Combining multiple correspondence analysis with association rule mining to conduct user-driven product design of wearable devices
Introduction
In recent years, market growth in smartphones is slow and becoming more and more flattening [25], [26]. Suppose you live in the developed country (i.e. West Europe, North America, and Asia Pacific regions), the future of smart phones is all about “upgrades” because people have already owned at least one by now. In contrast, wearable devices like smart watches, smart glasses, and wristbands are capturing much more eyes than before [24]. One of the most critical reasons is nothing but portability, especially in a hand-free user interface [6]. However, Google decided to terminate selling its smart glasses because of considering security, safety, and invasion of privacy. Although BI intelligence predicts an amount of over 30 billion in wearable devices can be reached before 2018 (see Fig. 1 [29]), a recent marketing survey indicates that almost 70% respondents may not be interested in buying Apple's new iWatch (http://www.reuters.com). In fact, consumers are not willing to purchase a high-priced iWatch because smartphones have already replaced the conventional watches and the low-end digital cameras for many years [30].
Today, a smart watch is well accepted to be an important area to move forwards [19]. In practice, one of the most critical concerns is learning how to lock smart-watch users into the smart-phone platform. It is equivalent to lock people into using an operating system so that when it comes to an upgrade they will not desert your brand. That means, to enlarge a market share, a manufacturer may produce a watch that is compatible with Apple's iOS, Google's Android, and Microsoft's Windows. For instance, Sony, Samsung, and Apple attempt to provide smart watches to lock their existing smartphone users. In contrast, Nike and other sport brands attempt to deliver watches that work with handsets and major in fitness or health features. Meanwhile, Epson (Seiko) and Casio has already developed a variety of sport watches for many years and they can easily switch to this field. Similar to the war in smartphones, the key point to a brand company is creating something magical and delivering its fans something to brag about [21].
In order to hold the crown as a key innovator in the consumer-electronics, only upgrading existing products is absolutely insufficient to inspire dynamically changing consumers. Needless to say, it is tough for brand companies to survive in a globally customized economy although they are always planning and launching new products for acquiring diverse customers [7], [17], [18]. One of the best examples is like Apple's iMac, iPod, iPhone, and iPad series, which have been heralded as an “aesthetic” paradigm in industrial design. Obviously, the front-end aspects like user preferences or user perceptions are not only critical to stimulating product sales, but also influential to constructing a firm's product image [11], [12], [16].
Inspired by Fig. 2, a novel framework is presented to bridge the research gaps between product customization, product differentiation, and product selection [14], [15], [22], [24]. In particular, the latent relationships among user profiles, design attributes, and product alternatives are systematically explored. Without loss of generality, this study focuses on three wearable devices, such as wristband, smart watch, and sport watch. In brief, several critical issues are addressed as follows:
- •
What are the impacts of user profiles on user preferences for wearable devices?
- •
What are the best portfolios of design features to configure various alternatives?
- •
How to incorporate user perceptions of design features into product selection?
Furthermore, managerial insights are offered to assist product planners in planning next-generation wearable devices. The rest of this paper is organized as follows. Section 2 overviews the concepts of product positioning (differentiation) and product recommendation (selection). The proposed framework integrating multiple correspondence analysis (MCA), association rule mining (ARM), with K nearest neighbor (KNN) is presented in Section 3. Section 4 illustrates a real example to develop user-driven wearable devices. Conclusions are drawn in Section 5.
Section snippets
Literature review
New product development (NPD) defined as a process of transforming an identified market opportunity into profitable product(s) for sale, in which a firm could employ it to accomplish the goal of customization, differentiation, and commercialization [5]. In general, the market is full of diverse customers who differ in usage preferences, buying behaviors, demographic profiles, and psychographic backgrounds [11], [17], [27]. Typical variables commonly used for carrying out market segmentation
The proposed techniques
A novel framework shown in Fig. 3 is presented to achieve product differentiation and product selection of three wearable devices, such as wristband, smart watch, and sport watch. In particular, user profiles and user perceptions are incorporated into the entire decision-making process. For convenience, the proposed framework is operated as follows:
- •
Initially, multiple correspondence analysis (MCA) is used to capture user profiles and user perceptions to form a basis of product differentiation,
- •
Product design and recommendation for wearable devices
In recent years, the popularity of smart phones has substantially changed our society because of powerful capability in processing multimedia information and acting as a social-media carrier [3], [25]. Although some experts argue that the social-media technology may detract from our face-to-face interactions with friends or colleagues, the wearable technology actually enhances our day-to-day lives [19]. Sony Inc. designed a wearable device which is compatible with most of the Android phones for
Concluding remarks
Today, to survive in a wide-spectrum of “buyer-dominated” market, industrial managers spend most of their time to make decisions on segmenting the marketplace, targeting prospective customers, and positioning attractive alternatives. In practice, a systematic approach to incorporate user profiles and user perceptions into the entire process of product recommendation is of importance to help product managers capture fast-changing market trends and diverse user requirements. Different from the
References (30)
- et al.
Recommender systems survey
Knowl.-Based Syst.
(2013) - et al.
Smart phone for mobile commerce
Comput. Stand. Interfaces
(2009) - et al.
Combining fuzzy AHP with MDS in identifying the preference similarity of alternatives
Appl. Soft Comput.
(2008) - et al.
Mass customization: a method for market segmentation and choice menu design
Int. J. Prod. Econ.
(2008) - et al.
Product portfolio identification based on association rule mining
Comput. Aided Des.
(2005) - et al.
Customer segmentation and strategy development based on customer lifetime value
A case study, Expert Systems with Applications 31
(2006) - et al.
Integrating AHP and data mining for product recommendation based on customer lifetime value
Inf. Manag.
(2005) - et al.
Hybrid approaches to product recommendation based on customer lifetime value and purchase preferences
J. Syst. Softw.
(2005) - et al.
Determining optimal levels of engineering characteristics in quality function deployment under multi-segment market
Comput. Ind. Eng.
(2010) - et al.
Integrating conjoint analysis with quality function deployment to carry out customer-driven concept development for ultrabooks
Comput. Stand. Interfaces
(2013)
Using the theory of inventive problem solving to brainstorm innovative ideas for assessing varieties of phone-cameras
Comput. Ind. Eng.
Effective personalized recommendation based on time-framed navigation clustering and association mining
Expert Syst. Appl.
Mining association rules between sets of items in large databases
New Products Management, McGraw-Hill, 9th, New York
Smartwatch Battle: What an Apple, Google and Samsung Smartwatch Might Mean for You Read More
Cited by (11)
Decision support system for safety improvement: An approach using multiple correspondence analysis, t-SNE algorithm and K-means clustering
2019, Computers and Industrial EngineeringCitation Excerpt :Moreover, in MCA association of variable categories can be visualized like in principal component analysis and association rules can be framed. Applications of MCA can be seen in product design (Wang & Nien, 2016), quality (Chatzipetrou & Moschidis, 2016), healthcare (Moritz et al., 2009; Costa, Santos, Cunha, Cotter, & Sousa, 2013), road safety (Das & Sun, 2016; Das et al., 2018) etc. Results obtained in MCA can be further mapped in lower dimensions using t-SNE and segregated using K-means into optimal clusters to find meaningful association among categories.
Incorporating data analytics into design science to predict user intentions to adopt smart TV with consideration of product features
2018, Computer Standards and InterfacesCitation Excerpt :Without considering users’ innate characteristics (product involvement, knowledge or variety seeking) and monetary factors (personal income or affordable price), this study considers the impacts of perceived benefits on user intentions [2,24,35]. To do so, this paper presents a novel framework that incorporates data analytics into design science and demonstrates the following merits [17,40]: Technical EFs and ergonomic GFs are defined to characterize two constructs of smart TV: perceived usefulness (PU) and perceived ease of use (PEOU),
A Detailed Observation on Association Rule Mining
2020, Proceedings of the 2nd International Conference on Inventive Research in Computing Applications, ICIRCA 2020Using the fuzzy weighted association rule mining approach to develop a customer satisfaction product form
2020, Journal of Intelligent and Fuzzy Systems