CN105022808A - Binary constant value interval matching method - Google Patents

Binary constant value interval matching method Download PDF

Info

Publication number
CN105022808A
CN105022808A CN201510386278.1A CN201510386278A CN105022808A CN 105022808 A CN105022808 A CN 105022808A CN 201510386278 A CN201510386278 A CN 201510386278A CN 105022808 A CN105022808 A CN 105022808A
Authority
CN
China
Prior art keywords
matching
eigenwert
sequence
scale
list
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
CN201510386278.1A
Other languages
Chinese (zh)
Inventor
程文举
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.)
Individual
Original Assignee
Individual
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 Individual filed Critical Individual
Priority to CN201510386278.1A priority Critical patent/CN105022808A/en
Publication of CN105022808A publication Critical patent/CN105022808A/en
Pending legal-status Critical Current

Links

Classifications

    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/20Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
    • G06F16/24Querying
    • G06F16/245Query processing
    • G06F16/2455Query execution
    • G06F16/24553Query execution of query operations
    • G06F16/24558Binary matching operations

Abstract

The invention relates to a binary constant value interval matching method. The method comprises the steps of: (1) establishing characteristic value tables A and B; (2) establishing a characteristic value matching table; (3) extracting characteristic values from the characteristic value matching table, and carrying out formatting; (4) performing constant value calculation on the characteristic values; (5) calculating an offset; and (6) matching calculation results. The binary constant value interval matching method has the beneficial effects that: the computation burden of a server during data or role matching is reduced; the memory occupation of the server during matching is reduced; complicated calculation methods and steps during matching are reduced; and the data matching degree is controllable, and the data matching degree for users of different needs can be controlled by controlling the offset to achieve the data precision required by the users. The method is high in matching efficiency. The method has matching flexibility, so that user states and corresponding characteristic values can be updated in real time, and changes of certain features of the users can be rapidly dealt with.

Description

The interval matching process of a kind of scale-of-two definite value
Technical field
The invention belongs to matching technique, be specifically related to the interval matching process of a kind of scale-of-two definite value.
Background technology
Universal along with the network life, quantity of information various, it is more and more difficult that people want to inquire about some paired or similar interactive informations.The coupling of the such information of major part is artificial treatment, that is: realize natural pairing by the supply and demand district of software setting itself or follow-up post district.If in the internet information of magnanimity automatic matching, then often or by Keywords matching, coupling poor accuracy.Such coupling general layout loses some potential users will to a lot of businessman.
Now also without any the products configuration that software or method recommend him to be concerned about according to the characteristic of user to it, or push based on information itself; Current software great majority are all browsing record to do and mate based on user, and the distortion of information is wherein very serious, and matching efficiency is very low.This patent mainly provide as: user needs anything just what to push; By user's request, judge whom potential buyer is, how to customize packaging scheme according to buyer, reduce a large amount of data grabber and computing in Data Matching.
Summary of the invention
In order to solve the problems referred to above that prior art exists, the invention provides the interval matching process of a kind of scale-of-two definite value.Be cured needing the Information Pull list of feature values extracting or mate; The eigenwert of scale-of-two control information is utilized to screen; Binary decimal convertibility is utilized to carry out fixed value calculation to information; Binary position (side-play amount calculated) is utilized to carry out the retractility of comformed information; Mated by fixed value calculation and side-play amount.
The technical solution adopted in the present invention is:
The interval matching process of a kind of scale-of-two definite value, its improvements are:
(1) list of feature values A and B is set up;
(2) eigenwert matching list is set up;
(3) extract eigenwert in eigenwert matching list, format;
(4) fixed value calculation is carried out to eigenwert;
(5) side-play amount is calculated;
(6) result of calculation is mated.
Preferably, described step (1) comprises according to website or searches plain demand and set up the list of feature values, and according to the priority descending sort of eigenwert.
Preferably, described step (2) comprises the list of feature values A and B, sets up user to user, user to information matches, adopts mapping relations one by one to set up eigenwert matching list to information or information.
Preferably, described step (3) comprises according to user behavior or existing feature, extracts eigenwert in the list of feature values, carries out format and generates binary sequence.
Preferably, described step (3) comprises
(3.1) select scale-of-two bit wide, scale-of-two bit wide is identical with eigenwert quantity in the list of feature values;
(3.2) generate a binary format sequence according to scale-of-two bit wide, default value corresponding to binary format sequence index is 0;
(3.3) eigenwert in the existing list of feature values, then set to 0 the position 1 in binary format sequence identical with its index, binary format sequence location default value is 0.
Preferably, described step (4) comprises
(4.1) retrieve in binary format and remove the position subscript that beginning is in the residue sequence of 0 each 0 entirely, and store subscript sequence, the sequence retrieved for empty, then retrieved the sequence of 1 and removes most significant digit replacement sequence in 1;
(4.2) carry out fixed value calculation, calculate the decimal value that binary format is corresponding.
Preferably, described step (5) comprises the value needing to be calculated by retrieve sequence side-play amount according to matching degree, is matched to power for controlling.
Preferably, described step (5) comprises and controls matching precision by side-play amount, result of calculation is mated the list of feature values A and B.
Beneficial effect of the present invention is:
The present invention can carry out the pre-service of role or information.As long as when user or information excite certain eigenwert, namely later coupling be can be used for interval once calculating of scale-of-two definite value of role or information.
The present invention reduces operand when server matches data or role.By calculating first independent for the mode of information or role's independent transformation, then only need to carry out counting coupling, the service pressure avoiding a large amount of string matchings to bring, is more applicable for small-sized server.
The EMS memory occupation of server when the present invention reduces coupling, relative to existing matching process, needs the numerical value that stores less during the method coupling, simultaneously can be real-time mutual with database, reduce the memory pressure of server.
Loaded down with trivial details computing method and step when the present invention reduces coupling: directly user and information are fixed into a particular value, each only needs mates value, and is not considered for role or data itself.Coupling is the mapping of relation, little with the content relation of user or information itself, so only need to process its eigenwert.
The present invention can utilize idle process data: after only needing to store the eigenwert of user or information, busy in service is temporary suspension data processing, starts data processing, more reasonably utilize service clearance in idle.
The Data Matching degree that the present invention is controlled, can carry out control data matching degree by controlling side-play amount for the different users that needs, reaching the data precision that user needs.
The efficient matching efficiency of the present invention.For all treated data or user, server only needs the query manipulation performing simple Numerical Range to complete matching task.
The present invention has the retractility of coupling, can the state of real-time update user and characteristic of correspondence value, the fast change of some characteristic of reply user.
Row control of the present invention is easier, as long as the method can be incorporated in corresponding website or other programs by the developer having programming basis.
Accompanying drawing explanation
Fig. 1 is the interval matching process process flow diagram of a kind of scale-of-two definite value of the present invention;
Fig. 2 is algorithm flow path switch figure in the interval matching process of a kind of scale-of-two definite value of the present invention.
Embodiment
As shown in Figure 1, the invention provides the interval matching process of a kind of scale-of-two definite value.
Explain and implication:
Eigenwert (Eigenvalue, hereinafter abbreviated as E), eigenwert ordered list (Ordered Eigenvalue, hereinafter abbreviated as OE), eigenwert matching list (Eigenvalue Mapping, hereinafter abbreviated as EM), index, the scale-of-two bit wide index of eigenwert, binary-formatted (Binary Format is called for short BF once,), retrieve sequence S, decimal system definite value D, the explanations of nouns such as side-play amount flag and explanation.
E: refer to and can represent or the description of representative of consumer (data).As: people, after 90s, student, the vocabulary liking billiard ball, pay close attention to the feature of the people that science and technology etc. can identify; And for example: Zhang San wants to buy a mobile phone, Zhang San's (people, Partial Feature of inheritor), buy (behavioural characteristic), the mobile phone feature of mobile phone (inherit).E can be that (Default, hereinafter abbreviated as Def for default value.Default value representative coupling extraneous features value, default value can be multiple, and also can set up mapping relations with any eigenwert, the priority of Def is lower than any eigenwert).
OE: the set (non-some specific users or information) being used for listing a certain class particular user or information (as student) eigenwert.Be used for setting up the eigenwert match-on criterion of a certain class user or information, priority descending sort carried out to eigenwert simultaneously.As: the OE for a certain class user be after 90s, university student, Anhui, undergraduate course, man, female, default value 1, default value 2, }.
EM: eigenwert matching list, set up user to user, user to information, information to the one by one mapping relations of the matching ways such as information from the former to the latter (based on the latter, the former will add Def until both numbers are equal at quantity to eigenwert number not in the former OE).Eigenwert matching list as a businessman of Ge Mai game products Taobao and its user following game key mouse, game optical disk, computer fittings, game is supplemented with money, game accessory, Def1}; Like game, focus on game experiencing, be concerned about the configuration that game needs, supplement purpose with money, and after 90s, university student }.
The index of eigenwert: the subscript value that in eigenwert ordered list, each eigenwert is corresponding, from priority minimum counting, initial index is 1;
Scale-of-two bit wide index: be subscript value corresponding to one digit number value in binary sequence, count from lowest order, initial index is 1;
BF: the binary-formatted sequence Sum fanction of user or information.Select suitable scale-of-two bit wide (as selected 64 or 8 etc., determine according to eigenwert size, 8bit is selected in unification for convenience of description hereinafter); By corresponding BF lowest order minimum for OE medium priority, corresponding BF the low level that priority second is little.By that analogy; If have the character in OE in assigned role or information, then by the correspondence position 1 (if the possible probability of a certain item character is greater than 50% namely think to have this character) in corresponding BF, otherwise set to 0, the value of Def is 0.
Retrieve sequence S: the position subscript retrieving each non-0 (namely occur for first 1 0) of taking the lead in BF below, and store subscript sequence S, if the S retrieved is for empty, then retrieves the sequence of 1 and remove most significant digit replacement S in 1, S arranges according to ascending order, sets up index simultaneously.
Decimal system definite value D: fixed value calculation, calculates decimal value corresponding to BF according to the conversion between scale-of-two with the decimal system.
Side-play amount flag: be used for controlling the efficiency of coupling and the accuracy of result; Side-play amount (flag) calculates, and the calculating of side-play amount need calculate according to retrieve sequence S, according to the needs of matching degree, chooses suitable subscript value in S, and the value calculating side-play amount (is i for the subscript value chosen; Flag=2^ (i-1)), be used for controlling being matched to power in the matching algorithm below;
Wherein, in coupling, side-play amount flag value is as follows on matching effect impact:
For a binary sequence B={10001001 ... 11101110};
Corresponding retrieve sequence is S={s1, s2, s3 ... sn}, computing method as illustrated above,
Selecting a value in S for calculating lower side-play amount, remembering that this value is i, the index of this value correspondence in S is that the subscript value in t, S adds up to n;
Then matching rate=(n-t+1)/n;
Wherein, the selection of i provided do following explanation:
If: the BF (u) (in BF, the index of first time appearance 1 is K) that user u is corresponding;
S(u)={s1,s2,s3,……,sn}(n<K) flag=2^(i-1) i∈S(u);
①2^sn>(2^(s1-1)+2^(s2-1)+2^(s3-1)+……+2^(sn-1));
②D(u)>∑2^(Si-1);
③2^(K-1)>∑2^(Si-1);
4. for any i ∈ S (u), if decimal number D ' ∈ (D (u)-flag, D (u)+flag) then: in the binary representation of D (u) and D ', i (not comprising) is identical to the binary sequence fragment on K position;
5. the value for i ∈ S (u), i is less, then D ' identical binary sequence fragment longer (namely matching rate is higher) upper with D (u), otherwise, more greatly.As i=1, then D ' is identical with D (u).
The interval matching process of a kind of scale-of-two definite value of the present invention, concrete grammar is as follows:
(1) list of feature values A and B is set up;
Determine user characteristics value, set up OE, according to website or search plain demand and set up effective OE.
In OE, the arrangement of E is according to descending sort, and the advantage about arrangement is described as follows:
If be 00110010 after a user or information scale-of-two (8bit) format (format rule can specifically describe below); Being converted into the decimal system according to priority descending sort is that 2^5+2^4+2 can find out that the value of figure place most significant digit to whole decimal data has the greatest impact.If larger little to the variable effect of whole data of priority can be found out according to ascending order (01001100) i.e. 2^6+2^3+2^2, simultaneously when setting up EM both sides' eigenwert quantity and not mating, priority descending controls very simple for the interpolation of Def.
(2) eigenwert matching list is set up;
Set up the EM of coupling both sides.
(3) extract eigenwert in eigenwert matching list, format;
The binary-formatted rule (Binary Format, hereinafter referred to as BF) of data is carried out according to the OE of role or information;
The rule of format is formulated according to the sequence of OE:
1, suitable scale-of-two bit wide (as selected 64 or 8 etc., determine according to eigenwert size, 8bit is selected in unification for convenience of description hereinafter) is selected;
2, by corresponding BF lowest order minimum for OE medium priority, corresponding BF the low level that priority second is little, by that analogy;
If have the character in OE in 3 assigned roles or information, then by the correspondence position 1 (if the possible probability of a certain item is greater than 50% namely think to have this character) in corresponding BF, otherwise set to 0, the value of Def is 0;
Such as: after 90s, university student, Anhui, undergraduate course, man, female, default value 1, default value 2, corresponding a kind of possible BF is 11010100.
(4) fixed value calculation and side-play amount calculating are carried out to eigenwert;
1, the position subscript that removing beginning in binary format is (1 later sequence appears in first time) each 0 in the residue sequence of 0 is entirely retrieved, and store subscript sequence, the sequence retrieved for empty, then retrieved the sequence of 1 and removes most significant digit replacement sequence in 1;
As: 01010101 S={2,4,6}
00011111 S={1,2,3,4}
2, fixed value calculation, calculates decimal value corresponding to BF (be abbreviated as D, D (u) is expressed as the decimal value for information or user u);
3, side-play amount (flag) calculates, and the calculating of side-play amount need calculate according to retrieve sequence S, needs according to matching degree the value calculating side-play amount, is used for controlling being matched to power in the matching algorithm below.
(5) control matching precision by side-play amount, result of calculation is mated;
By the interval A and side-play amount A ' of definite value calculating match party, by the B of match party, as long as meet B to belong to (A-A ', A+A '), then calculate that the match is successful.
The invention provides the scheme of different user coupling.As: businessman of Taobao coupling n the potential user specified to;
The invention provides Data Matching scheme, provide suitable data according to the demand of different user.As: user needs one about the data of iPhone6, and according to the products configuration that the custom of user recommends him to be concerned about to it, instead of full detail pushes;
The invention provides new search matching scheme.Push according to the definite value of algorithm interval coupling the information that user needs, instead of push (as existing character string pattern matching, or tree graph coupling etc.) according to the degree of correlation of information itself;
The invention provides precision customized, as: user needs anything just what to push;
The invention provides the pairing of product and potential user in ecommerce.As: I has an iphone, and who will be potential buyer;
The invention provides the flow custom in production.As: pack a collection of apple, whom potential buyer is, how to customize packaging scheme according to buyer;
The invention provides new market survey scheme.As: new product is for a certain particular group, how to avoid a large amount of non-user group personnel to participate in investigation, allows the investigator of all participations be all customer group personnel in other words;
The invention provides user's valid data pick in data mining and get control program.As: user wants to buy new cell-phone, how accurately to infer the type of cell phone that user needs, price, and these should based on the characteristic sum demand of user own instead of handset capability and price.
The invention provides Rapid matching scheme, reduce a large amount of data grabber and computing in Data Matching.
The invention provides information analysis method.
To the matching way of mass data, secondary definite value interval can be carried out by multiple OE, in the interval OE of secondary definite value, whether each meets and according to being that the no coupling of OE is determined in a definite value, if the match is successful in a definite value, the value of correspondence position need being set to 1, otherwise being 0.Subsequent match is according to mentioned above principle.
Embodiment
The hotel owner selling game Related product for one sets up EM with its user;
A:{ game key mouse, game optical disk, computer fittings, game is supplemented with money, other accessories, Def1};
B:{ likes game, focuses on game experiencing, is concerned about the configuration that game needs, supplements purpose with money, after 90s, university student };
Analyze OE: for hotel owner, take game products as the marketing system of radiation, the primary and secondary according to oneself product sets up the OE be applicable to.
For user, different users is different to the demand of product, but potential some characteristics of user determine the potential purchasing power of user.
Analyze EM: for EM above, O between two OE seems not obvious mapping relations, but by some specificity analysises to user, such as to buy game Related product for a people do not played games is normal, so whether user likes playing games to have limit priority, same for hotel owner, along with the characteristic of pc end is focused in game, key mouse etc. has limit priority, and the possibility that the people not focusing on game experiencing buys game Related product is also smaller etc.Certainly be not simple mathematical function relationship for the mapping between EM, this must meet the analysis of user personality and set up suitable eigenwert, and the foundation that behavior good equally maps is the precondition that the match is successful.
Be provided with a businessman M, its product meets: game key mouse, game optical disk, and game is supplemented with money, other, these four features;
Then BF (M): 00110110;
S(M)={1,4};
D(M)=2^5+2^4+2^2+2;
Selection for side-play amount: for when having multiple side-play amount available, select element in S maximum, then matching rate is lower, although but select least member matching rate high, but can mate with similar users again, so pair principle can self-defined setting as required, if can select most significant digit or a secondary high position to the suggestion of matching rate less demanding words.If instead be strict with matching rate then can select lowest order or secondary low level.Most significant digit is selected in this example.
flag(M)=2^3;
There is multiple user a, b, c, d etc.;
A: play once in a while, is indifferent to game experiencing, is concerned about the configuration that game needs, supplements purpose with money, after 90, and university student;
Then BF (a): 00001111;
S(a)={1,2,3};
D(a)=2^3+2^2+2+1 flag(a)=4;
B: like game, be indifferent to game experiencing, is concerned about the configuration that game needs, supplements purpose with money, after 90s, university student;
Then BF (b): 00101111;
S(b)={5};
D(b)=2^5+2^3+2^2+2+1 flag(b)=2^4;
C: like game, be indifferent to game experiencing, be indifferent to the configuration of game, do not supplement purpose with money, after 90s, university student;
Then BF (c): 00100011;
S(c)={3,4,5};
D(c)=2^5+2^1+1 flag(c)=2^4;
D: like game, is concerned about game experiencing, is concerned about the configuration that game needs, supplements purpose with money, non-after 90s, non-university student;
Then BF (d): 00111100;
S(a)={1,2};
D(d)=2^5+2^4+2^3+2^2 flag(d)=2;
Coupling: M{a, b, c, d};
k∈{a,b,c,d};
if(D(k)∈(D(M)-flag(M),D(M)+flag(M))){//D(k)∈(46,62);
Obtain { the e|e ∈ user that k} // the match is successful set;
}
Then: obtain k={b, d}.
The present invention is not limited to above-mentioned preferred forms; anyone can draw other various forms of products under enlightenment of the present invention; no matter but any change is done in its shape or structure; every have identical with the application or akin technical scheme, all drops within protection scope of the present invention.

Claims (8)

1. the interval matching process of scale-of-two definite value, is characterized in that:
(1) list of feature values A and B is set up;
(2) eigenwert matching list is set up;
(3) extract eigenwert in eigenwert matching list, format;
(4) fixed value calculation is carried out to eigenwert;
(5) side-play amount is calculated;
(6) result of calculation is mated.
2. the interval matching process of a kind of scale-of-two definite value according to claim 1, is characterized in that: described step (1) comprises according to website or searches plain demand and set up the list of feature values, and according to the priority descending sort of eigenwert.
3. the interval matching process of a kind of scale-of-two definite value according to claim 1, it is characterized in that: described step (2) comprises the list of feature values A and B, set up user to user, user to information matches, adopts mapping relations one by one to set up eigenwert matching list to information or information.
4. the interval matching process of a kind of scale-of-two definite value according to claim 1, is characterized in that: described step (3) comprises according to user behavior or existing feature, extracts eigenwert in the list of feature values, carries out format and generates binary sequence.
5. the interval matching process of a kind of scale-of-two definite value according to claim 1, is characterized in that: described step (3) comprises
(3.1) select scale-of-two bit wide, scale-of-two bit wide is identical with eigenwert quantity in the list of feature values;
(3.2) generate a binary format sequence according to scale-of-two bit wide, default value corresponding to binary format sequence index is 0;
(3.3) eigenwert in the existing list of feature values, then set to 0 the position 1 in binary format sequence identical with its index, binary format sequence location default value is 0.
6. the interval matching process of a kind of scale-of-two definite value according to claim 1, is characterized in that: described step (4) comprises
(4.1) retrieve in binary format and remove the position subscript that beginning is in the residue sequence of 0 each 0 entirely, and store subscript sequence, the sequence retrieved for empty, then retrieved the sequence of 1 and removes most significant digit replacement sequence in 1;
(4.2) carry out fixed value calculation, calculate the decimal value that binary format is corresponding.
7. the interval matching process of a kind of scale-of-two definite value according to claim 1, is characterized in that: described step (5) comprises the value needing to be calculated by retrieve sequence side-play amount according to matching degree, is matched to power for controlling.
8. the interval matching process of a kind of scale-of-two definite value according to claim 1, is characterized in that: described step (5) comprises and controls matching precision by side-play amount, result of calculation is mated the list of feature values A and B.
CN201510386278.1A 2015-06-29 2015-06-29 Binary constant value interval matching method Pending CN105022808A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201510386278.1A CN105022808A (en) 2015-06-29 2015-06-29 Binary constant value interval matching method

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201510386278.1A CN105022808A (en) 2015-06-29 2015-06-29 Binary constant value interval matching method

Publications (1)

Publication Number Publication Date
CN105022808A true CN105022808A (en) 2015-11-04

Family

ID=54412778

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201510386278.1A Pending CN105022808A (en) 2015-06-29 2015-06-29 Binary constant value interval matching method

Country Status (1)

Country Link
CN (1) CN105022808A (en)

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102682033A (en) * 2011-03-17 2012-09-19 环达电脑(上海)有限公司 Method for querying words by matching binary characteristic values
CN103116588A (en) * 2011-11-17 2013-05-22 腾讯科技(深圳)有限公司 Method and system for personalized recommendation
CN103530377A (en) * 2013-10-16 2014-01-22 北京理工大学 Scene information searching method based on binary feature codes
CN103678431A (en) * 2013-03-26 2014-03-26 南京邮电大学 Recommendation method based on standard labels and item grades

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102682033A (en) * 2011-03-17 2012-09-19 环达电脑(上海)有限公司 Method for querying words by matching binary characteristic values
CN103116588A (en) * 2011-11-17 2013-05-22 腾讯科技(深圳)有限公司 Method and system for personalized recommendation
CN103678431A (en) * 2013-03-26 2014-03-26 南京邮电大学 Recommendation method based on standard labels and item grades
CN103530377A (en) * 2013-10-16 2014-01-22 北京理工大学 Scene information searching method based on binary feature codes

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
CHOI,SEUNG-SEOK等: "A Survey of Binary Similarity and Distance Measures", 《SYSTEMICS,CYBERNETICS AND INFORMATIONS》 *

Similar Documents

Publication Publication Date Title
CN109101620B (en) Similarity calculation method, clustering method, device, storage medium and electronic equipment
US9898773B2 (en) Multilingual content based recommendation system
CN103678518B (en) Method and device for adjusting recommendation lists
JP5691735B2 (en) CONTENT RECOMMENDATION DEVICE, RECOMMENDED CONTENT SEARCH METHOD, AND PROGRAM
Selke et al. Pushing the boundaries of crowd-enabled databases with query-driven schema expansion
CN106168980A (en) Multimedia resource recommends sort method and device
CN107885852B (en) APP recommendation method and system based on APP usage record
WO2015148422A1 (en) Recommendation system with dual collaborative filter usage matrix
TW201415261A (en) Method and system for recommending search phrases
CN102622399A (en) Search apparatus, search method and program
CN103886081A (en) Information sending method and system
CN111708934A (en) Knowledge content evaluation method and device, electronic equipment and storage medium
CN109299383A (en) Generate method, apparatus, electronic equipment and the storage medium for recommending word
US20160140213A1 (en) Search device, search method, program, and storage medium
CN112100396B (en) Data processing method and device
CN103532565A (en) Code processing technique
CN103309869A (en) Method and system for recommending display keyword of data object
CN116894711A (en) Commodity recommendation reason generation method and device and electronic equipment
KR20140015653A (en) Contents recommendation system and contents recommendation method
KR20120007446A (en) Information processing device, information processing method, and program
CN106919693A (en) It is a kind of to improve the method and apparatus that hot word exposes coverage rate
CN108595395B (en) Nickname generation method, device and equipment
CN105022808A (en) Binary constant value interval matching method
CN110971973A (en) Video pushing method and device and electronic equipment
CN112966177B (en) Method, device, equipment and storage medium for identifying consultation intention

Legal Events

Date Code Title Description
C06 Publication
PB01 Publication
C10 Entry into substantive examination
SE01 Entry into force of request for substantive examination
RJ01 Rejection of invention patent application after publication
RJ01 Rejection of invention patent application after publication

Application publication date: 20151104