Yukinobu Hamuro

Mining Pharmacy Data Helps to Make Profits

Pharma, a drugstore chain in Japan, has been remarkably successful in the effective use of data mining. From over one tera bytes of sales data accumulated in databases, it has derived much interesting and useful knowledge that in turn has been applied to produce profits. In this paper, we shall explain several interesting cases of knowledge discovery at Pharma. We then discuss the innovative features of the data mining system developed in Pharma that led to meaningful knowledge discovery.

Data mining oriented CRM systems based on MUSASHI: C-MUSASHI

MUSASHI is a set of commands which enables us to efficiently execute various types of data manipulations in a flexible manner, mainly aiming at data processing of huge amount of data required for data mining. Data format which MUSASHI can deal with is either an XML table written in XML or plain text file with table structure. In this paper we shall present a business application system of MUSASHI, called C-MUSASHI, dedicated to CRM oriented systems. Integrating a large amount of customer purchase histories in XML databases with the marketing tools and data mining technology based on MUSASHI, C-MUSASHI offers various basic tools for customer analysis and store management based on which data mining oriented CRM systems can be developed at extremely low cost. We apply C-MUSASHI to supermarkets and drugstores in Japan to discover useful knowledge for their marketing strategy and present possibility to construct useful CRM systems at extremely low cost by introducing MUSASHI.

Data Mining Oriented System for Business Applications

This paper proposes a new concept called historybase that helps one to mining data for deriving useful knowledge from a huge amount of data, in particular aiming at business applications. At first we shall point out that data accumulated in databases for daily routine operation is not usually enough for data mining, and that we need to record much more detailed data for the knowledge discovery. Distinguished feature of historybase is to record a stream of events that occur during the course of information processing regardless of the necessity for routine operation, hoping that they are helpful for future data mining. Historybase is already implemented in real business, and has been successfully utilized to improve the business quality

Prediction Model Using Micro-clustering

Abstract This study proposes a method of clarifying the purchase consciousness of customers by conceptualizing their awareness as consumers. Specifically, the method addresses the purchase record data of the customer, uses micro-clustering based on the data polishing technique to conceptualize the customers mind according to the items that the customer has purchased, and uses a regularized regression model to build a prediction model based on the conceptualization. Micro-clustering is an algorithm for clustering graphs, and the data polishing technique clarifies the unclear hidden dense structures in the graph so that we can exhaustly enumerate with simple methods. By this method, we can obtain clusters of strongly correlated items, which are commonly purchased, are obtained. The clusters represent the customers’ minds, and thus we used them to build a classification model in an application; a model with the predictor variables representing the customers of health-conscious.

The Future Direction of New Computing Environment for Exabyte Data in the Business World

With the rapid spread of the Internet and the computerization of trading a huge amount of data on the Internet and of transaction database in enterprises has been accumulated. The purpose of this paper is to explain the significance of the technology to process of exabyte-scale data and presents the business application, CODIRO, which will make it possible to integrate various types of large scale data. CODIRO is a consumer research system which discovers new knowledge by integrating the huge amount of different types of data both on the Internet and within companies. This paper will demonstrate the business implications for exabyte-scale information technology research, by explaining an example of the analysis of the sales effectiveness of television commercials using CODIRO.

Business Application for Sales Transaction Data by Using Genome Analysis Technology

We have recently developed an E-BONSAI (Extended BONSAI) for discovering useful knowledge from time-series purchase transaction data, developed by improving and adding new features to a machine learning algorithm for analyzing string pattern such amino acid sequence, BONSAI, proposed by Shimozono et al. in 1994. E-BONSAI we developed can create a good decision tree to classify positive and negative data for records whose attributes are either numerical, categorical or string patterns while other methods such as C5.0 and CART cannot deal with string patterns directly. We shall demonstrate advantages of E-BONSAI over existing methods for forecasting future demands by applying the methods to real business data. To demonstrate an advantage of E-BONSAI for business application, it is significant to evaluate it from the two perspectives. The first is the objective and technical perspective such as the prediction accuracy. The second is the management perspective such as the interpreterability to create new business action. Applying the E-BONSAI to forecast how long new products survive in instant noodle market in Japan, we have succeeded in attaining high prediction ability and discovering useful knowledge for domain experts.

Discovering Interpretable Rules that Explain Customers' Brand Choice Behavior

The problem about brand choice or brand switching has been discussed for a long time in a marketing research field [1][2][3][6]. They focus on revealing a probability of brand switching and what factors are related to the brand switching. However, brand choice behavior of individual customer has been neglected in most of existing literature. In this study, we consider the problem of finding an optimal distribution strategy of discount coupon that determines to which customers and at what price coupons should be distributed, using detailed customer information.

Treatment pathways of Japanese prostate cancer patients - A retrospective transition analysis with administrative data

Background Limited availability of real-world data that describe treatment patterns of Japanese prostate cancer (PCA) patients. Methods A biweekly transition analysis of PCA treatment was performed for patients with PCA diagnosis and a specific treatment between 2010 and 2015. To account for different cancer stages, two patient populations were analyzed. The first group consisted of patients on medication for hormone-sensitive prostate cancer (HSPC). The second group is comprised of patients who ended up receiving specific therapy for castration-resistant prostate cancer (CRPC). For each treatment, the average of treatment duration and the portion of patients transitioning to a consecutive treatment was calculated. Results We identified 59,626 patients from the Japanese administrative database with a PCA diagnosis and specific treatment. In the first year of our observational study 786 patients commenced a HSPC treatment and 695 received a CRPC specific therapy Among the HSPC group, we found that combination hormonal therapy, comprised of a gonadotrophin releasing hormone agonist or antagonist with an antiandrogen was more common than monotherapy. The results of the CRPC group indicated that chemotherapy administration was for a shorter time period in a real-world setting as compared to published clinical studies. Conclusion Utilizing a novel method to visualize real-world treatment pathways for PCA patients we found that real treatment pathways are in line with international guidelines.

A Classification Model Using Emerging Patterns Incorporating Item Taxonomy

By extracting frequent patterns efficiently, it is possible to enhance some existing algorithms. Using many candidate patterns causes the results of the classification model to be more powerful. Moreover, aggregating similar items within patterns increases the possibility of creating more powerful patterns. In our method, we define some taxonomies and extract more powerful frequent patterns to incorporate such taxonomies and items. Our aim is to improve Classification by Aggregating Emerging Patterns(CAEP) by using more promising patterns with taxonomy. Using certain computational experiments as a source of practical data, we show that our performance is better than the one that does not use taxonomy. By identifying the reason behind our performance, we show that our method can extract better candidate patterns incorporating taxonomy.

Decision tree-based classifier incorporating contrast patterns

In the last ten years, studies that focus on the extraction of patterns among contrast classes and reveal the differences among these classes have been conducted. There exist two major streams in such studies, namely, emerging patterns (EPs) and contrast patterns (CPs), and related works concerning both have been proposed. In this field of study, the main problems pertain to extracting the efficiency of EPs or CPs and constructing smart classifiers on their basis. In this study, we propose a decision tree-based classifier using contrast patterns extracted by LCM. We also propose a method to construct a decision tree model that incorporates contrast patterns. Contrast patterns are extracted by LCM efficiency, and diverse scenarios are indicated by decision tree models in terms of business applications. Further, an example of our studies is illustrated using practical case data.