Download NAME: Vipin Saini STUDENT ID: M964011062 Hsu Po Sung 許博淞

Survey
yes no Was this document useful for you?
   Thank you for your participation!

* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project

Document related concepts

First-mover advantage wikipedia , lookup

Retail wikipedia , lookup

Global marketing wikipedia , lookup

Planned obsolescence wikipedia , lookup

Pricing strategies wikipedia , lookup

Marketing channel wikipedia , lookup

Sensory branding wikipedia , lookup

Customer satisfaction wikipedia , lookup

Product placement wikipedia , lookup

Marketing strategy wikipedia , lookup

Product lifecycle wikipedia , lookup

Predictive engineering analytics wikipedia , lookup

Product planning wikipedia , lookup

Transcript
NAME: Vipin Saini
STUDENT ID: M964011062
Hsu Po Sung 許博淞
M964020009
Chen Yun-Chih 陳昀志
M964020043
DATA MINING PROJECT 1 REPORT
MINING PRODUCT MAPS FOR NEW PRODUCT DEVELOPMENT
April (2008)
College of Management,
National Sun Yat-sen University
Kaohsiung.
Abstract:
This paper presents the product map obtained from data mining results, which investigates the
relationships among customer demands, product characteristics, and transaction records, using
the Apriori algorithm as a methodology of association rules for data mining.
Introduction:
New product development (NPD) is critical for long-term firm performance and competitive
advantages. Because new products take a substantial time to develop, suppliers should anticipate
changes in what customers will value as early as possible (Flint, 2002). Therefore, a strong
market orientation and customer knowledge competence are vital to the success of new products
(Bonner, 2005). It is available but not easily accessible, and there is little possibility to explore
the full volume of data that should be collected for its potential value. Inefficient utilization can
render the data collected useless, causing databases to become ‘data dumps’ (Keim, Pansea,
Sipsa, & Northb,2004).
Similarly, there are also many application methods, including association rules, sequential
patterns, grouping analysis, classification analysis and probability heuristic analysis (Fan, Lu,
Mad-nick, & Cheung, 2002; Hui & Jha, 2000; Goodwin, Dyne,Lin, & Talbert, 2003; Mehta &
Bhattacharyya, 2004;Musaev, 2004; Tsechansky, Pliskin, Rabinowitz, & Porath,1999). So, How
to effectively process and use customer data is becoming increasingly important.
There are many data mining models in the literature such as classification, estimation, predictive
modeling, clustering/segmentation, affinity grouping or association rules, description and
visualization, as well as sequential modeling. Based on this idea, this research implements a
product map to illustrate that new product development is essentially the function that matches
the enterprise’s offers to customers’ demands.
This paper investigates the following research issues in the development of new cosmetic
products: 1.What exactly are the customers’ ‘‘needs’’ and ‘‘wants’’ for cosmetic products?
2.Can product design and planning for product lines/product collection be integrated with the
knowledge of customers?
3.Can the knowledge of customers be transformed into knowledge assets of the enterprises
during the stage of new product development (NPD)?
Research design
2.1 Research Problems
First, cosmetics product development is not only concerned with market reasons but also
with safety concerns. Secondly, cosmetic products are a kind of biomedical and truly personal
product, which is popular with both woman and man on different age scales (Chang & Chang,
2003).Customer-orientation is the main concern for investigating market segmentation according
to customers’ attributes, preferences, and personal factors. Thirdly, cosmetic products are a kind
of collection products, which include different cosmetic products and functions on specific usage
stages in order to fulfill customer needs (Santoro & Oliveira, 2005).
Fourthly, in a demand chain, the focus is clearly customer-centric, as defined early by Brace
(1989), who explained the concept of a demand chain as ‘‘...the whole manufacturing and
distribution process may be seen as a sequence of events with but one end in view: it exists to
serve the ultimate consumer.’’ Customers have a new position in the demand chain and assume a
leading role in bringing greater benefit for enterprises, especially in the cosmetic product
industry.
Accordingly, with demand side consideration, this paper focuses on integration of data, including
customer data, transaction data, and product data, on a database for data mining in order to
develop a product map of customers’ demands and purchase patterns as collection product for
new product development and marketing.
2.2 Data collection
1. Data are collected from customers and channel/sales stores based on NPD and marketing
questionnaire design, including (1) customer basic data table; (2) customer product preference
data table; (3) customer skin attitudes data table; (4) product function data table; (5) customer
suggestion data table; (6) product data table; and (7) transaction data table. All data are collected
by interviewing customers who have purchased and used cosmetic products from cosmetic
product channels/stores.
Data collection was then Data collection was then conducted between May and October 2004 on
channel/sales locations in Taipei, Taichung, Kaoshiung and Hsin-chu cities, which are the four
largest cities in Taiwan. A total of 1400 questionnaires were sent, and 1207 completed
questionnaires were returned. There were 1009 valid responses, for a response rate of about 72%,
and the relational database construction was completed in December 2004.
2.3. System framework
This paper proposes the association rules for data mining to extract customer and market
knowledge on product and transaction data from database.
2.4. Relational database design
Some research articles have shown that the association rules of a relational data-base can provide
a useful method for mining knowledge on different application areas. In this study, the relational
database contains 8 entities, 7 relationships, and 47 attributes.
2.5. Physical database design
The structure of physical database design in this paper is described in Fig. 2.
This study established relational tables on MS Access 2002 and transferred them on MS SQL
Server within OBDC environment in order to implement the data table on SPSS Clementine.
3. Data mining
3.1. Association rules
For the exploration of the association rules, many researchers usually use the Apriori algorithm
(Agrawal et al., 1993). In order to reduce the possible biases incurred.When using these measure
standards, the simplest way to judge the standard is to use the lift judgment. Lift is defined as:
Lift = Confidence(X! Y)/Sup(Y) (Wang et al., 2004).
3.2. Data mining tool – SPSS Clementine
There are several enterprise data mining workbenches such as Clementine (SPSS), Darwin
(Oracle), EnterpriseMiner (SAS), IntelligentMiner (IBM). Clementine is one of the most
powerful enterprise data mining workbenches, enabling the user to quickly develop predictive
models using business expertise and deploy them into business operations to improve decisionmaking. The data processing in Clementine is done through the use of nodes,which are then
connected together to form a stream frame.In addition, data visualization can be presented to
users after the mining process has been done.
3.3. Data mining process
It can be separated into product knowledge and marketing knowledge, which then can be applied
to NPD and marketing. The database is accessed receive or search for the product knowledge
needed in the NPD process (see Fig. 3).
Consumer experiences or feedback information should be continually collected, and repeatedly
added into the database for future use. Especially in the NPD stage, understanding the voice of
the customer during new product development (NPD) leads to the development of superior
products that meet customer needs better than the competitors’ products (Bonner, 2005).
4. Data mining results
4.1 New product development (NPD) analysis
New product development (NPD) analysis is developed by considering various decision
variables ( such as customer sex, age, residential area, shin type, product style), and purchasing
tendencies of customer’s to obtain knowledge about customers product preferences.
(A) Personal care products
1. Basic personal care – complicated skin type
In the basic personal care item association rules, consumers’ experience is based on the female
customers between 25 and 29 years old, indicating that those who purchase cleansing powder,
makeup remover and facial cleanser would most possibly have the complicated skin type
(Table 1). The diagram in Fig. 4 shows the complex relationships between all the decision
variables in the association rules. The collections of all the decision variables can be shown with
this type of graph. All the lines in the graph represent the sign on records of the customers in the
database; the thickness of the lines represents whether the extent of the relationship between the
two decision variables is high or low.
2. For women with oily skin type, those who tend to use the basic personal care production
collection of delicate care toner, pure white moisturizer, moisture rich toner and facial cleanser
products show a great possibility to use cleansing powder (Table 2).
3. Basic personal care – normal skin type
For women with normal skin type who use the basic personal care products, the mining result
shows that refreshing moisturizer and makeup remover have the largest association with
cleansing powder (Table 3).
4. Special personal care
In the special personal care item association rules, it can be seen that women who live in Taipei
would more frequently consider the products of acne mask, mask, and black-head treatment
altogether (Table 4).
(B) Cosmetics products
For women using these cosmetics, those who prefer to purchase the products of thickening type
mascara, lipstick and nail enamel show great associations to purchase lip color (Table 5).
4.2 Product map
According to data mining results, this report investigates the relationship among those decision
variables and results by illustrating the product map. A product map, which describes association
relationship on customer data, product data, and transaction data, proposes a complete picture of
customer demands for NPD. Thus, this report presents several NPD knowledge patterns and
rules of personal care and cosmetics collections for implementing new product development and
its pricing design.
4.2.1. Mining results for new product development
4.2.1.1. NPD pattern A: basic personal care-oriented product collection.
Fig. 5 shows a product map for a personal care-oriented product collection for people with
different skin types. In the product map, there are totally nine different cells, which correspond to
diverse decision variables or product collections. After specific correlation analysis, certain
decision variables will be combined and collected in the form of a pattern. Different patterns will
develop different rules. As shown in Fig. 5, there are six rules available for basic personal careoriented product collection. In the process of basic personal care new product development, it is
suggested to develop at least six product collections. Thus, the following product collections
should be the most preferable considerations for consumers between 25 and 29 years old
(Table 6).
4.2.1.2. NPD pattern B: specific personal care-oriented product collection.
For special person care products, the data mining results, thus, indicate that it is suitable to
develop at least four product collections for consumers between 25 and 29 years old (Table 7).
Among them, product collections from B1 to B3 apply to female customers, while item B4 will
be developed to meet men’s requirements.
4.2.1.3. NPD pattern C: cosmetics-oriented product collection.
In the cosmetic section, at least five product collection rules are implemented for developing
product sets for female customers (Table 8). The proposed suggestions are rules C1–C4 for
women between 25 and 29 years old and rule C5 is for man from 25 to 29 years old.
4.2.2. Mining results for price effects
Product price is one of the main factors to affect the customer’s purchasing behaviors. Using
budget segmentation, the cosmetics industry should make commitments to developing various
price-based product collections so as to provide more customized or specialized commodities for
different requirements. In the following, when price effects are considered, several rules are
proposed.
4.2.2.1. NPD pattern D: price effects for basic personal care oriented product collection.
To customize the products, several product collections are proposed based on different budget
plans for purchasing. Thus, the following basic personal
care product collections should be the most preferable considerations for different customer
groups (Table 9).
4.2.2.2. NPD pattern E: price effects for cosmetic-oriented product collection.
Similar marketing strategies should be applied to cosmetic product collections. Thus, the
following product collections should be the most preferable considerations for different customer
groups (Table10).
5. Discussions and future works
5.1 New Product Development
Fashion trends, greater demand for convenience and multi-functional products together with
higher demand among selected specialty consumer segments, are the key drivers in the cosmetics
industry. Since there is a growing trend towards industry globalization, industry leaders have to
focus on diverse product lines and effective marketing strategies to meet the specific customers’
needs and preferences and to reach the enterprise’s sales goals. Therefore, this study suggests
that the cosmetics industry should pay more attention, not only to biomedical or bioscience
experiments, but also to consumer factors for new product development.
5.2 Demand Chain Management
Demand chain management, including customer satisfaction, involvement, and customization, is
attained through identifying specific needs of groups of customers and developing appropriate
offers to certain groups of customers or market segments on products and service. The cosmetics
industry should extract customer knowledge from the demand chain and use it as a knowledge
resource on its supply chain. Integration of demand chain and supply chain not only offers better
capabilities for understanding its market, but also enhance its manufacturing and product
innovation capabilities by extending its product lines.
5.3 Database approach
Information technology can make dramatic changes to business practice, and one major trend is
the coming of age of database technology. Enterprises are realizing that the most fundamental
benefits of IT lie at the customer interface, through interactive exchange of information with
suppliers and customers, and leading to analyze and information-enhanced products. Mining
customer knowledge for new product development is an example of implementing a database
approach for increased connectivity and using information technology, thus creating internal data
mining capability for analysis and providing decision supports.
5.4 Product Map
Map display is a multi-category mapping approach,which allows the analyst to see the whole
picture of a problem in an attribute space from analysis results. When analysis results from
different attributes or categories are integrated together on a map, this means that the analyst can
initiate strategic and tactical plans with complete information or knowledge for decision-making
or problem solving.
This report Presents the product map with the association rules. The product map shows that
different knowledge patterns and rules are extracted for new cosmetic product development
suggestions and solutions. In addition, the product map illustrates a visualization data analysis
result, and this is what data mining approach works for to develop. Therefore, this report
suggests that the map display approach could be implement for multi-category or multi-attribute
data analysis in other research problems.
5.5 Data Mining Methodologies
This research used SPSS Clementine data mining functions under the Windows XP environment
with Intel Pentium 4 2.2 GHz CPU and 8 GHz RAM. However, due to system limitations, this
research could not integrate more decision variables and levels for data mining, and this may
have limited the scope of classification and clustering results. This paper proposes Apriori
algorithms methodology of association rules for data mining, which is implemented for mining
product map knowledge from customers.
6. Conclusion
Customer needs and wants are sensitive and complex. If a firm can understand them and make
efforts to fulfill customer demands and provide friendly service, then customers will be more
supportive and loyal to the enterprise, it will increase the enterprise’s competitiveness and it is an
essential criterion to earning higher loyalties and profits. Knowledge extraction from data mining
results is illustrated as knowledge patterns and rules in order to propose suggestions and
solutions for NPD and possible marketing solutions.
7. Critics to the work
After understanding what is the customer really want and redesign the product or service for
target customer can make the firm to have more profit or loyalties from customer. A complete
new product design is doing so for that reason. A product can’t fulfill all need, but can support
some part of requirement. In this paper, choose some attributes to do the data mining tasks by
association rules technique. Suggesting that, would choosing more attributes or to have more
dataset from another area in Taiwan not just the questionnaire form Taipei. The more diverse
dataset, comes with the more unknown knowledge mining out, but should take it seriously about
the higher error rate of the result
8. References
Liao, S.H., Hsieh, C.L., & Huang, S.P. (2008). Mining product maps for new product
development. Expert Systems with Applications, Volume 34, Issue 1, January 2008, Pages 50-62.
9. Appendix