KR101680195B1

KR101680195B1 - Method for analyzing relationship between human personality and favored location

Info

Publication number: KR101680195B1
Application number: KR1020150089257A
Authority: KR
Inventors: 송하윤; 김승연; 이은별
Original assignee: 홍익대학교 산학협력단
Priority date: 2015-06-23
Filing date: 2015-06-23
Publication date: 2016-12-06

Abstract

The method comprising the steps of: obtaining, for each of a plurality of persons, at least one toughness value indicating toughness and a staying frequency value indicating a degree of staying in a specific place; and determining a staying frequency value for the specific place as a dependent variable, And calculating a regression coefficient for each of the toughness values by performing a regression analysis using the toughness value as an independent variable and discloses a method for analyzing the correlation between the toughness and the preferred place.

Description

A method for analyzing the relationship between personality and preference place (human personality and favored location)

TECHNICAL FIELD The present invention relates to computing technology and relates to techniques for analyzing the impact of humanity on a preferred location.

Personality tests are carried out for various reasons, such as to find out the type of personality of oneself, to make learning more efficient, or for employment. In order to know these personality types, it is common to make surveys such as MBTI in order to perform personality tests in general. However, this personality test can be a bit of a hassle.

When a personality test is needed, a team may be required to conduct an inspection to refer to the personality of the other person in order to know the other person in the same community. In this case, there may be people who do not want to be tested for personality.

On the other hand, due to the proliferation of smartphones, users can easily obtain location data and location data of smartphone users using various applications. In other words, the functions of many smartphones, including GPS, GLONASS and indoor location acquisition systems, can be used for location acquisition. Smartphone applications can also be used to easily check in wherever a user goes.

The present invention provides a technique for quantifying the influence of toughness on a preferred place.

According to one aspect of the present invention, there is provided a method for analyzing a correlation between a toughness and a preferred place, the method comprising: acquiring, for each of a plurality of persons, one or more toughness values indicating toughness and a staying frequency value indicating a degree of staying in a specific place step; And a step of calculating a regression coefficient for each toughness value by performing a regression analysis using a value of a staying frequency for the specific place as a dependent variable and the one or more toughness values as independent variables.

In this case, the toughness value is a toughness factor according to a big five factor, and the property value may be normalized to have a value between 0 and 1.

At this time, the staying frequency value may be a value between 0 and 1, which is presented as a ratio of a time or a frequency of staying at the specific place among the time or frequency of staying in a plurality of places where each of the plurality of persons remains.

According to another aspect of the present invention, there is provided a method for analyzing a correlation between toughness and a preferred place, the method comprising the steps of: obtaining, for each of a plurality of persons, one or more toughness values indicating toughness and a staying frequency value indicating a degree of staying in a specific place step; Providing the toughness value and the staying frequency value obtained for the plurality of persons to the input layer and the output layer of the artificial neural network model, respectively, to learn the artificial neural network model; Sequentially inputting a predetermined set of toughness values to an input layer of the artificial neural network model and successively obtaining a set of staying frequency values from an output layer of the artificial neural network model; And a set of retention frequency values as dependent variables and a set of toughness values corresponding to each set of retention frequency values as independent variables, And calculating a regression coefficient.

The predetermined set of toughness values may be set to have a plurality of values each having five toughness factor values different from each other depending on a big five factor and then composed of combinations of values that the five factors may have .

The computing device provided according to another aspect of the present invention may be configured to perform the steps according to the method of analyzing the correlation between the toughness and the preferred place.

A computer readable medium provided according to another aspect of the present invention may be a program for causing the computing device to perform steps according to a correlation analysis method between the toughness and the preferred place.

According to the present invention, it is possible to provide a technique for quantifying the influence of toughness on a preferred place.

FIG. 1 is a table showing personality data of an experimenter using five personality types (BFF) necessary for implementing an embodiment of the present invention.
2 shows a part of the place data of the experimenter 1 as a table.
3 shows another part of the place data of the experimenter 1 as a table.
FIG. 4 shows regression analysis data of an experimenter according to an embodiment of the present invention.
Fig. 5 is a table showing the degrees of relevance determined by the decision coefficients.
FIG. 6 is a table showing weight values of positional data presented to have a weight value between 0 and 1 over 24 hours of the day.
FIGS. 7 to 13 show tables showing CoD values according to a regression analysis result.
FIG. 14 is a summary of main combinations that can explain the relationship with the preferred position among the combinations found in FIGS. 7 to 13.
FIG. 15 is a flowchart illustrating a method of analyzing a correlation between a toughness and a preferred place according to an embodiment of the present invention.
FIG. 16 is a flowchart illustrating a method of analyzing a correlation between a toughness and a preferred place according to another embodiment of the present invention.

Hereinafter, embodiments of the present invention will be described with reference to the accompanying drawings. However, the present invention is not limited to the embodiments described herein, but may be implemented in various other forms. The terminology used herein is for the purpose of understanding the embodiments and is not intended to limit the scope of the present invention. Also, the singular forms as used below include plural forms unless the phrases expressly have the opposite meaning.

Time, and personality of the person. There are many ways to express human personality. In the present invention, BFF (Big Five Factors) is used, and the study on this can be found in the following paper.

[Paper]

PT Costa and RR McCrae, "" Four ways five factors are basic, "" Personality and individual differences, vol. 13, no. 6, 1992, pp. 653-665.

&Lt; Example 1 >

FIG. 1 is a table showing personality data of an experimenter using five personality types (BFF) according to an embodiment of the present invention.

The columns in the table represent the five types of BFFs, and the rows are the numerical values of the characteristics of each experimenter according to the personality type of each column. At this time, the first row of each experimenter represents the BFF value in the linear distribution, and the second row represents the BFF value in the normal distribution. In the present invention, it is desired to use the BFF value in the normal distribution.

For example, the first row may have a value between 0 and 5, and the second row may be modified to have a value between 0 and 1 by modifying the value between 0 and 5 by a normal division. This makes it possible to convert the BFF value into a value that can be easily applied to the regression analysis. The values thus provided can be used for independent variables in regression analysis.

Personality data can be expressed in five types using 'BFF' as 'Openness', 'Integrity', 'Enthusiastic', 'Synonymy', and 'Nervousness'. At this time, the above-mentioned five types can be expressed numerically. The advantage of digitizing personality data is that you can easily apply the data to the algorithm. Therefore, data can be easily applied to regression analysis.

For example, according to the volunteer personality data, the experimenter with the highest openness is the experimenter 4, the experimenter with the highest sincerity (C) is the experimenter 5, and the person with the highest enthusiasm (E) .

For example, Experiment 3, which has the highest openness (O), shows that creativity and intellectual curiosity of art have higher human tendencies than other experimenters. Experient 6, who has the highest enthusiasm (E), is energetic and has a tendency to be human with positive thoughts.

The regression analysis can be a useful tool for verifying the relationship between BFF and location data. The BFF data in the normal distribution can be input and used as independent variables in the regression analysis.

2 shows a part of the place data of the experimenter 1 as a table.

The place data is obtained through the Swarm application, and the place data is collected using the Check_In function for each place visited by the experimenter. From the source data collected by the Swarm application, the number of visits, place names, and category names in each place can be extracted through parsing code. The place data indicates a part of the extracted data.

It can be seen that Experient 1 visited 73 times in 'Hongik University T-Dong' and the category name of 'Hongik University T-Dong' is 'University building'.

3 shows another part of the place data of the experimenter 1 as a table.

FIG. 3 is data obtained by extracting from the source data and combining the number of visits of the same category name, and shows the data of the experimenter 1, which is one of the 15 experimenters. That is, it shows the non-overlapping category name of Experiment 1 and the number of visits of the category name.

Experient 1 visited 'House' for '149', 'University Building' for '114', and 'Korean Restaurant' for '62'

In the embodiment of the present invention, 15 experimenters are taken as an example, so that a total of 14 place data corresponding to each experimenter can be obtained as shown in FIG. At this time, all of the 15 category names except the overlapping category names are combined.

According to an embodiment of the present invention, the category name may be linked with the industry code of the standard industry classification of the National Tax Service and listed in the order of the business type codes. Then, when the total number of 15 visits by the experimenter and each visit to the place, the place where the sum of the number of visits to the place is less than 10, is classified as etc and the data can be combined. In this way, in this embodiment, the number of category names of 15 experimenters is reduced to 57.

Additional steps can be performed to use the data combined in the category name for regression analysis. At this time, the following equation 1 can be used in the present invention.

[Equation 1]

Equation 1 relates to a value obtained by dividing the number of visits of a place of an experimenter by the total number of visits of the experimenter. In this case, loc means the place of the place data.

For example, Experient 1 has a total number of visits of 881, and Experient 1 has visited a university building 114 times. Therefore, the value for the university building of Experient 1 is 0.129398. In this way, you can organize data for each of the 15 experimenter sites to create place data that can be used for regression analysis. As a result of this procedure, the place data can be entered as a dependent variable used for regression analysis.

FIG. 4 shows regression analysis data of an experimenter according to an embodiment of the present invention. And Fig. 5 is a table showing the degrees of association determined by the decision coefficients.

FIG. 4 is a regression analysis of the personality data values of FIG. 1 as independent variables and 57 place data as dependent variables. That is, the coefficient of determination (CoD) according to the place where the regression analysis is performed is shown. Regression analysis can tell you which places you visit because of which elements of personality data.

Referring to FIG. 5, it can be judged that there is a meaning between personality and place when the determination coefficient value of each place is 0.4 or more as a result of the regression analysis. Therefore, FIG. 4 shows 28 locations determined to be significant among the 57 place data.

FIG. 4 is a graph showing the relationship between the coefficient of determination according to the place where the regression analysis is performed and the coefficients and characteristics of the personality openness (O), integrity (C), passion (E), synonymity (A) And how it affects the performance of the system.

At this time, if the absolute value of the personality factor is more than 0.01, it can be judged that it affects the place. When the absolute value of the coefficient is less than 0.01, it is divided into a case where the absolute value of the coefficient is not less than 0.01, a case where the absolute value of the coefficient is less than 0.1, a case where the absolute value is not less than 0.1, If it is less than 0.01, it is judged that the influence is hardly given. If it is 0.01 or more but less than 0.1, it is considered that it gives an artistic effect. For example, if it is openness, If the absolute value of the coefficient is 0.1 or more and less than 0.2, it can be judged that it has a great influence on the place, and in this case, it can be expressed as OO. If it is 0.2 or more, it can be judged that it is highly influential, and it can be expressed as OOO. The above description is shown in 'RELATION' of FIG. It can be easily understood that the sign of 'RELATION' in FIG. 4 can also be changed according to the sign of the coefficient.

In other embodiments, the degree of influence by the number of + or - may be indicated instead of indicating the degree of influence by the number of first letters. For example, if the absolute value of the personality factor coefficient is more than 0.1, it is judged to have a large effect on the place. If the personality factor coefficient value is positive, ++ is displayed. At this time, if the count value is a positive number, it has a positive effect, and if it is a negative number, it can be judged to have a negative influence.

For example, the personality coefficients of openness (O), integrity (C), enthusiasm (E), synonymity (A) and neuroticism (N) for university buildings are -0.23943, 0.13319, 0.01566 , 0.043851, 0.23505. Openness (O) can be expressed as -OOO since the absolute value is 0.2 or more and the original count value is negative. Integrity (C) can be expressed as CC because the absolute value is 0.1 or more and 0.2 or less and the original count value is positive. Both passivity and tangency are less than 0.01 and less than 0.1, and since the original coefficient values are positive, they can be expressed as E and A, respectively. It can be seen that the absolute value of the neurotic coefficient is 0.2 or more. Therefore, it can be displayed as NNN. Therefore, university buildings can be represented as -OOO, CC, E, A, NNN.

Figure 4 shows that the coefficient of determination of the university building is the highest at 0.89101 and that all personality factors such as openness (O), integrity (C), enthusiasm (E), synonymity (A) have. Among them, openness (O), integrity (C), and neuroticism (N) are more influential. This is centered on universities, but similar BFF patterns can be found in the company.

The coefficient of determination of the university laboratory is 0.58485. Although all five personality factors have an effect, openness (O) and synchronicity (A) have the greatest influence. At this time, it can be known that a person who is positive to openness and negative to harmony frequently visits the university lab.

At this time, where the column of influence is empty, the coefficient of the personality factor is less than 0.01, so the place and personality are related, but each personality factor is not related to one personality factor.

&Lt; Example 2 >

There may be patterns in BFF toughness data and preference position data collected from a plurality of people. This pattern can be analyzed using BPN. For this purpose, the BPN can be learned using the BFF toughness data and the preferred position data. The collected BFF toughness data may be used as learning data to be provided to an input layer of the BPN, and the collected preference position data may be provided as a learning data to the output layer of the BPN to complete the learning.

In regression analysis, only a subset of the five BFFs can be used as independent variables. For example, two combinations, three combinations, or four combinations of BFFs can be used as independent variables. By combining all of the above two combinations, three combinations, and four combinations, a total of 25 combinations can be obtained. These 25 combinations can form 25 subsets as independent variables. The CoD can then be analyzed for each subset. This can explain the ability of toughness to affect the preferred position.

Regression analysis between location data and toughness data can be performed with toughness data as independent variables. Toughness data are a subset of BFF combinations. The subset may consist of two or more of the five toughness data.

The toughness data subsets are independent variables in regression analysis using position data extracted by BPN. CoD is the key value of the regression analysis because CoD is the result factor value which shows the explanatory power of the independent variable on the dependent variable. In other words, the greater the CoD, the more relevant the tangible data is with the location data.

A BFF and a location data set provided by a plurality of persons, for example, five experts, can be used.

FIG. 6 is a table showing weight values of positional data presented to have a weight value between 0 and 1 over 24 hours of the day. The weight is a different value from the probability. In FIG. 6, only the weight values of each experimenter in a place called a school are shown by time.

The pattern is extracted from the collected raw data. The BPN is used for this because the BPN learns the input data and provides a patterned output. This patterning process creates a neural network. Through BPN, possible pairs of position and toughness can be patterned.

This can create an artificial position-toughness combination. At least 271 testers are required to obtain low-, medium-, and high-values of toughness factors. With BPN, however, 16,807 pairs of position-toughness patterns can be generated from the data of five experimans. Since there are 25 subsets, 25 regression analyzes are performed.

FIGS. 7 to 13 show tables showing CoD values according to a regression analysis result. Combinations of toughness factors are presented in each row and column.

For example, Figure 7 shows a table showing CoD values for two combinations of toughness parameters as independent variables. The CoD is 0.1023 for the combination of openness (O) and entanglement (E). The normal threshold level of CoD for describing natural phenomena is 0.3. The combination of openness (O) and enthusiasm (E) can not explain the location data of school because CoD value by combination of openness (O) and passivity (E) is 0.1023. In other words, the combination of integrity (C) and neuroticism (N), because of the highest value of 0.4731 according to the combination of integrity (C) and neuroticism (N) Best describes the relationship with the place.

Figures 8-12 show CoD values for three combinations of toughness parameters as independent variables. A combination of toughness parameters that best describe the relationship with the preferred position can be found using CoD in the same manner as described above. Figure 8 shows the CoD for three toughness factors including openness (O). Figure 9 shows the CoD for three toughness factors including integrity (C). Figure 10 shows the CoD for three toughness factors including passivity (E). Figure 11 shows the CoD for three toughness factors including (A). Figure 12 shows the CoD for three toughness factors including neurotic (N).

Figure 13 shows CoD values for four combinations of toughness parameters as independent variables. A combination of toughness parameters that best describe the relationship with the preferred position can be found using CoD in the same manner as described above.

FIG. 14 is a summary of main combinations that can explain the relationship with the preferred position among the combinations found in FIGS. 7 to 13. Figure 14 summarizes the toughness subsets with CoD values greater than 0.4. In Table 10, X indicates that five toughness factors are excluded. For example, OCEAX represents a subset except N.

&Lt; Example 3 >

A method for analyzing a correlation between toughness and a preferred place provided according to an embodiment of the present invention will be described with reference to FIG.

FIG. 15 is a flowchart illustrating a method of analyzing a correlation between a toughness and a preferred place according to an embodiment of the present invention.

The analysis method may include a step (S11) of acquiring, for each of the plurality of persons, at least one toughness value indicating toughness and a staying frequency value indicating a degree of staying at a specific place. For example, the plurality of persons may be five persons. The toughness may be, for example, five OCEAN robustness factors according to the Big Five model. And the one or more toughness values may mean one or more toughness factors of the OCEAN. The specific place may mean any one of a plurality of places where a plurality of people can stay, for example, a school building. The value of the staying frequency may be given by Equation (1).

Next, the analysis method calculates a regression coefficient for each toughness value by performing a regression analysis in which the value of the staying frequency for the specific place is used as a dependent variable and the toughness value is used as an independent variable Step S12 may be included. For example, if the one or more toughness values refer to two toughness factors out of the five toughness factors of OCEAN, then the two regression coefficients may be derived.

A method of analyzing the correlation between the toughness and the preferred place provided according to another embodiment of the present invention will be described with reference to FIG.

16 is a flowchart illustrating a method of analyzing a correlation between a toughness and a preferred place according to another embodiment of the present invention.

The analyzing method may include the step (S21) of acquiring, for each of the plurality of persons, one or more toughness values indicating toughness and a staying frequency value indicating a degree of staying in a specific place. The step S21 may be the same as the step S11.

And then (S22) learning the artificial neural network model by providing the toughness value and the retention frequency value obtained for the plurality of persons to the input layer and the output layer of the artificial neural network model, respectively. For example, when a plurality of persons are five persons, five sets of toughness values may be prepared.

(S23) sequentially inputting a predetermined set of toughness values to the input layer of the artificial neural network model, and successively obtaining a set of staying frequency values from the output layer of the artificial neural network model have. Where a predetermined set of toughness values may be, for example, 16,807 sets. This specific number is the number of combinations generated by arbitrarily designating 7 (= b) values for each toughness factor using 5 (= a) kinds of toughness factors according to the OCEAN toughness factor as the toughness values = a < b >). If the values of a and b are changed, the concrete values of 16,807 may be changed.

Then, a regression analysis is performed in which the settling frequency value of the set is used as a dependent variable and the set of toughness values corresponding to each set of the staying frequency values are set as independent variables, (S24) of calculating a regression coefficient with respect to the regression coefficient. Here, the toughness values used as the independent variables may be two, three, four, or five toughness factors selected from the five toughness factors according to the OCEAN.

It will be apparent to those skilled in the art that various modifications and variations can be made in the present invention without departing from the essential characteristics thereof. The contents of each claim in the claims may be combined with other claims without departing from the scope of the claims.

Claims

delete

A processing unit of a computing device, comprising: obtaining, for each of a plurality of people, one or more toughness values indicative of toughness and a stay frequency value indicating a degree of staying in a particular place;
Providing the toughness values and the staying frequency values obtained for the plurality of people to the input layer and the output layer of the artificial neural network model, respectively, so as to learn the artificial neural network model;
Wherein the processing unit is adapted to determine the number of toughness values used as independent variables for regression analysis and the number of sets of staying frequency values used as the dependent variable for the regression analysis, Determining a set of toughness values that are greater than the number of toughness values obtained for the plurality of persons so that the number of toughness values is greater than the number of toughness values and the number of staying frequency values obtained for the plurality of persons, Sequentially inputting the set retention frequency values to the input layer of the artificial neural network model, and sequentially obtaining the set retention frequency values from the output layer of the learned neural network model; And
Wherein the processing unit performs regression analysis with the retention frequency value of the one set as a dependent variable and the set of toughness values corresponding to each set of the retention frequency values as independent variables, Lt; RTI ID = 0.0 > a < / RTI >
/ RTI >
How to analyze the correlation between personality and preference place.

5. The method according to claim 4, wherein the predetermined set of toughness values are set so that each of the five toughness factor values according to the big five factor has seven different values, Characterized in that it comprises a combination of toughness and preference.

delete

A computing device adapted to analyze a correlation between toughness and a preference location, the computing device including a storage and a processing unit,
Wherein,
Obtaining, for each of a plurality of persons, at least one toughness value indicating toughness and a staying frequency value indicating a degree of staying in a specific place;
Providing the toughness value and the staying frequency value obtained for the plurality of persons to the input layer and the output layer of the artificial neural network model, respectively, to learn the artificial neural network model;
The number of toughness values of a set used as an independent variable for regression analysis and the number of the set of staying frequency values used as a dependent variable for the regression analysis are respectively the values of the toughness values The number of toughness values obtained for the plurality of persons is set to be greater than the number of toughness values obtained for the plurality of persons and the number of the staying frequency values obtained for the plurality of persons, Sequentially inputting the set retention frequency values from the output layer of the learned neural network model; And
Performing a regression analysis on the set of retention frequency values as dependent variables and the set of toughness values corresponding to each set of retention frequency values as independent variables, Calculating a coefficient
, &Lt; / RTI >
Wherein the storage unit stores the one or more toughness values and the staying frequency values,
Computing device.

A computing device adapted to analyze a correlation between toughness and a preference location,
Obtaining, for each of a plurality of persons, at least one toughness value indicating toughness and a staying frequency value indicating a degree of staying in a specific place;
Providing the toughness value and the staying frequency value obtained for the plurality of persons to the input layer and the output layer of the artificial neural network model, respectively, to learn the artificial neural network model;
The number of toughness values of a set used as an independent variable for regression analysis and the number of the set of staying frequency values used as a dependent variable for the regression analysis are respectively the values of the toughness values The number of toughness values obtained for the plurality of persons is set to be greater than the number of toughness values obtained for the plurality of persons and the number of the staying frequency values obtained for the plurality of persons, Sequentially inputting the set retention frequency values from the output layer of the learned neural network model; And
Performing a regression analysis on the set of retention frequency values as dependent variables and the set of toughness values corresponding to each set of retention frequency values as independent variables, Calculating a coefficient
To do
A computer-readable recording medium on which a program is recorded.