KR20160133835A - Social network-based User rating bias correction apparatus and method therefor - Google Patents

Social network-based User rating bias correction apparatus and method therefor Download PDF

Info

Publication number
KR20160133835A
KR20160133835A KR1020150066875A KR20150066875A KR20160133835A KR 20160133835 A KR20160133835 A KR 20160133835A KR 1020150066875 A KR1020150066875 A KR 1020150066875A KR 20150066875 A KR20150066875 A KR 20150066875A KR 20160133835 A KR20160133835 A KR 20160133835A
Authority
KR
South Korea
Prior art keywords
evaluation
score
evaluator
data
evaluators
Prior art date
Application number
KR1020150066875A
Other languages
Korean (ko)
Inventor
하정락
정영호
김순철
최범석
홍진우
Original Assignee
한국전자통신연구원
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 한국전자통신연구원 filed Critical 한국전자통신연구원
Priority to KR1020150066875A priority Critical patent/KR20160133835A/en
Publication of KR20160133835A publication Critical patent/KR20160133835A/en

Links

Images

Classifications

    • 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—Systems or methods specially adapted for specific business sectors, e.g. utilities or tourism
    • G06Q50/01—Social networking
    • 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—Systems or methods specially adapted for specific business sectors, e.g. utilities or tourism
    • G06Q50/10—Services
    • 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—Systems or methods specially adapted for specific business sectors, e.g. utilities or tourism
    • G06Q50/30—Transportation; Communications

Landscapes

  • Business, Economics & Management (AREA)
  • Tourism & Hospitality (AREA)
  • Engineering & Computer Science (AREA)
  • General Health & Medical Sciences (AREA)
  • General Business, Economics & Management (AREA)
  • Human Resources & Organizations (AREA)
  • Marketing (AREA)
  • Primary Health Care (AREA)
  • Strategic Management (AREA)
  • Physics & Mathematics (AREA)
  • Economics (AREA)
  • General Physics & Mathematics (AREA)
  • Health & Medical Sciences (AREA)
  • Theoretical Computer Science (AREA)
  • Operations Research (AREA)
  • Computing Systems (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

The present invention relates to a social network-based user rating bias correction apparatus and a method therefor which may include a rating data acquisition part, a score bias correction part, and a score storage part. In order to prevent the score of an object from being biased by an active rater, the score of the rater who evaluates an appropriate rating number or more is corrected. So, the score of the object is multiplied by raters.

Description

[0001] The present invention relates to a social network-based user evaluation bias correction apparatus and a method thereof,

The present invention relates to an evaluation system that can be evaluated by a user on an object to be evaluated in a social network. More particularly, the present invention relates to an evaluation system in which an evaluation of a plurality of evaluators is fairly applied And the contents to be corrected.

Currently, the evaluation of individuals, organizations, products, and places is done by various people regardless of internet, intranet, online, offline.

However, as the number of evaluators participating in the evaluation of the positive tendency is relatively higher than that of the passive tendency, the evaluation results are biased according to the opinions of the evaluators who are actively inclined according to the tendency of the evaluators.

In addition, as part of commercial marketing in recent years, it has been known that intentionally malicious intentions such as advertising intention, intentional intentional intentions, such as providing money, inducing a high evaluation of a particular product, or intentionally providing a bad evaluation of a product, There are many cases where the evaluation result is distorted, so that the evaluation results are more frequently biased.

In order to prevent the evaluation score of the evaluation object from being biased by the active evaluators, the evaluation score of the evaluator to be evaluated is corrected by a plurality of evaluators It has its purpose.

According to an embodiment of the present invention, a social network based user evaluation bias correction apparatus includes an evaluation data acquisition unit for obtaining evaluation data for each evaluation object from at least one user; An evaluation score deflection correcting unit for calculating evaluation scores by analyzing the evaluation data, correcting the evaluation scores according to the number of evaluations for each evaluator, and generating or updating evaluation score data including the deflection corrected evaluation results; And an evaluation score storage unit for outputting the evaluation score data to a user as a result of storage and evaluation.

1 is a diagram showing a relationship between an evaluator and an evaluation object in a general evaluation system.
2 is a configuration diagram of a user evaluation bias correction apparatus according to an embodiment of the present invention.
Fig. 3 is a diagram showing evaluation score calculation in an evaluation system not subjected to evaluation deflection correction.
4 is a diagram showing evaluation score calculation using the evaluation deflection correcting apparatus according to the embodiment of the present invention.
5 is a flowchart illustrating a user evaluation bias correction method according to an embodiment of the present invention.

Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings so that those skilled in the art can easily carry out the present invention. The present invention may, however, be embodied in many different forms and should not be construed as limited to the embodiments set forth herein.

In order to clearly illustrate the present invention, parts not related to the description are omitted, and similar parts are denoted by like reference characters throughout the specification.

Throughout the specification, when an element is referred to as "comprising ", it means that it can include other elements as well, without excluding other elements unless specifically stated otherwise.

Hereinafter, an apparatus and method for correcting a user evaluation bias based on a social network according to an embodiment of the present invention will be described with reference to the drawings.

1 is a diagram showing a relationship between an evaluator and an evaluation object in a general evaluation system.

Referring to FIG. 1, a plurality of evaluators evaluate a plurality of evaluation objects in a general evaluation system.

Generally, in a social network service (SNS), a person who can evaluate an object to be disclosed, such as a post by a person on the other end, more precisely, a post, a music, an artwork, a person, an organization, a product, Service.

Such an evaluation service may affect other users viewing it for purchasing, using, visiting, or personal evaluation of the subject.

Referring to FIG. 1, in this general evaluation system, a plurality of evaluators can evaluate a plurality of evaluation objects, and when there are I persons in the evaluation system and J evaluation objects exist, each evaluator performs i 1 , i 2 , i 3 , ... i I , and the object to be evaluated is j 1 , j 2 , ... , j J , respectively.

Here, the evaluator may be an individual or organization as the subject who gives the evaluation opinion on the evaluation subject. The evaluation subject may be a variety of evaluation items including writing, music, artwork, person, organization, product, Object. ≪ / RTI >

1, the evaluator i 1 evaluates the evaluation objects j 1 and j 3 , the evaluator i 2 evaluates the evaluation objects j 1 , j 2 , and j J , and the evaluator i 3 evaluates the evaluation object j 3 And the evaluator i I can evaluate the evaluation object j J.

In this evaluation system, active evaluators such as i2, which conducts a large number of evaluations on various evaluation subjects positively according to the tendency of evaluators, and passive evaluators such as evaluator i3, whose evaluation frequency is relatively lower than the positive evaluator, can be distinguished.

An evaluation score for a specific evaluation object on the SNS used by a large number of people may be deflected according to the opinions of the positive evaluators.

2 is a configuration diagram of a user evaluation bias correction apparatus according to an embodiment of the present invention.

Referring to FIG. 2, the user evaluation bias correcting apparatus according to an embodiment of the present invention may include an evaluation data obtaining unit 100, an evaluation score deflection correction unit 200, and an evaluation score storage unit 300.

The evaluation data obtaining unit 100 may obtain the evaluation data for each evaluation object from at least one user.

According to the embodiment of the present invention, the evaluation data is not limited to the evaluation comment for each evaluation object or the evaluation score such as the likelihood, the likelihood, and the water quality, but can be used without restriction as long as it is data that can be quantified by reflecting the opinion of the evaluator.

The evaluation score deflection correction unit 200 analyzes the acquired evaluation data to calculate an evaluation score, corrects the calculated evaluation score according to the number of evaluations for each evaluator, generates or updates the evaluation score data, The deviation of the score can be corrected.

According to the embodiment of the present invention, the evaluation score is calculated from the evaluation data received to correct the deviation of the evaluation score by the specific evaluators actively evaluating, and the evaluation result is calculated by performing the deviation correction on the calculated evaluation score .

According to an embodiment of the present invention, the evaluation score calculated from the received evaluation data can be normalized for each evaluator.

According to an embodiment of the present invention, the evaluation result can be calculated by normalizing the deflection-corrected evaluation score according to the evaluation subject.

The deflection correction formula that can calculate the corrected score here is described with reference to Table 1.

sign Item Contents I Number of overall evaluators Individual evaluators are marked with i J Number of total evaluation subjects Individual evaluation subject is marked with j k (i, j) Evaluator i's evaluation of evaluation object j As an evaluation of j for any i

Figure pat00001
Normalize to have a value between K (i, j) Evaluator-specific normalization of evaluator i's evaluation target j Calculated by reflecting the average of k (i, j) for all evaluated objects j which i has evaluated
Figure pat00002
 M Number of appropriate evaluations of evaluator i When a specific evaluator i evaluates M or more evaluation objects, the sum of the evaluation scores is calculated to be M
(
Figure pat00003
) m Number of evaluators evaluated by evaluator i When a specific evaluator i evaluates m evaluation objects whose number is M or more, the sum of the evaluation scores is calculated to be M
Figure pat00004
)
E (i, j) The bias correction of the evaluation score for the evaluation object j of the evaluator i The score where each i evaluated each j
Figure pat00005
 V (j) Total of all evaluators' evaluations received by subject j
Figure pat00006
 A (j) The evaluator evaluates the evaluation results of all evaluators received by the evaluation subject j according to the evaluation subject
Figure pat00007

Referring to Table 1, the total number of evaluators is defined as I, and the individual evaluators are denoted by i, and each individual evaluator is represented by i1, i2, ... i, the total number of subjects to be evaluated is defined as J, the individual evaluation subject is represented by j, and each individual evaluation subject is represented as j1, j2, ... jJ.

According to the embodiment of the present invention, the evaluation score calculated from the evaluation of evaluation made by the evaluator i on the evaluation subject j can be defined as k (i, j). For convenience of description, k (i, j) (I, j) < / = 1, but it is not limited to this. According to another embodiment, k (i, j) .

According to the embodiment using the range where 0? K (i, j)? 1, according to the evaluation method, the affinity between the superscripts is 1.0, 0.8, 0.6, 0.4, 0.2 points, , 0.1 for 10 points, ... , One point is calculated as 1 point for "good" and 0 point for "dislike", and the evaluation score can be calculated from the evaluation opinion.

(I, j) = k (i, j) -meanj (k (i, j)) is calculated by normalizing the evaluation scores evaluated for the respective objects by the evaluator according to an embodiment of the present invention. )) + 0.5), that is, normalization by the evaluator.

K Applying the evaluator by normalization to the (i, j) in the example of evaluating a score of 0 to to 1.0, the evaluator i 1 the evaluation object j 1 of the k (i, j) in accordance with one embodiment of the present invention, j 2 , and j 3 are evaluated as 0.1, 0.2, and 0.3, and the evaluator i 2 is evaluated as 0.7, 0.8, and 0.9 for the evaluation subjects j 1 , j 2 , and j 3 . Is adjusted to the middle 0.5, these scores can be normalized to 0.4, 0.5, and 0.6, respectively.

At this time, it is possible to selectively normalize the evaluation score obtained by evaluating each object. In an embodiment that does not normalize, k (i, j) can be defined as K (i, j).

According to an embodiment of the present invention, K (i, j) obtained after the normalization process described above can be corrected according to the number of evaluations of the evaluator i.

At this time, the bias correction method judges whether the evaluation number of the evaluator exceeds the maximum evaluation number, and if M> m, the evaluator i only evaluates the evaluation score of the evaluator i If M <= m, the evaluator i has evaluated the number of evaluation (M) or more, so that the evaluation score of evaluator i is reduced so as not to overestimate so that the total is M

Figure pat00008
).

This adds up the evaluation scores of all evaluators for each bias-corrected evaluation subject (

Figure pat00009
), Summed up
Figure pat00010
Is normalized by dividing it by I, which is the total number of evaluators, to generate or update evaluation score data as a result of the calculation, and to provide a bias correction function to the evaluation system on the SNS.

In this case, the normalization per subject can be selectively applied. If not applied, A (j) = V (j) may be used.

According to an embodiment of the present invention, when the evaluation score data for the evaluation subject does not exist, the evaluation score data is generated, and when the evaluation score data for the evaluation subject already exists, the evaluation score data existing is updated can do.

The evaluation score storage unit 300 may output the evaluation score data to the user as a result of storage and evaluation.

According to the embodiment of the present invention, the method of outputting the evaluation result can be used in any format as long as the method of conveying the opinion on evaluation such as numerical value, star rating, likelihood,

Fig. 3 is a diagram showing evaluation score calculation in an evaluation system not subjected to evaluation deflection correction.

Referring to FIG. 3, it is shown that four evaluators calculate evaluation scores when evaluating three evaluation objects individually.

According to the embodiment in which the evaluation score for the evaluation of the three evaluation objects by the four evaluators is calculated through the evaluation system including the evaluation comment of "good ", the evaluators i1, i2, i3, (1, 1), k (1,2), k (2,1), k (i, j) (3, 3), k (3, 1), k (3, 3), and k (4, 3)

In this case, if the evaluation score is calculated without using the evaluation bias correction apparatus according to the embodiment of the present invention, the calculated score is not subjected to the normalization by the evaluator but the evaluation score of each evaluator is added as it is, It is possible to output the evaluation result calculated through the normalization process for each evaluation subject, which does not go through bias correction but divides the sum of the scores by the number of evaluators.

4 is a diagram showing evaluation score calculation using the evaluation deflection correcting apparatus according to the embodiment of the present invention.

4, evaluators i1, i2, i3, and i4 select evaluation objects j1, j2, i3, and i4 in an evaluation system that is selected by " j3, the evaluator i1 presents the evaluation opinion of "good" to j1, j2, the evaluator i2 presents the evaluation opinion of "good" to j1, j3, the evaluator i3 evaluates the evaluation opinion of "good" to j1, j3 , And the evaluator i4 presents an evaluation comment of "good" to j3.

At this time, for example, when the predetermined number M of evaluations is 1, the evaluator i4 indicates the number of evaluations one time or less, so that it can be distinguished as a passive evaluator. Evaluators i1, i2, i3 are evaluated twice. .

In this case, the evaluation comment of "good" is yes or no. Therefore, if the evaluation opinion of the evaluator is expressed as "yes" or "no", the evaluator's evaluation comment "good" can be numerically converted to 1 and the evaluation opinion is merely " There is only one case in which the evaluation score is one, and therefore even if the calculated evaluation score is normalized by the evaluator, there is no change in the evaluation score (

Figure pat00011
1 + 0.5 + 0.5 = 1).

After normalization, if bias correction is performed for each evaluator, evaluation comments of passive evaluator i4 can be quantified as it is, and one point can be added to j4.

On the other hand, the positive evaluators i1, i2, and i3 each have two evaluation times,

Figure pat00012
Formula for calculating
Figure pat00013
in
Figure pat00014
M = 1 and m = 2, so a larger value of 2 is selected
Figure pat00015
Lt; RTI ID = 0.0 &gt; 1/2 &
Figure pat00016
Respectively.

The biased corrected score

Figure pat00017
, And adding them to each evaluation object (
Figure pat00018
, V (1) can be 1/2 + 1/2 + 1/2 = 1.5, V (2) can be 1/2 and V (3) can be 1/2 + 1/2 + 1 = 2.

When the normalization is performed on the subject V (j) subjected to the bias correction by the above-described method

Figure pat00019
), The total number of evaluators is 4, and the total number of evaluators I is 4. Therefore, the normalized values of A (1) and A (2) are 3/8 (3/2 X 1/4) (1/2 X 1/4) and A (3) can be calculated as 1/2 (2 X 1/4).

5 is a flowchart illustrating a user evaluation bias correction method according to an embodiment of the present invention.

Evaluation data for each evaluation object is obtained from users (510).

Save the score data and output it to the user as the evaluation result

According to the embodiment of the present invention, the evaluation data is not limited to the evaluation comment for each evaluation object or the evaluation score such as the likelihood, the likelihood, and the water quality, but can be used without restriction as long as it is data that can be quantified by reflecting the opinion of the evaluator.

The acquired evaluation data is analyzed to calculate an evaluation score (520).

According to the embodiment of the present invention, the evaluation score is calculated from the evaluation data received to correct the deviation of the evaluation score by the specific evaluators actively evaluating, and the evaluation result is calculated by performing the deviation correction on the calculated evaluation score .

According to an embodiment of the present invention, the evaluation score calculated from the received evaluation data can be normalized for each evaluator.

The calculated evaluation score is deflected according to the number of evaluations (530).

According to the embodiment of the present invention, the bias correction formula which can calculate the corrected score can follow the formula of Table 1. [

The evaluation score data is generated or updated with the bias-corrected evaluation score (540).

In addition, according to the embodiment of the present invention, the evaluation result can be calculated by normalizing the deflection-corrected evaluation score by the evaluation object.

Here, the evaluation score data may mean that the evaluation value that is biased or biased corrected is a normalized evaluation score.

According to an embodiment of the present invention, when the evaluation score data for the evaluation subject does not exist, the evaluation score data is generated, and when the evaluation score data for the evaluation subject already exists, the evaluation score data existing is updated can do.

The evaluation score data is output to the user as a result of storage and evaluation (550).

The score data may be stored in a database, a proxy server, a main server, or the like on a social network, but not limited thereto.

The format or method of output to the user is not limited, and any method can be used without limitation as long as the user can recognize the information by visual, auditory, or the like.

While the present invention has been particularly shown and described with reference to exemplary embodiments thereof, it is to be understood that the invention is not limited to the disclosed exemplary embodiments, but, on the contrary, Various modifications and improvements of those skilled in the art using the basic concept of the present invention are also within the scope of the present invention.

100: evaluation data acquisition unit 200: evaluation score deflection correction unit
300: score storage unit

Claims (1)

An evaluation data obtaining unit that obtains evaluation data for each evaluation object from at least one user;
An evaluation score deflection correcting unit for calculating evaluation scores by analyzing the evaluation data, correcting the evaluation scores according to the number of evaluations for each evaluator, and generating or updating evaluation score data including the deflection corrected evaluation results; And
And an evaluation score storage unit for outputting the evaluation score data to a user as a result of storage and evaluation.
KR1020150066875A 2015-05-13 2015-05-13 Social network-based User rating bias correction apparatus and method therefor KR20160133835A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
KR1020150066875A KR20160133835A (en) 2015-05-13 2015-05-13 Social network-based User rating bias correction apparatus and method therefor

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
KR1020150066875A KR20160133835A (en) 2015-05-13 2015-05-13 Social network-based User rating bias correction apparatus and method therefor

Publications (1)

Publication Number Publication Date
KR20160133835A true KR20160133835A (en) 2016-11-23

Family

ID=57541488

Family Applications (1)

Application Number Title Priority Date Filing Date
KR1020150066875A KR20160133835A (en) 2015-05-13 2015-05-13 Social network-based User rating bias correction apparatus and method therefor

Country Status (1)

Country Link
KR (1) KR20160133835A (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20190050782A1 (en) * 2017-08-14 2019-02-14 ScoutZinc, LLC System and method for rating of personnel using crowdsourcing in combination with weighted evaluator ratings

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20190050782A1 (en) * 2017-08-14 2019-02-14 ScoutZinc, LLC System and method for rating of personnel using crowdsourcing in combination with weighted evaluator ratings
US11816622B2 (en) * 2017-08-14 2023-11-14 ScoutZinc, LLC System and method for rating of personnel using crowdsourcing in combination with weighted evaluator ratings

Similar Documents

Publication Publication Date Title
CN109690608B (en) Extrapolating trends in trust scores
Ali Impact of consumer animosity, boycott participation, boycott motivation, and product judgment on purchase readiness or aversion of Kurdish consumers in Iraq
Savage et al. Detection of opinion spam based on anomalous rating deviation
Valentine et al. The ethical decision making of men and women executives in international business situations
Shaari et al. Dimension of halal purchase intention: A preliminary study
A. Clottey et al. Non-response bias assessment in logistics survey research: use fewer tests?
Hung et al. Examining the cultural moderation on the acceptance of mobile commerce
Zhang et al. The effect of management response similarity on online hotel booking: Field evidence from Expedia
US20140172545A1 (en) Learned negative targeting features for ads based on negative feedback from users
Rifon et al. Don’t hate me because I am beautiful: Identifying the relative influence of celebrity attractiveness and character traits on credibility
Lappeman et al. Firestorm response: Managing brand reputation during an nWOM firestorm by responding to online complaints individually or as a cluster
US20180089318A1 (en) Two-sided network growth optimization in social networks
Tiwari et al. Corporate social responsibility and supply chain: A study for evaluating corporate hypocrisy with special focus on stakeholders
US20140172544A1 (en) Using negative feedback about advertisements to serve advertisements
JP6566515B2 (en) Item recommendation system and item recommendation method
Phau et al. To buy or not to buy a “counterfeit” Ralph Lauren polo shirt: The role of lawfulness and legality toward purchasing counterfeits
AU2019203701A1 (en) Weighted Voting System
Feng et al. Effects of review spam in a firm-initiated virtual brand community: Evidence from smartphone customers
Primabudi et al. Analysis of factors influencing purchase decision in online store. Case study: Game sales in online store
KR20160133835A (en) Social network-based User rating bias correction apparatus and method therefor
Nikolopoulos Forecasting with quantitative methods: The impact of special events in time series
Yuniarinto et al. Retail attribute’s effect on shopping motivation and customer loyalty: Age as a moderating variable
Oboler et al. Hate speech: A quality of service challenge
US20230032053A1 (en) Monitoring of a project by video analysis
Ramesh et al. Secure recommendation system for E-commerce website