JP5948292B2 - Improvement factor extraction apparatus and method - Google Patents

Description

  The present invention relates to an improvement factor extracting apparatus and method, and more particularly to an improvement factor extracting apparatus and method for extracting an improvement factor in a technique for evaluating customer satisfaction of an information communication service.

  In the information communication service, in order to provide a service with high customer satisfaction, it is important to extract satisfaction constituent elements with respect to the requirements required by the user, and perform service development and function addition focusing on the factors. Therefore, there is a need for a method for extracting satisfaction component factors.

  A Japanese customer satisfaction index (JCSI) (see Non-Patent Document 1) exists as a method for extracting satisfaction component factors.

  JCSI specifies survey and analysis methods to clarify the user satisfaction with services and the impact of multiple satisfaction components on satisfaction. The survey method uses a questionnaire survey and evaluates the customer satisfaction, perceived quality, perceived value, word-of-mouth, review, and loyalty of the 18 items of satisfaction component factors and service satisfaction ratings in 10 stages. Let The analysis method clarifies the influence of each dimension and satisfaction constituent factors on satisfaction by covariance structure analysis using the evaluation value as an input value.

Minami, Ogawa, "Development of the Japanese Customer Satisfaction Index (JCSI) Model and Its Theoretical Basis" Quarterly Marketing Journal 117. pp.4-19, June 2010.

  However, in order to extract satisfaction component factors and reflect them in service improvement, it is necessary to solve the following problems and extract factors that need improvement.

  As a first problem, satisfaction component factors vary depending on usage scenes (before use / during use / after use). JCSI divides satisfaction components into five dimensions: customer expectations, perceived quality, perceived value, word of mouth, and loyalty, and is not classified based on usage scenes. Can not be extracted.

  As a second problem, it is necessary to set the granularity of satisfaction constituent factors at a granularity that can be reflected in service improvement. In JCSI, the granularity of 18 factors is coarse and it is difficult to reflect in service improvement. For example, the constituent factors of customer expectation are the whole, reliability, and needs. Even if it becomes clear that the user's evaluation value for the whole is low, it is unclear which point of service should be improved.

  As a third problem, it is necessary to clarify which satisfaction constituent factor should be preferentially solved. In JCSI, the influence of each satisfaction component on the satisfaction and the user's evaluation value for each satisfaction component are clarified, but the improvement factor cannot be determined. For example, when it becomes clear that the evaluation value from the user with respect to reliability and variation in perceptual quality is low, it is unclear which factor should be preferentially improved.

  As a fourth problem, the influence of each satisfaction component on satisfaction is different for each user. In JCSI, since the covariance structure analysis is performed using the evaluation values of all users as input values, the impact on satisfaction of each satisfaction component that differs from user to user cannot be captured. For example, when covariance structure analysis is performed using evaluation value data including users who consider the entire customer expectation and users who do not consider the input, the satisfaction of the overall satisfaction of the customer expectation in the user who considers the entire customer expectation The influence is estimated to be smaller than when the analysis is performed only by the user who considers the entire customer expectation.

  The present invention has been made in view of the above points, and with a granularity that can be fed back to service improvement, the influence on satisfaction of satisfaction components that differ for each use scene and for each user is clarified. An object of the present invention is to provide an improvement factor extracting apparatus and method for clarifying priorities.

According to one aspect, in the information communication service, the improvement factor is extracted based on the evaluation value of the user with respect to the satisfaction degree to the service and the factor affecting the satisfaction degree (hereinafter referred to as “satisfaction factor”). An improvement factor extraction device,
Use scene classification means for acquiring the user's evaluation value for the satisfaction degree and the satisfaction component, and classifying the user evaluation value for each satisfaction component into use scenes before use, during use, and after use; ,
User classification means for acquiring satisfaction component as user attribute information, and classifying users considering the same satisfaction component into the same group based on the user attribute information;
With the obtained evaluation value of the user as an input, an influence calculation means for performing covariance structure analysis for each user classification and calculating the influence on the satisfaction for each satisfaction constituent factor,
Priority calculation means for calculating priority from the evaluation value and the influence of satisfaction for each satisfaction component calculated by the influence calculation means for each user classification and for each satisfaction component When,
Calculated using the average value of the priorities of all satisfaction components included in the same usage scene plus the standard deviation of the priorities of all satisfaction components included in the same usage scene as a threshold value Threshold calculation means for
There is provided an improvement factor extraction device having improvement factor extraction means for extracting and outputting a factor having the priority equal to or higher than the threshold as an improvement factor for each use scene.

  According to one aspect, the following effects can be obtained.

  First, since the influence of the satisfaction component on the satisfaction level can be clarified for each use scene (before use / during use / after use), the satisfaction component to be improved in the use scene desired by the user Can be revealed.

  Second, the satisfaction component can be extracted with a granularity that can reflect the satisfaction component in service improvement. JCSI's satisfaction component factors are “overall”, “variation”, etc. Even if the satisfaction component factors that strongly affect satisfaction are clear, it is unclear how the service should be improved. However, in the present invention, since “communication speed”, “service area” and the like are defined as satisfaction constituent factors, for example, when the communication speed strongly affects the satisfaction, a service such as a measure for improving the communication speed is proposed. It is clear how to improve.

  Third, when there are a plurality of satisfaction constituent factors, the influence of each satisfaction constituent factor on the satisfaction and the priority of improvement of each satisfaction constituent factor can be clarified.

  Fourth, by classifying users with the same satisfaction constituent factors to be considered when determining satisfaction into the same group and clarifying the impact of satisfaction constituent factors on satisfaction for each group, It is possible to clarify the influence on satisfaction of each satisfaction component that is different for each.

  That is, according to the present invention, improvement factors can be extracted and reflected in service improvement.

It is a figure which shows the relationship between the improvement factor extraction apparatus and user terminal in one embodiment of this invention. It is an example of a structure of the improvement factor extraction apparatus in one embodiment of this invention. It is a flowchart of the improvement factor extraction device in one embodiment of the present invention. It is an example of the covariance structure analysis result in one embodiment of this invention.

  Hereinafter, embodiments of the present invention will be described with reference to the drawings.

  The present invention is based on the evaluation value of the user with respect to the satisfaction degree of service and satisfaction component factors (for example, communication speed, service area, etc.) for each use scene (before use / during use / after use) in the information communication service. This is to extract improvement factors.

  FIG. 1 shows the relationship between an improvement factor extraction device and a user terminal according to an embodiment of the present invention.

  The improvement factor extraction apparatus 100 determines, for the user terminal (for example, a smartphone) 200, an evaluation value for satisfaction, an evaluation value for a satisfaction component factor, a use scene flag (before use / during use / after use). ) And user attribute information is output.

  FIG. 2 shows an example of the configuration of the improvement factor extraction apparatus in one embodiment of the present invention.

  The improvement factor extraction device 100 shown in the figure includes a use scene classification unit 110, a user classification unit 120, an influence calculation unit 130, a priority calculation unit 140, an improvement factor extraction unit 150, a priority list update unit 160, and a priority list. A storage unit 170 and a memory (not shown) are included, and factors with high priority for improvement are extracted in order to improve user satisfaction based on the acquired satisfaction level and the user's evaluation value for the satisfaction level configuration factor. To do. A memory (not shown) stores the following Tables 1 to 3 and a priority list.

  The process of the improvement factor extraction apparatus 100 in said structure is shown.

  FIG. 3 is a flowchart of the improvement factor extraction apparatus according to the embodiment of the present invention.

  Step 1) The usage scene classification unit 110 acquires user evaluation values for satisfaction and satisfaction constituent factors, and uses the user evaluation values for each satisfaction component as usage scenes (before, during, and after use). Classify into: Table 1 shows the results of classifying user evaluation values for each satisfaction component into use scenes.

表１において、各満足度構成要因の利用シーンへの分類基準は、満足度構成要因に利用シーンフラグ（例えば、"広告"という満足度構成要因においては、"利用前"というフラグ）を予め設定し、当該利用シーンフラグに基づき分類してもよいし、アンケート調査などを用いて分類してもよい。

In Table 1, the classification criteria for each satisfaction component constituting the usage scene is set in advance with a usage scene flag (for example, a flag “before use” for the satisfaction component “advertising”) as the satisfaction component. Then, it may be classified based on the use scene flag, or may be classified using a questionnaire survey or the like.

  Step 2) The user classifying unit 120 acquires user attribute information and classifies users based on the user attribute information. The user attribute information is information indicating a satisfaction component that is considered when each user determines the satisfaction and a satisfaction component that is not considered. For user attribute information, a multiple regression analysis is performed with the evaluation value for satisfaction as the objective variable and the evaluation value for the satisfaction component as the candidate for the explanatory variable, and the satisfaction component is determined for the satisfaction component as the explanatory variable. It is extracted and generated as a satisfaction constituent factor to be taken into consideration. Alternatively, the user may input a satisfaction component that is considered when determining the satisfaction. An example of user attribute information is shown in Table 2.

上記の表２の場合、ユーザ（１）とユーザ（２）が同じグループ、ユーザ（３）とユーザ（４）が同じグループに分類される。

In the case of Table 2 above, user (1) and user (2) are classified into the same group, and user (3) and user (4) are classified into the same group.

  Step 3) For each user classification in Table 2 above, the influence calculator 130 determines each satisfaction degree by covariance structure analysis using the user's evaluation value for the satisfaction degree and the satisfaction constituent factor obtained in Step 1 as input values. Calculate the influence of constituent factors on satisfaction. An example of the evaluation value to be acquired is shown in Table 3, and an example of the covariance structure analysis result is shown in FIG. Here, each satisfaction degree component factor and the satisfaction degree are evaluated in 10 stages as “10. bad” and “1. good”.

ステップ４） 優先度計算部１４０は、上記のユーザ分類ごとに、満足度構成要因ごとに評価値の平均値を計算し、上記の各満足度構成要因の満足度への影響力と各満足度構成要因の評価値の平均値の積を、優先度として満足度構成要因ごとに計算する。計算した各満足度構成要因の評価値の平均値及び優先度の一例を表１の評価値列及び優先度列に示す。

Step 4) The priority calculation unit 140 calculates the average value of the evaluation values for each satisfaction constituent factor for each of the user classifications described above, and influences each satisfaction constituent factor on the satisfaction and each satisfaction degree. The product of the average evaluation values of the constituent factors is calculated for each satisfaction constituent factor as the priority. An example of the average value and priority of the calculated evaluation value of each satisfaction degree component factor is shown in the evaluation value column and priority column of Table 1.

  Step 5) The improvement factor extraction unit 150 extracts a satisfaction component that needs improvement according to the following procedure.

  (A) The satisfaction degree constituent factors are rearranged in descending order of priority for each user classification and each usage scene. Table 1 shows an example of the result of rearranging the satisfaction degree constituent factors in one user classification.

  (B) The standard deviation of the priorities of all satisfaction constituent factors in the same usage scene is added to the average value of the priorities of all satisfaction constituent factors included in the same usage scene for each user classification and usage scene. The added value is calculated as a threshold value. An example of the calculated threshold value is shown in the threshold value column of Table 1.

  (C) Satisfaction constituent factors having the above-mentioned priority level equal to or higher than the threshold obtained in (a) are extracted as improvement factors. In the improvement factor column of Table 1, “○” is given to the constituent factors extracted as improvement factors. In the example of Table 1, the satisfaction component “advertisement” of the usage scene “before use” is the satisfaction component “terminal” of the usage scene “in use”, and the satisfaction component “use” of the usage scene “after use” is “ “Competitive comparison”, “discount conditions”, and “content appeal” are the improvement factors.

  Step 6) The priority list update unit 160 determines the influence of each satisfaction component on the satisfaction, the average value of the evaluation values, the priority, the threshold value, and the improvement factor extraction result (“◯” in Table 1). And the like are stored in the priority list, and the priority list is stored in the priority list storage unit 170.

  As described above, the improvement factor extraction apparatus 100 outputs to the user terminal 200 the evaluation value for the satisfaction value obtained by the above processing, the evaluation value for the satisfaction degree configuration factor, the use scene flag, and the user attribute information.

  In addition, the operation | movement of each component of the improvement factor extraction apparatus 100 in said embodiment is constructed | assembled as a program, installed in the computer utilized as an improvement factor extraction apparatus, or making it distribute | circulate through a network. Is possible.

  The present invention is not limited to the above-described embodiments, and various modifications and applications are possible within the scope of the claims.

100 improvement factor extraction device 110 use scene classification unit 120 user classification unit 130 influence calculation unit 140 priority calculation unit 150 improvement factor extraction unit 160 priority list update unit 170 priority list storage unit 200 user terminal

Claims (8)

In an information communication service, an improvement factor extraction device that extracts improvement factors based on a user's evaluation value with respect to satisfaction with services (hereinafter referred to as “satisfaction constituent factors”). ,
Use scene classification means for acquiring the user's evaluation value for the satisfaction degree and the satisfaction component, and classifying the user evaluation value for each satisfaction component into use scenes before use, during use, and after use; ,
User classification means for acquiring satisfaction component as user attribute information, and classifying users considering the same satisfaction component into the same group based on the user attribute information;
With the obtained evaluation value of the user as an input, an influence calculation means for performing covariance structure analysis for each user classification and calculating the influence on the satisfaction for each satisfaction constituent factor,
Priority calculation means for calculating priority from the evaluation value and the influence of satisfaction for each satisfaction component calculated by the influence calculation means for each user classification and for each satisfaction component When,
Calculated using the average value of the priorities of all satisfaction components included in the same usage scene plus the standard deviation of the priorities of all satisfaction components included in the same usage scene as a threshold value Threshold calculation means for
An improvement factor extracting means for extracting and outputting a factor of which the priority is equal to or higher than the threshold as an improvement factor for each use scene;
An improvement factor extraction apparatus characterized by comprising:   Priorities for storing the influence on the satisfaction of satisfaction factors for each use scene, the average value of the evaluation values, the priority, the threshold value, and the improvement factor extraction result in the priority list storage means The improvement factor extracting device according to claim 1, further comprising a list updating unit. The usage scene classification means includes:
A user evaluation value for satisfaction, a user evaluation value for the satisfaction constituent factor, a usage scene flag indicating any of the satisfaction constituent factors before, during, or after use are acquired, and the usage scene flag is 3. The improvement factor extracting apparatus according to claim 1, further comprising means for classifying the satisfaction degree constituent factor into a use scene based on the basis. The user classification means includes
Means for acquiring satisfaction constituent factors to be considered when each user determines satisfaction and satisfaction constituent factors not to be considered as the user attribute information, and classifying users considering the same satisfaction constituent factors into the same group The improvement factor extraction apparatus of Claim 1 or 2 containing these. The priority calculation means includes:
An average value of user evaluation values is calculated for each user classification and for each satisfaction component, and the average value and the satisfaction for each satisfaction component calculated by the influence calculation means are calculated. 3. The improvement factor extraction device according to claim 1, further comprising means for calculating a product of influence for each satisfaction constituent factor as the priority. The threshold calculation means includes
The improvement factor extracting device according to claim 1 or 2, further comprising means for rearranging the priorities calculated by the priority calculating means and calculating the threshold value for each user classification and each use scene. In an information communication service, an improvement factor extraction method for extracting improvement factors based on a user's evaluation value with respect to satisfaction with a service and an influence on the satisfaction (hereinafter referred to as “satisfaction constituent factor”). ,
In an apparatus having use scene classification means, user classification means, influence calculation means, priority calculation means, threshold value calculation means, improvement factor extraction means,
The usage scene classification means obtains the user's evaluation value for the satisfaction degree and the satisfaction component, and sets the user evaluation value for each satisfaction component to the usage scene before use, during use, and after use. Use scene classification step to classify,
A user classification step in which the user classification means obtains a satisfaction factor as user attribute information, and classifies users considering the same satisfaction factor into the same group based on the user attribute information;
The influence calculating means performs the covariance structure analysis for each user classification, using the acquired evaluation value of the user as an input, and calculates the influence on the satisfaction for each satisfaction constituent factor; ,
The priority calculation means determines the priority from the evaluation value and the influence of the satisfaction for each satisfaction component calculated in the influence calculation step for each user classification and for each satisfaction component. Priority calculation step to calculate,
The threshold value calculation means adds the standard deviation of the priorities of all satisfaction constituent factors included in the same usage scene to the average value of the priorities of all satisfaction constituent factors included in the same usage scene. A threshold calculation step for calculating the calculated value as a threshold;
An improvement factor extracting step in which the improvement factor extracting means extracts and outputs a factor of which the priority is equal to or higher than the threshold as an improvement factor for each use scene;
The improvement factor extraction method characterized by performing. In an apparatus further comprising priority list storage means and priority list update means,
The priority list update means stores the influence on the satisfaction of satisfaction constituent factors for each use scene, the average value of the evaluation values, the priority, and the threshold value in the priority list storage means. The improvement factor extraction method according to claim 7, wherein a priority list update step is performed.

Images