to each information of registering of acquisition of information of registering after treatment in step B2Register the time, the register scope at time place of judgement, the time of registering that each is registered in information by this scope in all segmentationsIn position replace, and amendment record. As the time of registering be 8:03, the scope at this time place is 6:30-9:00, changes the timeThe position of section in all segmentations is 2, so this time of registering replaces with 2, and amendment record.
All information of registering that step B4 obtains B3 are added up, and the record of registering after adding up B3 and processing, in phaseWith register number sum record in time period and identical cluster, record form (time period at the time place of registering, placeThe center position of cluster, number of times).
Step B5 processes the information of registering that participates in the user who recommends screening according to step B1-B4, obtain eachParticipate in result the record of recommending the user of screening to register after Information Statistics.
The result that step B6 comparison B4 and B5 obtain, adds up each and participates in user and the needs that recommendation is screenedThe number of identical point in all sign-in desks of user of commending friends, the number of similitude and the number of difference, wherein identical point isRefer to register time period at time place is the same with the central point of place cluster, and similitude refers to that the time period at the time place of registering is identicalBut the central point of the cluster at place is different, the central point that difference refers to the register time period at time place and the cluster at place allDifferent. In the time that whether the central point of the cluster at place is relatively the same, be that distance by calculating two central points sees whether exceededThe cluster radius of giving in step B2, if not, both are identical, otherwise different. At the identical point that obtains recommended user kN1, similitude N2, difference N3Number, and the number N of the information of registering obtained of B1, adopts formula two to calculate groundReason weights score, wherein d1，d2，d3It is corresponding geographical weight.
s c o r e = d 1 N 1 N + d 2 N 2 N + d 3 N 3 N (formula two)
Step B7 is obtaining, after each geographical weight that participates in the user who recommends screening, passing through formula by step B6One carries out stipulations, and [N, M]=[0,10], stipulations interval obtains the geographical score that the user of screening is recommended in each participation
The calculating of the hobby score in step S101 as shown in Figure 4, comprises the steps C1-C5:
Step C1 is that the type of the pushed information of requirement to each user is the same, and gives the information of every kind of propelling movementThe initial propelling movement amount that type is certain, records and backs up the primary quantity of each pushed information type, as the initial number of times 50 that pushes of living,Diet initially pushes number of times 80 times.
Step C2 record is pushed to user's information type and on the record of step C1, corresponding propelling movement result is carried outAmendment, the information giving a discount as pushed daily necessities to user, has checked if user clicks, the information type propelling movement amount of lifeAdd 1, if user has refused, propelling movement amount reduces half, to ensure the validity of each pushed information.
Step C3 obtains the propelling movement amount of all pushed information types that need the user of commending friends, right by propelling movement amountNeed all pushed information types of user of commending friends to sort, obtain the type of front 3 kinds of pushed information.
Step C4 obtains after first three type of planting pushed information at step C3, obtains respectively each and participates in recommending screeningThe propelling movement amount m of the corresponding three kinds of pushed information types of useri, wherein i is information type, and in the backup of step C1, obtain rightThe primary quantity M of three kinds of pushed information types of answeringi, by calculating the m of three kinds of pushed information typesi/MiAnd, obtain correspondingHobby weights
Step C4 calculates all participations and recommends, after user's the hobby weights of screening, to carry out stipulations by formula one,[N, M]=[0,10], stipulations interval, obtains final preference information score
Content described in this description embodiment is only enumerating of way of realization to inventive concept, protection of the present inventionScope should not be regarded as only limiting to the concrete form that embodiment states, protection scope of the present invention is also and in art technologyPersonnel conceive the equivalent technologies means that can expect according to the present invention.

Claims (5)

1. the friend recommendation method based on integrate score, is characterized in that, comprising:
1) personal information of filling according to user, determines the information score that participates in the user who recommends screening, comprises the steps:
A) age is carried out to segmentation, determine and need between user's the age location of commending friends.
B) according to above-mentioned gained interval, different age brackets is composed to power.
C), by needing user's the hobby of commending friends, calculate participation in different intervals and recommend the user's of screening weights.
D) recommend the user profile weights of screening to carry out stipulations processing to all participations that calculate, obtain each and participate in the user profile score that recommendation is screened.
2) according to user's the information of registering, determine the geographical score that participates in the user who recommends screening, comprise the steps:
A) obtain the information of registering needing in the user of commending friends certain hour, the information spinner of registering of the i time will comprise the time T of registeringi, position Addr registersi。
B) adopt clustering algorithm, pass through AddriIn through, latitude information, cluster is carried out in the user's who needs commending friends the position of registering, obtain and need the register clustering information in the place of registering in information of the user of commending friends, the cluster centre point position under the position of registering for the i time is addri。
C) time of one day is divided, determine different the registering the time period at the time place of registering each time, the time period at the place of wherein registering for the i time is ti, the user's who needs commending friends the information table of registering for the i time is shown to (ti，addri)。
D) identical processing is carried out to the time of registering in the position of registering of registering in information that participates in the user who recommends screening, obtain all information of registering that participate in the user who recommends screening, wherein the k time attendance sheet is shown (tk，addrk)。
E) by recommending the user's of screening the information of registering to compare, calculate and obtain the geographical weights that participate in recommending the user who screens to the user who needs commending friends and participation.
F) recommend the user's of screening score to carry out stipulations processing to all participations that calculate, obtain the user's of final participation recommendation screening geographical score.
3) pushed information of clicking according to user, determines the hobby score that participates in recommending screening user, and step is as follows:
A) the propelling movement amount of the information type pushing is initialized.
B) information type that record pushes to user at every turn and the corresponding result that pushes, revise corresponding propelling movement amount.
C) user's who needs commending friends the propelling movement number of times of all information types is sorted, obtain first three and plant the type of recommendation information.
D) calculate respectively the weights that participate in three kinds of recommendation information types corresponding to user of recommending screening, wherein the propelling movement amount of i kind recommendation information type is mi, and calculate the corresponding user preferences weights that participate in pushing screening.
E) the recommendation weights of the user to all participations propelling movement screenings that calculate carry out stipulations processing, obtain final participation and push the user preferences score of screening.
4) calculate integrate score:
For ensureing to consider the impact of different marks on result, three kinds of marks are weighted to summation, obtain the participation the most close with the user who needs commending friends and recommend the user of screening, and these users are recommended to the user who needs commending friends.
2. the friend recommendation method based on integrate score according to claim 1, is characterized in that: in step (1), adopt to recommend the interval right weight at user place the highest to the division of age bracket, then reduce gradually to two ends. When calculating, adopt identical hobby number between contrast hobby calculated recommendation user and recommended user, hobby number is multiplied each other and obtains score with corresponding weights. In the time of stipulations, adopt following formula:
Wherein [N, M] is the scope of stipulations, and x is the data of stipulations, xmaxData maximum in stipulations data, xminBe data minimum in stipulations data, y is the result after x stipulations.
3. the friend recommendation method based on integrate score according to claim 1, is characterized in that: in step (2), adopt DBSCAN clustering algorithm, reduced and calculated and raising matching degree by the cluster to geographical position. Time is being divided, avoiding the amount of calculation problems of too causing because of time factor. In the time calculating geographical score, adopt following formula:
Wherein d1，d2，d3Be respectively corresponding weighting, N is all number of times of registering of recommended user k, N1Be in all sign-in desks of recommended user k and recommend time and all numbers of not identical point of position between user, N2Be in all sign-in desks of recommended user k and recommend between user the time identical, but the number of the not identical point in position, N3Be in all sign-in desks of recommended user k and recommend time and all numbers of identical point of position between user.
4. the friend recommendation method based on integrate score according to claim 1, it is characterized in that: step initializes pushed information in (3), by the information of each propelling movement being given to certain propelling movement number of times, do like this validity that can effectively ensure pushed information. Push when result at record, increase by 1 if user has clicked, adopt minimizing method by half if user does not click, the half of each less current quantity, is not the round numbers part of integer, when result is 1 to be 0 in the time reducing. In the time calculating hobby weights, use following formula:
Wherein miThe hobby weights that participate in three kinds of recommendation information types corresponding to user i of recommending screening, vkThe dissimilar propelling movement amount of current recommendation information, MkIt is the initialization propelling movement amount of the information type of three kinds of propelling movements.
5. the friend recommendation method based on integrate score according to claim 1, it is characterized in that: in step (4), consider that different factors is to the Different Effects of recommending, obtain final integrate score by different scores is weighted to evaluation, thereby recommend by the integrate score order of obtaining best commending friends that sorts.
CN201510980588.6A 2015-12-24 2015-12-24 Friend recommendation method based on composite scores Pending CN105589966A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201510980588.6A CN105589966A (en) 2015-12-24 2015-12-24 Friend recommendation method based on composite scores

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201510980588.6A CN105589966A (en) 2015-12-24 2015-12-24 Friend recommendation method based on composite scores

Publications (1)

Publication Number Publication Date
CN105589966A true CN105589966A (en) 2016-05-18

Family

ID=55929545

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201510980588.6A Pending CN105589966A (en) 2015-12-24 2015-12-24 Friend recommendation method based on composite scores

Country Status (1)

Country Link
CN (1) CN105589966A (en)

Cited By (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106204300A (en) * 2016-07-23 2016-12-07 杨跃龙 A kind of interaction systems based on position, motion conditions and End-user relevance
CN106649682A (en) * 2016-12-15 2017-05-10 咪咕数字传媒有限公司 Book friend recommendation method and device
CN106951515A (en) * 2017-03-17 2017-07-14 上海衡修信息科技有限公司 A kind of contact person's matching process and device based on social software
CN113761391A (en) * 2021-09-09 2021-12-07 北京北大方正电子有限公司 Data search method, apparatus, medium and product
CN114900554A (en) * 2022-04-28 2022-08-12 北京北春园商贸有限责任公司 Social media information accurate pushing system and device based on big data

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101847226A (en) * 2009-12-17 2010-09-29 广州市盈海文化传播有限公司 Method and system for recommending opposite sex friend in social network service SNS community
US20130151527A1 (en) * 2011-11-15 2013-06-13 Sean Michael Bruich Assigning social networking system users to households
CN103488678A (en) * 2013-08-05 2014-01-01 北京航空航天大学 Friend recommendation system based on user sign-in similarity

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101847226A (en) * 2009-12-17 2010-09-29 广州市盈海文化传播有限公司 Method and system for recommending opposite sex friend in social network service SNS community
US20130151527A1 (en) * 2011-11-15 2013-06-13 Sean Michael Bruich Assigning social networking system users to households
CN103488678A (en) * 2013-08-05 2014-01-01 北京航空航天大学 Friend recommendation system based on user sign-in similarity

Cited By (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106204300A (en) * 2016-07-23 2016-12-07 杨跃龙 A kind of interaction systems based on position, motion conditions and End-user relevance
CN106649682A (en) * 2016-12-15 2017-05-10 咪咕数字传媒有限公司 Book friend recommendation method and device
CN106951515A (en) * 2017-03-17 2017-07-14 上海衡修信息科技有限公司 A kind of contact person's matching process and device based on social software
CN113761391A (en) * 2021-09-09 2021-12-07 北京北大方正电子有限公司 Data search method, apparatus, medium and product
CN114900554A (en) * 2022-04-28 2022-08-12 北京北春园商贸有限责任公司 Social media information accurate pushing system and device based on big data

Similar Documents

Publication Publication Date Title
CN105589966A (en) Friend recommendation method based on composite scores
Iregui Efficiency gains from the elimination of global restrictions on labour mobility: An analysis using a multiregional CGE model
Wang Foreign direct investment and domestic investment in the host country: evidence from panel study
Hashim Income inequality and poverty in Malaysia
CN109102371A (en) Method of Commodity Recommendation, device, computer equipment and storage medium
Yamarik et al. A sensitivity analysis of the gravity model
Anikin et al. Income stratification: Key approaches and their application to Russia
Krämer et al. Consumer integration into innovation processes*: A new approach for creating and enhancing innovations for the base of the pyramid?
CN109426985A (en) Information processing method, device, computer equipment and storage medium
CN110009417A (en) Target customer's screening technique, device, equipment and computer readable storage medium
Athukorala Asian trade flows: trends, patterns and projections
Palfrey et al. Warranties, performance, and the resolution of buyer-seller disputes
CN106056413A (en) Interest point recommendation method based on space-time preference
CN107153985A (en) A kind of excavation of user side multi-energy data and analysis method
Muštra et al. Regional inequalities in the European Union and the role of institutions
CN112767046B (en) Method and device for constructing non-standardized service product
Klemp et al. Picking winners? The effect of birth order and migration on parental human capital investments in pre-modern England
CN104484818A (en) Mobile App optimal advertising auction method based on keyword auction
Clarke et al. Vulnerability to what? Multidimensional poverty in Melanesia
Li et al. New service development using GAP-based QFD: a mobile telecommunication case
Stats et al. Non-profit institutions satellite account: 2018
Choudhury et al. Changing Landscape of Professional Higher Education in India: What Do We Know and What Do Recent Data Tell Us?
CN108510353A (en) A kind of e-commerce system that intelligence degree is high
Fuest et al. Is a flat tax politically feasible in a grown-up welfare state?
CN106682205A (en) Device and method for data processing

Legal Events

Date Code Title Description
C06 Publication
PB01 Publication
C10 Entry into substantive examination
SE01 Entry into force of request for substantive examination
WD01 Invention patent application deemed withdrawn after publication
WD01 Invention patent application deemed withdrawn after publication

Application publication date: 20160518