WO2017213278A1 - Attribute analysis system and method for analysis target, and recording medium having, stored thereon, computer program for method - Google Patents

Abstract

The present invention comprises: multiple panel terminals for assessing attributes of an analysis target provided in the form of a word or an image; an attribute analysis server for providing an attribute analysis word for assessment of the attributes of the analysis target and analyzing the attributes of the analysis target on the basis of the assessment results of panels; and a client terminal for receiving the analyzed attribute assessment results for the analysis target from the attribute analysis server, receiving from a client the input of two keywords, the similarity of which is desired to be received from the attribute analysis server, and providing same to the attribute analysis server, wherein the attribute analysis server comprises: a word storage unit in which attribute analysis word pairs are stored; a word provision unit for providing the analysis target and information on the attribute analysis word pairs to the panel terminals; an assessment result reception unit for receiving and storing the assessment results and response time information of a user via the panel terminals; an attribute analysis unit for analyzing the attributes and sensibilities of the analysis target by using the assessment results; an attribute database for storing attribute results for analysis targets; a similarity calculation unit for calculating the similarity between the two keywords; and a request processing unit for providing the attribute assessment results and similarity to the client terminal.

Description

Property analysis system and method for analysis target and recording medium storing computer program for method

The present invention relates to a system and method for analyzing an attribute of an object to be analyzed, and a recording medium storing a computer program for the method. The present invention relates to a property of a word or an image in a multi-dimensional manner, so that an image for each attribute of an object felt by a consumer can be identified. The present invention relates to an attribute analysis system and method for an analysis target, and a recording medium storing a computer program for the method.

Companies in modern society are implementing various strategies to secure the competitiveness of their products. In particular, the advertising marketing strategy focuses on informing the consumer of the attribute image of the product. Recently, through advertising marketing, it is known that it is a product that meets the mental and psychological needs of consumers.

This is called emotional marketing, and in the case of such emotional marketing, an analysis should be conducted in accordance with the mental and psychological needs of consumers, and the human emotion should be quantitatively and qualitatively measured. In addition, the measured results can be scientifically analyzed and applied to the object of emotional engineering measurement.

The trend of modern consumption is not only changing the primary physical and material satisfaction of products but also the tendency to think about the emotions, satisfaction, and emotions that consumers feel by consuming and using products. Accordingly, researches related to emotional engineering have been conducted in various ways in developed countries such as the United States and Japan, and large Korean companies have also set up an emotional engineering research organization for emotional design, and have developed technology for emotional measurement in the long term. Coming.

In order to find such a measure of emotion, that is, an attribute image of a product that a consumer feels, there is a limit to the existing direct and surface question-and-answer form of question and answer.

An object of the present invention is to analyze the property of the target object to solve the problem and to give a participant to evaluate the degree of the attribute of the object to provide a solution that can specify the property of the object to derive a marketing strategy analysis method and method Then, a recording medium storing a computer program for the method is provided.

It is also an object of the present invention to provide an attribute analysis system and method for analyzing an object to measure the clarity of the connection between the object and the property through the attribute score and response time, and a recording medium storing a computer program for the method .

It is also an object of the present invention to provide a system and method for analyzing an attribute of an analysis target, which can extract an object (for example, an entertainer or an animal) similar to the attribute information, and to derive insights about a media advertisement. The stored recording medium is provided.

It is also an object of the present invention to implement a means for providing similarity between two keywords.

The present invention includes a plurality of panel terminals for collecting the property evaluation results of consumers for the analysis target provided in the form of words or images; An attribute analysis server for storing the attribute evaluation result for the analysis target and analyzing the attribute for the analysis target according to the evaluation result of the panel; A client terminal that receives a keyword (object) to be analyzed from a client and provides the attribute analysis server to the attribute analysis server, and receives a property evaluation result for the input analysis object from the attribute analysis server, wherein the attribute analysis server includes: A word storage unit for storing analysis word pairs, a word providing unit for providing analysis object and attribute analysis word pair information to a panel terminal, an evaluation result receiving unit for receiving and storing evaluation results and response time information of a user through the panel terminal; An attribute analyzer for analyzing attributes and emotions for the analysis target using the evaluation results, an attribute database for storing attribute results for the analysis targets, a similarity calculator for calculating similarity between two keywords, and the client It includes a request processing unit that provides the terminal with the result of the attribute evaluation and the degree of similarity. That may be characterized.

The word storage unit stores 30 attribute analysis word pairs, the attribute analysis unit analyzes the primary attributes of the analysis target using the evaluation results, and derives 28 secondary attributes using the primary analysis results, 58 dimensions of the analysis can be characterized.

The secondary attribute may be obtained by linearly combining the primary attributes.

The attribute analysis server may further include a time adder configured to measure clarity of the attribute connection by using the response time information of the user, and an object to extract and provide another analysis target having attributes similar to any analysis target in the attribute database. And a compensator configured to compensate the user of the panel terminal.

The compensation unit may compare the attribute score with other respondents and perform higher rewards as the scores are similar.

The word providing unit provides an attribute evaluation window for the attribute evaluation of the attribute analysis word to the panel terminal, and provides a test attribute analysis word for testing the reliability of the panel evaluation, thereby evaluating the attribute of the panel with respect to the provided test attribute analysis word. If the result is different from the pre-stored test answer, the evaluation may be terminated by exiting the attribute evaluation window, and points may be deducted to the panel that has incorrectly evaluated the test answer.

The similarity calculator may calculate one or more of similarity between attributes of two keywords and integrated similarity between two keywords.

The similarity for each attribute may be calculated using similarity of attribute distribution between two keywords.

When the two keywords a similarity are calculated, a is the first keyword, b is the second keyword, and S _{<a, b>} is the similarity for each attribute, TS _{<a, b>} is the integrated similarity,

Can be calculated by

In another aspect, the present invention provides an analysis object providing step of the attribute analysis server provides the evaluator with the analysis object and the attribute analysis word for the attribute analysis to the panel terminal, and providing two keywords to the panel terminal as the analysis target for the attribute analysis; An evaluation result receiving step in which the attribution analysis server receives an evaluation result and input time information of an attribution analysis word for an analysis target by an evaluator from the panel terminal; An attribute analysis step of analyzing, by the attribute analysis server, primary and secondary attributes of an analysis target using the evaluation result; An attribute information storage step of storing, by the attribute analysis server, attribute information about the analyzed object, and additionally storing the attribute information on the previously stored information; An attribute information providing step of providing, by the attribute analysis server, attribute information on the analysis target to a client; And a similarity calculation step of the attribute analysis server calculating a similarity between two keywords received from the client terminal using the attribute information.

The analyzing of the primary and secondary attributes may include analyzing 54 primary attributes by using the evaluation result and analyzing 45 secondary attributes by using the primary attribute analysis result. .

The similarity may include at least one of similarity between attributes of two keywords and integrated similarity between two keywords.

The recording medium of the present invention may be a recording medium storing a computer program for a method of analyzing an attribute of an analysis subject, wherein the computer program is stored by an attribute analyzing method.

The present invention can derive a marketing strategy by providing a solution that allows the participant to evaluate the degree of an attribute possessed by the target, thereby realizing the attribute of the target.

Attribute scores and response times can also measure the clarity of linking objects to attributes.

In addition, it is possible to derive insight into media advertisements by extracting other objects (for example, entertainers or animals) similar to the information on the attributes of the objects.

In addition, because the standardized questionnaire is prepared in advance, databased and stored, longitudinal comparison is possible.

In addition, through the similarity of the two keywords can obtain the information desired by the client.

1 is a block diagram of an attribute analysis system of an analysis target according to an embodiment of the present invention.

2 is a block diagram of a property analysis server according to an embodiment of the present invention.

3 is a screen provided for attribute analysis of the object of the attribute analysis server according to an embodiment of the present invention.

4 is a view showing a property evaluation window according to an embodiment of the present invention.

5 is a view showing a property evaluation window for the reliability test according to an embodiment of the present invention.

6 is a table comparing each calculated similarity between two keywords according to an embodiment of the present invention.

7 is a flowchart illustrating a method of analyzing an attribute of an analysis target, according to an exemplary embodiment.

Hereinafter, with reference to the accompanying drawings will be described an embodiment of the present invention in more detail. However, the present invention is not limited to the embodiments disclosed below, but will be implemented in various forms, and only the embodiments are intended to complete the disclosure of the present invention, and to those skilled in the art to fully understand the scope of the invention. It is provided to inform you. Like numbers refer to like elements in the figures.

It is intended to clarify that the division of the components in the present specification is only divided by the main function of each component. That is, two or more components to be described below may be combined into one component, or one component may be provided divided into two or more for each function. Each of the components to be described below may additionally perform some or all of the functions of other components in addition to the main functions of the components, and some of the main functions of each of the components are different. Of course, it may be carried out exclusively by. Therefore, the presence or absence of each component described through this specification will be to be functionally interpreted. For this reason, it is evident that the configuration of the system and method for attribute analysis of words or images of the present invention and the components of the recording medium storing the computer program for the method may be different to the extent that the object of the present invention can be achieved. Reveal.

1 is a block diagram of an attribute analysis system of an analysis target according to an embodiment of the present invention. 2 is a block diagram of an attribute analysis server according to an exemplary embodiment. 3 is a screen provided for attribute analysis of an analysis target of an attribute analysis server according to an exemplary embodiment.

As shown in FIGS. 1 and 2, the attribute analysis system according to the present embodiment includes a plurality of panel terminals 100 evaluating attributes of an analysis target provided in a word or image form through a network, and the analysis target for the analysis target. Providing an attribute analysis word for the attribute evaluation, the attribute analysis server 200 for analyzing the attribute for the analysis target according to the evaluation result of the panel, and the client terminal 300 for checking the attribute evaluation result for the analyzed analysis target It includes.

The panel terminal 100 provides a user with an attribute analysis word pair for an analysis target provided through a wired or wireless network. Then, the attribute evaluation result for the analysis target evaluated by the user is provided to the attribute analysis server 200.

Various types of terminals capable of wired and wireless communication may be used as the panel terminal 100. For example, a PC, a notebook, a mobile terminal (smartphone, PDA, etc.) may be used.

In the present embodiment, the attribute analysis server 200 includes a client terminal 300 for requesting an attribute analysis for an analysis target and receiving the result.

Like the panel terminal 100, the client terminal 1000 may be a PC, a notebook computer, a mobile terminal (smartphone, PDA, etc.).

The client terminal 300 receives and displays an attribute evaluation result for the analyzed analysis target from the attribute analysis server.

In particular, the client terminal 300 of the present invention receives two keywords from the client to receive the similarity from the attribute analysis server and provides the attribute analysis server. For example, if a first client, a cosmetics company, wants to use his cosmetics company and talent Kim Tae-hee as a cosmetic model, in order to grasp the similarity between the talented Kim Tae-hee (first keyword) and the first client company name (second keyword) , Talent Kim Tae-hee (first keyword) may provide the first client company name (second keyword) to the attribute analysis server 200 as a keyword for determining similarity.

As shown in FIG. 2, the attribute analysis server 200 may include a word storage unit 210 that stores attribute analysis word pairs and secondary attribute element information, and a word system that provides analysis object and attribute analysis word pair information to a panel terminal. Study 220, an evaluation result receiving unit 230 that receives and stores the evaluation (decision) result and response time information of the user through the panel terminal, and an attribute for analyzing the attributes and emotions of the analysis target using the evaluation results An analysis unit 240, a time adding unit 250 for measuring clarity of attribute connection using the response time information of the user, an attribute database 260 for storing attribute results for the analysis targets, and An object comparator 270 that extracts and provides another analysis object having properties similar to one analysis object in the property database 260, a compensator 280 that compensates a user of the panel terminal, and two A similarity calculating unit 295 for calculating similarity between the keywords of and a request processing unit 290 for processing the request of the client terminal 300 to provide the attribute evaluation result or similarity to the client terminal 300.

The attribute analysis server 200 according to the present embodiment provides an attribute analysis word for an analysis target to a user of the panel terminal 100 connected through a network, evaluates the attribute for the analysis target, and uses the evaluation result to market. Make sure to draw a strategy.

In this case, the analysis target may include a product (product), a brand, a multimedia work (movie, etc.), a literary work, a person, an animal, a design. Of course, in addition to a variety of tangible and intangible objects that can be displayed as words or video images may be selected as an analysis target.

The word storage unit 210 stores an attribute analysis word for evaluating and analyzing an attribute of an analysis target. Preferably, the words stored by the present embodiment are effectively words having about 30 attributes each.

Preferably, it is effective to have 30 attribute analysis words in this embodiment. Here, thirty attribute analysis words are shown in Table 1 below.

Good / bad	Fast / Slow	Big / small	Strong / weak	Familiar / Unfamiliar	Trust / Distrust
Advanced / Low	Beautiful / ugly	Interesting / boring	Carefree / fearful	Glad / sad	Calm / Furious
Masculine / feminine	Emotional / rational	White / Black	Rough / Soft	Light / dark	Loud / quiet
Hot / Cold	Sweet / Write	Ambiguous / clear	Smart / dumb	Latest	High / low
Rich / lack	Korean / Exotic	Easy / difficult	Plain / unusual	Near / far	Short term / long term

The word provider 220 sequentially provides the analysis target and the attribute analysis word to the panel terminal 100. As shown in FIG. 3, the analysis target may be a word representing the name of the target or an image photograph representing the analysis target. Of course, the present invention is not limited thereto, and everything that can be expressed can be expressed.

In addition, the word provider 220 provides an attribute analysis word, and also provides evaluation score information on the word to obtain an evaluation value for the attribute analysis word. At this time, the evaluation score information may be a value expressed as a percentage (%) of the feeling of the evaluator (user) for the two words (ie, word pairs) provided. Of course, the present invention is not limited thereto and may be expressed as a score capable of scoring 1 to 100 points. In addition, it may be possible to represent up to -100 to 100.

Also, the word providing unit may provide an attribute evaluation window for attribute evaluation of an attribute analysis word to the panel terminal, and provide a test attribute analysis word for testing the reliability of the panel evaluation, thereby providing the attribute of the panel with respect to the provided test attribute analysis word. If the evaluation result is different from the pre-stored test answer, the evaluation is terminated by exiting the attribute evaluation window, and points are deducted to the panel that made the wrong answer with the test answer.

An example of an attribute evaluation window is shown in FIG. The analysis target here is Kia Motors, and the attribute analysis word is cold / hot.

In addition, the word provider provides a test attribute analysis word, which is a question that can objectize the panel's evaluation, in order to improve data reliability of the panels. The test answer, which is the correct answer of the test attribute analysis word, is set in advance and stored. If the panel answers another answer without the correct test answer, the test is terminated by closing the attribute evaluation window. In addition, points will be deducted from panels with incorrect test results.

For example, as shown in FIG. 5, when the test property analysis word “furnace” is provided and the panel evaluates the property as “cold”, the panel considers the answer unfaithfully and closes the property evaluation window. Ends and deducts the points on the panel.

The evaluation result receiving unit 230 receives and stores the evaluation result for the analysis target evaluated by the evaluator. At this time, 71% closer to the concept that the property evaluation result of the evaluator for the spider is more beautiful than that shown in FIG. In this case, the provided value is 29% information that is beautiful, and 71% information that is ugly may be provided. Of course, the information provided at this time may be represented by various types of values. For example, beautiful can be -100 and ugly can be +71.

Such information is received by the evaluation result receiver 230 and stored in a separate storage means. Of course, various types of evaluation results may be provided according to evaluation score information provided by the word provider 220 mentioned above. In this case, 54 evaluation results are provided and stored. Of course, the 54 evaluation results may be reduced or increased according to the customer's request by default. At this time, the expansion of the evaluation result can be secured by providing additional attribute analysis words.

Here, the evaluation result receiver 230 is also provided with response time information on the one evaluation screen. For example, after FIG. 3 is provided, time information required to complete the evaluation of the analysis target (spider) is provided. In this case, the time information is preferably provided with time information in milliseconds.

The attribute analyzer 240 performs a primary evaluation on 30 dimensions according to the evaluation result, and derives about 28 secondary attributes by using the result of the primary evaluation. Through this, a total of 58 dimensions are analyzed. 28 secondary attributes are shown in Table 2 below.

Secondary attribute	Element 1	Element2	Element3	Element4	Element5	Element6
Loyalty	trust	trust	interesting	good	-	-
Reliability	trust	trust	good	-	-	-
Upper respiratory tract	close	Normal	-	-	-	-
Awareness	Familiar	Familiar	close	-	-	-
Degree of discrimination	unique	-	-	-	-	-
Refinement	brand-new	Advanced	beautiful	-	-	-
S-W	brand-new	strong	trust	smart	Familiar	-
O-T	Long term	scarce	ambiguous	-	-	-
High-Low	interesting	difficult	-	-	-	-
Rational-emotional	Rational	Rational	ambiguous	-	-	-
Quality Competitiveness	brand-new	brand-new	strong	Long term	good	-
Brand Competitiveness	unique	Familiar	trust	-	-	-
Service Competitiveness	Advanced	trust	abundant	calm	-	-
Model Competitiveness	Advanced	Familiar	trust	Interesting	good	-
Emotional Competitiveness	Masculine	Emotional	Advanced	Happy	beautiful	-
Global Competitiveness	Exotic	beautiful	-	-	-	-
Sustainable Competitiveness	strong	trust	ambiguous	ambiguous	-	-
Potential Competitiveness	unfamiliar	ambiguous	scarce	good	-	-
lead	strong	fast	-	-	-	-
society	fast	abundant	hot	soft	Familiar	-
stability	slow	trust	obvious	Long term	-	-
discretion	Rational	slow	low	-	-	-
Impulsivity	high	strong	fast	-	-	-
Rigidity	fast	ambiguous	hot	soft	Familiar	-
Controllability	slow	trust	obvious	Long term	-	-
Responsive	Rational	slow	smart	Long term	-	-
Positivity	Happy	tranquil	-	-	-	-
Arousal	Happy	tranquil	-	-	-	-

The attribute analyzer 240 analyzes the secondary attributes of the analysis target by combining the evaluation results stored through the evaluation result receiver 230.

It is possible to score new attributes through attribute scores for the analysis target. In other words, it is possible to score new attributes such as global competitiveness, potential competitiveness, and SWOT analysis. For example, the attribute of loyalty is not evaluated, but a new attribute can be evaluated by synthesizing the attribute score such as newest, superior or reliable.

In the case of secondary attributes, it is derived using the sum of the attribute scores of the elements / number of elements. The response time is also calculated equally by changing the attribute score only to the response time, and overlapping attributes are weighted.

The attribute analyzer 240 may evaluate scores for 99 attributes and provide the score results to the client terminal. In addition, the attribute analyzer 240 may store evaluation results by a plurality of evaluators (users) and calculate and store an average value. In this case, the analysis result may be calculated according to the user's personal information. The attribute evaluation result calculated in this way is stored in the attribute database 260.

That is, the secondary attribute is calculated through the combination of the primary attributes. For example, the attribute 'Marketability' is not an attribute evaluated by a survey participant, but an attribute such as 'latest', 'relevant', 'excellent' and 'trusted' corresponding to an attribute evaluated by a survey participant. It is further calculated through the score for. The primary attribute is no longer broken by the consumer intuitively, and the secondary attribute is an attribute determined by combining several items of the primary attribute.

In addition, the attribute evaluated for each analysis target may be converted into a numerical value and provided as a graph. In this way, not only the attributes of words but also the standard of judgment or value can be measured.

In addition, an attribute of each analysis target may be represented as an emotion image, and this may be represented on a separate scale. Of course, the present invention is not limited thereto, and various types of images may be represented by a 2D chart of response time and attribute scores for all attributes of the analysis target.

In this embodiment, through the above-described investigation, even the in-depth psychology that the evaluators may not perceive consciously can be grasped.

It is possible to grasp the in-depth psychology of consumers through the result of attribute score and response time. And, in the case of this system, a quick judgment is induced. This provides an intuitive, in-depth response. Of course, it induces quick judgment, but compares attribute scores with other respondents so that they do not answer silently, and gives high points when scores are similar. The reaction time is then determined by the clarity of the connection between the object and the property. This is a difficult part of autonomous control, so it is possible to understand in-depth psychology. This can increase the credibility of the survey because the question is simple and the same form is repeated and the intention of the question cannot be predicted.

In the present embodiment, the time adder 250 may be provided to measure the clarity of the connection between the analysis target and the property by using the response time information of the user.

In the present embodiment, the attribute score means the strength of the analysis object having a property, and the response time (response time) means the clarity of the connection between the analysis object and the property.

In other words, when a survey participant evaluated 'feminine-male' for the analysis target of 'Kim Il-sung', he gave 90 points to 'male', and if he thought more than 3 seconds to determine this, Although it has a high attribute score for this 'male', it is considered that the clarity of the connection between the object and the attribute is low because it takes a long time in the process of assigning this score. However, in the process of evaluating 'good-evil' for 'Kim Il-sung', if you gave 70 points to 'evil' and thought it to be less than 1 second to judge this, this means that Kim Jong-il's attribute of 'evil' It has a lower attribute score than 'male', but the judgment on this does not require much concern and has a very high connection clarity.

And, when the response time is within 0.1 second to 1 second, it is possible to recognize that the attribute recognized by the evaluator is firm and to give a weight of 6 to 10% to the evaluation score. In addition, in the case of 1 second to 3 seconds, a weight of about 1 to 5% is given. In addition, in the case of 3 seconds or more, the weight may not be given.

Of course, it is desirable to avoid integrating with attribute scores, since response time is another dimension of the clarity of the link between analyte and attribute. Of course, when the response time is high, the participants have a lot of worries about the evaluation, and therefore, the clarity of the link between the analysis object and the property is low, and the modification of the evaluation of the analysis object and the property may be possible in marketing.

In this way, the weighted attribute results are used to convert and store the new attribute evaluation, graph form transformation, and emotional image. At this time, the weighted attribute analysis result is stored in the attribute database 260.

Alternatively, a newly analyzed result value may be added to a value of the previously stored attribute database 260. This enables multiple cumulative investigations and analysis of a single analysis target, which has the advantage of accurately identifying attribute images of the analysis target.

Of course, this embodiment is not limited to this, and may be given a negative weight according to the response time. For example, if the response time is more than 30 seconds, the user may determine that the attribute of the analysis target is not accurately recognized and may give a weight of -1 to -5%.

The attribute analysis server 200 of the present embodiment includes a target comparator 270. Through this, the client terminal may provide an entertainer or a plant or animal having an attribute similar to that of the analysis target requested by the client terminal. The target comparator 270 selects an analysis target having a similar property analysis result or an emotional image value for the analysis target. Preferably, a plurality of analysis objects are analyzed to select similar analysis objects within 10% of attribute analysis result values of the request analysis objects.

That is, a similar analysis target may be extracted by comparing the degree of the attribute of the analysis target with the attribute information stored in the attribute database 260.

In addition, in the present embodiment, the compensation unit 280 may be provided to compensate for the evaluation result of the evaluator connected to the panel terminal 100, thereby improving participation of the evaluator.

The compensator 280 compares the evaluator's attribute evaluation result with the previously evaluated attribute evaluation result and provides a higher reward if similar.

The reward calculation method is as follows.

If the average of the A-B scores for word A and property B is N,

'1>' N is upper, middle, lower,

If N and input score are in the range of 0 ~ 48, 49 ~ 52, 53 ~ 100, match: 2 points

Difference from '2>' N

If the absolute value of (Input-N) is less than 5: 2

If the absolute value of (Input-N) is less than 20: 1

If the absolute value of (Input-N) is greater than 20: 0

Score obtained is '1> score' + '2> score'

The scorecard according to this is as follows.

4 points: EXC, 3 points: GOOD, 2 points: NORMAL, 1 point: BAD, 0 points: POOR

Compensation may be performed through the reward points as described above. In this case, the reward may be cash and may be points or products corresponding to cash. In this case, it is preferable that the third reward is a basic reward, the second reward is larger than the third reward, and the first reward is larger than the second reward. In this case, the second reward may be given 5 to 20% more reward than the third reward, and the first reward may be given 5 to 20% more reward than the second reward.

Accordingly, the present invention may derive insights on media advertisements by extracting other analysis targets (for example, celebrities or animals) that are similar to the attribute information of the analysis target. In addition, because the standardized questionnaire is prepared in advance, databased and stored, longitudinal comparison is possible.

The similarity calculator includes a similarity calculator that calculates similarity between two keywords. The similarity calculator calculates one or more of the similarity for each attribute between the two keywords and the integrated similarity between the two keywords.

The similarity calculation for each attribute may be calculated by the similarity calculator using similarity of attribute distribution between two keywords. An example of an equation calculated using the similarity of attribute distribution between two keywords will be described below.

i = keyword index

j = attribute index

N _ij = total number of responses for attribute j of keyword i

X _ij = distribution of scores for attribute j of keyword i

x _ij is defined as 10 (k-1) <X _ij <10k, where k = 1,2,3, ....., 10

Defined as

Where a is the first keyword and b is the second keyword, the similarity S < _{a, b>, j} of the attribute j for the keyword <a, b>

Similarity S < _{a, b>, j} has a value between 0 and 1, the closer to 0, the lower the similarity, and the closer to 1, the higher the similarity.

For reference, it is described in the following [Table 3] as an example table of the similarity using the similarity of the attribute distribution between the two keywords by the above formula. Table 3 below shows a distribution of response frequencies of two keywords by inspection intervals to calculate a similarity index of two keywords for the attribute 'hot'.

division	Keyword a: joinability	Keyword b: Kia Motors
0 <x≤10	5	3
10 <x≤20	7	5
20 <x≤30	13	2
30 <x≤40	17	7
40 <x≤50	23	10
50 <x≤60	15	27
60 <x≤70	5	13
70 <x≤80	6	16
80 <x≤90	3	13
90 <x≤100	2	9
system	96	105

Table 4 below is a table showing the response frequency as a ratio in the table of Table 3 above.

division	Keyword a: fan	Keyword b: air conditioner	∥ difference ∥
0 <x≤10	0.052	0.029	0.024
10 <x≤20	0.073	0.048	0.025
20 <x≤30	0.135	0.019	0.116
30 <x≤40	0.177	0.067	0.110
40 <x≤50	0.240	0.095	0.144
50 <x≤60	0.156	0.257	0.101
60 <x≤70	0.052	0.124	0.072
70 <x≤80	0.063	0.152	0.090
80 <x≤90	0.031	0.124	0.093
90 <x≤100	0.021	0.086	0.065
system	1.000	1.000	0.840

The similarity index of the attribute 'hot' is calculated as "1- (0.840) /2=0.58. If there are 30 attributes, the similarity as described above may be calculated for each attribute.

On the other hand, when two keywords for which similarity are calculated are a first keyword and b is a second keyword, and S _{<a, b>} is similarity for each attribute, the integrated similarity TS _{<a, b>} is,

to be.

For example, assuming that the number of evaluation attributes is 30, the integrated similarity TS _{< a, b >} calculates the average by dividing the similarity of the individual attributes by 30 after adding all the similarities of the individual attributes. Just do it.

The request processor 290 provides the client terminal with an attribute evaluation result and similarity. In particular, the similarity is provided to the client terminal so that the client can know the similarity between the two keywords.

For example, the similarity with two keywords of cosmetic companies and a female actor can be seen as shown in FIG. Therefore, emotional similarity between keywords is calculated by objective similarity index so that emotional similarity can be seen at a glance. This can be used for various purposes, such as advertising model, advertising copy, product name selection suitable for the product.

Hereinafter, a method of analyzing an attribute of an analysis target according to the present embodiment will be described.

7 is a flowchart illustrating a method of analyzing an attribute of an analysis target, according to an exemplary embodiment. As shown in FIG. 7, in the method for analyzing an attribute of an analysis target according to the present embodiment, a client first provides an attribute analysis server that wants to perform an attribute analysis through its terminal to the attribute analysis server. Subsequently, the attribute analysis server performs attribute analysis on the analysis target.

The evaluator accesses the attribution analysis server through his panel terminal. In this case, a separate login step may be performed.

Subsequently, the attribute analysis server has an analysis target for analyzing and providing an analysis target for providing the attribute analysis word to the panel terminal (S110).

In this case, the attribute analysis word includes about 54 attribute words.

Subsequently, the attribute analysis server receives the evaluation result information of the word corresponding to the attribute that the evaluator feels and recognizes about the analysis target provided through the panel terminal, and input time information to obtain the evaluation result from the panel terminal. It has a receiving step (S120).

Subsequently, the attribute analysis server stores the provided evaluation result information as the primary attribute, and calculates the secondary attribute using the primary attribute to analyze the attribute for the analysis target (S130). That is, first, the primary attribute is analyzed using the input result, and the secondary attribute is analyzed using the primary attribute.

Subsequently, the attribute analysis server has an attribute information storing step of storing attribute information on the analyzed object. In this case, if there is previously stored information, the result information is additionally stored in the corresponding information. Here, the average value of the above information is preferably stored.

Subsequently, rewards are made for the evaluator's evaluation, and the reward is set higher as the evaluator's evaluation result is closer to the attribute mean value for the analysis target (S140). Subsequently, there is an attribute information providing step of providing attribute average information on the analysis target to the client (S150).

Subsequently, it is determined whether there is a request for entertainer or flora and fauna information from the client with attributes similar to those of the analysis object. If there is a request as a result of the determination, a list of entertainers and plants and plants similar to the attributes of the analysis target is provided (S160).

In addition, the attribute analysis server may further have a similarity calculating step S170 of calculating similarity between two keywords received from the client terminal using attribute information.

The client terminal receives two keywords to be provided with similarity from the client and provides the same to the attribute analysis server, and the attribute analysis server calculates the similarity between the two keywords using the attribute information and provides the same to the client terminal (S170). ).

Such similarity includes at least one of similarity between attributes of two keywords and integrated similarity between two keywords, and examples of calculating similarity and integrated similarity of attributes are omitted because they have been described above.

Although the technical spirit of the present invention described above has been described in detail in a preferred embodiment, it should be noted that the above-described embodiment is for the purpose of description and not of limitation. In addition, the present invention will be understood by those skilled in the art that various embodiments are possible within the scope of the technical idea of the present invention.

Claims (13)

1. A plurality of panel terminals evaluating attributes of an analysis target provided in a word or image form;

   An attribute analysis server for providing an attribute analysis word for attribute evaluation of the analysis target and analyzing the attribute for the analysis target according to the evaluation result of the panel; And

   And a client terminal for receiving an attribute evaluation result of the analyzed analysis object from the attribute analysis server, receiving two keywords from the client, and providing the two keywords to be provided with similarity from the attribute analysis server to the attribute analysis server. Attribute Analysis Server,

   A word storage unit for storing attribute analysis word pairs, a word providing unit for providing analysis objects and attribute analysis word pair information to the panel terminal, an evaluation result receiving unit for receiving and storing the evaluation result and response time information of the user through the panel terminal; A property analysis unit for analyzing the attributes and emotions of the analysis target using the evaluation results, an attribute database storing the attribute results for the analysis targets, a similarity calculator for calculating the similarity between two keywords, And a request processing unit for providing the client terminal with an attribute evaluation result and a similarity degree.
2. The method of claim 1,

   The word storage unit stores 30 attribute analysis word pairs,

   The attribute analyzer analyzes the primary attributes of the object to be analyzed using the evaluation result and derives 28 secondary attributes by using the primary analysis results.

   Attribute analysis system of the analysis object characterized in that the analysis of 58 dimensions.
3. The method of claim 2,

   The secondary attribute is the attribute analysis system of the analysis object, characterized in that derived by using a value obtained by dividing the sum of the attribute score of the elements by the number of elements.
4. The method of claim 1,

   The attribute analysis server may further include a time adder configured to measure clarity of the attribute connection by using the response time information of the user, and an object to extract and provide another analysis target having attributes similar to any analysis target in the attribute database. And a comparator configured to compensate the user of the panel terminal.
5. The method of claim 4, wherein

   And the compensation unit compares the attribute score with other respondents and performs a higher reward as the score is similar.
6. The method of claim 1,

   The word providing unit provides an attribute evaluation window for the attribute evaluation of the attribute analysis word to the panel terminal, and provides a test attribute analysis word for testing the reliability of the panel evaluation, thereby evaluating the attribute of the panel with respect to the provided test attribute analysis word. If the result is different from the pre-stored test answer, the property evaluation window is terminated, the evaluation is terminated, and the analysis object attribution analysis system to deduct the point to the panel that was incorrectly evaluated.
7. The method of claim 1,

   The similarity calculator, the attribute analysis system of the analysis target to calculate at least one of the similarity for each attribute between the two keywords, the combined similarity between the two keywords.
8. The method of claim 7, wherein

   The similarity for each attribute is calculated using the similarity of attribute distribution between two keywords.
9. The method of claim 7, wherein

   When two keywords, similarity, are calculated, a is the first keyword, b is the second keyword, and S _{<a, b>} is the similarity for each attribute, TS _{< a, b >} is the integrated similarity,

   

   Attribute analysis system of the analysis target calculated by the.
10. An analysis object providing step of providing, by the attribute analysis server, an analysis object for attribute analysis and an attribute analysis word to the panel terminal to the evaluator, and providing two keywords to the panel terminal as analysis object for the attribute analysis;

    An evaluation result receiving step in which the attribution analysis server receives an evaluation result and input time information of an attribution analysis word for an analysis target by an evaluator from the panel terminal;

    An attribute analysis step of analyzing, by the attribute analysis server, primary and secondary attributes of an analysis target using the evaluation result;

    An attribute information storage step of storing, by the attribute analysis server, attribute information about the analyzed object, and additionally storing the attribute information on the previously stored information;

    An attribute information providing step of providing, by the attribute analysis server, attribute information on the analysis target to a client; And

    A similarity calculation step of calculating, by the attribute analysis server, a similarity between two keywords received from the client terminal using the attribute information;

    Property analysis method of the analysis target comprising a.
11. The method of claim 10, wherein analyzing the primary and secondary attributes,

    30 primary attributes are analyzed using the evaluation results, and 28 secondary attributes are analyzed using the primary attribute analysis results.
12. The method of claim 10,

    The similarity is an attribute analysis method of the analysis target including at least one of the similarity for each attribute between the two keywords, integrated similarity between the two keywords.
13. A computer program according to any one of claims 10 to 12, wherein the computer program is stored according to the property analysis method of the analysis object.

Images