Takeshi Kurosawa

Service-demand-forecasting method using multiple data sources

A service-demand-forecasting method that uses multiple data sources for improved accuracy is presented in this paper. We presented a advanced scenario simulation framework to analyze each customer service choice behavior and total service demand under an assumed condition. To estimate the demand on service, especially for a new service, scenario simulation is executed based on stated preference data (SP). However, there is usually a considerable gap between an estimate based on SP and the actual outcome. Furthermore, the distribution of attributes of users collected in market surveys is not equal to that of actual market users, and there is a bias between collected data and attributes of users in the market. In this paper, we propose an adjusting method that can re-estimate demands by analyzing the difference between actual and estimated results. We consider the problem of choosing an Internet access line service to evaluate the method. An application example shows that distribution differences between areas with respect to customer attributes strongly affect actual market shares

Mobile Carrier Choice Modeling under Competitive Conditions

This paper presents a mobile-carrier choice modeling to analyze customer preference and understand customer choice behavior in the mobile phone market. Due to severe competitive conditions, there are a few differences among mobile phone services provided by mobile-carriers. We propose a new mobile-carrier choice modeling considering incentive factors and restrictive factors as decision-making factors. Web survey is carried out to obtain the sample data for this model. We show the model estimated by using the survey data to analyze mobile-carrier choice behavior.

Mobile-carrier Choice Modeling Framework Under Competitive Conditions

This paper presents a mobile-carrier choice modeling framework to analyze customer preference and understand customer choice behavior in the mobile phone market. Due to severe competitive conditions, there are few differences between the mobile phone services provided by mobile-carriers. We propose a new mobile-carrier choice modeling that takes into account incentive factors and restrictive factors as decision-making factors. A Web survey was carried out to obtain the sample data for this model. We show the model estimated from the survey data to analyze mobile-carrier choice behavior.

Customer Segmentation in Mobile Carrier Choice Modeling

Service choice behavior of customers is becoming complicated and diversified in the mobile phone market. The aim of this paper is to analyze mobile carrier choice behavior and clarify the customer segments and their decision-making factors. We classify mobile phone customers into two groups: churning customers and stable customers. Customer segmentation is carried out based on survey data for each customer group in order to clarify the decision-making factors. We propose a new service choice modeling with regard to mobile carriers based on the customer segmentation results.

Demand Analysis by Modeling Choice of Internet Access and IP Telephony

In Japan, demands for broadband Internet access and IP telephony have increased dramatically in recent years. According to official sources, as of September 2009, there are 30.9 million users of broadband Internet access and 20.9 million of IP telephony. This study evaluates and estimates the market share of fiber-optic Internet connection, which is becoming the major player in broadband services, paying specific attention to IP telephony. A comprehensive combined choice model of Internet access line, IP telephony, and awareness of IP telephony is presented. The most suitable parameters for the model were determined by using an original market research survey conducted in Japan during 2004 with stated-preference choice experiments of both Internet access and IP telephony. The results indicate that increasing awareness has the potential to dramatically increase the penetration of IP telephony. The results also indicate a great variability in price sensitivity across income groups for the choices of Internet access line and IP telephony. The fiber optic share is shown to change with a change in its own monthly usage charge, indicating that market share gains are possible through reduced usage fees.

Evaluation of HTTP video classification method using flow group information

Recently, the traffic volume of HTTP video applications, such as YouTube, is rapidly growing on the Internet. To support the quality of service requirements of HTTP video applications, network carriers need to design bandwidth by taking into account the traffic volume of HTTP video applications. However, since most HTTP video applications are provided in a web browser, it is difficult to classify HTTP video applications with other web applications by the port number. We propose an HTTP video application classification method by using a machine learning method with traffic-flow features such as packet size. We propose a new feature that is useful in classifying HTTP video applications. We can improve the accuracy of traffic classification of HTTP video applications by 12%. Furthermore, we compare the accuracy and calculation time among three machine learning methods.

Mobile-Carrier Choice Behavior Analysis Around Smart Phone Market

The rapid growth of the smart phone market has caused various changes in customer preference and carrier choice behavior in the mobile phone market. We analyzed the characteristics and the change in relation to the usage patterns of mobile phones and mobile-carrier choice behavior. This paper presents the results based on the two sets of the original survey data collected from customers of major three mobile-carriers: docomo, au and Soft Bank. We show the estimation results of a new mobile-carrier choice model considering new factors.

Econometric approach for broadband market in Japan

We investigated the growth in the Japanese broadband market. We analyzed price elasticity and changes in the broadband market in 2009 by comparing those of the market in 2005, which we consider the growth stage of Fiber-To-The-Home (FTTH) service in terms of econometrics. Through analysis, we found that users feel FTTH service fees are reasonable and that the service is used among high- and low-income households. We also believe that FTTH service is no longer for users who use the Internet for heavy downloading activities, such as P2P file sharing. Furthermore, FTTH fees have becomes less elastic than in 2005.

Regression correlation coefficient for a Poisson regression model

This study examines measures of predictive power for a generalized linear model (GLM). Although many measures of predictive power for GLMs have been proposed, most have limitations. Hence, we focus on the regression correlation coefficient (RCC) (Zheng and Agresti, 2000), which satisfies the four requirements of (i) interpretability, (ii) applicability, (iii) consistency, and (iv) affinity. The RCC is a population value that is defined by the correlation between a response variable and the conditional expectation of the response variable. Its sample value is defined by the sample correlation between the observed response values and estimated values of the response variable. For an arbitrary GLM, we do not always have an explicit form of the RCC. However, for a Poisson regression model, assuming that the predictor variables have a multivariate normal distribution, we can find the explicit form of the RCC (true value). Therefore, it is possible to compare the estimators (sample values) of the RCC in terms of bias and RMSE (root of the mean square error) by using the true value. Furthermore, by using the explicit form, we propose a new estimator of the RCC for the Poisson regression model. We then compare the new estimator with the sample correlation estimator, the jack-knife estimator, and the leave-one-out cross validation estimator in terms of bias and RMSE. The leave-one-out cross validation estimator has large negative bias and large RMSE. Although the remaining three estimators show similar behavior for a large sample size, for a small sample size the new estimator shows the best behavior in terms of bias and RMSE.

Mobile-Carrier Churning Behavior Modeling Based on Customer Satisfaction

The Japanese mobile phone market is in the matured stage. Under competitive conditions, the most important objective of a mobile-carrier is to increase the market share. The aim of this research is to understand mobile-carrier choice behavior under assumed market conditions. We propose the binary choice models of which alternatives are change to other carrier or not. Three models are constructed based on survey data for three mobile-carriers: NTT docomo, au by KDDI, and Soft Bank. Each model has eight factors expressed by the satisfaction levels for mobile-service characteristics. It is shown that important decision-making factors vary from carrier to carrier. It implies that a mobile carrier should keep and improve the satisfaction level of the important factors in order to increase the market share.