Int J Adv Manuf Technol (2006) 30: 166–173 DOI 10.1007/s00170-005-0048-3
O R I G I N A L A RT I C L E
Muh-Cherng Wu . Ying-Fu Lo . Shang-Hwa Hsu
A case-based reasoning approach to generating
new product ideas
Received: 4 November 2004 / Accepted: 17 March 2005 / Published online: 10 November 2005 # Springer-Verlag London Limited 2005
Abstract This paper presents an approach to facilitate the generation of new product ideas. In this approach, a product is represented by a vector, which involves a number of product attributes. These attributes are defined based on the user-centered design (UCD) paradigm. That is, the context of use and interactions with users for the product is described by these attributes. By the product representation, a product database can be established. For a benchmarking product, we use the case-based reasoning (CBR) technique to retrieve those products, which are“sufficiently relevant” to the benchmarking one, from the database. The functions of the retrieved products are the candidate new ideas to be added on the benchmarking product. These retrieved products are screened manually to identify creative and valuable ideas. A prototype system of this approach has been implemented. Experiments show that the proposed approach outperforms the brainstorming technique, which has been widely used in industry.
Keywords User-centered design . Case-based reasoning . New product development . Product attributes
1 Introduction
In this competitive era, the product life cycle is getting shorter and shorter. New products have to be constantly developed and timely placed to market. Enhancing the pro-ductivity of new product development is therefore a very important issue.
The tasks of new product development can be cat-egorized into four stages [1]. The first stage involves the idea exploration of new products; that is, which new func-tions shall be included for a particular product. The second stage is to screen and rank the new ideas, generated in the
first stage, according to their potential market capacity. The third stage is to design the prototypes for the selected ideas. The fourth stage is to perform a marketing test for the new products to get customers’ feedbacks.
Among the above four stages, the first stage—idea exploration—is the most important because the other three stages are essentially to justify the effectiveness of the new product ideas. Therefore, how to create good and new ideas is a critical factor for successfully developing new prod-ucts. In creating new ideas for products, designers often use creative problem solving (CPS) techniques. More than one hundred CPS methods have been developed [2]. Among these CPS techniques, the brainstorming method proposed by Osborn [3] has been most widely used in industry [4].
Using the brainstorming (BS) method for generating new product ideas has three drawbacks. First, the BS meth-od in nature is developed for a team and not suitable for an individual. Second, the performance of the BS method highly depends on the capacities of team members. Third, the BS method facilitates team members to generate new ideas through a free-association mechanism. The mecha-nism is not customer-oriented or user-centered. Therefore, some ideas good for users might be ignored in the BS method.
Kankainen and her team use a user-centered design (UCD) paradigm to generate new product ideas [5]. In their UCD approach, they first analyzed the demands of users. Based on the users’ demands, a “social map” is drawn to illustrate the scenario of using products. Based on the so-cial map, two designers emulate the activities of users and create new product ideas by discussion. The social map approach is essentially a human-based approach. The quan-tity and quality of new product ideas, in a human-based approach, are undoubtedly limited by the capacities of designers.
To overcome the above drawbacks, this research pro-poses a computer-aided technique to create new product ideas. Based on the UCD paradigm, a product is repre-sented by a vector consisting of 87 elements. Each element represents a product attribute, which relates the product to users from a user-centered perspective. By such a product M.-C. Wu (*) . Y.-F. Lo . S.-H. Hsu
Department of Industrial Engineering and Management, National Chiao Tung University,
Hsin-Chu, Taiwan, ROC e-mail: mcwu@cc.nctu.edu.tw Fax: +886-35-720610
representation scheme, a large amount of products are coded in vectors and stored in a database. To create new ideas for an existing product (called benchmarking prod-uct), we use the case-based reasoning (CBR) technique to retrieve other products that are“sufficiently relevant” to the benchmarking product. The functions of the retrieved products are the candidate new ideas to be added on the benchmarking product.
The remainder of this paper is organized as follows. Section 2 reviews the literature on brainstorming (BS) techniques, user-centered deign (UCD), and case-based reasoning (CBR). Section 3introduces the product repre-sentation scheme. Section 4 describes how to create new product ideas by using the proposed approach. Section 5 uses two benchmarking products to compare the perfor-mance of the proposed approach and the BS technique. Concluding remarks are presented in Section6.
2 Relevant literature
This section briefly describes the basic idea of brainstorm-ing (BS) and its applications to product innovation. Ap-plications of user-centered design (UCD) and case-based reasoning (CBR) techniques are also presented.
2.1 Brainstorming
The basic idea of the BS technique is to provide an en-couraging and inspiring environment for facilitating a team of participants to generate new ideas [3]. The encouraging environment is established by temporarily withholding any criticisms on the new ideas raised by the participants. The inspiring mechanism is achieved by encouraging the par-ticipants to modify, extend, or associate the ideas raised by others. Only at the end of a BS process, the new ideas thus generated will be sorted and evaluated.
According to two empirical surveys [6,7], the BS tech-nique, compared to some other idea generation techniques, has received the highest awareness and has been most widely used in industry. For the processes of developing new products, the BS technique is mostly used in the stage of generating new product ideas [6]. A comprehensive review of brainstorming literature has been presented by Isaksen [4].
2.2 User-centered design
User-centered design (UCD) had its origins with the sem-inal work of Norman and Draper [8]. UCD has been defined as“a philosophy based on the needs and interests of the user, with an emphasis on making products usable and understandable” [9]. The basic idea of UCD is to ex-plore the context of use for the design of a new product in order to enhance the product usability [10].
The process of implementing UCD has been developed by much literature [11,12]. The ISO 13407 standard also
provides a framework for applying UCD [13]. According to the standard, the following four UCD activities need to start at the earliest stage of developing a product: (1) un-derstand and specify the context of use, (2) specify the user and organizational requirements, (3) produce design so-lutions, and (4) evaluate designs against requirements.
UCD has been widely used in industry [10,14] and two handbooks have been published [15, 16]. Some example applications of UCD involve the design of modern micro-electronic products [12], the design of a website [17], the design of a user interface of a welding machine [18], the user-interface design of a chemical process [19], the design of software [20], and the design of new product concept [5] as discussed in Section1.
2.3 Case-based reasoning
Case-based reasoning (CBR) is the process of using past cases to solve the new problem case [21–23]. The basic idea of CBR is to retrieve the past cases that are similar enough to the new case. These retrieved past cases and their associated solutions are then used to develop the solution of the new case. Two survey papers of CBR can be referred in [24,25].
A CBR system involves the following three compo-nents: (1) a case representation scheme, (2) a similarity metric, and (3) a case-retrieval mechanism. A case repre-sentation scheme is to model a case by a set of attributes, which are intended to characterize the case for a particular application. A similarity metric is designed to measure the similarity between any two cases that are described by the case-representation scheme. A case-retrieval mechanism is to retrieve the past cases that are similar to the new case, according to a predetermined similarity threshold.
CBR has been applied in a wide variety of design prob-lems [26]. A review paper on the application of CBR to architecture designs and mechanical designs has been pub-lished [27, 28]. In the domain of product design and development, CBR has been applied to retrieve existing product data or product design [29, 30], to manage the assignment of physical structures to the functions of a product [31], to integrate the structure–behavior–function model of products [32], to develop a decision support tool for product design in a concurrent engineering environment [26].
The previous CBR studies relevant to product design focus on the engineering aspect. That is, given a product idea (a functional requirement), they investigated how to apply CBR to effectively and efficiently develop a design for fulfilling the functional requirement. Yet, how to apply CBR to create new and meaningful functional requirements is rarely studied.
3 Product representation
In this research, a product is modeled by a vector consisting of a number of product attributes. Based on the paradigm of
UCD, these product attributes are developed from five dimensions: (1) interface modality, (2) task, (3) physical feature, (4) environment, and (5) user. These five di-mensions are further deployed into 20 groups (or called sub-dimensions) that are subsequently used to yield 87 attributes (Table1).
3.1 Interface modality
The first dimension—interface modality is to identify the medium through which the product interacts with the user. Here, the medium denotes a particular portion of the user’s body. This research classifies the interface modality into three groups: (1) sensory modality, (2) response modality, and (3) positioning location.
Sensory modality denotes the mediums or sensors through which the user receives the messages sent from the product. In modeling the sensory modality, the fol-lowing five human sensors are taken as product attributes: (a) visualization(or sight), (b) audio (or hearing), (c) gus-tatory (or tasting), (d) olfactory (or smelling), and (e) tactile. For example, the sensory modality of a watch is visual and that of a radio is audio. A product may have two or more sensory modality; for example, the sensory modality of television involves audio and visual.
Response modality denotes the mediums through which the user responds to or manipulates the product. The re-sponse modality is composed of seven attributes: (a) eye, (b) mouth, (c) face, (d) neck, (e) hand, (f) foot, and (g) finger. For example, a pair of glasses involves two response modality attributes: eye and face. A saxophone involves three response modality attributes: mouth, hand, and finger. Positioning location denotes the medium through which the user carries the product. The positioning location in-volves six attributes: (a) head, (b) neck–shoulder, (c) chest– waist, (d) back, (e) hand, and (f) foot. For example, the positioning location of a typical watch involves hand, while that of a pocket-watch involves chest-waist. Some products may have no positioning location; for example, a refrigerator is typically not carried by user.
3.2 Task
The second dimension—task is to model the tasks to be performed by the user through using the product. Accord-ing to users’ needs, the tasks are categorized into seven groups: (1) eating, (2) clothing, (3) living, (4) transporta-tion, (5) education/entertainment, (6) working, and (7) health care.
The first group, which describes the tasks relevant to eating, is characterized by the following four processes: (a) preparation, (b) processing, (c) serving, and (d) cleaning. For example, a cooking pot is for the processing of food; a dish is for the serving of food; and a brush is for cleaning purpose. The second group, describing the tasks relevant to clothing, is characterized by three processes: (a) prepara-tion, (b) modificaprepara-tion, and (c) cleaning. For example, a
cloth closet is for the preparation purpose; and a washing machine is for cleaning purpose.
The third group, which denotes the tasks relevant to living, is characterized by the following four processes: (a) cleaning, (b) maintenance, (c) monitoring, and (d) decorat-ing. For example, a hammer is for the purpose of main-tenance; a vacuum cleaner is for cleaning purpose.
The fourth group, which describes the tasks relevant to transportation, is characterized by the five processes: (a) planning, (b) navigating, (c) monitoring, (d) emergency handling, and (e) vehicle. For example, GPS (global po-sitioning system) is used in the tasks of path planning and path navigation. Cell phone is used for the purpose of emergency handling.
The fifth group—education/entertainment is character-ized by the following five processes: (a) searching, (b) understanding/integrating, (c) creating, (d) storing, and (e) distributing. For example, a tape recorder is for the purpose of storing; and an e-mail system is for distributing purpose. The sixth group, which describes the tasks relevant to working, is characterized by five processes: (a) planning, (b) execution, (c) inspection/review, (d) decision, and (e) communication. For example, telephone is for the purpose of communication; and a thermometer is for inspection purpose.
The seventh group—health care is characterized by the following processes: (a) daily maintenance, (b) detection/ diagnosis, (c) remedy, and (d) rehabilitation. For example, soap is for the daily maintenance purpose; and wheelchair is for the rehabilitation purpose.
3.3 Physical feature
The third dimension of the product representation is called physical feature. This dimension is characterized by three features: (1) physical size, (2) movability, and (3) scale-ability. For the feature of physical size, products are clas-sified into five types: (a) mini, (b) small, (c) medium, (d) large, and (e) extra-large. For the feature of movability, products are classified into two types: (a) portable and (b) stationary. For the feature of scalability, products, accord-ing to the extent of change, are classified into three types: (a) small, (b) medium, and (c) large. For example, a note-book has the following attributes: medium in size, portable, and medium level in scalability.
3.4 Environment
The fourth dimension of the product representation de-scribes the environment where the product is used. The dimension is characterized by the following three sub-di-mensions: (1) number of users, (2) indoor/outdoor place, and (3) the harshness of environment.
The sub-dimension“number of users” is classified into two types: (a) single user and (b) a group of users. For example, PDA (personal digital assistant) product is for single user; and television is for a group of users.
The sub-dimension “indoor/outdoorplace” is classified into two types: (a) indoor and (b) outdoor. For example, a walkie-talkie is for both indoor and outdoor places; a printer is for indoor use.
The sub-dimension “the harshness of environment” is characterized by the following six attributes: (a)
illumina-tion, (b) temperature, (c) humidity, (d) noise/vibraillumina-tion, (e) dust, and (f) pressure. For example, an industrial com-puter can be used in a high-temperature and high-noise environment.
Table 1 Shows the vector re-presentation of two products, cell phone and ball pen
ID Dimension Group Attribute Cell phone Ball pen
1 Interface Modality sensory modality visualization (or sight) 0.5 1.0
2 audio/hearing 1.0 0
3 gustatory (or tasting) 0 0
4 olfactory (or smelling) 0 0
5 tactile 0.5 0.5
6 response modality eye 0.5 1.0
7 mouth 1.0 0 8 face 0.5 0.5 9 neck 0.5 0 10 hand 1.0 1.0 11 foot 0 0 12 finger 0.5 0.5
13 positioning location head 0 0.25
14 neck-shoulder 0.25 0
15 chest-waist 1.0 1.0
16 back 0 0
17 hand 0.5 0.5
18 foot 0 0
19 Tasks eating preparation 0 0
20 processing 0 0 21 serving 0 0 22 cleaning 0 0 23 clothing preparation 0 0 24 modification 0 0 25 cleaning 0 0 26 living cleaning 0 0 27 maintenance 0 0 28 monitoring 0 0 29 decorating 0 0 30 transportation planning 0.5 1.0 31 navigating 0 0 32 monitoring 0 0 33 emergency handling 1.0 0 34 vehicle 0 0 35 education/entertainment searching 0.5 0 36 understanding/integrating 0 0 37 creating 0.5 1.0 38 Storing 0.25 0 39 distributing 0.25 0.75 40 working planning 0.25 1.0 41 execution 0.5 0.25 42 inspection/review 0 0.5 43 decision 0.25 0 44 communication 1 0.25
45 health care daily maintenance 0 0
46 detection/diagnosis 0 0
47 remedy 0 0
3.5 Users
The fifth dimension of the product representation is about users. Users are characterized by the following demograph-ic features: (1) gender, (2) age group, (3) working industry, and (4) job position. The feature gender denotes the population of users in terms of gender. The feature age group involves six clusters: (a) infant, (b) child, (c) teen-ager, (d) young adult, (e) adult, and (f) the aged. The feature “working industry” involves five clusters: (a) agriculture, (b) manufacturing, (c) trade, (d) service/education, and (e) government. The feature “job position” involves six clusters: (a) top management, (b) middle management, (c) supervisor, (d) engineer, (e) office clerk, and (f) operator.
Based on the representation scheme discussed above, a product can be modeled by a vector consisting 87 at-tribute values. Each atat-tribute of a product can be valued in five scale points: extreme relevance (1.0), high relevance (0.75), relevance (0.5), low relevance (0.25), no relevance (0). Table 1shows the vector representation of two prod-ucts, cell phone and ball pen.
4 Generation of new product ideas
Based on the proposed product representation scheme, this research applies the CBR (case-based reasoning) technique to create new product ideas. The idea-generation
mech-Table 1 (continued) ID Dimension Group Attribute Cell phone Ball pen
49 Physical Feature physical size mini 0.25 0.25
50 small 0.75 0.75 51 medium 0 0 52 large 0 0 53 extra-large 0 0 54 movability portable 1.0 1.0 55 stationary 0 0 56 scaleability small 0 0.5 57 medium 1.0 0.5 58 large 0 0
59 Environment number of users single user 1.0 1.0
60 group of users 0 0
61 indoor/outdoorplace indoor 0.5 0.5
62 outdoor 0.5 0.5
63 the harshness of environment illumination 0.75 0.75
64 temperature 0.5 0.5
65 humidity 0.5 0.5
66 noise/vibration 0.25 0.5
67 dust 0.5 0.5
68 pressure 0.5 0.5
69 Users gender male 0.5 0.5
70 female 0.5 0.5
71 age group infant 0 0
72 child 0.25 0 73 teenager 0.5 0.25 74 young adult 0.5 0.5 75 adult 0.5 0.5 76 the aged 0.25 0.25 77 workingindustry agriculture 0.5 0.5 78 manufacturing 0.5 0.5 79 trade 0.5 0.5 80 service/education 0.5 0.5 81 government 0.5 0.5
82 job position top management 0.5 0.5
83 middle management 0.5 0.5
84 supervisor 0.5 0.5
85 engineer 0.5 0.5
86 office clerk 0.5 0.5
anism starts with a benchmarking product, an existing product on which new functions are to be added. For the benchmarking product, its high-value attributes (≥0.75) are used as keys to retrieve other products that are“sufficiently relevant” to the benchmarking product. These retrieved products are then screened and subsequently rated manual-ly to determine the quality of new ideas. Here, the quality of new ideas implies how creative and valuable it is to add the main function of a retrieved product to the bench-marking product.
The method for computing the relevance metric between two product vectors is first described. Then, the CBR pro-cedure for retrieving the“sufficiently relevant” products is explained.
4.1 Computing relevance metric
Suppose a product database has been established, which involves a large amount of products and each product is modeled by the proposed representation scheme, a vector form. Let P = {Pi,i = 1,…, K} represent the database, where Pi= [pij] 1≤ j ≤ 87 denotes the i-th product and pijis the value of its j-th attribute. Let B = [bj] 1≤ j ≤ 87 denote the benchmarking product.
Let AR(bj, pij) represent the relevance metric of j-th attribute for product B and Pi. AR(bj, pij) is defined below. AR b j; pij ¼ 1 jbj pijj (1) Let S ¼ jjb j 0:75 represent the set of high-value attributes of product B. Some attributes in S are selected to form a key set, which are used to retrieving “sufficiently relevant” products. Define T ⊂ S as the key set. Let PRTi represent the product relevance metric between products Pi and B, with respect to the set of attributes T. PRTi is defined below, where Wj∈ [0, 1] denotes the weighting factor of j‐th attribute given by users. That is, users can intentionally gives their preferences on the attributes in T.
PRT i ¼ ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi P jT bj AR bj; pij Wj 2 P jT bj Wj 2 v u u u u u t (2) 4.2 Procedure of CBR
The concepts presented above can be used to develop a product innovation retrieval system, called PIRS. The pro-cedure for using the PIRS to generate new product ideas is presented below.
Step 1: Initialization
. Assign R_Set = ∅, A_SET = ∅
. Ask user to determine the benchmarking product B
Step 2: DetermineS ¼ jjb j 0:75, the set of impor-tant attributes for product B.
Step 3: Ask users to select attributes in S to form several key sets Tk(k = 1,…, n) and give Wjfor each attribute in Tk.
Step 4: Retrieve “sufficiently relevant” products from the product database
. Ask users to determine a relevance threshold h ∈ [0, 1]. . Retrieve and place the “sufficiently relevant” products
in R_Set; that is,R Set ¼ Sn
k¼1 PijPR T
i h for i 2 Tk
Step 5: Ask users to select creative and valuable product ideas from R_Set and place them in A_Set.
5 Experiment
To justify the proposed PIRS (product innovation retrieval system), this research performs an empirical study to compare the effectiveness between PIRS and the brain-storming (BS) approach. Two experiments are tested. One experiment uses a cell phone as the benchmarking product and the other uses a ball pen as the benchmarking product. In each experiment, the two approaches aim to find creative and valuable functions to be added to the benchmarking product. The procedures for the experiments are presented below.
Table 2 The results of the two experiments are shown in
Group Cell phone experiment≥4.0 Ball pen experiment≥4.0
B.S. PIRS B.S. PIRS
Q’ty T-score Q’ty T-score Q’ty T-score Q’ty T-score
Group 1 12 49.2 13 53.6 4 16.4 8 33.0 Group 2 11 45.4 14 61.7 5 20.8 9 37.1 Group 3 12 49.3 15 61.4 6 24.7 9 37.7 Group 4 10 41.5 13 45.4 5 20.6 10 41.4 Group 5 9 37.0 14 58.0 6 20.4 9 37.7 average 10.8 44.48 13.8 56.02 5.2 20.58 9 37.38
Table 3 Product ideas for cell phone generated by PIRS
Language Translator/Dictionary Radio and Walkman
Remote Controller for Door/Car
Television
USB Flash Memory Drive Stun Gun for Security
Electronic Pet (Game) Projector
Mosquito prevention set Language Learning Machine
E-Map PDA (personal digital assistant)
Alcohol tester Pedometer
MP3 e-Book
Anti-camera detector Bar-code detector (for
shopping memo)
5.1 Subjects for experiments
Each experiment involves 25 subjects who are sophomores of a university. These subjects are randomly clustered into five groups. In each group, one subject is randomly chosen to use the PIRS approach and the other four subjects are requested to use the BS approach to generate new product ideas for the benchmarking products.
5.2 Product database
A prototype product database involves 500 products is established for the experiments. The product database is coded by Microsoft Access; the user interface is developed by Macromedia Dreamwaver; the retrieval mechanism is coded by ASP (Active Server Page); and Microsoft Internet Explorer is used as the vehicle for web browsing.
5.3 Experiment procedures
The experiment time for each of the two approaches is 1 hour. The BS approach, which has been widely known, follows the procedure described by Osborn [3]. The PIRS approach follows the procedure described in Section 4. Using the benchmarking product ball pen as an example, each step of the PIRS approach is explained below.
Step 1, claims that new functions of a ball pen are to be searched. In step 2, of the 87 product attributes of a ball pen, only 12 attributes are important attributes, which are highlighted in Table1and are with the following IDs {1, 6, 10, 15, 30, 37, 39, 40, 50, 54, 59, 63}. In step 3, subjects can freely form several key sets Tkfrom the 12 attributes. In step 4, the relevance threshold h = 0.5 is defined, and new product ideas are generated by the PIRS system. In step 5, of these new product ideas generated, subjects by their own judgments select the creative and valuable ones.
5.4 Evaluation method
Three experts familiar with new product developments are invited to evaluate the new product ideas generated. In evaluating a new product idea, three criteria—originality, valuableness, and usefulness [33] are used. Each criterion is rated in a five-point scale. The higher the point, the more appreciated is the new idea. The creativity of a new product idea is measured by averaging the points of the three criteria.
5.5 Result comparison
The results of the two experiments are shown in Table2. This table analyzes the new product ideas that are regarded as creative; that is, their creativity values are grater than or equal to 4.0. For the cell phone experiment, the BS ap-proach creates 10.8 creative ideas per group, and the PIRS
approach creates 13.8 creative ideas per group. The asso-ciated t-test, withα=0.05, t-value=−3.005, P-value=0.017, shows that the PIRS approach is statistically better than the BS approach. For the ball pen experiment, the BS approach creates 5.2 creative ideas per group, and the PIRS approach creates 9.0 creative ideas per group. The associated t-test, withα=0.05, t-value=−8.970, P-value=0.0 shows that the PIRS approach is also statistically better than the BS ap-proach. Table3shows some product ideas generated by the PIRS approaches for the cell phone experiment.
6 Concluding remarks
This research presents a cased-based reasoning (CBR) ap-proach to generate new product ideas. A product is modeled by 87 attributes based on the user-centered design (UCD) paradigm. For each attribute, a product is assigned a 0–1 value. The higher the value, the more important is the attribute. For a benchmarking product, some of its important attributes are manually selected as keys for retrieving from database the products that are“sufficiently relevant” to the benchmarking product. The main functions of these re-trieved products are taken as candidate new functions of the benchmarking product. A prototype system of the proposed approach, called PIRS (product innovation retrieval system), has been implemented. Experiments show that the PIRS outperforms the brainstorming approach in generating crea-tive and valuable product ideas.
An extension of this approach is applying the PIRS to enhance the productivity of the brainstorming technique. That is, during a brainstorming process, a BS participant is inspired by other participants as well as by the PIRS sys-tem. Another extension is the development of an idea-propagation mechanism. That is, for the products retrieved, we can select some of them as the benchmarking products to retrieved more new ideas.
References
1. Kuczmarski TD (1992) Managing New Products: The Power of Innovation: Prentice–Hall, pp 33–35
2. Higgins JM (1994) 101 Creative Problem Solving Techniques: The Handbook of New Ideas for Business: New Management Pub. Co.
3. Osborn A (1963) Applied Imagination, Charle Scribners Son, New York
4. Isaksen SG (1998) A review of brainstorming research: six critical issues for inquiry. Monograph #302, Creative Research Unit, Creative Problem Solving Group, Buffalo NY
5. Kankainen A (2003) UCPCD: User-Centered Product Concept Design. Proceedings of the 2003 Conference on Design for User Experience. San Francisco, California, USA, pp1–13 6. Nijssen EJ, Lieshout KFM (1995) Awareness, Use, and
Ef-fectiveness of Models and Methods for New Product Devel-opment. Eur J of Mark 29(10):27–43
7. Geschka H (1983) Creativity techniques in product planning and development: a view from West Germany. R&D Manage-ment 13(3):8–25
8. Norman DA, Draper SW (1986) User-Centered System Design: New Perspectives on Human–Computer Interaction. Lawrence Earlbaum Associates, Hillsdale, NJ
9. Norman DA, (1988) Design of everyday things. New York, NY: Currency Doubleday
10. Vredenburg K, Mao JY, Smith PW, Carey T (2002) A Survey of User-Centered Design Practice. CHI 2002 Conference on Human Factors in Computing Systems, Minnesota, USA, pp471–478
11. Gould JD, Lewis C (1985) Designing for usability: Key Principles and what designers think. Commun ACM 28:300– 311
12. Buurman RD (1997) User-centered design of smart products. Ergonomics 40(10):1159–1169
13. Bevan N (2003) Usability Net Methods for User Centered Design: Human–Computer Interaction: Theory and Practice (part 1), Proceeding of HCI Int. 2003, Lawrence Erlbaum, pp434–438
14. Vredenburg K, Isensee S, Righi C (2002) User-Centered Design: An Integrated Approach, Prentice–Hall
15. Daly–Jones O, Bevan N and Thomas C (1999) INUSE: Handbook of User-Centered Design: Serco Usability Services, Teddington
16. Poulson D, Ashby M, Richardson S (1996) USERfit: A prac-tical handbook on user-centered design for assistive technology. Tide European Commission
17. Corry DM, Frick TW, Hansen L (1997) User-Centered Design and Usability Testing of a Web Site: An illustrative Case Study, 45(4):65–76
18. Burmester M, Beu A, Hackl H, Niederer F (2002) User centered design for a digital welding machine. Behav Inf Technol 21(5):321–326
19. Kontogiannis T, Embrey D (1997) A user-centered design approach for introducing computer-based process information systems. Appl Ergon 28(2):109–119
20. Vredenburg K (2003) Building ease of use into the IBM User experience. IBM Syst J 42(4):517–531
21. Marling C, Sqalli M, Rissland E, Munoz–Avila H, Aha D (2002) Case-Based Reasoning Integrations. AI Mag 23(1): 69–86
22. Aamodt A, Plaza E (1994) Case-based reasoning: foundational issues, methodological variations, and system approaches. AI Commun 7(1):39–59
23. Kolodner JL, Leake DB (1996) A Tutorial Introduction to Case-Based Reasoning: Experiences, Lessons, and Future Directions, ed. Leake DB, 31–66. Menlo park , Calif: AAAIPress 24. Watson I, Marir F (1994) Case-based reasoning: a review. The
Knowl Eng Rev 9(4):327–354
25. Mantaras RL, Plaza E (1997) Case-based reasoning: An overview. AI Commun J 10(1):21–29
26. Belecheanu R, Pawar KS, Barson RJ, Bredehorst B, Weber F (2003) The application of case-based reasoning to decision support in new product development. Integr Manuf 14(1): 36–45
27. Trousse B, Visser W (1993) Use of case-based reasoning techniques for intelligent computer–aided–design systems. In, Proceedings of the IEEE Int Conference on Systems, Man and Cybernetics 1993, 3:pp513–517
28. Maher ML, Garza AGS (1997) Case-Based Reasoning in design. IEEE Expert, March–April:34–41
29. Bilgic T, Fox MS (1996) Constraint-based retrieval of engineering design cases: Context as constraints: Gero J and Sudweeks F (eds) Artificial Intelligence in Design‘96, Kluwer Academic Publishers, pp269–288
30. Maher ML, Balachandran MB, Zhang DM (1995) Case-Based Reasoning in Design, Lawrence Erlbaum, Hillsdale, NJ 31. Nedeß C, Jacob U (1997) A case-based reasoning approach
towards learning from experience connecting design and shop floor. Comput Ind 33:127–37
32. Goel AK, Bhatta SR, Stroulia E (1997) KRITIK: an early case-based design system, in Maher ML and Pu P.(Eds), Issues and Applications of Case- based Reasoning to Design, Lawrence Erlbaum Associates, Hillsdale, NJ, pp87–132
33. Besemer SP, O’Quin K (1986) Analysis of creative products: Refinement and test of a judging instrument. J Creat Behav 20 (2):115–126
