- Email: [email protected]
Implementing a community web site: a scenario-based methodology
Information and Software Technology 46 (2004) 17–33 www.elsevier.com/locate/infsof
Implementing a community web site: a scenario-based methodology Choongseok Leea, Woojong Suhb, Heeseok Leea,* a Corporate Information System Laboratory, Department of Management Information Systems, Graduate School of Management, Korea Advanced Institute of Science and Technology, 207-43, Cheongryangri-dong, Dongdaemun-gu, Seoul 130-012, South Korea b Division of Business Administration, College of Business Administration, Inha University, 253, Yonghyeun-dong, Nam-gu, Incheon 402-751, South Korea
Received 12 February 2003; revised 15 April 2003; accepted 29 April 2003
Abstract The use of the Internet to facilitate commerce promises vast benefits. An important challenge in the age of Internet business is to fine-tune the Internet business system with customers. For this alignment, this paper proposes a scenario-based object-oriented methodology for developing a community web site. The methodology consists of five phases such as customer analysis, value analysis, web design, implementation design, and construction. Scenarios are used to analyze customers’ needs in a natural fashion. A meta-data scheme is proposed for supporting our methodology. A real-life community web is illustrated to demonstrate the usefulness of the methodology. Scenario thinking can help companies use the Internet to buttress and extend their values. q 2003 Elsevier B.V. All rights reserved. Keywords: Web site development; Scenario-based methodology; Meta-data
1. Introduction The use of the Internet to facilitate commerce among companies promises vast benefits. Many companies have attempted to develop Internet business systems for virtual companies, markets, or communities [25,35]. Internet businesses can enable commercial exchanges that cross physical, temporal, cultural, and legal boundaries on a scale that was technically complicated [33]; it can help companies strengthen the links between customers and suppliers [18,30]. As customers gain control over the design of products or services, competition takes on a whole new shape. Because customers can compare products and services with a variety of rich information, their role shifts from passive recipient to active designer [5]. As a result, companies tend to be more customer-oriented [10,16]. Companies spend a lot of time and money fine-tuning their offerings with customers. Business web systems need to be in perfect alignment with ever-changing customers’ needs. Companies should first analyze customers’ scenarios, and then respond to their needs [49,60]. Typically, Internet business systems are implemented by the use of web technologies. Web technologies employ * Corresponding author. Tel.: þ 82-2-958-3615; fax: þ82-2-958-3604. E-mail address: [email protected] (H. Lee). 0950-5849/03/$ - see front matter q 2003 Elsevier B.V. All rights reserved. doi:10.1016/S0950-5849(03)00094-6
hypermedia as a major technical principle; Internet business is one of the most popular applications of hypermedia [46]. This implementation requires several steps such as information modeling for customers, navigation design to help customers find information, user interface design for web page layout, and construction. Some implementation features such as navigation, contents, and user interfaces are not common in conventional software development [29]. Building business web sites is challenging because of the lack of developers’ experiences, poorly defined processes, and unrealistic schedules [6,52]. To solve these problems, a Web Engineering approach has been proposed. Web Engineering employs methods from several areas such as software engineering (SE), human– computer interaction (HCI), project management, usability engineering, testing, and simulation [24,25]. Internet business development methodology can be investigated from these perspectives. A variety of business web development methodologies have been proposed. They can be categorized into formal and informal [17]. The informal methodologies do not specify detailed activities for each phase. These informal methodologies are flexible for formulating business strategies [61]. Formal methodologies describe not only the development activities but also linkages of inputs
18
C. Lee et al. / Information and Software Technology 46 (2004) 17–33
Table 1 Classification of Internet business development methodologies Classification Researcher (Methodology) Informal
Artz [1], Abels et al. [3,4], Chen and Heath [15], Standing (ICDM) [61]
Formal
E-R Model Based
OO ModelBased
Garzotto et al. (HDM) [23], Isakowitz et al. (RMM) [32,33], Takahashi and Liang [63], Atzeni et al. [2], Lee et al. (VHDM) [43], Fraternali and Paolini (HDM-Lite) [22] Lange (EORM) [40–42], Schwabe and Rossi (OOHDM) [57,58], Bichler and Nusser (W3DT) [8,9], Lee et al. (SOHDM) [45], Troyer and Leune (WSDM) [67]
ICDM: Internet Commerce Development Methodology; HDM: Hypermedia Design Method; RMM: Relationship Management Model; VHDM: View based Hypermedia Design Methodology; EORM: Enhanced Object Relationship Model; OOHDM: Object Oriented Hypermedia Design Model; W3DT: World Wide Web Design Technique; SOHDM: Scenariobased Object-oriented Hypermedia Design Methodology; WSDM: Web Site Design Method.
and outputs. Most formal methodologies adopt one of the two data modeling techniques such as Entity-Relationship (E-R) model and Object-Oriented (OO) model. The OO model provides semantic richness, reusability, and flexibility [65,69]. Interestingly, WHDM [46] allows designers to choose between the two data models. The classification of Internet business development methodologies can be summarized as shown in Table 1. HDM, EORM, RMM, and OOHDM were originally proposed for hypertext or hypermedia applications. Most past methodologies focus on the analysis of business information and its relationships based on a data model; they fail to see how this analysis fits into the real lives of their customers. Artz [1], Troyer and Leune [67], Abel et al. [3,4], Chen and Heath [15], Standing [61], and Troyer and Leune [67] emphasized the importance of the analysis of customers’ needs. However, they do not provide systematic and robust steps. As customers’ needs change continuously, reflecting them is critical for sustaining competitive advantages. Most corporate web sites fall short of managers’ high expectations because of a fundamental mismatch. The destination business web simply does not suit the needs of most companies or their customers. A methodology is required to support design activities ranging from the analysis of customers’ needs to system implementation. This paper attempts to sharpen scenario-based techniques SOHDM [45] by focusing on aligning customers’ needs with implementation details. Scenario thinking can help companies use the Internet to extend their customers’
values. The outline of the paper is as follows. The next section provides an overview of the methodology. Section 3 describes the methodology in further detail using a reallife case. Section 4 describes a meta-data scheme supporting our methodology. Section 5 compares our methodology with others. The paper concludes in Section 6.
2. Methodology The architecture of our methodology is depicted in Fig. 1. It consists of five phases: customer analysis, value analysis, web design, implementation design, and construction. Customer analysis, value analysis, and web design are independent of any specific technical platforms and details, while implementation design and construction depend on them. For the sake of clear presentation, the feedback among phases is not depicted. However, feedback is of importance in our methodology because iterative design is essential for Internet business system development [15,25,55]. The customer analysis phase analyzes customers’ needs. Customers are categorized according to their shared needs. Customer groups and the corresponding needs are summarized in the customer need analysis table; their correspondence is analyzed by the use of a correspondence analysis technique. The value analysis phase models value activities. Customers’ requirements are explored in the form of events. In the web design phase, the Internet business to be implemented is logically designed. This phase consists of five sub-phases: scenario design, object modeling, view design, navigation design, and page design. In the scenario design, scenarios are designed in the form of natural language to describe customers’ interactions with an Internet business system from the customers’ perspective. These scenarios lead to object modeling. In the view design, information units that customers want to find are designed. The navigation design subphase designs navigational paths for customers to access information conveniently. Finally, in the page design, users’ information windows and the flow from one page to another are defined according to the views and their navigational paths. During the implementation design, page specifications are determined and the object model is transformed into the physical database schema. Finally, the construction phase implements a physically running Internet business system.
3. Methodology details Here, each phase of our methodology is described in further detail by the use of a real-life case. This Internet business case will be referred to as “I site.” I site (www.iloveschool.co.kr), an Internet alumni association site, has provided a school-based community
C. Lee et al. / Information and Software Technology 46 (2004) 17–33
19
Fig. 1. Methodology architecture.
service since October 1999. Its members can find any school friend who is also a member by locating the name of his or her school. They can leave messages via bulletin board and exchange e-mail. It has shown a dramatic network effect because its service becomes valuable as more people use it. Drawn in by a rapid expansion of the Internet population, I site has turned into one of the top Korean community web sites. In market surveys conducted by major private research institutes such as Samsung Economic Research Institute (www.koreaeconomy.org) and Yahoo! Korea (kr.yahoo.com), the site was picked as one of the best hit products in 2000. About 10 million customers are using the site and about 11,000 (almost 100 percent) Korean universities and schools are registered. I site has 400 million page views per month and its Alexa (www.alexa.com) ranking is the 47th in the world and the seventh in Korea as of April 2001. Its main service areas include schoolmate search, school Bulletin
Board System (BBS), school club service, and club service. I site laid out plans to become a serious player in the Internet community market. Customer loyalty has become more important than customer acquisition. Accordingly, our methodology was employed to help developers improve I site. We will illustrate this application in details as follows. 3.1. Customer analysis Customer analysis includes customer need analysis and correspondence analysis. This phase can find creative ways to fulfill supplementary needs that may not involve the current company’s offerings by building a detailed understanding of common customer needs. The customer need analysis sub-phase analyzes customers on the basis of customers’ data. First, data may be collected by observation, survey, or experimentation.
20
C. Lee et al. / Information and Software Technology 46 (2004) 17–33
Customers may be categorized according to their geographic, demographic, psychographic, or behavioral variables. Because behavioral variables are based on actual customer behavior toward products or service, they are more likely to be closely related to Internet business than other variables [38]. Our methodology adopts behavioral variables. In the case of I site, customers are categorized into five groups according to their roles: school club manager, club manager, school club member, club member, and member. Communities can be categorized into school club and club. A school club is a community that belongs to a school. An example of school club is “soccer club in K school.” Only the alumni of “K school” can join this school club. However, a club is not related to a particular school. To join a school club or club is not member’s obligation. Accordingly, there may be members who join neither of these two clubs. A school club member is a customer who joins the school club, and a school club manager manages this school club. A club manager manages the club; a club member joins the club. To analyze customers’ needs, we looked at 19,259 suggestions, which were made in the BBS by 11,240 volunteer customers between December 2000 and April 2001. I site ran the BBS to listen to its customers. Because 18,956 suggestions are related to the existing services, remaining 303 suggestions (from 291 volunteer customers) are analyzed. Customers’ needs are grouped into nine categories according to their service characteristics, as shown in Table 2. For example, the schoolmate segmentation is most frequently suggested. This segmentation can divide schoolmates into various groups by their majors and graduation dates. Different customer groups may have different needs. It is important to analyze the correspondence between customer groups and their needs. However, the previous customer need analysis provides the frequencies only. Typically, the correspondence analysis [27] is adopted for finding this correspondence. Third quantification method, reciprocal
averaging, appropriate scoring, homogeneity analysis, dual scaling, and scalogram are all synonyms of correspondence analysis. The correspondence analysis investigates simple two-way or multi-way tables containing some measure of correspondence between the rows and columns. It is similar to factor analysis techniques in statistics; it can explore the structure of categorical variables included in the table. For more information, refer to Refs. [27,28]. The analysis results in the chi-square value, 93.2 (degree of freedom ¼ 32). This value confirms the correspondence. Fig. 2 shows the customer correspondence diagram drawn in a Euclidian space. In this diagram, the horizontal axis and vertical axis describe row and column coordinates, respectively. For example, the needs for school club members and members include teacher community, schoolmate segmentation, and post-graduate community. 3.2. Value analysis Business webs are inventing new value propositions. The objective of the value analysis is to identify the value for customers’ needs and then describe the activities between customers and web system. First, customer values are explored on the basis of customers’ activities with the system. The value derivation sub-phase employs a value table to analyze customers’ values, value activities, and implementation priorities. Customer value is defined as a bundle of benefits customers expect from a given product or service [39]. Customer value can be measured from financial, perceptual, and behavioral perspectives [68]. Value activity is a customers’ distinct event or integrated set of events to create customer value. Because customers may not be able to understand all of the new opportunities by new technologies or environments, implementation priorities may be decided on the basis of technical feasibility, business capabilities, and strategic importance. Table 3 shows the result of value derivation. For example, school club members and members highly value
Table 2 Customer need analysis table Customer Group Customer Need
School Club Manager
Club Manager
School Club Member
Club Member
Member
Total
Teacher Community Schoolmate Segmentation Post-graduate Community Album Service Schoolmate e-mail Wireless Internet Army Community Off-line Support Cooperative Purchase Total
14 28 12 20 10 11 1 3 0 99
2 3 2 0 1 5 1 1 0 15
16 36 19 5 3 0 4 0 1 84
4 4 3 0 0 0 3 0 0 14
12 42 18 5 4 0 4 3 3 91
48 113 54 30 18 16 13 7 4 303
C. Lee et al. / Information and Software Technology 46 (2004) 17–33
21
Fig. 2. Customer correspondence diagram.
“Friendliness among teachers and students” in the teacher community according to customers’ perceptual perspective. “Search for teachers” and “Join teacher communities” are activities for achieving this value. The implementation priorities of the teacher community and the album service are fairly high according to I site’s current capabilities and strategic importance. Next, it is important to investigate the activities between customers and the system to be implemented. At the same time, the scope of the system is determined. A value activity
diagram can capture these value activities. The diagram is based on events, customer groups, and sub-systems. Events are identified for each customer group. An event is a trigger that starts the system [12]. In the value activity diagram, the needs of each customer group are depicted for the corresponding sub-system. Each customer group triggers events for this sub-system via value activities. The value activity diagram is similar to a use case diagram in standard Unified Modeling Language (UML) notations [21]. Use cases describe the actions the system
Table 3 Value table Customer Group
Customer Need
Customer Value
Value Activity
Implementation Priority
Member
Teacher Community
Friendliness among teachers and students
Search for teachers Join teacher communities
WWW
School Club Member
Schoolmate Segmentation
Friendliness among schoolmates
Search for schoolmate segments Join schoolmate segments
WW
Post-graduate Community
Friendliness among post-graduate schoolmates
W
Album Service
Sharing of photos
Search through post-graduate communities Join post-graduate communities Manage on-line albums
WWW
Schoolmate e-mail
Communication with all schoolmates Ease of management
Mail for all schoolmates
W
Manage community using wireless Internet Search through army communities Join army communities
WW
School Club Manager
Club Manager
Wireless Internet
Club member
Army Community
Friendliness among army colleagues
W
22
C. Lee et al. / Information and Software Technology 46 (2004) 17–33
Fig. 3. Value activity diagram.
takes to deliver to the actor [34]. The emphasis of our value activity diagram is on finding detailed event activities between customer groups and the business web system. The value activity diagram is thus simpler to depict a variety of activities between customer groups and system. Fig. 3 shows the value activity diagram in our case. For the sake of simple presentation, it highlights two items for customer needs with top implementation priority (“Teacher Community” and “Album Service”). These two items are depicted as subsystems. Although “Search for teachers” activity is directly transformed into the “Search for teachers” event, this activity can be subdivided into several activities such as “Search for teachers by teacher name” and “Search for teachers by school name” to provide alternative search facilities for the members and the school club members. “Join teacher communities” activity is transformed into seven events. 3.3. Web design The web design phase consists of five sub-phases: scenario design, object modeling, view design, navigation design, and page design. Our methodology employs scenarios to identify business system requirements at the earliest opportunity. Task analysis may be adopted for capturing user activities in the human computer interaction studies [19,55]. A scenario analysis is less formal than task analysis [55]. Scenarios can be categorized into three types: descriptive, exploratory, and explanatory [54]. Descriptive scenarios capture requirements. Exploratory scenarios are
used to find possible solutions for satisfying given system requirements through the exploration and evaluation of alternatives. Explanatory scenarios are intended to support the explanation and argumentation of drawbacks, inefficiencies, or system performance problems. Our scenarios are descriptive to capture customers’ requirements. A scenario is a description of customers’ interactions with a sub-system from the perspective of customers. Although it is difficult to infer object classes or behaviors, our scenarios can better describe customers’ interactions. Scenarios correspond to key business processes and thus can capture customers’ requirements in a natural fashion. In the scenario design sub-phase, the events drawn from the value activity diagram are described in the form of natural language. The natural language can enhance ease of use and understandability [49,53]. In the I site case, 12 scenarios are generated from the 12 events. For example, Fig. 4 through 6 show three scenarios from three events such as “Search for teachers,” “Search for photos,” and “Post photo.” Next, we need to model a variety of information the web provides for customers. An object-oriented modeling technique has several advantages (comprehensiveness, understandability, changeability, adaptability, and reusability) [12,31]. Our methodology adopts the object-oriented modeling technique to inherit these advantages. Notations for our object modeling technique are depicted in Fig. 7. An object includes its name, attributes, and responsibilities. Attributes contain the properties of the object. Responsibilities are behavioral properties of
C. Lee et al. / Information and Software Technology 46 (2004) 17–33
Fig. 4. “Search for teachers” scenario.
Fig. 5. “Search for photos” scenario.
Fig. 6. “Post photo” scenario.
the object [12,69]. Among objects, four types of object relationships are described: superclass/subclass, association, collaboration, and component. Any subclass can inherit or use the attributes and responsibilities of its superclass. The association relationship describes a linkage among objects with relationship properties: one-to-one (1:1), one-to-many (1:N), or many-to-many (N:N) [7]. The collaboration relationship describes a dynamic reference among objects through responsibilities [48]. The object that can be used as an attribute in other objects is described by the use of the component relationship. Objects are derived from scenarios. Customer groups and information used in a subsystem can be candidates for
Fig. 7. Notations for the object model.
23
24
C. Lee et al. / Information and Software Technology 46 (2004) 17–33
Fig. 8. View specification.
objects. Additional attributes and responsibilities of objects and relationships among objects can be obtained. For example, in case of the “Search for teachers” scenario, objects such as “Member,” “School Club Member,” and “School Club Manager” are generated. The resulting object model is depicted in Appendix A. The view design subphase reorganizes information contents in the form of navigational units. A view is the information unit that customers want to find. These views
may be referred to as OO (Object-Oriented) views. OO views [37] are designed on the basis of responsibilities, attributes, and relationships in the object model. From our object model (eight objects), five views are generated as shown in Fig. 8. Views are categorized into three types: base view, association view, and collaboration view. (1) a base view is generated from a single object. It is a sub-set of responsibilities and attributes in an object. For example, in
C. Lee et al. / Information and Software Technology 46 (2004) 17–33
Fig. 8, “Teacher Number by School” is a base view, (2) an association view is extracted from the objects that have association relationship(s). It is a subset of responsibilities and attributes in these objects. “Teacher List by School” and “Teacher Profile” are association views, (3) a collaboration view is generated from the objects having collaboration relationships. It is used to change information. In the case of collaboration view, the responsibilities required for collaboration are of importance. “School Club Album Update” is a collaboration view extracted from the collaboration relationship between “School Club Album” and “School Club.” Collaboration views are rarely generated in contentsoriented Internet business systems, like newspaper, since information is less dynamically updated according to customers’ interaction. These three types of OO views can be used in any Internet business system. The navigation design subphase builds navigational links between nodes. The OO view and Access Structure Node (ASN) are adopted for navigational units. The ASN differs from the OO view in that ASN contains access paths to OO views. The OO view contains actual information that customers want to obtain. These OO views and ASNs correspond to nodes. A link denotes the relationship between the source node and the destination node. These source and destination nodes may be OO views or ASNs. Web pages are implemented on the basis of these OO views and ASNs in the subsequent page design subphase. From the scenarios, ASNs are found and then navigational links are determined. First, an event that begins with a scenario can become an event ASN. An event ASN starts navigation. For example, “Search for teachers,” “Search for photos,” and “Enroll photo” events become “Search for teachers ASN,” “Search for photos ASN,” and “Enroll photo ASN,” respectively. Second, a selection ASN can be generated when a customer should select the next activities.
25
For example, from the “Search for photos” scenario, “Condition for search for photos ASN” is built. Third, an input ASN can be generated when a customer needs to input contents. From the “Enroll photo” scenario, “Input photo ASN” is built. Navigation specifications for I site are depicted in Fig. 9. Linking ASNs and views together on the basis of scenario flows results in the navigation path. For example, according to the “Search for teachers” scenario, “Search for teachers ASN,” “Teacher Number by School View,” “Teacher List by School View,” “Profile, memo, e-mail ASN,” and “Teacher Profile View” are linked sequentially. The page design subphase specifies web pages. A page is a window having information and a navigational guide. Web pages should be designed for customers to obtain necessary information in a convenient fashion. A page specification may be composed of many views and ASNs. Pages are designed according to the following guidelines. (1) a view can be a page, (2) an input ASN and a selection ASN can be an individual page, (3) when an ASN follows a view, a page can include both this ASN and the following view. Fourth, an event ASN can be used with an anchor in other pages. Pages are specified by organized anchors, OO views, and additional description details (e.g. embedded components, text, images, sounds, etc.). Appendix B shows the page specification entitled “My Teacher.” 3.4. Implementation design The implementation design phase generates: (1) physical database schema and (2) user interface (UI) specification for construction. An Internet business system can be developed under a variety of system environments. Implementation decisions are made according to the choice of technology. These technologies include DBMSs such as Relational
Fig. 9. Navigation specifications.
26
C. Lee et al. / Information and Software Technology 46 (2004) 17–33
Fig. 10. User interface component.
DBMS (RDBMS) or OODBMS, and development tools such as CGI, HTML, JAVA, Shockwave, Flash, DHTML, XML, and SGML. If OODBMS is used, our object model can be transformed directly. However, in many real-life cases, RDBMSs are most popular. Therefore, the transformation of the object model into relational schema is required. Here, designers need guidelines for this transformation. These transformation guidelines can be found in Refs. [11,20,45]. In addition to these guidelines, more views, Structured Query Language (SQL) stored procedures, or SQL triggers are required for the transformation of collaboration relationships. An SQL stored procedure is a user-defined program module that is stored at the database server and can be invoked by client applications. An SQL trigger is usually in the form of SQL stored procedure and is automatically invoked by the data-related event. It is typically used to perform tasks related to changes in a table [50]. For
example, the collaboration relationship between “School Club Album” and “School Club” can be transformed into a trigger as depicted in Appendix C. Finally, our object model results in ten tables. For actual construction, the page specification should be extended to incorporate data location, interface component choice, and component properties. The user interface design sub-phase maps the page specification to user interface components like choice box, list box, button, check box, and scroll bar, which are supported in various implementation tools. Fig. 10 summarizes the notation for user interface components and transitions. A caption is a description of components. An action is an interactive procedure. Items are sets of data included in the user interface component. Two notations are proposed for transitions. The key-based transition refers to the current state of the component while the simple transition does not. These user interface components may be determined according to the access structure in the navigational design. All anchors may be transformed into a set of HTML anchors. The user interface specification for “My Teacher” page specification is illustrated in Appendix D. 3.5. Construction The construction phase implements a physically running Internet business system. All of the outputs during the implementation design phase should be mapped to physical components. For example, Fig. 11 shows the “My Teacher” screen that is constructed using the UI specification. In this screen, customers can execute the “Search for teachers” scenario. Our design results in seven newly implemented pages.
Fig. 11. “My Teacher” screen.
C. Lee et al. / Information and Software Technology 46 (2004) 17–33
27
Fig. 12. A meta-data schema for development system.
4. A meta-data schema To improve business webs according to customers’ needs continuously, reuse of documents is of particular importance. A variety of models and diagrams can be drawn via various drawing facilities. Object model and views can be automatically transformed to physical database schema.
Furthermore, it can provide simulated web pages. Such simulation can enhance navigational capability before construction. Our methodology is likely to be successful when it is consistently used within an organization. A development support system can strengthen this usability. A system entitled eBizBench can support our methodology [44].
28
Criteria
Classification Informal methodology
Formal methodology E-R Model Based
OO Model Based
Isakowitz Takahashi and Atzeni et al. [32,33] Liang [63] et al. [2] RMM
Fraternali and Paolini [22] HDM-Lite
Schwabe and Rossi [57]; Schwabe et al. [58] OOHDM
Bichler and Nusser [8,9] W3DT
Troyer and Leune [67]; Casteleyn and Troyer [14] WSDM
This Methodology
User User Involvement Involvement
N/A
N/A
N/A
N/A
N/A
N/A
Audience Class Hierarchy
Correspondence Analysis
Objectives User Criteria of Web Applications
Use Case
GRSs
N/A
Scenario
N/A
N/A
N/A
N/A
Audience Class Hierarchy
Scenario
Source of Navigation Design
N/A
N/A
N/A
E-R E-R Relationship Relationship
E-R Relationship
E-R Relationship
OO Relationship
OO Relationship
Navigation Track
Scenario and OO Relationship
Focus
Development Development
Development and Maintenance
Development
Development and Maintenance
Artz [1]
Abels et al. [3,4]
Chen and Heath [15]
Customer Analysis
Target Audience
Users’ Task-related Information and Use Behavior
System Requirement Analysis
Structure, Linkage, and Search
Standing [61] ICDM
Development Development
Development Development Development and Maintenance
Development Development
C. Lee et al. / Information and Software Technology 46 (2004) 17–33
Table 4 Methodology comparison
Development Supporting System Methodology Phases
N/A
N/A
Problem Statement
Information Gathering
Basic System Strategy Analysis and Design Architecture MetaDecision Development Strategy Build Basic Component Functionality Strategy
Constraining Development Requirements Conceptual Model
Evaluation
N/A
Derived Implementation Deployments Component Requirements Strategy Detailed Analysis Detailed Design Development
RMCase
WebArchitect N/A
ER Design
ER Analysis
Slice Design Scenario Analysis Navigation Design
Architecture Design
Conversion Protocol Design UI Screen Design
Attribute Definition
Logical Construction Functional Requirements Semi-Physical Run-time Maintenance Architecture Behavior Design Design Construction and Testing Implementation Evolution
AutoWeb
OOHDM-Web WebDesigner
Database Conceptualize Conceptual Conceptual Design Design Database Generate Navigational Logical Design Database Design Implement Hypertext and Deploy Conceptual Design Hypertext Logical Design
Abstract UI Design
N/A
eBizBench
Customer Analysis
Information Structuring
Mission Statement Specification User Classification
Navigational Design
User Class Description
Requirement Analysis
Implementation Organizational Object Design Modeling
Presentation Design
Interface Design
Hypertext to DB Mapping
Implementation Implementation Design Introduction and Maintenance
Navigational Design
Implementation
Value Analysis Web Design
Implementation Design Construction
C. Lee et al. / Information and Software Technology 46 (2004) 17–33
N/A
29
30
C. Lee et al. / Information and Software Technology 46 (2004) 17–33
This design support system requires a sound understanding of the way in which data (inputs or outputs) should be actually managed. Meta-data can provide a foundation for managing this data effectively. Meta-data is generally known as data about data or information about information [62,64]. The meta-data for eBizBench is depicted as shown in Fig. 12. This meta-data includes 29 entities. The meta-data consists of five parts: customer analysis, value analysis, web design, implementation design, and construction. Clearly, each part of this scheme corresponds to each phase of our methodology. This meta-data can sharpen the reusability and interoperability of the resulting development support system.
5. Methodology comparison This section compares our methodology with other Internet business implementation methodologies. For the comparison, several criteria are of interest: customer analysis, system requirement analysis, source of navigation design, development supporting system, focus, and methodology phases. This comparison is highlighted in Table 4. Customer analysis is a starting point for any system development [56]. Artz [1] places an emphasis on the analysis of customer groups for the top-down development of corporate Web applications. Abels et al. [3,4] identify the criteria that can influence a particular use of the web. They propose that web site should be designed for the user and user input be solicited throughout the process. Chen and Heath [15] and Standing [61] explore user involvement in the evolutionary design process. Troyer and Leune [67] propose a user-modeling phase for user-centered Internet business system development. Our methodology includes systematic methods for customer analysis like the correspondence analysis. It helps developers capture customers’ needs at an early stage and improve the system according to their needs. To implement a usable system, system requirement analysis is important [47]. Artz [1] suggests the exploration of the objectives of a web application. Abels et al. [3,4] propose user criteria that consist of use, content, structure, linkage, search, and appearance. These criteria can be used to provide specific suggestions for the Internet business system design. Standing [61] uses Group Requirements Sessions (GRSs) to obtain the detailed requirements with users’ involvement. GRSs can lead to a compromised solution by striking balances between heterogeneous stakeholders. Takahashi and Liang [63] use a scenario for the dynamic analysis. Their scenario does not describe interactions between the system and its customers. Lee et al. [45] adopt a Scenario Activity Chart (SAC) to describe activities between users and Internet business system in a graphical fashion. Similarly, our methodology adopts scenarios that can describe processes between the business
web and its customers in a natural fashion. It can enhance the ease of use and understandability. The success of an Internet business system is associated with navigation [59,66,70]. Most of the proposed methodologies use the relationships in data model as a major source of navigation design. For example, Abel et al. [3,4] use information derived from user criteria. Our methodology uses scenarios and relationships in object-oriented models to capture navigational requirements. The use of scenarios is likely to better reflect customers’ navigational requirements. In Internet businesses, another important challenge is to reflect on customers’ needs. Bichler and Nusser [8,9] and Atzeni et al. [2] propose methodologies for a new development as well as maintenance. Standing [61] emphasizes the continual evolution of Internet business systems from a managerial perspective. Similarly, our methodology can improve Internet business systems by monitoring customers’ needs continuously. A critical success factor for wider use of design methodologies is the availability of software tools [6,36]. RMCase, OOHDM-Web, WebArchitect, AutoWeb, and WebDesigner have been developed to support the corresponding methodologies. For our methodology, eBizBench is developed.
6. Conclusion Our methodology helps align customers’ needs with implementation details for developing business webs. Scenarios are employed for analyzing customers’ requirements in a natural fashion. Scenarios are better able to monitor customers’ requirements, and thus enhance the feasibility of our methodology. By using our methodology, companies are more likely to capture the context in which customers interact with their Internet business systems. This is especially true if shifts in customer needs mandate some sort of reintegration of the system. Our methodology has its limitations. The following future research avenues will be able to alleviate these possible weaknesses. First, it is noted our scenarios are descriptive. They can be extended to cover exploratory or explanatory designs. Furthermore, the direct incorporation of customers’ experiences into the web site may be able to make our methodology more powerful [60]. Second, adaptive web design is of interest. Analyzing customer profiles will be able to enhance the adaptability of presentation or navigation. Web server logs on the basis of customers’ profiles may help find more convenient navigation [51]. Finally, our current mechanism for determining the implementation priority is not systematic. Quality Function Deployment (QFD) may be applied to evaluating this priority. QFD can transform what the customer wants into how the company plans to provide it [13,26]. QFD can help analyze the value via crossfunctional teams in a collaborative fashion.
C. Lee et al. / Information and Software Technology 46 (2004) 17–33
31
Appendix A. Object model
Appendix C. Collaboration relationship transformation Appendix B. Page specification
32
C. Lee et al. / Information and Software Technology 46 (2004) 17–33
Appendix D. User interface specification
References [1] J.M. Artz, A top-down methodology for building corporate web applications, Internet Research: Electronic Networking Applications and Policy 6 (2/3) (1996) 64–74. [2] P. Atzeni, G. Mecca, P. Merialdo, Design and maintenance of dataintensive web sites, EDBT’98 Lecture Notes in Computer Science (1998) 436–450. [3] E.G. Abels, M.D. White, K. Hahn, Identifying user-based criteria for web pages, Internet Research: Electronic Networking Applications and Policy 7 (4) (1997) 252 –262. [4] E.G. Abels, M.D. White, K. Hahn, A user-based design process for web sites, OCLS Systems and Services (1) (1999) 35–44. [5] A. Afuah, C.L. Tucci, Internet Business Models and Strategies: Text and Cases, McGraw-Hill, New York, 2000. [6] C. Barry, M. Lang, A survey of multimedia and web development techniques and methodology usage, IEEE Multimedia 8 (2) (2001) 52–60. [7] C. Batini, S. Ceri, S. Navathe, Conceptual Database Design: An Entity-Relationship Approach, The Benjamin/Cummings, Menlo Park, 1992. [8] M. Bichler, S. Nusser, SHDT: the structured way of developing WWW-sites, Proceedings of the Fourth European Conference on Information Systems, Lisbon, 1996, pp. 1093–1101 [9] M. Bichler, S. Nusser, Modular design of complex web-applications with W3DT, Proceedings of Fifth International Workshops on Enabling Technologies: Infrastructure for Collaborative Enterprises (WET ICE’96), 1996, pp. 328 –333 [10] B. Bishop, The Strategic Enterprise, Stoddart Publishing, Toronto, 2000. [11] M.R. Blaha, W.J. Premerlani, J.E. Rumbaugh, Relational database design using an object-oriented methodology, Communication of the ACM 31 (4) (1998) 414 –427. [12] G. Booch, Object-Oriented Analysis and Design with Application, The Benjamin/Cummings, Melno Park, 1994. [13] J. Brady, Systems engineering and cost as an independent variable, Systems Engineering 4 (4) (2001) 233 –241. [14] S. Casteleyn, O.D. Troyer, Structuring web sites using audience class hierarchies, Proceedings of 20th International Conference on Conceptual Modeling, 2001
[15] J.Q. Chen, R.D. Heath, Building web applications: challenges, architectures, and methods, Information Systems Management 18 (1) (2001) 68 –79. [16] D.W. Cravens, G. Greenley, N.F. Piercy, S. Slater, Integrating contemporary strategy management perspective, Long Range Planning 3 (4) (1997) 493– 506. [17] D. Cunliffe, Developing usable web sites: a review and model, Internet Research: Electronic Networking Applications and Policy 10 (4) (2000) 295–307. [18] K.M. Delargy, The new business landscape, Manufacturing Engineering 80 (4) (2001) 169 –174. [19] A.J. Dix, J.E. Finlay, G.D. Abowd, R. Beale, Human–Computer Interaction, Prentice-Hall, Englewood Cliffs, 1998. [20] J. Fong, Mapping extended entity relationship model to object modeling technique, SIGMOD Record 24 (3) (1995) 18–22. [21] M. Fowler, K. Scott, U.M.L. Distilled:, A. Brief, Guide to the Standard Object Modeling Language, second ed., Addison-Wesley, Reading, MA, 1999. [22] P. Fraternali, P. Paolini, Model-driven development of web applications: the Autoweb system, ACM Transaction on Information Systems 28 (4) (2000) 323–382. [23] F. Garzotto, P. Paolini, D. Schwabe, HDM: A model-based approach to hypertext application design, ACM Transactions on Information Systems 11 (1) (1993) 1–26. [24] H. Gellersen, R. Wicke, M. Gaedke, WebComposition: an objectoriented support system for the Web engineering lifecycle, Computer Networks and ISDN Systems 29 (1997) 1429–1437. [25] A. Ginige, S. Murugesan, Web engineering: an introduction, IEEE Multimedia 8 (1) (2001) 14 –18. [26] C.P.M. Govers, QFD not just a tool but a way of quality management, International Journal of Production Economics 69 (2001) 151–159. [27] M.J. Greenacre, Theory and Applications of Correspondence Analysis, Academic, Academic Press, New York, 1984. [28] M.J. Greenacre, T.J. Hastie, The geometric interpretation of correspondence analysis, Journal of the American Statistical Association 82 (1987) 437– 447. [29] V. Gruhn, L. Schope, Software processes for the development of electronic commerce systems, Information and Software Technology 44 (2002) 891–901.
C. Lee et al. / Information and Software Technology 46 (2004) 17–33 [30] G. Hackbarth, W.J. Kettinger, Building an e-business strategy, Information Systems Management 17 (30) (2000) 78– 93. [31] A.T.F. Hutt, Object Analysis and Design: Description of Methods, Wiley, London, 1994. [32] T. Isakowitz, E.A. Stohr, P. Balasubramanian, RMM: a methodology for structured hypermedia design, Communications of the ACM 38 (8) (1995) 34 –44. [33] T. Isakowitz, A., Kamis, M. Koufaris, Extending RMM: Russian dolls and hypertext, Proceedings of the 30th Hawaii International Conference on System Sciences, 1997, pp. 177 –186 [34] I. Jacobson, M. Ericsson, A. Jacobson, The Object Advantage: Business Process Reengineering with Object Technology, AddisonWesley, Reading, MA, 1995. [35] S.L. Jarvenpaa, E.H. Tiller, Integrating market, technology, and policy opportunities in e-business strategy, Journal of Strategic Information Systems 8 (3) (1999) 235–249. [36] A.J. Killander, Why design methodologies are difficult to implement, International Journal Technology Management 21 (3/4) (2001) 271–276. [37] W. Kim, Modern Database Systems: The Object Model, Interoperability, and Beyond, Addison-Wesley, Reading, MA, 1995. [38] P. Kotler, G. Armstrong, Principles of Marketing, Prentice-Hall, Englewood Cliffs, NJ, 1996. [39] P. Kotler, Marketing Management Analysis, Planning, and Control, Prentice Hall, Englewood Cliffs, NJ, 1977. [40] D.B. Lange, Enhanced relationships in object-oriented database modeling. Proceedings of Information Science, 1993, pp. 296–304 [41] D.B. Lange, An object-oriented design method for hypermedia information systems, Proceedings of the 27th Hawaii International Conference on System Sciences, 1994, pp. 380 –389 [42] D.B. Lange, An object-oriented design approach for developing hypermedia information systems, Journal of Organizational Computing 6 (3) (1996) 269 –293. [43] H. Lee, J. Kim, Y. Kim, S. Cho, A view-based hypermedia design methodology, Journal of Database Management 10 (2) (1999) 3–13. [44] C. Lee, H. Lee, B. Choi, A system for developing customer-oriented Internet business: eBizBench, Proceedings of the ITS 14th Biennial Conference, Seoul, Korea, 2002 [45] H. Lee, C. Lee, C. Yoo, A scenario-based object-oriented hypermedia design methodology, Information and Management 36 (3) (1999) 121–138. [46] H. Lee, W. Suh, A workflow-based methodology for developing hypermedia information systems, Journal of Organizational Computing and Electronic Commerce 11 (2) (2001) 77 –196. [47] G.P. Marquis, Application of traditional system design techniques to web site design, Information and Software Technology 44 (2002) 507–512. [48] G. Maughan, B. Durnota, MON: an object relationship model incorporating roles, classification, publicity and assertions, Proceedings of OOIS’94, International Conference on Object Oriented Information Systems (1993). [49] T.P. Novak, D.L. Hoffman, Y-F. Yung, Measuring the customer experience in online environments: a structural modeling approach, Marketing Science 19 (1) (2000) 43–62.
33
[50] R. Orfali, D. Harkey, J. Edwards, The Essential Client/Server Survival Guide, Wiley, London, 1996. [51] M. Perkowitz, O. Etzioni, Towards adaptive web sites: Conceptual framework and case study, Artificial Intelligence 118 (1–2) (2000) 245–275. [52] T.A. Powell, Web Design: The Complete Reference, Osborne/ McGraw-Hill, New York, 2002. [53] C.R. Rolland, C.B. Anchour, Guiding the construction of textual use case specifications, Data and Knowledge Engineering (1) (1998) 125–160. [54] C. Rolland, E. Dubois, P. Heymans, C. Ben Achour, C. Cauvet, J. Ralyte, A. Sutcliffe, N. Maiden, M. Jarke, P. Haumer, K. Pohl, A proposal for a scenario classification framework, Requirements Engineering 3 (1) (1998) 23–47. [55] J. Preece, H. Sharp, D. Benyon, S. Holland, T. Carey, Human– Computer Interaction, Addison-Wesley, Reading, MA, 1994. [56] K.L. Smart, M.E. Whiting, Designing systems that support learning and use: a customer-centered approach, Information and Management 39 (3) (2001) 177– 190. [57] D. Schwabe, G. Rossi, The object oriented hypermedia design model, Communications of the ACM 38 (8) (1995) 45–46. [58] D. Schwabe, R.A. Pontes, I. Moura, OOHDM-Web: an environment for implementation of hypermedia applications in the WWW, SigWEBO News Letter (1999). [59] D. Schwabe, L. Esmeraldo, G. Rossi, F. Lyardet, Engineering web applications for reuse, IEEE Multimedia 8 (1) (2001) 20–31. [60] P.B. Seybold, Get inside the lives of your customers, Harvard Business Review May (2001) 81–88. [61] C. Standing, Methodologies for developing Web applications, Information and Software Technology 44 (2002) 151–159. [62] W. Suh, H. Lee, Hypermedia document management: a metadata and meta-information system, Journal of Database Management 12 (2) (2001) 25–35. [63] K. Takahashi, E. Liang, Analysis and design of web-based information systems, Proceedings of Sixth World Wide Web Conference, 1997 [64] A. Tannenbaum, Implementing a Corporate Repository, Wiley, London, 1994. [65] D.A. Taylor, Business Engineering with Object Technology, Wiley, London, 1995. [66] M.J. Taylor, J. McWilliam, H. Forsyth, S. Wade, Methodologies and website development: a survey of practice, Information and Software Technology 44 (6) (2002) 381–391. [67] M.F.D. Troyer, C.J. Leune, WSDM: a user centered design method for web sites, Proceedings of the Seventh International WWW Conference, Brisbane, Australia, 1998 [68] B.C. Wheeler, NEBIC: a dynamic capabilities theory for assessing net-enablement, Information Systems Research 13 (2) (2002) 125–146. [69] R. Wirfs-Brock, B. Wilkerson, L. Wiener, Designing Object-oriented Software, Prentice-Hall, Englewood Cliffs, NJ, 1990. [70] P. Zang, G.M. Dran, User expectation and rankings of quality factors in different web site domains, International Journal of Electronic Commerce 6 (2) (2002) 9–34.
Implementing a community web site: a scenario-based methodology Choongseok Leea, Woojong Suhb, Heeseok Leea,* a Corporate Information System Laboratory, Department of Management Information Systems, Graduate School of Management, Korea Advanced Institute of Science and Technology, 207-43, Cheongryangri-dong, Dongdaemun-gu, Seoul 130-012, South Korea b Division of Business Administration, College of Business Administration, Inha University, 253, Yonghyeun-dong, Nam-gu, Incheon 402-751, South Korea
Received 12 February 2003; revised 15 April 2003; accepted 29 April 2003
Abstract The use of the Internet to facilitate commerce promises vast benefits. An important challenge in the age of Internet business is to fine-tune the Internet business system with customers. For this alignment, this paper proposes a scenario-based object-oriented methodology for developing a community web site. The methodology consists of five phases such as customer analysis, value analysis, web design, implementation design, and construction. Scenarios are used to analyze customers’ needs in a natural fashion. A meta-data scheme is proposed for supporting our methodology. A real-life community web is illustrated to demonstrate the usefulness of the methodology. Scenario thinking can help companies use the Internet to buttress and extend their values. q 2003 Elsevier B.V. All rights reserved. Keywords: Web site development; Scenario-based methodology; Meta-data
1. Introduction The use of the Internet to facilitate commerce among companies promises vast benefits. Many companies have attempted to develop Internet business systems for virtual companies, markets, or communities [25,35]. Internet businesses can enable commercial exchanges that cross physical, temporal, cultural, and legal boundaries on a scale that was technically complicated [33]; it can help companies strengthen the links between customers and suppliers [18,30]. As customers gain control over the design of products or services, competition takes on a whole new shape. Because customers can compare products and services with a variety of rich information, their role shifts from passive recipient to active designer [5]. As a result, companies tend to be more customer-oriented [10,16]. Companies spend a lot of time and money fine-tuning their offerings with customers. Business web systems need to be in perfect alignment with ever-changing customers’ needs. Companies should first analyze customers’ scenarios, and then respond to their needs [49,60]. Typically, Internet business systems are implemented by the use of web technologies. Web technologies employ * Corresponding author. Tel.: þ 82-2-958-3615; fax: þ82-2-958-3604. E-mail address: [email protected] (H. Lee). 0950-5849/03/$ - see front matter q 2003 Elsevier B.V. All rights reserved. doi:10.1016/S0950-5849(03)00094-6
hypermedia as a major technical principle; Internet business is one of the most popular applications of hypermedia [46]. This implementation requires several steps such as information modeling for customers, navigation design to help customers find information, user interface design for web page layout, and construction. Some implementation features such as navigation, contents, and user interfaces are not common in conventional software development [29]. Building business web sites is challenging because of the lack of developers’ experiences, poorly defined processes, and unrealistic schedules [6,52]. To solve these problems, a Web Engineering approach has been proposed. Web Engineering employs methods from several areas such as software engineering (SE), human– computer interaction (HCI), project management, usability engineering, testing, and simulation [24,25]. Internet business development methodology can be investigated from these perspectives. A variety of business web development methodologies have been proposed. They can be categorized into formal and informal [17]. The informal methodologies do not specify detailed activities for each phase. These informal methodologies are flexible for formulating business strategies [61]. Formal methodologies describe not only the development activities but also linkages of inputs
18
C. Lee et al. / Information and Software Technology 46 (2004) 17–33
Table 1 Classification of Internet business development methodologies Classification Researcher (Methodology) Informal
Artz [1], Abels et al. [3,4], Chen and Heath [15], Standing (ICDM) [61]
Formal
E-R Model Based
OO ModelBased
Garzotto et al. (HDM) [23], Isakowitz et al. (RMM) [32,33], Takahashi and Liang [63], Atzeni et al. [2], Lee et al. (VHDM) [43], Fraternali and Paolini (HDM-Lite) [22] Lange (EORM) [40–42], Schwabe and Rossi (OOHDM) [57,58], Bichler and Nusser (W3DT) [8,9], Lee et al. (SOHDM) [45], Troyer and Leune (WSDM) [67]
ICDM: Internet Commerce Development Methodology; HDM: Hypermedia Design Method; RMM: Relationship Management Model; VHDM: View based Hypermedia Design Methodology; EORM: Enhanced Object Relationship Model; OOHDM: Object Oriented Hypermedia Design Model; W3DT: World Wide Web Design Technique; SOHDM: Scenariobased Object-oriented Hypermedia Design Methodology; WSDM: Web Site Design Method.
and outputs. Most formal methodologies adopt one of the two data modeling techniques such as Entity-Relationship (E-R) model and Object-Oriented (OO) model. The OO model provides semantic richness, reusability, and flexibility [65,69]. Interestingly, WHDM [46] allows designers to choose between the two data models. The classification of Internet business development methodologies can be summarized as shown in Table 1. HDM, EORM, RMM, and OOHDM were originally proposed for hypertext or hypermedia applications. Most past methodologies focus on the analysis of business information and its relationships based on a data model; they fail to see how this analysis fits into the real lives of their customers. Artz [1], Troyer and Leune [67], Abel et al. [3,4], Chen and Heath [15], Standing [61], and Troyer and Leune [67] emphasized the importance of the analysis of customers’ needs. However, they do not provide systematic and robust steps. As customers’ needs change continuously, reflecting them is critical for sustaining competitive advantages. Most corporate web sites fall short of managers’ high expectations because of a fundamental mismatch. The destination business web simply does not suit the needs of most companies or their customers. A methodology is required to support design activities ranging from the analysis of customers’ needs to system implementation. This paper attempts to sharpen scenario-based techniques SOHDM [45] by focusing on aligning customers’ needs with implementation details. Scenario thinking can help companies use the Internet to extend their customers’
values. The outline of the paper is as follows. The next section provides an overview of the methodology. Section 3 describes the methodology in further detail using a reallife case. Section 4 describes a meta-data scheme supporting our methodology. Section 5 compares our methodology with others. The paper concludes in Section 6.
2. Methodology The architecture of our methodology is depicted in Fig. 1. It consists of five phases: customer analysis, value analysis, web design, implementation design, and construction. Customer analysis, value analysis, and web design are independent of any specific technical platforms and details, while implementation design and construction depend on them. For the sake of clear presentation, the feedback among phases is not depicted. However, feedback is of importance in our methodology because iterative design is essential for Internet business system development [15,25,55]. The customer analysis phase analyzes customers’ needs. Customers are categorized according to their shared needs. Customer groups and the corresponding needs are summarized in the customer need analysis table; their correspondence is analyzed by the use of a correspondence analysis technique. The value analysis phase models value activities. Customers’ requirements are explored in the form of events. In the web design phase, the Internet business to be implemented is logically designed. This phase consists of five sub-phases: scenario design, object modeling, view design, navigation design, and page design. In the scenario design, scenarios are designed in the form of natural language to describe customers’ interactions with an Internet business system from the customers’ perspective. These scenarios lead to object modeling. In the view design, information units that customers want to find are designed. The navigation design subphase designs navigational paths for customers to access information conveniently. Finally, in the page design, users’ information windows and the flow from one page to another are defined according to the views and their navigational paths. During the implementation design, page specifications are determined and the object model is transformed into the physical database schema. Finally, the construction phase implements a physically running Internet business system.
3. Methodology details Here, each phase of our methodology is described in further detail by the use of a real-life case. This Internet business case will be referred to as “I site.” I site (www.iloveschool.co.kr), an Internet alumni association site, has provided a school-based community
C. Lee et al. / Information and Software Technology 46 (2004) 17–33
19
Fig. 1. Methodology architecture.
service since October 1999. Its members can find any school friend who is also a member by locating the name of his or her school. They can leave messages via bulletin board and exchange e-mail. It has shown a dramatic network effect because its service becomes valuable as more people use it. Drawn in by a rapid expansion of the Internet population, I site has turned into one of the top Korean community web sites. In market surveys conducted by major private research institutes such as Samsung Economic Research Institute (www.koreaeconomy.org) and Yahoo! Korea (kr.yahoo.com), the site was picked as one of the best hit products in 2000. About 10 million customers are using the site and about 11,000 (almost 100 percent) Korean universities and schools are registered. I site has 400 million page views per month and its Alexa (www.alexa.com) ranking is the 47th in the world and the seventh in Korea as of April 2001. Its main service areas include schoolmate search, school Bulletin
Board System (BBS), school club service, and club service. I site laid out plans to become a serious player in the Internet community market. Customer loyalty has become more important than customer acquisition. Accordingly, our methodology was employed to help developers improve I site. We will illustrate this application in details as follows. 3.1. Customer analysis Customer analysis includes customer need analysis and correspondence analysis. This phase can find creative ways to fulfill supplementary needs that may not involve the current company’s offerings by building a detailed understanding of common customer needs. The customer need analysis sub-phase analyzes customers on the basis of customers’ data. First, data may be collected by observation, survey, or experimentation.
20
C. Lee et al. / Information and Software Technology 46 (2004) 17–33
Customers may be categorized according to their geographic, demographic, psychographic, or behavioral variables. Because behavioral variables are based on actual customer behavior toward products or service, they are more likely to be closely related to Internet business than other variables [38]. Our methodology adopts behavioral variables. In the case of I site, customers are categorized into five groups according to their roles: school club manager, club manager, school club member, club member, and member. Communities can be categorized into school club and club. A school club is a community that belongs to a school. An example of school club is “soccer club in K school.” Only the alumni of “K school” can join this school club. However, a club is not related to a particular school. To join a school club or club is not member’s obligation. Accordingly, there may be members who join neither of these two clubs. A school club member is a customer who joins the school club, and a school club manager manages this school club. A club manager manages the club; a club member joins the club. To analyze customers’ needs, we looked at 19,259 suggestions, which were made in the BBS by 11,240 volunteer customers between December 2000 and April 2001. I site ran the BBS to listen to its customers. Because 18,956 suggestions are related to the existing services, remaining 303 suggestions (from 291 volunteer customers) are analyzed. Customers’ needs are grouped into nine categories according to their service characteristics, as shown in Table 2. For example, the schoolmate segmentation is most frequently suggested. This segmentation can divide schoolmates into various groups by their majors and graduation dates. Different customer groups may have different needs. It is important to analyze the correspondence between customer groups and their needs. However, the previous customer need analysis provides the frequencies only. Typically, the correspondence analysis [27] is adopted for finding this correspondence. Third quantification method, reciprocal
averaging, appropriate scoring, homogeneity analysis, dual scaling, and scalogram are all synonyms of correspondence analysis. The correspondence analysis investigates simple two-way or multi-way tables containing some measure of correspondence between the rows and columns. It is similar to factor analysis techniques in statistics; it can explore the structure of categorical variables included in the table. For more information, refer to Refs. [27,28]. The analysis results in the chi-square value, 93.2 (degree of freedom ¼ 32). This value confirms the correspondence. Fig. 2 shows the customer correspondence diagram drawn in a Euclidian space. In this diagram, the horizontal axis and vertical axis describe row and column coordinates, respectively. For example, the needs for school club members and members include teacher community, schoolmate segmentation, and post-graduate community. 3.2. Value analysis Business webs are inventing new value propositions. The objective of the value analysis is to identify the value for customers’ needs and then describe the activities between customers and web system. First, customer values are explored on the basis of customers’ activities with the system. The value derivation sub-phase employs a value table to analyze customers’ values, value activities, and implementation priorities. Customer value is defined as a bundle of benefits customers expect from a given product or service [39]. Customer value can be measured from financial, perceptual, and behavioral perspectives [68]. Value activity is a customers’ distinct event or integrated set of events to create customer value. Because customers may not be able to understand all of the new opportunities by new technologies or environments, implementation priorities may be decided on the basis of technical feasibility, business capabilities, and strategic importance. Table 3 shows the result of value derivation. For example, school club members and members highly value
Table 2 Customer need analysis table Customer Group Customer Need
School Club Manager
Club Manager
School Club Member
Club Member
Member
Total
Teacher Community Schoolmate Segmentation Post-graduate Community Album Service Schoolmate e-mail Wireless Internet Army Community Off-line Support Cooperative Purchase Total
14 28 12 20 10 11 1 3 0 99
2 3 2 0 1 5 1 1 0 15
16 36 19 5 3 0 4 0 1 84
4 4 3 0 0 0 3 0 0 14
12 42 18 5 4 0 4 3 3 91
48 113 54 30 18 16 13 7 4 303
C. Lee et al. / Information and Software Technology 46 (2004) 17–33
21
Fig. 2. Customer correspondence diagram.
“Friendliness among teachers and students” in the teacher community according to customers’ perceptual perspective. “Search for teachers” and “Join teacher communities” are activities for achieving this value. The implementation priorities of the teacher community and the album service are fairly high according to I site’s current capabilities and strategic importance. Next, it is important to investigate the activities between customers and the system to be implemented. At the same time, the scope of the system is determined. A value activity
diagram can capture these value activities. The diagram is based on events, customer groups, and sub-systems. Events are identified for each customer group. An event is a trigger that starts the system [12]. In the value activity diagram, the needs of each customer group are depicted for the corresponding sub-system. Each customer group triggers events for this sub-system via value activities. The value activity diagram is similar to a use case diagram in standard Unified Modeling Language (UML) notations [21]. Use cases describe the actions the system
Table 3 Value table Customer Group
Customer Need
Customer Value
Value Activity
Implementation Priority
Member
Teacher Community
Friendliness among teachers and students
Search for teachers Join teacher communities
WWW
School Club Member
Schoolmate Segmentation
Friendliness among schoolmates
Search for schoolmate segments Join schoolmate segments
WW
Post-graduate Community
Friendliness among post-graduate schoolmates
W
Album Service
Sharing of photos
Search through post-graduate communities Join post-graduate communities Manage on-line albums
WWW
Schoolmate e-mail
Communication with all schoolmates Ease of management
Mail for all schoolmates
W
Manage community using wireless Internet Search through army communities Join army communities
WW
School Club Manager
Club Manager
Wireless Internet
Club member
Army Community
Friendliness among army colleagues
W
22
C. Lee et al. / Information and Software Technology 46 (2004) 17–33
Fig. 3. Value activity diagram.
takes to deliver to the actor [34]. The emphasis of our value activity diagram is on finding detailed event activities between customer groups and the business web system. The value activity diagram is thus simpler to depict a variety of activities between customer groups and system. Fig. 3 shows the value activity diagram in our case. For the sake of simple presentation, it highlights two items for customer needs with top implementation priority (“Teacher Community” and “Album Service”). These two items are depicted as subsystems. Although “Search for teachers” activity is directly transformed into the “Search for teachers” event, this activity can be subdivided into several activities such as “Search for teachers by teacher name” and “Search for teachers by school name” to provide alternative search facilities for the members and the school club members. “Join teacher communities” activity is transformed into seven events. 3.3. Web design The web design phase consists of five sub-phases: scenario design, object modeling, view design, navigation design, and page design. Our methodology employs scenarios to identify business system requirements at the earliest opportunity. Task analysis may be adopted for capturing user activities in the human computer interaction studies [19,55]. A scenario analysis is less formal than task analysis [55]. Scenarios can be categorized into three types: descriptive, exploratory, and explanatory [54]. Descriptive scenarios capture requirements. Exploratory scenarios are
used to find possible solutions for satisfying given system requirements through the exploration and evaluation of alternatives. Explanatory scenarios are intended to support the explanation and argumentation of drawbacks, inefficiencies, or system performance problems. Our scenarios are descriptive to capture customers’ requirements. A scenario is a description of customers’ interactions with a sub-system from the perspective of customers. Although it is difficult to infer object classes or behaviors, our scenarios can better describe customers’ interactions. Scenarios correspond to key business processes and thus can capture customers’ requirements in a natural fashion. In the scenario design sub-phase, the events drawn from the value activity diagram are described in the form of natural language. The natural language can enhance ease of use and understandability [49,53]. In the I site case, 12 scenarios are generated from the 12 events. For example, Fig. 4 through 6 show three scenarios from three events such as “Search for teachers,” “Search for photos,” and “Post photo.” Next, we need to model a variety of information the web provides for customers. An object-oriented modeling technique has several advantages (comprehensiveness, understandability, changeability, adaptability, and reusability) [12,31]. Our methodology adopts the object-oriented modeling technique to inherit these advantages. Notations for our object modeling technique are depicted in Fig. 7. An object includes its name, attributes, and responsibilities. Attributes contain the properties of the object. Responsibilities are behavioral properties of
C. Lee et al. / Information and Software Technology 46 (2004) 17–33
Fig. 4. “Search for teachers” scenario.
Fig. 5. “Search for photos” scenario.
Fig. 6. “Post photo” scenario.
the object [12,69]. Among objects, four types of object relationships are described: superclass/subclass, association, collaboration, and component. Any subclass can inherit or use the attributes and responsibilities of its superclass. The association relationship describes a linkage among objects with relationship properties: one-to-one (1:1), one-to-many (1:N), or many-to-many (N:N) [7]. The collaboration relationship describes a dynamic reference among objects through responsibilities [48]. The object that can be used as an attribute in other objects is described by the use of the component relationship. Objects are derived from scenarios. Customer groups and information used in a subsystem can be candidates for
Fig. 7. Notations for the object model.
23
24
C. Lee et al. / Information and Software Technology 46 (2004) 17–33
Fig. 8. View specification.
objects. Additional attributes and responsibilities of objects and relationships among objects can be obtained. For example, in case of the “Search for teachers” scenario, objects such as “Member,” “School Club Member,” and “School Club Manager” are generated. The resulting object model is depicted in Appendix A. The view design subphase reorganizes information contents in the form of navigational units. A view is the information unit that customers want to find. These views
may be referred to as OO (Object-Oriented) views. OO views [37] are designed on the basis of responsibilities, attributes, and relationships in the object model. From our object model (eight objects), five views are generated as shown in Fig. 8. Views are categorized into three types: base view, association view, and collaboration view. (1) a base view is generated from a single object. It is a sub-set of responsibilities and attributes in an object. For example, in
C. Lee et al. / Information and Software Technology 46 (2004) 17–33
Fig. 8, “Teacher Number by School” is a base view, (2) an association view is extracted from the objects that have association relationship(s). It is a subset of responsibilities and attributes in these objects. “Teacher List by School” and “Teacher Profile” are association views, (3) a collaboration view is generated from the objects having collaboration relationships. It is used to change information. In the case of collaboration view, the responsibilities required for collaboration are of importance. “School Club Album Update” is a collaboration view extracted from the collaboration relationship between “School Club Album” and “School Club.” Collaboration views are rarely generated in contentsoriented Internet business systems, like newspaper, since information is less dynamically updated according to customers’ interaction. These three types of OO views can be used in any Internet business system. The navigation design subphase builds navigational links between nodes. The OO view and Access Structure Node (ASN) are adopted for navigational units. The ASN differs from the OO view in that ASN contains access paths to OO views. The OO view contains actual information that customers want to obtain. These OO views and ASNs correspond to nodes. A link denotes the relationship between the source node and the destination node. These source and destination nodes may be OO views or ASNs. Web pages are implemented on the basis of these OO views and ASNs in the subsequent page design subphase. From the scenarios, ASNs are found and then navigational links are determined. First, an event that begins with a scenario can become an event ASN. An event ASN starts navigation. For example, “Search for teachers,” “Search for photos,” and “Enroll photo” events become “Search for teachers ASN,” “Search for photos ASN,” and “Enroll photo ASN,” respectively. Second, a selection ASN can be generated when a customer should select the next activities.
25
For example, from the “Search for photos” scenario, “Condition for search for photos ASN” is built. Third, an input ASN can be generated when a customer needs to input contents. From the “Enroll photo” scenario, “Input photo ASN” is built. Navigation specifications for I site are depicted in Fig. 9. Linking ASNs and views together on the basis of scenario flows results in the navigation path. For example, according to the “Search for teachers” scenario, “Search for teachers ASN,” “Teacher Number by School View,” “Teacher List by School View,” “Profile, memo, e-mail ASN,” and “Teacher Profile View” are linked sequentially. The page design subphase specifies web pages. A page is a window having information and a navigational guide. Web pages should be designed for customers to obtain necessary information in a convenient fashion. A page specification may be composed of many views and ASNs. Pages are designed according to the following guidelines. (1) a view can be a page, (2) an input ASN and a selection ASN can be an individual page, (3) when an ASN follows a view, a page can include both this ASN and the following view. Fourth, an event ASN can be used with an anchor in other pages. Pages are specified by organized anchors, OO views, and additional description details (e.g. embedded components, text, images, sounds, etc.). Appendix B shows the page specification entitled “My Teacher.” 3.4. Implementation design The implementation design phase generates: (1) physical database schema and (2) user interface (UI) specification for construction. An Internet business system can be developed under a variety of system environments. Implementation decisions are made according to the choice of technology. These technologies include DBMSs such as Relational
Fig. 9. Navigation specifications.
26
C. Lee et al. / Information and Software Technology 46 (2004) 17–33
Fig. 10. User interface component.
DBMS (RDBMS) or OODBMS, and development tools such as CGI, HTML, JAVA, Shockwave, Flash, DHTML, XML, and SGML. If OODBMS is used, our object model can be transformed directly. However, in many real-life cases, RDBMSs are most popular. Therefore, the transformation of the object model into relational schema is required. Here, designers need guidelines for this transformation. These transformation guidelines can be found in Refs. [11,20,45]. In addition to these guidelines, more views, Structured Query Language (SQL) stored procedures, or SQL triggers are required for the transformation of collaboration relationships. An SQL stored procedure is a user-defined program module that is stored at the database server and can be invoked by client applications. An SQL trigger is usually in the form of SQL stored procedure and is automatically invoked by the data-related event. It is typically used to perform tasks related to changes in a table [50]. For
example, the collaboration relationship between “School Club Album” and “School Club” can be transformed into a trigger as depicted in Appendix C. Finally, our object model results in ten tables. For actual construction, the page specification should be extended to incorporate data location, interface component choice, and component properties. The user interface design sub-phase maps the page specification to user interface components like choice box, list box, button, check box, and scroll bar, which are supported in various implementation tools. Fig. 10 summarizes the notation for user interface components and transitions. A caption is a description of components. An action is an interactive procedure. Items are sets of data included in the user interface component. Two notations are proposed for transitions. The key-based transition refers to the current state of the component while the simple transition does not. These user interface components may be determined according to the access structure in the navigational design. All anchors may be transformed into a set of HTML anchors. The user interface specification for “My Teacher” page specification is illustrated in Appendix D. 3.5. Construction The construction phase implements a physically running Internet business system. All of the outputs during the implementation design phase should be mapped to physical components. For example, Fig. 11 shows the “My Teacher” screen that is constructed using the UI specification. In this screen, customers can execute the “Search for teachers” scenario. Our design results in seven newly implemented pages.
Fig. 11. “My Teacher” screen.
C. Lee et al. / Information and Software Technology 46 (2004) 17–33
27
Fig. 12. A meta-data schema for development system.
4. A meta-data schema To improve business webs according to customers’ needs continuously, reuse of documents is of particular importance. A variety of models and diagrams can be drawn via various drawing facilities. Object model and views can be automatically transformed to physical database schema.
Furthermore, it can provide simulated web pages. Such simulation can enhance navigational capability before construction. Our methodology is likely to be successful when it is consistently used within an organization. A development support system can strengthen this usability. A system entitled eBizBench can support our methodology [44].
28
Criteria
Classification Informal methodology
Formal methodology E-R Model Based
OO Model Based
Isakowitz Takahashi and Atzeni et al. [32,33] Liang [63] et al. [2] RMM
Fraternali and Paolini [22] HDM-Lite
Schwabe and Rossi [57]; Schwabe et al. [58] OOHDM
Bichler and Nusser [8,9] W3DT
Troyer and Leune [67]; Casteleyn and Troyer [14] WSDM
This Methodology
User User Involvement Involvement
N/A
N/A
N/A
N/A
N/A
N/A
Audience Class Hierarchy
Correspondence Analysis
Objectives User Criteria of Web Applications
Use Case
GRSs
N/A
Scenario
N/A
N/A
N/A
N/A
Audience Class Hierarchy
Scenario
Source of Navigation Design
N/A
N/A
N/A
E-R E-R Relationship Relationship
E-R Relationship
E-R Relationship
OO Relationship
OO Relationship
Navigation Track
Scenario and OO Relationship
Focus
Development Development
Development and Maintenance
Development
Development and Maintenance
Artz [1]
Abels et al. [3,4]
Chen and Heath [15]
Customer Analysis
Target Audience
Users’ Task-related Information and Use Behavior
System Requirement Analysis
Structure, Linkage, and Search
Standing [61] ICDM
Development Development
Development Development Development and Maintenance
Development Development
C. Lee et al. / Information and Software Technology 46 (2004) 17–33
Table 4 Methodology comparison
Development Supporting System Methodology Phases
N/A
N/A
Problem Statement
Information Gathering
Basic System Strategy Analysis and Design Architecture MetaDecision Development Strategy Build Basic Component Functionality Strategy
Constraining Development Requirements Conceptual Model
Evaluation
N/A
Derived Implementation Deployments Component Requirements Strategy Detailed Analysis Detailed Design Development
RMCase
WebArchitect N/A
ER Design
ER Analysis
Slice Design Scenario Analysis Navigation Design
Architecture Design
Conversion Protocol Design UI Screen Design
Attribute Definition
Logical Construction Functional Requirements Semi-Physical Run-time Maintenance Architecture Behavior Design Design Construction and Testing Implementation Evolution
AutoWeb
OOHDM-Web WebDesigner
Database Conceptualize Conceptual Conceptual Design Design Database Generate Navigational Logical Design Database Design Implement Hypertext and Deploy Conceptual Design Hypertext Logical Design
Abstract UI Design
N/A
eBizBench
Customer Analysis
Information Structuring
Mission Statement Specification User Classification
Navigational Design
User Class Description
Requirement Analysis
Implementation Organizational Object Design Modeling
Presentation Design
Interface Design
Hypertext to DB Mapping
Implementation Implementation Design Introduction and Maintenance
Navigational Design
Implementation
Value Analysis Web Design
Implementation Design Construction
C. Lee et al. / Information and Software Technology 46 (2004) 17–33
N/A
29
30
C. Lee et al. / Information and Software Technology 46 (2004) 17–33
This design support system requires a sound understanding of the way in which data (inputs or outputs) should be actually managed. Meta-data can provide a foundation for managing this data effectively. Meta-data is generally known as data about data or information about information [62,64]. The meta-data for eBizBench is depicted as shown in Fig. 12. This meta-data includes 29 entities. The meta-data consists of five parts: customer analysis, value analysis, web design, implementation design, and construction. Clearly, each part of this scheme corresponds to each phase of our methodology. This meta-data can sharpen the reusability and interoperability of the resulting development support system.
5. Methodology comparison This section compares our methodology with other Internet business implementation methodologies. For the comparison, several criteria are of interest: customer analysis, system requirement analysis, source of navigation design, development supporting system, focus, and methodology phases. This comparison is highlighted in Table 4. Customer analysis is a starting point for any system development [56]. Artz [1] places an emphasis on the analysis of customer groups for the top-down development of corporate Web applications. Abels et al. [3,4] identify the criteria that can influence a particular use of the web. They propose that web site should be designed for the user and user input be solicited throughout the process. Chen and Heath [15] and Standing [61] explore user involvement in the evolutionary design process. Troyer and Leune [67] propose a user-modeling phase for user-centered Internet business system development. Our methodology includes systematic methods for customer analysis like the correspondence analysis. It helps developers capture customers’ needs at an early stage and improve the system according to their needs. To implement a usable system, system requirement analysis is important [47]. Artz [1] suggests the exploration of the objectives of a web application. Abels et al. [3,4] propose user criteria that consist of use, content, structure, linkage, search, and appearance. These criteria can be used to provide specific suggestions for the Internet business system design. Standing [61] uses Group Requirements Sessions (GRSs) to obtain the detailed requirements with users’ involvement. GRSs can lead to a compromised solution by striking balances between heterogeneous stakeholders. Takahashi and Liang [63] use a scenario for the dynamic analysis. Their scenario does not describe interactions between the system and its customers. Lee et al. [45] adopt a Scenario Activity Chart (SAC) to describe activities between users and Internet business system in a graphical fashion. Similarly, our methodology adopts scenarios that can describe processes between the business
web and its customers in a natural fashion. It can enhance the ease of use and understandability. The success of an Internet business system is associated with navigation [59,66,70]. Most of the proposed methodologies use the relationships in data model as a major source of navigation design. For example, Abel et al. [3,4] use information derived from user criteria. Our methodology uses scenarios and relationships in object-oriented models to capture navigational requirements. The use of scenarios is likely to better reflect customers’ navigational requirements. In Internet businesses, another important challenge is to reflect on customers’ needs. Bichler and Nusser [8,9] and Atzeni et al. [2] propose methodologies for a new development as well as maintenance. Standing [61] emphasizes the continual evolution of Internet business systems from a managerial perspective. Similarly, our methodology can improve Internet business systems by monitoring customers’ needs continuously. A critical success factor for wider use of design methodologies is the availability of software tools [6,36]. RMCase, OOHDM-Web, WebArchitect, AutoWeb, and WebDesigner have been developed to support the corresponding methodologies. For our methodology, eBizBench is developed.
6. Conclusion Our methodology helps align customers’ needs with implementation details for developing business webs. Scenarios are employed for analyzing customers’ requirements in a natural fashion. Scenarios are better able to monitor customers’ requirements, and thus enhance the feasibility of our methodology. By using our methodology, companies are more likely to capture the context in which customers interact with their Internet business systems. This is especially true if shifts in customer needs mandate some sort of reintegration of the system. Our methodology has its limitations. The following future research avenues will be able to alleviate these possible weaknesses. First, it is noted our scenarios are descriptive. They can be extended to cover exploratory or explanatory designs. Furthermore, the direct incorporation of customers’ experiences into the web site may be able to make our methodology more powerful [60]. Second, adaptive web design is of interest. Analyzing customer profiles will be able to enhance the adaptability of presentation or navigation. Web server logs on the basis of customers’ profiles may help find more convenient navigation [51]. Finally, our current mechanism for determining the implementation priority is not systematic. Quality Function Deployment (QFD) may be applied to evaluating this priority. QFD can transform what the customer wants into how the company plans to provide it [13,26]. QFD can help analyze the value via crossfunctional teams in a collaborative fashion.
C. Lee et al. / Information and Software Technology 46 (2004) 17–33
31
Appendix A. Object model
Appendix C. Collaboration relationship transformation Appendix B. Page specification
32
C. Lee et al. / Information and Software Technology 46 (2004) 17–33
Appendix D. User interface specification
References [1] J.M. Artz, A top-down methodology for building corporate web applications, Internet Research: Electronic Networking Applications and Policy 6 (2/3) (1996) 64–74. [2] P. Atzeni, G. Mecca, P. Merialdo, Design and maintenance of dataintensive web sites, EDBT’98 Lecture Notes in Computer Science (1998) 436–450. [3] E.G. Abels, M.D. White, K. Hahn, Identifying user-based criteria for web pages, Internet Research: Electronic Networking Applications and Policy 7 (4) (1997) 252 –262. [4] E.G. Abels, M.D. White, K. Hahn, A user-based design process for web sites, OCLS Systems and Services (1) (1999) 35–44. [5] A. Afuah, C.L. Tucci, Internet Business Models and Strategies: Text and Cases, McGraw-Hill, New York, 2000. [6] C. Barry, M. Lang, A survey of multimedia and web development techniques and methodology usage, IEEE Multimedia 8 (2) (2001) 52–60. [7] C. Batini, S. Ceri, S. Navathe, Conceptual Database Design: An Entity-Relationship Approach, The Benjamin/Cummings, Menlo Park, 1992. [8] M. Bichler, S. Nusser, SHDT: the structured way of developing WWW-sites, Proceedings of the Fourth European Conference on Information Systems, Lisbon, 1996, pp. 1093–1101 [9] M. Bichler, S. Nusser, Modular design of complex web-applications with W3DT, Proceedings of Fifth International Workshops on Enabling Technologies: Infrastructure for Collaborative Enterprises (WET ICE’96), 1996, pp. 328 –333 [10] B. Bishop, The Strategic Enterprise, Stoddart Publishing, Toronto, 2000. [11] M.R. Blaha, W.J. Premerlani, J.E. Rumbaugh, Relational database design using an object-oriented methodology, Communication of the ACM 31 (4) (1998) 414 –427. [12] G. Booch, Object-Oriented Analysis and Design with Application, The Benjamin/Cummings, Melno Park, 1994. [13] J. Brady, Systems engineering and cost as an independent variable, Systems Engineering 4 (4) (2001) 233 –241. [14] S. Casteleyn, O.D. Troyer, Structuring web sites using audience class hierarchies, Proceedings of 20th International Conference on Conceptual Modeling, 2001
[15] J.Q. Chen, R.D. Heath, Building web applications: challenges, architectures, and methods, Information Systems Management 18 (1) (2001) 68 –79. [16] D.W. Cravens, G. Greenley, N.F. Piercy, S. Slater, Integrating contemporary strategy management perspective, Long Range Planning 3 (4) (1997) 493– 506. [17] D. Cunliffe, Developing usable web sites: a review and model, Internet Research: Electronic Networking Applications and Policy 10 (4) (2000) 295–307. [18] K.M. Delargy, The new business landscape, Manufacturing Engineering 80 (4) (2001) 169 –174. [19] A.J. Dix, J.E. Finlay, G.D. Abowd, R. Beale, Human–Computer Interaction, Prentice-Hall, Englewood Cliffs, 1998. [20] J. Fong, Mapping extended entity relationship model to object modeling technique, SIGMOD Record 24 (3) (1995) 18–22. [21] M. Fowler, K. Scott, U.M.L. Distilled:, A. Brief, Guide to the Standard Object Modeling Language, second ed., Addison-Wesley, Reading, MA, 1999. [22] P. Fraternali, P. Paolini, Model-driven development of web applications: the Autoweb system, ACM Transaction on Information Systems 28 (4) (2000) 323–382. [23] F. Garzotto, P. Paolini, D. Schwabe, HDM: A model-based approach to hypertext application design, ACM Transactions on Information Systems 11 (1) (1993) 1–26. [24] H. Gellersen, R. Wicke, M. Gaedke, WebComposition: an objectoriented support system for the Web engineering lifecycle, Computer Networks and ISDN Systems 29 (1997) 1429–1437. [25] A. Ginige, S. Murugesan, Web engineering: an introduction, IEEE Multimedia 8 (1) (2001) 14 –18. [26] C.P.M. Govers, QFD not just a tool but a way of quality management, International Journal of Production Economics 69 (2001) 151–159. [27] M.J. Greenacre, Theory and Applications of Correspondence Analysis, Academic, Academic Press, New York, 1984. [28] M.J. Greenacre, T.J. Hastie, The geometric interpretation of correspondence analysis, Journal of the American Statistical Association 82 (1987) 437– 447. [29] V. Gruhn, L. Schope, Software processes for the development of electronic commerce systems, Information and Software Technology 44 (2002) 891–901.
C. Lee et al. / Information and Software Technology 46 (2004) 17–33 [30] G. Hackbarth, W.J. Kettinger, Building an e-business strategy, Information Systems Management 17 (30) (2000) 78– 93. [31] A.T.F. Hutt, Object Analysis and Design: Description of Methods, Wiley, London, 1994. [32] T. Isakowitz, E.A. Stohr, P. Balasubramanian, RMM: a methodology for structured hypermedia design, Communications of the ACM 38 (8) (1995) 34 –44. [33] T. Isakowitz, A., Kamis, M. Koufaris, Extending RMM: Russian dolls and hypertext, Proceedings of the 30th Hawaii International Conference on System Sciences, 1997, pp. 177 –186 [34] I. Jacobson, M. Ericsson, A. Jacobson, The Object Advantage: Business Process Reengineering with Object Technology, AddisonWesley, Reading, MA, 1995. [35] S.L. Jarvenpaa, E.H. Tiller, Integrating market, technology, and policy opportunities in e-business strategy, Journal of Strategic Information Systems 8 (3) (1999) 235–249. [36] A.J. Killander, Why design methodologies are difficult to implement, International Journal Technology Management 21 (3/4) (2001) 271–276. [37] W. Kim, Modern Database Systems: The Object Model, Interoperability, and Beyond, Addison-Wesley, Reading, MA, 1995. [38] P. Kotler, G. Armstrong, Principles of Marketing, Prentice-Hall, Englewood Cliffs, NJ, 1996. [39] P. Kotler, Marketing Management Analysis, Planning, and Control, Prentice Hall, Englewood Cliffs, NJ, 1977. [40] D.B. Lange, Enhanced relationships in object-oriented database modeling. Proceedings of Information Science, 1993, pp. 296–304 [41] D.B. Lange, An object-oriented design method for hypermedia information systems, Proceedings of the 27th Hawaii International Conference on System Sciences, 1994, pp. 380 –389 [42] D.B. Lange, An object-oriented design approach for developing hypermedia information systems, Journal of Organizational Computing 6 (3) (1996) 269 –293. [43] H. Lee, J. Kim, Y. Kim, S. Cho, A view-based hypermedia design methodology, Journal of Database Management 10 (2) (1999) 3–13. [44] C. Lee, H. Lee, B. Choi, A system for developing customer-oriented Internet business: eBizBench, Proceedings of the ITS 14th Biennial Conference, Seoul, Korea, 2002 [45] H. Lee, C. Lee, C. Yoo, A scenario-based object-oriented hypermedia design methodology, Information and Management 36 (3) (1999) 121–138. [46] H. Lee, W. Suh, A workflow-based methodology for developing hypermedia information systems, Journal of Organizational Computing and Electronic Commerce 11 (2) (2001) 77 –196. [47] G.P. Marquis, Application of traditional system design techniques to web site design, Information and Software Technology 44 (2002) 507–512. [48] G. Maughan, B. Durnota, MON: an object relationship model incorporating roles, classification, publicity and assertions, Proceedings of OOIS’94, International Conference on Object Oriented Information Systems (1993). [49] T.P. Novak, D.L. Hoffman, Y-F. Yung, Measuring the customer experience in online environments: a structural modeling approach, Marketing Science 19 (1) (2000) 43–62.
33
[50] R. Orfali, D. Harkey, J. Edwards, The Essential Client/Server Survival Guide, Wiley, London, 1996. [51] M. Perkowitz, O. Etzioni, Towards adaptive web sites: Conceptual framework and case study, Artificial Intelligence 118 (1–2) (2000) 245–275. [52] T.A. Powell, Web Design: The Complete Reference, Osborne/ McGraw-Hill, New York, 2002. [53] C.R. Rolland, C.B. Anchour, Guiding the construction of textual use case specifications, Data and Knowledge Engineering (1) (1998) 125–160. [54] C. Rolland, E. Dubois, P. Heymans, C. Ben Achour, C. Cauvet, J. Ralyte, A. Sutcliffe, N. Maiden, M. Jarke, P. Haumer, K. Pohl, A proposal for a scenario classification framework, Requirements Engineering 3 (1) (1998) 23–47. [55] J. Preece, H. Sharp, D. Benyon, S. Holland, T. Carey, Human– Computer Interaction, Addison-Wesley, Reading, MA, 1994. [56] K.L. Smart, M.E. Whiting, Designing systems that support learning and use: a customer-centered approach, Information and Management 39 (3) (2001) 177– 190. [57] D. Schwabe, G. Rossi, The object oriented hypermedia design model, Communications of the ACM 38 (8) (1995) 45–46. [58] D. Schwabe, R.A. Pontes, I. Moura, OOHDM-Web: an environment for implementation of hypermedia applications in the WWW, SigWEBO News Letter (1999). [59] D. Schwabe, L. Esmeraldo, G. Rossi, F. Lyardet, Engineering web applications for reuse, IEEE Multimedia 8 (1) (2001) 20–31. [60] P.B. Seybold, Get inside the lives of your customers, Harvard Business Review May (2001) 81–88. [61] C. Standing, Methodologies for developing Web applications, Information and Software Technology 44 (2002) 151–159. [62] W. Suh, H. Lee, Hypermedia document management: a metadata and meta-information system, Journal of Database Management 12 (2) (2001) 25–35. [63] K. Takahashi, E. Liang, Analysis and design of web-based information systems, Proceedings of Sixth World Wide Web Conference, 1997 [64] A. Tannenbaum, Implementing a Corporate Repository, Wiley, London, 1994. [65] D.A. Taylor, Business Engineering with Object Technology, Wiley, London, 1995. [66] M.J. Taylor, J. McWilliam, H. Forsyth, S. Wade, Methodologies and website development: a survey of practice, Information and Software Technology 44 (6) (2002) 381–391. [67] M.F.D. Troyer, C.J. Leune, WSDM: a user centered design method for web sites, Proceedings of the Seventh International WWW Conference, Brisbane, Australia, 1998 [68] B.C. Wheeler, NEBIC: a dynamic capabilities theory for assessing net-enablement, Information Systems Research 13 (2) (2002) 125–146. [69] R. Wirfs-Brock, B. Wilkerson, L. Wiener, Designing Object-oriented Software, Prentice-Hall, Englewood Cliffs, NJ, 1990. [70] P. Zang, G.M. Dran, User expectation and rankings of quality factors in different web site domains, International Journal of Electronic Commerce 6 (2) (2002) 9–34.