CN105718535A - Online scoring method and system - Google Patents

Online scoring method and system Download PDF

Info

Publication number
CN105718535A
CN105718535A CN201610029729.0A CN201610029729A CN105718535A CN 105718535 A CN105718535 A CN 105718535A CN 201610029729 A CN201610029729 A CN 201610029729A CN 105718535 A CN105718535 A CN 105718535A
Authority
CN
China
Prior art keywords
data
user
commodity
parameter
scoring
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.)
Granted
Application number
CN201610029729.0A
Other languages
Chinese (zh)
Other versions
CN105718535B (en
Inventor
廖好
沈婧
曾安
毛睿
王毅
许红龙
李荣华
刘刚
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 University
Original Assignee
Shenzhen 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 University filed Critical Shenzhen University
Priority to CN201610029729.0A priority Critical patent/CN105718535B/en
Priority to PCT/CN2016/081859 priority patent/WO2017121054A1/en
Publication of CN105718535A publication Critical patent/CN105718535A/en
Application granted granted Critical
Publication of CN105718535B publication Critical patent/CN105718535B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

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
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
    • G06F16/24—Querying
    • G06F16/245—Query processing
    • G06F16/2458—Special types of queries, e.g. statistical queries, fuzzy queries or distributed queries
    • G06F16/2465—Query processing support for facilitating data mining operations in structured databases
    • 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
    • G06Q10/00—Administration; Management
    • 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
    • G06Q20/00—Payment architectures, schemes or protocols
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q30/00—Commerce

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Databases & Information Systems (AREA)
  • General Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Data Mining & Analysis (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)
  • Fuzzy Systems (AREA)
  • Mathematical Physics (AREA)
  • Probability & Statistics with Applications (AREA)
  • Software Systems (AREA)
  • Computational Linguistics (AREA)
  • Information Retrieval, Db Structures And Fs Structures Therefor (AREA)

Abstract

The invention provides an online scoring method and system. The method comprises the steps of predefining user data and commodity data; inputting a set of original scoring data; converting the original scoring data to form a set of new scoring data; adjusting all data in the set of new scoring data through a preset algorithm and performing iteration until a steady state is achieved; and obtaining optimal values of all the data in the steady state and obtaining preference settings of the original scoring data according to the optimal values. The invention furthermore provides an online scoring system. According to the online scoring method and system provided by the invention, the accuracy of a sorting algorithm can be effectively improved.

Description

A kind of online methods of marking and system thereof
Technical field
The present invention relates to Data Mining, particularly relate to a kind of online methods of marking and system thereof.
Background technology
At big data age how from the existing valuable information of big extracting data, this is a very vital problem for the on-line system that data are extremely enriched.On-line system allows substantial amounts of user to interact, and provides thousands of film, books etc. online commodity.But this has created a lot of incoherent online commodity, in order to filter out these incoherent commodity, it is recommended that system has taked very a lot of method based on dependency, for instance collaborative filtering method obtains vast application.Except the consideration to dependency, the quality of commodity is important too to online user, and therefore, much online website such as Amazon and Netflix etc. has introduced so-called online marking system.
In these online marking systems, user can assess this commodity by the grading system value that commodity are set.Online marking system can help to the user discover that real high-quality commodity.After obtaining score data, it would be desirable to adopt some algorithms that these commodity are ranked up.
Most current electricity business adopts five-pointed star scoring (such as 1 to 5 corresponding five kinds of different brackets scoring) as main standards of grading, but this scoring tends not to the horizontal service concrete manifestation embodying each businessman, also cannot specifically embody the true subjective feeling of consumer.Additionally, due to the particularity of the network platform, the technology such as some businessmans application network especially improves the grading system of self by improper approach, cheats consumer.Therefore, the reliability disadvantages of current marking system is urgently to be resolved hurrily.
Meanwhile, online credit system has a long-time unheeded key issue, and namely the score value in the online credit system of great majority is all discrete and linear separation.For example it is known that Amazon and Netflix website adopt be five-pointed star marking system, commodity are marked by user by these five integers of setting 1 (worst) to 5 (best).But, the scoring of commodity is in practice likely to be nonlinear by user between two successive values.Such as, in five-pointed star marking system, the difference between score value 3 and 4 is likely to the difference being not equivalent between 4 and 5.Based on the consideration of this point, therefore, need a kind of brand-new scoring mapping method of proposition badly and grading system value is carried out definition again, to improve the accuracy of sort algorithm.
Summary of the invention
In view of this, the purpose of the embodiment of the present invention is in that to provide a kind of online methods of marking and system thereof, it is intended to solve the problem that not high and sort algorithm the accuracy of accuracy of five-pointed star marking system of the prior art is relatively low.
The embodiment of the present invention is achieved in that a kind of online methods of marking, including:
User data and commodity data are carried out predefined process；
Input one group of original score data；
Change to form one group of new score data to described original score data；
Adjust all data in described one group of new score data by preset algorithm, and utilize iterative manner until reaching stable state；
Under described stable state, obtain the optimum of described all data, and show that the preference of described original score data is arranged according to described optimum.
Preferably, the described step that user data and commodity data carry out predefined process specifically includes:
Labelling user data and marked articles data, wherein, be labeled as U by user's collection, and indication of goods is O, and the scoring of commodity α is labeled as r by user iiα, the commodity aggregated label that user i selects is Oi, the user's aggregated label selecting commodity α is Uα, the degree of user i and commodity α is respectively labeled as kiAnd kα, the prestige of user i and the quality of commodity α are respectively labeled as RiAnd Qα。
Preferably, described one group of original score data includes 1,2,3,4 and 5, is respectively used for representing the fine or not degree that commodity are evaluated by user, and wherein, the described step that described original score data is changed to be formed one group of new score data specifically includes:
Convert scoring 2 and 4 to the R that newly marks respectively2And R4, wherein, R2=1+p1* 2, R4=3+p2* 2, the first parameter P1∈ (0,1), the second parameter P2∈ (0,1), and adjust P1And P2Step-length be 0.01；
By 1, R2、3、R4Form one group of new score data.
Preferably, the described all data adjusted by preset algorithm in described one group of new score data, and utilize iterative manner until the step reaching stable state includes:
Described new scoring R is adjusted according to predetermined order algorithm2And R4In described first parameter p1With described second parameter p2, and utilize iterative manner until reaching stable state.
Preferably, the described optimum obtaining described all data under described stable state, and show that the step that the preference of described original score data is arranged specifically includes according to described optimum:
Described first parameter p is obtained under described stable state1With described second parameter p2Optimum；
According to described first parameter p1With described second parameter p2Optimum judge the higher or on the low side of original scoring.
On the other hand, the present invention also provides for a kind of online marking system, including:
Pretreatment module, for carrying out predefined process to user data and commodity data；
Data input module, is used for inputting one group of original score data；
Data conversion module, for changing to form one group of new score data to described original score data；
Data iteration module, for adjusting all data in described one group of new score data by preset algorithm, and utilizes iterative manner until reaching stable state；
Preference arranges module, for obtaining the optimum of described all data under described stable state, and show that the preference of described original score data is arranged according to described optimum.
Preferably, described pretreatment module is specifically for labelling user data and marked articles data, wherein, user's collection is labeled as U, and indication of goods is O, and the scoring of commodity α is labeled as r by user iiα, the commodity aggregated label that user i selects is Oi, the user's aggregated label selecting commodity α is Uα, the degree of user i and commodity α is respectively labeled as kiAnd kα, the prestige of user i and the quality of commodity α are respectively labeled as RiAnd Qα。
Preferably, described one group of original score data includes 1,2,3,4 and 5, is respectively used for representing the fine or not degree that commodity are evaluated by user, and wherein, described data conversion module, specifically for converting scoring 2 and 4 to the R that newly marks respectively2And R4, wherein, R2=1+p1* 2, R4=3+p2* 2, the first parameter P1∈ (0,1), the second parameter P2∈ (0,1), and adjust P1And P2Step-length be 0.01；And by 1, R2、3、R4Form one group of new score data.
Preferably, described data iteration module is specifically for adjusting described new scoring R according to predetermined order algorithm2And R4In described first parameter p1With described second parameter p2, and utilize iterative manner until reaching stable state.
Preferably, described preference arranges module specifically for obtaining described first parameter p under described stable state1With described second parameter p2Optimum, and according to described first parameter p1With described second parameter p2Optimum judge the higher or on the low side of original scoring.
In embodiments of the present invention, technical scheme provided by the invention for one based on 5 star marking systems, scoring mapping remains 1,3,5 scorings, represent worst, neutral and best respectively, and scoring 2 and 4 has been carried out definition again, the score data of gained is redistributed, scoring is mapped and combines from each sort algorithm being in harmony with current, thus expressing the quality of commodity more accurately, and then drastically increase the accuracy of five-pointed star marking system and improve the accuracy of sort algorithm.
Accompanying drawing explanation
Fig. 1 is online methods of marking flow chart in an embodiment of the present invention；
Fig. 2 is online marking system structural representation in an embodiment of the present invention.
Detailed description of the invention
In order to make the purpose of the present invention, technical scheme and advantage clearly understand, below in conjunction with drawings and Examples, the present invention is further elaborated.Should be appreciated that specific embodiment described herein is only in order to explain the present invention, is not intended to limit the present invention.
The specific embodiment of the invention provides a kind of online methods of marking, mainly comprises the steps:
S11, user data and commodity data are carried out predefined process；
S12, input one group of original score data；
S13, change to form one group of new score data to described original score data；
S14, adjusted all data in described one group of new score data by preset algorithm, and utilize iterative manner until reaching stable state；
S15, under described stable state, obtain the optimum of described all data, and show that the preference of described original score data is arranged according to described optimum.
The online methods of marking of one provided by the present invention, improves the accuracy of five-pointed star marking system and improves the accuracy of sort algorithm.
Hereinafter the online methods of marking of one provided by the present invention will be described in detail.
Refer to Fig. 1, for methods of marking flow chart online in an embodiment of the present invention.
In step s 11, user data and commodity data are carried out predefined process.
In the present embodiment, the step S11 that user data and commodity data carry out predefined process specifically includes:
Labelling user data and marked articles data, wherein, be labeled as U by user's collection, and indication of goods is O, and the scoring of commodity α is labeled as r by user iiα, the commodity aggregated label that user i selects is Oi, the user's aggregated label selecting commodity α is Uα, the degree of user i and commodity α is respectively labeled as kiAnd kα, the prestige of user i and the quality of commodity α are respectively labeled as RiAnd Qα。
In step s 12, one group of original score data is inputted.
In the present embodiment, inputting one group of original score data is 1 in five-pointed star marking system, 2,3,4,5 these five scoring marks, it is respectively used for representing the fine or not degree that commodity are evaluated by user, namely 1,2,3,4,5 represents difference worst, secondary respectively, neutrality, secondary good and best.
In step s 13, change to form one group of new score data to described original score data.
In the present embodiment, the described step S13 that described original score data is changed to be formed one group of new score data specifically includes:
Convert scoring 2 and 4 to the R that newly marks respectively2And R4, wherein, R2=1+p1* 2, R4=3+p2* 2, the first parameter P1∈ (0,1), the second parameter P2∈ (0,1), and adjust P1And P2Step-length be 0.01；
By 1, R2、3、R4Form one group of new score data.
In step S14, adjust all data in described one group of new score data by preset algorithm, and utilize iterative manner until reaching stable state.
In the present embodiment, each data have the parameter of oneself, it is actually the parameter that all data are corresponding here by what preset algorithm adjusted, when original score data and new score data existence part repeat, parameter owing to repeating data corresponding is not changed in, now adjusting all data is then adjust the parameter that more new data is corresponding, when original score data and new score data are absent from part repetition or repeat completely, then each has the data of renewal to adjust accordingly, and namely adjusts the parameter that each data is corresponding.
In the present embodiment, the described all data adjusted by preset algorithm in described one group of new score data, and utilize iterative manner until the step S14 reaching stable state includes:
Described new scoring R is adjusted according to predetermined order algorithm2And R4In described first parameter p1With described second parameter p2, and utilize iterative manner until reaching stable state.
In the present embodiment, predetermined order algorithm include on average (Mean) algorithm, iteration optimization (IterativeRefinement, IR) algorithm, based on relevance ranking (Correlation-basedRanking, CR) algorithm, prestige reallocation sequence (Reputationredistributionranking, RR) algorithm.
In the present embodiment, illustrate for IR algorithm, in this step, first by the quality Q of commodityαIt is defined as:
Q α = Σ i ∈ U α R i r i α Σ i ∈ U α R i , Wherein, initial setting up Qα=1；
Secondly, by the prestige R of user iiIt is defined as:
R i = ( 1 | O i | Σ α ∈ O i ( r i α - Q α ) 2 + ϵ ) - β ;
Then, iterative manner is utilized and at QαAnd RiAfter reaching stable state, terminate IR algorithm.
In the present embodiment, illustrate for CR algorithm, in this step, first by the quality Q of commodityαIt is defined as:
Secondly, by the prestige R of user iiIt is defined as:
R i = 1 k i Σ α ∈ O i ( r i α - r l ‾ σ r i ) ( Q α - Q l ‾ σ Q i ) ;
Then, the prestige R being initially set to user of CR algorithmi=ki/ | O |, wherein σriAnd σQiRepresent the standard deviation of all scorings of user i and the standard deviation of all commercial qualities of his evaluation respectively, andWithIt is the meansigma methods of the meansigma methods of all scorings of user i and all commercial qualities that he evaluates, if RiValue less than 0, then the prestige of user i will become 0, the prestige R of the user of this CR algorithmiValue interval be [0,1], iteration at QαAnd RiTerminate after reaching stable state.
In step S15, under described stable state, obtain the optimum of described all data, and show that the preference of described original score data is arranged according to described optimum.
In the present embodiment, the described optimum obtaining described all data under described stable state, and show that the step S15 that the preference of described original score data is arranged specifically includes according to described optimum:
Described first parameter p is obtained under described stable state1With described second parameter p2Optimum；
According to described first parameter p1With described second parameter p2Optimum judge the higher or on the low side of original scoring.
In the present embodiment, if p1< 0.5, then illustrate that scoring 2 is relatively scoring 1, original scoring is higher；P1> 0.5, then illustrating that scoring 2 is relatively scoring 3, original scoring is on the low side；If p2< 0.5, then illustrate that scoring 4 is relatively scoring 3, original scoring is higher；P2> 0.5, then illustrating that scoring 4 is relatively scoring 5, original scoring is on the low side.
In the present embodiment, carrying out digital simulation and test in systems in practice, and achieve considerable effect, present embodiment selects two conventional true marking system: Netflix and MovieLens.Wherein, MovieLens is provided by University of Minnesota (UniversityofMinnesota) GroupLens project team, the present invention uses a subset of complete data, in this subset, have 1,000,000 data marked based on the integer between 6040 users and 1 to the 5 of 3706 commodity；Netflix is according to collection by the googol that DVD leasing company Netflix is its Netflix prize issue, the present invention is by randomly choosing one less data set of extraction, this data set contains 5000 users and 16195 films that they were evaluated, here having ultimately generated 1070000 scorings based on five-pointed star, these basic features present invention is aggregated into as shown in table 1 below:
Table 1 show the basic feature of Netflix and MovieLens marking system
In the present embodiment, in order to test the performance of ranking, the present invention uses the Rank scores index RS (Rankingscore accurately weighing ranking of a standard, Rank scores), first the present invention selects one group by the high-quality benchmark project E (in Movielens in 203 and Netflix 293) of Oscar nomination, the commodity prize-winning for each are designated as α, and its ranking is designated as Dα, all commodity collection are designated as M, RS and are defined as:
R S = 1 | E | &Sigma; &alpha; &Element; E D &alpha; M
Wherein, according to this definition, an accurate ranking should have a less RS.
In the present embodiment, the present invention chooses RS at parameter space (P1, P2) hotspot's distribution figure, show according to this hotspot's distribution figure, scoring mapping method in, parameter p2Occupy an leading position, more specifically, as p in Movielens2<0.5 and in Netflix during p2>0.5, minimum RS can be obtained, this result that the present invention obtains when using CR algorithm with RR algorithm is consistent, as can be seen here, in Movielens, when a film is provided score value 4 by user, this value is relatively 3, and in Netflix, when a film is provided score value 4 by user, this value is relatively 5.In other words, p1Impact relatively small in RS, work as p1> 0.5 time, RS slightly improves, best p1And p2It is worth in the two real system respectively (0.75,0.25) and (0.85,1).
In the present embodiment, for embodying the specifically improvement of scoring mapping method, the RS based on the algorithm of scoring mapping after original sort algorithm and optimization is compared by the present invention, as shown in table 2 below:
Table 2 is the contrast table of several algorithm
As can be seen here, the RS employing the sort algorithm after scoring maps obtains certain improvement.In Movielens, scoring mapping pair average algorithm optimizes 2%, IR algorithm optimization 2%, CR algorithm optimization 2%, and RR algorithm optimization 6%.In Netflix, scoring mapping pair average algorithm optimizes 5%, IR algorithm optimization 4%, CR algorithm optimization 4%, and RR algorithm optimization 11%.Legitimate reading shows, the scoring of score value 4 map it is critical that.
The online methods of marking of one provided by the present invention, for one based on 5 star marking systems, scoring mapping remains 1,3,5 scoring, represent worst, neutral and best respectively, and scoring 2 and 4 has been carried out definition again, the score data of gained is redistributed, scoring is mapped and combines from each sort algorithm being in harmony with current, thus expressing the quality of commodity more accurately, and then drastically increase the accuracy of five-pointed star marking system and improve the accuracy of sort algorithm.
The specific embodiment of the invention also provides for a kind of online marking system 10, specifically includes that
Pretreatment module 11, for carrying out predefined process to user data and commodity data；
Data input module 12, is used for inputting one group of original score data；
Data conversion module 13, for changing to form one group of new score data to described original score data；
Data iteration module 14, for adjusting all data in described one group of new score data by preset algorithm, and utilizes iterative manner until reaching stable state；
Preference arranges module 15, for obtaining the optimum of described all data under described stable state, and show that the preference of described original score data is arranged according to described optimum.
The online marking system 10 of one provided by the present invention, improves the accuracy of five-pointed star marking system and improves the accuracy of sort algorithm.
Refer to Fig. 2, it is shown that for the structural representation of marking system 10 online in an embodiment of the present invention.In the present embodiment, online marking system 10 includes pretreatment module 11, data input module 12, data conversion module 13, data iteration module 14 and preference and arranges module 15.
Pretreatment module 11, for carrying out predefined process to user data and commodity data.
In the present embodiment, described pretreatment module 11 is specifically for labelling user data and marked articles data, wherein, user's collection is labeled as U, and indication of goods is O, and the scoring of commodity α is labeled as r by user iiα, the commodity aggregated label that user i selects is Oi, the user's aggregated label selecting commodity α is Uα, the degree of user i and commodity α is respectively labeled as kiAnd kα, the prestige of user i and the quality of commodity α are respectively labeled as RiAnd Qα。
Data input module 12, is used for inputting one group of original score data.
In the present embodiment, in the present embodiment, inputting one group of original score data is 1 in five-pointed star marking system, 2,3,4,5 these five scoring marks, it is respectively used for representing the fine or not degree that commodity are evaluated by user, namely 1,2,3,4,5 represents difference worst, secondary respectively, neutrality, secondary good and best.
Data conversion module 13, for changing to form one group of new score data to described original score data.
In the present embodiment, described one group of original score data includes 1,2,3,4 and 5, is respectively used for representing the fine or not degree that commodity are evaluated by user, and wherein, described data conversion module 13, specifically for converting scoring 2 and 4 to the R that newly marks respectively2And R4, wherein, R2=1+p1* 2, R4=3+p2* 2, the first parameter P1∈ (0,1), the second parameter P2∈ (0,1), and adjust P1And P2Step-length be 0.01；And by 1, R2、3、R4Form one group of new score data.
Data iteration module 14, for adjusting all data in described one group of new score data by preset algorithm, and utilizes iterative manner until reaching stable state.
In the present embodiment, each data have the parameter of oneself, it is actually the parameter that all data are corresponding here by what preset algorithm adjusted, when original score data and new score data existence part repeat, parameter owing to repeating data corresponding is not changed in, now adjusting all data is then adjust the parameter that more new data is corresponding, when original score data and new score data are absent from part repetition or repeat completely, then each has the data of renewal to adjust accordingly, and namely adjusts the parameter that each data is corresponding.
In the present embodiment, described data iteration module 14 is specifically for adjusting described new scoring R according to predetermined order algorithm2And R4In described first parameter p1With described second parameter p2, and utilize iterative manner until reaching stable state.
Preference arranges module 15, for obtaining the optimum of described all data under described stable state, and show that the preference of described original score data is arranged according to described optimum.
In the present embodiment, described preference arranges module 15 specifically for obtaining described first parameter p under described stable state1With described second parameter p2Optimum, and according to described first parameter p1With described second parameter p2Optimum judge the higher or on the low side of original scoring.
In the present embodiment, if p1< 0.5, then illustrate that scoring 2 is relatively scoring 1, original scoring is higher；P1> 0.5, then illustrating that scoring 2 is relatively scoring 3, original scoring is on the low side；If p2< 0.5, then illustrate that scoring 4 is relatively scoring 3, original scoring is higher；P2> 0.5, then illustrating that scoring 4 is relatively scoring 5, original scoring is on the low side.
The online marking system 10 of one provided by the present invention, for one based on 5 star marking systems, scoring mapping remains 1,3,5 scoring, represent worst, neutral and best respectively, and scoring 2 and 4 has been carried out definition again, the score data of gained is redistributed, scoring is mapped and combines from each sort algorithm being in harmony with current, thus expressing the quality of commodity more accurately, and then drastically increase the accuracy of five-pointed star marking system and improve the accuracy of sort algorithm.
It should be noted that in above-described embodiment, included unit is carry out dividing according to function logic, but is not limited to above-mentioned division, as long as being capable of corresponding function；It addition, the concrete title of each functional unit is also only to facilitate mutually distinguish, it is not limited to protection scope of the present invention.
Additionally, one of ordinary skill in the art will appreciate that all or part of step realizing in the various embodiments described above method can be by the hardware that program carrys out instruction relevant and completes, corresponding program can be stored in a computer read/write memory medium, described storage medium, such as ROM/RAM, disk or CD etc..
The foregoing is only presently preferred embodiments of the present invention, not in order to limit the present invention, all any amendment, equivalent replacement and improvement etc. made within the spirit and principles in the present invention, should be included within protection scope of the present invention.

Claims (10)

1. an online methods of marking, it is characterised in that described online methods of marking includes:
User data and commodity data are carried out predefined process；
Input one group of original score data；
Change to form one group of new score data to described original score data；
Adjust all data in described one group of new score data by preset algorithm, and utilize iterative manner until reaching stable state；
Under described stable state, obtain the optimum of described all data, and show that the preference of described original score data is arranged according to described optimum.
2. online methods of marking as claimed in claim 1, it is characterised in that the described step that user data and commodity data carry out predefined process specifically includes:
Labelling user data and marked articles data, wherein, be labeled as U by user's collection, and indication of goods is O, and the scoring of commodity α is labeled as r by user iiα, the commodity aggregated label that user i selects is Oi, the user's aggregated label selecting commodity α is Uα, the degree of user i and commodity α is respectively labeled as kiAnd kα, the prestige of user i and the quality of commodity α are respectively labeled as RiAnd Qα。
3. online methods of marking as claimed in claim 2, it is characterized in that, described one group of original score data includes 1,2,3,4 and 5, it is respectively used for representing the fine or not degree that commodity are evaluated by user, wherein, the described step that described original score data is changed to be formed one group of new score data specifically includes:
Convert scoring 2 and 4 to the R that newly marks respectively2And R4, wherein, R2=1+p1* 2, R4=3+p2* 2, the first parameter P1∈ (0,1), the second parameter P2∈ (0,1), and adjust P1And P2Step-length be 0.01；
By 1, R2、3、R4Form one group of new score data.
4. online methods of marking as claimed in claim 3, it is characterised in that the described all data adjusted by preset algorithm in described one group of new score data, and utilize iterative manner until the step reaching stable state includes:
Described new scoring R is adjusted according to predetermined order algorithm2And R4In described first parameter p1With described second parameter p2, and utilize iterative manner until reaching stable state.
5. online methods of marking as claimed in claim 4, it is characterised in that the described optimum obtaining described all data under described stable state, and show that the step that the preference of described original score data is arranged specifically includes according to described optimum:
Described first parameter p is obtained under described stable state1With described second parameter p2Optimum；
According to described first parameter p1With described second parameter p2Optimum judge the higher or on the low side of original scoring.
6. an online marking system, it is characterised in that described online marking system includes:
Pretreatment module, for carrying out predefined process to user data and commodity data；
Data input module, is used for inputting one group of original score data；
Data conversion module, for changing to form one group of new score data to described original score data；
Data iteration module, for adjusting all data in described one group of new score data by preset algorithm, and utilizes iterative manner until reaching stable state；
Preference arranges module, for obtaining the optimum of described all data under described stable state, and show that the preference of described original score data is arranged according to described optimum.
7. online marking system as claimed in claim 6, it is characterised in that described pretreatment module is specifically for labelling user data and marked articles data, wherein, is labeled as U by user's collection, and indication of goods is O, and the scoring of commodity α is labeled as r by user iiα, the commodity aggregated label that user i selects is Oi, the user's aggregated label selecting commodity α is Uα, the degree of user i and commodity α is respectively labeled as kiAnd kα, the prestige of user i and the quality of commodity α are respectively labeled as RiAnd Qα。
8. online marking system as claimed in claim 7, it is characterised in that described one group of original score data includes 1,2,3,4 and 5, it is respectively used for representing the fine or not degree that commodity are evaluated by user, wherein, described data conversion module, specifically for converting scoring 2 and 4 to the R that newly marks respectively2And R4, wherein, R2=1+p1* 2, R4=3+p2* 2, the first parameter P1∈ (0,1), the second parameter P2∈ (0,1), and adjust P1And P2Step-length be 0.01；And by 1, R2、3、R4Form one group of new score data.
9. online marking system as claimed in claim 8, it is characterised in that described data iteration module is specifically for adjusting described new scoring R according to predetermined order algorithm2And R4In described first parameter p1With described second parameter p2, and utilize iterative manner until reaching stable state.
10. online marking system as claimed in claim 9, it is characterised in that described preference arranges module specifically for obtaining described first parameter p under described stable state1With described second parameter p2Optimum, and according to described first parameter p1With described second parameter p2Optimum judge the higher or on the low side of original scoring.
CN201610029729.0A 2016-01-15 2016-01-15 A kind of online methods of marking and its system Active CN105718535B (en)

Priority Applications (2)

Application Number Priority Date Filing Date Title
CN201610029729.0A CN105718535B (en) 2016-01-15 2016-01-15 A kind of online methods of marking and its system
PCT/CN2016/081859 WO2017121054A1 (en) 2016-01-15 2016-05-12 Online scoring method and system

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201610029729.0A CN105718535B (en) 2016-01-15 2016-01-15 A kind of online methods of marking and its system

Publications (2)

Publication Number Publication Date
CN105718535A true CN105718535A (en) 2016-06-29
CN105718535B CN105718535B (en) 2018-03-02

Family

ID=56147090

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201610029729.0A Active CN105718535B (en) 2016-01-15 2016-01-15 A kind of online methods of marking and its system

Country Status (2)

Country Link
CN (1) CN105718535B (en)
WO (1) WO2017121054A1 (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109711875A (en) * 2018-12-19 2019-05-03 口碑(上海)信息技术有限公司 Content recommendation method and device

Families Citing this family (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111258327B (en) * 2020-04-01 2023-04-11 昆山天科智控无人机科技有限公司 Unmanned aerial vehicle automatic control method based on maximum scoring state updating
CN112269924B (en) * 2020-10-16 2024-08-09 北京师范大学珠海校区 Ranking-based comment method, ranking-based comment device, electronic equipment and medium

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
KR20050013350A (en) * 2003-07-28 2005-02-04 주식회사파이소프트 Searching method for online score
CN103020784A (en) * 2012-12-27 2013-04-03 智比特信息技术（镇江）有限公司 Object evaluation system
CN103679507A (en) * 2013-12-06 2014-03-26 冯志刚 Method and system for grading and ranking works or commodities

Family Cites Families (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
KR20010091480A (en) * 2000-03-15 2001-10-23 강병수 Cyber traveling apparatus in the internet
JP2002342639A (en) * 2001-05-17 2002-11-29 Takuo Koyama Method for publicly opening user voice to merchandise via communication network
CN100568243C (en) * 2007-09-04 2009-12-09 华为技术有限公司 The method and system of a kind of data mining and modeling
CN103544610A (en) * 2012-07-15 2014-01-29 常州白相相文化传媒有限公司 Evaluation system based on online shopping
CN102915501A (en) * 2012-10-29 2013-02-06 江苏乐买到网络科技有限公司 Method for optimizing online shopping evaluating information
CN103309990A (en) * 2013-06-18 2013-09-18 上海晶樵网络信息技术有限公司 User multidimensional analysis and monitoring method based on public information of Internet user
CN103778235A (en) * 2014-01-26 2014-05-07 北京京东尚科信息技术有限公司 Method and device for processing commodity assessment information
CN104133833A (en) * 2014-05-27 2014-11-05 汉柏科技有限公司 Data mining method and system
CN104765852B (en) * 2015-04-21 2018-01-23 成都博元时代软件有限公司 Data digging method based on fuzzy algorithmic approach under big data background

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
KR20050013350A (en) * 2003-07-28 2005-02-04 주식회사파이소프트 Searching method for online score
CN103020784A (en) * 2012-12-27 2013-04-03 智比特信息技术（镇江）有限公司 Object evaluation system
CN103679507A (en) * 2013-12-06 2014-03-26 冯志刚 Method and system for grading and ranking works or commodities

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109711875A (en) * 2018-12-19 2019-05-03 口碑(上海)信息技术有限公司 Content recommendation method and device
CN109711875B (en) * 2018-12-19 2021-04-20 口碑(上海)信息技术有限公司 Content recommendation method and device

Also Published As

Publication number Publication date
CN105718535B (en) 2018-03-02
WO2017121054A1 (en) 2017-07-20

Similar Documents

Publication Publication Date Title
CN102508907B (en) Dynamic recommendation method based on training set optimization for recommendation system
CN103617157B (en) Based on semantic Text similarity computing method
CN104809243B (en) It is a kind of that method is recommended based on the mixing excavated to user behavior composite factor
CN102236693A (en) Method and device for determining similarity between documents
CN105975459B (en) A kind of the weight mask method and device of lexical item
CN104850567A (en) Method and device for identifying association between network users
CN107330727A (en) A kind of personalized recommendation method based on hidden semantic model
CN106126751A (en) A kind of sorting technique with time availability and device
CN102982107A (en) Recommendation system optimization method with information of user and item and context attribute integrated
CN105022754A (en) Social network based object classification method and apparatus
CN104077415A (en) Searching method and device
Gao et al. SeCo-LDA: Mining service co-occurrence topics for recommendation
CN105718535A (en) Online scoring method and system
CN108573041A (en) Probability matrix based on weighting trusting relationship decomposes recommendation method
CN106919699A (en) A kind of recommendation method for personalized information towards large-scale consumer
CN108920647B (en) Low-rank matrix filling TOP-N recommendation method based on spectral clustering
WO2013122205A1 (en) Dictionary generation device, dictionary generation method, dictionary generation program and computer-readable recording medium storing same program
CN104820702A (en) Attribute weighting method based on decision tree and text classification method
Yao et al. Online deception detection refueled by real world data collection
CN107193916A (en) Method and system are recommended in a kind of personalized variation inquiry
CN104462597B (en) A kind of positive negativity of synthetic user scores and the collaborative filtering method of scoring preference heterogeneity
CN106204153A (en) A kind of two-staged prediction Top N proposed algorithm based on attribute proportion similarity
CN106682151A (en) Personalized educational-resource recommendation method and system
Luo et al. Discrimination-Aware Classifiers for Student Performance Prediction.
CN104731819A (en) Advertisement keyword extraction method and advertisement keyword extraction device

Legal Events

Date Code Title Description
C06 Publication
PB01 Publication
C10 Entry into substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant