داده کاوی برای بهبود استانداردهای صنعتی و افزایش تولید و بازاریابی: یک مطالعه تجربی در صنعت پوشاک
|کد مقاله||سال انتشار||مقاله انگلیسی||ترجمه فارسی||تعداد کلمات|
|22155||2009||7 صفحه PDF||سفارش دهید||4730 کلمه|
Publisher : Elsevier - Science Direct (الزویر - ساینس دایرکت)
Journal : Expert Systems with Applications, Volume 36, Issue 3, Part 1, April 2009, Pages 4185–4191
Apparel production is a high value-added industry in the global textile manufacturing chain. Standard size charts are crucial industrial standards for high-tech apparel industries to maintain competitive advantages in knowledge economy era. However, these industries suffering from production management and marketing often find it hard to obtain the accurate standard size charts. In addition to conventional experience approaches, there is an urgent need to develop effective mechanism to find the industrial standards that are the most suitable to their own industries. This study aims to fill the gap by developing a data mining framework based on two-stage cluster approach to generate useful patterns and rules for standard size charts. The results can provide high-tech apparel industries with industrial standards. An empirical study was conducted in an apparel industry to support their manufacturing decision for production management and marketing with various customers’ needs. The results demonstrated the practical viability of this approach. Moreover, since the anthropometric database must be repeatedly updated, standard size charts may also be continuously renewed via application of the proposed data mining framework. By applying the proposed framework for solving industrial problems, these industrial standards will remain continually beneficial for both production planning and reducing inventory costs, while facilitating production management and marketing.
Apparel manufacturing produces products with the highest added value in the global textile manufacturing chain. Standard size charts provide very crucial industrial standards, and play an important role in high-tech apparel industries (Burns and Bryant, 2000 and Jongsuk and Jasper, 1993). For large-scale production, effective production management and marketing are very important factors for apparel manufacturers seeking to reduce cost and increase marketplace competitiveness. Apparel manufacturers seek to produce the best designs that meet standard size charts for fitting customers’ body types and needs (Regan, Kincade, & Sheldon, 1998). Furthermore, standard size charts can correctly predict numbers of items and ratio of sizes to be produced, resulting in accurate inventory control and production planning for facilitating production management and marketing (Chung and Wang, 2006 and Dai, 2004). Due to the lack of up-to-date standard size charts, many manufacturers cannot develop their own size charts for production; as a result, very often the overdue size charts do not fit the customers’ body types, so consumers are forced to choose suitable apparel by trial and error, resulting in enormous inconvenience, not to mention wasted time and money (Burns and Bryant, 2000 and Hsu and Jing, 1999). Owing to current variations in body type, thus, the developments of standard size charts that accurately conform to the body types of people are crucial for improving production management and marketing (Gupta and Gangadhar, 2004 and Laing et al., 1999). Thus, an issue of importance is to have the current standard size charts of the customers’ body types, in order to predict the production and marketplace demand of different sizes of apparel for the apparel industries. Standard size charts originated from the experienced tailors in the late 18th century. Tailors measured the body dimensions of each customer, and then drew and cut patterns. After many original patterns had been accumulated, tailors gradually developed patterns into a system for storing apparel, which could be utilized to make clothes for people with similar body types. For the conventional approach of establishing standard size charts, Emanuel, Alexander, Churchill, and Truett (1959) developed a set of procedures to formulate standard size charts for all body types. According to this approach, people were first classified by body weight into four shape categories. Within the four shape categories, they were subdivided into two body height: tall and short. Thus, eight size groups were classified, and each category had similar body height and weight. The sizing classifications of other countries were also similar, and the classification was based on two or three sizing variables (Winks, 1997). In related studies, Tryfos (1986) suggested an integer programming approach to optimize the number of sizes in order to maximize expected sales. Chen (1998) focused on women to measure anthropometric data for developing standard size charts. McCulloch, Paal, and Ashdown (1998) proposed a non-linear optimization technique to derive a set of standard size charts from anthropometric data. Chang and Shen (1999) applied decision analysis to develop standard size charts for Taiwanese women. Laing et al. (1999) used statistical analysis to develop standard size charts for protective apparel used by New Zealand firemen. Moon and Nam (2003) measured the anthropometric data of Korean women to classify lower trunk figure types. Gupta and Gangadhar (2004) used a statistical method to develop standard size charts for young Indian women, and Chung and Wang (2006) applied statistical analysis to establish standard size charts for Taiwanese students. Human body types can be distinguished by taking various approaches. As an alternative to catering to the individual consumer in classifying apparel sizes, apparel manufacturers limit their production to a few standard sizes. Consumers are offered more limited choices, but manufacturers can avoid production and inventory problems. Therefore, it would be helpful to develop standard size charts, which have the fewest number of sizes to fit the largest number of body types, for the majority of consumers (Chung and Wang, 2006 and McCulloch et al., 1998). Data mining has been successfully applied in many domains, such as health insurance (Chas, Ho, Cho, Lee, & Ji, 2001), biomedicine (Maddour & Elloumi, 2002), human resource management (Min & Emam, 2003), semiconductor manufacturing (Chien, Hsiao, & Wang, 2004), production schedule (Sha & Liu, 2005), knowledge management (Hou & Yang, 2006), education (Chang, 2007) and course planning (Hsia, Shie, & Chen, 2008). However, there is a lack of research in developing industrial standards by using the data mining approach. This study aims to develop a data mining framework for industrial standards to explore useful patterns and rules from anthropometric data. By applying the proposed framework, body types can be classified. These industrial standards can then be developed to facilitate apparel production and marketing. Thus, production management and marketing will be enhanced with these standard size charts. An empirical study for industrial standards in one of the largest apparel company in Taiwan is studied to demonstrate the validity of this approach.
نتیجه گیری انگلیسی
This study applied a data mining framework, using anthropometric data, to develop industrial standards for adult females. By applying the proposed framework, body types can be accurately classified. The newly industrial standards can then be developed. The percentage of females was made available in these standard size charts corresponding to each size group, body types and their distribution. Such standards can predict the proportional quantities necessary for each size, resulting in enhanced production, economic material control and accurate production planning for specific marketplace. The proposed framework combines traditional statistical methods and data mining techniques to explore the anthropometric data. Based on the empirical results, we validate that the proposed framework has practical viability. It helps domain experts find out body types and provides information to understand how to develop industrial standards by the analysis framework. The proposed data mining framework by this study will continually update the database and continually develop the latest industrial standards for facilitating production management and marketing.