DESRIST 2008 Doctoral Consortium – May 7, 2008 (Georgia State)

Supported by National Science Foundation and Pennsylvania State University

PE00: Doctoral Consortium (8:30 am – 5:30 pm)

Chairs:

Sandeep Purao (Penn State University)

Ying Zhu (Georgia State University)

Faculty Mentors

Carliss Baldwin (Harvard Business School)

Al Hevner (University of South Florida)

Jan Pries-Heje (Roskilde University, Denmark)

Brian Smith (Penn State University)

Veda Storey (Georgia State University)

The Animated Work Environment: A Vision for Working Life in a Digital Society, Henrique Houayek (Clemson University)

Abstract Dramatic transformations in the nature, place and organization of working life in the United States suggest a redesign of the work environment as a socially and technologically responsive system occupying both home and office. Emerging characteristics for new ways of labor are defined less by standardized behavior and more by multiple shared group activities with information exchange and human interaction. Work activities have become more than just the completion of specific tasks; instead, these are now characterized by an overlapping process of work, study, and leisure.

In response to these conditions, our transdisciplinary team (from Architecture, Robotics, Sociology and Human Factors Psychology) asks: (1) How can Intelligent Systems, Information Technology and Robotics become design elements in the creation of work environments? (2) How to design technological spaces supporting positive human interaction? (3) What are the metrics to evaluate such a project?

Our “Animated Work Environment” [AWE] is an environment-as-responsive-robot: an articulated, programmable, interior environment accommodating a range of digital technologies that facilitates productivity, connectedness and innovation across fluid assemblages of people working with printed and digital materials in a variety of locations and settings. AWE sits technologically at the interface between computer technology, robotics, and architectural design, where the physical environment is also subject to manipulation. The AWE project involves designing, prototyping, demonstrating and evaluating a fully operational workspace which uses robotic components to achieve certain levels of responsiveness and adaptation to support different work configurations.

In developing AWE, the team investigates why, how and what computing possibilities and robotics can contribute to work spaces towards generating a new vision for working life in a digital society. Our research is in part inspired by William Mitchell’s vision that our buildings will “become…more like us. We will continually interact with them, and increasingly think of them as robots for living in.” Featuring a continuous, morphing surface controlled by a user-friendly interface, AWE is characterized as adaptive and robust when applied to a range of work activities and working populations currently not accommodated by robotics, architecture or interface-design alone. The AWE prototype can reasonably be expanded to the scale of a room having walls and ceiling altered in response to users’ demands or environmental stimuli.

The effectiveness and continuity of enterprise architecture, Marlies van Steenbergen (Sogeti Netherlands and Utrecht University)

Abstract Research question. The topic of the Ph.D. research is the effectiveness and continuity of enterprise architecture in the field of Information Systems. Enterprise architecture, the application of principles and models to guide the design and realization of processes, information systems and technological infrastructure, is seen by many as a means to make complexity in IS manageable. For this promise to come true, sound architectural practices, by which we mean the whole of activities, responsibilities and actors involved in the development and application of enterprise architecture, have to be implemented. We aim to answer the question “How can architectural practices achieve effectiveness and continuity?”. This is translated into the following sub-questions: (1) What goals, tasks and responsibilities is the architecture practice comprised of? (2) Do different types of organizations require different approaches to the architecture practice? (3) How can the effectiveness of the architecture practice be measured? (4) What factors provide continuity of the architecture practice for various types of organizations? and (5) What kinds of interventions lead to improvements in architecture effectiveness?

Methodology and results so far. A mixed method approach is adopted, including design research, case studies, action research, and surveys. Design research is applied in the development of instruments to measure various aspects of architectural practices. The instruments are tested and refined in one or more case studies. They are applied to conduct an in-depth action research study of the architecture practice within a multinational financial institution. A survey will be done to complement this in-depth insight with an overview of the state of architecture maturity over many organizations.

Thus far, design research has been conducted delivering two instruments: the Architecture Maturity Matrix and the Architecture Effectiveness Model. A case study exploring the relationship between organizational aspects, architecture maturity and architecture effectiveness is in progress. The action research in architecture practice is in the initiating stage.

Learning Profiles from User Interactions, Pelin Atahan (The University of Texas at Dallas)

Abstract Personalization and recommendation systems play an important role in online business settings. They increase customer retention and loyalty by providing enhanced personalized services, and improve the profitability of firms by deploying targeted marketing strategies and exploiting cross-sell and up-sell opportunities. An important determinant of the performance of these systems is the reliability of the user profile available to the system. Prior research on learning user profiles has focused on learning sets of terms (i.e., feature vectors) that best describe documents or web pages liked by a user. However, such profiles are not particularly useful in identifying products and services that a firm should advertise to its users. Profiles consisting of demographic, psychographic, and behavioral characteristics of users are more relevant for such purposes.

We develop a technique that a firm can use to implicitly learn such a profile based on a user’s web traversals. We compare the proposed model’s performance with well-accepted classification techniques using real-world data on sites visited by users. We find that the performance of the proposed model is as good as, and often better than, other established classification techniques.

We then extend our analysis to show how a firm can expedite learning a user’s profile by dynamically determining the set of links to be made available to the user, referred to as the offer set, at each interaction. We refer to this learning approach as active profile learning. The set of links that is more informative in terms of inferring the profile attribute values are determined and made available to the user at each page. We present a technique for determining the optimal offer set for each page visited by the user. We also develop a heuristic approach to determine the offer sets in an efficient manner for when the number of potential offer sets to evaluate are large.

A Conceptual Modeling Method to Use Agents in Systems Analysis, Kafui Monu (University of British Columbia)

Abstract There are many system analysis and design methodologies that represent events and processes. However, none include or analyse the assumptions behind the processes or the context of the events. It has been proposed that the conceptual agent may act as a new concept that system analysts can use to better represent the domain. Unfortunately, there is much confusion about how agents interact, and the mechanism of their behaviour.

We propose that by conceptualizing the agent as a feedback system, that a new system analysis methodology, called the conceptual agent model (CAM) methodology, can be developed. We propose that this new methodology will be able to aid modellers in explaining the processes in a domain and why certain events occur in a domain. However, we must not only develop this design artifact but also test it. To create the methodology and validate it we divide our proposed study into three essays which will: provide a precise definition of agents (essay 1), create a methodology of using the conceptual agent concepts (essay 2), and test the methodology’s usability, usefulness, and quality (essay 3).

We have begun our study and currently have developed conceptual models of conceptual agents, analysed the constructs in relation to other agent methodologies, and compared the CAM constructs to those developed by novice modellers to create conceptual agents. We have found that the novice modeller's understanding of an actor in a domain is similar to the CAM constructs and that the CAM constructs explicitly represent agents as defined by other researchers. We have also developed part of the methodology for CAM and tested it.

For future research, we can conduct a larger empirical test of the usefulness of the method . We are also interested in using this work to analyse work systems in non-business areas such as government.

Developing Computational Tools that Facilitate Individual and Group Creativity, Brittany Smith (University of Illinois at Urbana-Champaign)

Abstract A considerable amount of research has been conducted in order to explore and define creativity, and there has been much work exploring how computers are and can be used to support the creative process. I am interested in working within the intersection of these spaces by developing computational tools that help facilitate both individual and collocated group creativity within two particular domains: design and computer programming. Designers typically use pencil and paper when generating early ideas because paper can be easily and quickly accessed and is very flexible; however, computers provide some affordances that pencil, paper, and other traditional mediums do not effortlessly provide such as the ability to interact with one’s ideas, easily modify one’s ideas, and easily store and transmit one’s ideas to others if so desired. Such interactions have the potential to positively affect the creative process. Computer programmers, on the other hand, use computers extensively. However, many of the tools that have been developed to aid in the programming process are designed with the idea that only one user will be utilizing the tool at once. This is increasingly not the case with the adoption of paired programming and other team software development techniques, and this lack of proper support could be affecting programmers. With the above notions in mind, I submit that there are many ways yet to be actualized in which computers can help individuals and groups to not only record ideas and solutions but aid in creating and better conveying these ideas and solutions as well.

Both the use and the functions of sketching for designers have been explored, and earlier work shows that sketching is a useful tool during the early stages of design. I am interested in utilizing and expanding upon this research by augmenting and studying the use of computer- based tools that provide sketching capabilities in addition to other functionalities that assist and enhance the creative process. One such functionality, in a tool with which I am currently working, is the ability to work on multiple ideas in parallel, which allows designers to better re- interpret existing ideas and thus generate better ideas and products.

My focus is a little different in the computer programming domain. As noted above, programmers are increasingly utilizing the paired programming development framework, which means that not one, but two individuals are using one machine and one integrated development environment (IDE) at once. There have yet to be many software tools developed that seek to make this setup less awkward. There is, consequently, a need for computer tools, and specifically IDEs, to help make the programming environment more conducive to paired work, that aid in improving communication and work flow between the programmers, and that enhance the programmers’ creative output. I am currently exploring how best to modify these IDEs in order to both retain and add to the advantages of the pair programming practice.

Modeling Service Systems to Support their Analysis and Design, Lysanne Lessard (University of Toronto)

Abstract Services, once considered a residual economic activity of the manufacturing sector, have come to play a major role in our economy. Researchers have started to investigate the paradigmatic changes at play in the service sector, but these efforts have yet to result in strong theoretical constructs. There is thus a need for integrated theories and models of modern services that can act as an effective bridge between the understanding of services throughout their lifecycle and the design of Information Systems able to facilitate stakeholder interactions and service adaptation. In order to explain the principles governing modern services beyond existing management and economic frameworks, theories need to be inductively generated from empirical research.

My research project aims to use methods and theoretical insights from social sciences to develop conceptual models of services suitable for Service-Oriented Information Systems design. This aim is structured around three objectives: 1) analyze real-world instances of service design projects using a framework derived from social sciences, 2) incorporate knowledge gathered from the analysis into descriptive and analytical conceptual models of services, and 3) assess the ability of the models to lead to the design of Information Systems able to support the creation and evolution of services.

Services are often described as being co-created at the time of their delivery to end users by a service provider. But this interaction is extended when the service provider designs its service with the help of a consulting firm. A chain of services is then recognizable, from the contractual service between the consulting firm and the service provider to the transactional service between the later and the service end user. While this complete chain of co-creation must be accounted for in order to understand how and why services are designed and evolve, my study focuses on services as they are being co-created by the consulting firm and its customer. As such, focusing on service design projects may shed light on the complete lifecycle of services.

To adequately express the process of service design, the nature of the resulting service and its perceived value by those involved, current modeling approaches such as agent orientation will be adapted. The method of grounded theory will guide a systematic analysis of service design projects, ensuring a strong cohesion between the conceptual models and the phenomena they try to express. Participant and peer validation will be used to refine conceptual models and reflect on their descriptive, analytical and design capabilities. The resulting conceptual models of service systems will contribute to the field of Information Systems analysis and design by proposing extensions to current conceptual modeling approaches leading to Service-Oriented Information Systems design.

Improving Design Intent Research for Software Maintenance, Paul Grisham (The University of Texas at Austin)

Abstract Managing system evolution is one of the fundamental problems of software engineering. A system is the result of a long design process during which various design trade-offs and decisions about both functional and non-functional requirements are made. In the process of abstracting actual requirements into models and converting those models into code, only the end result of the design process is reflected in the code, leaving the designers’ intent to be represented, if at all, in some form of external documentation.

A descriptive software architecture is a model of the current, implemented architecture of a system in a non-prescriptive, non-normative manner. Although certain emergent properties of a particular design, such as an architectural style, can be explicitly expressed in some architecture languages, in practice, such properties are often only implicitly represented. A prescriptive architecture provides guidance for implementation of the system. The prescriptive architecture for a system captures the intentional design qualities for the system and may express a normative view of subsequent design considerations.

We define architectural drift as the divergence of the implementation from the prescriptive architecture. Architectural drift may begin early in the initial implementation and can only become exacerbated during post-installation maintenance and system evolution. In practice, it is often only the code itself that is kept current with respect to the requirements and the design. Architectural erosion occurs when changes in the descriptive architecture violate critical constraints in the prescriptive architecture. Violating these constraints erodes the coherence of the design and may yield errors or make additional changes more expensive. Architectural drift often leads to architectural erosion, which means that without up-to-date documentation, starting with a good initial design only delays erosion.

To ameliorate the problem of design erosion, research into explicit models of Design Rationale has been increasing. Moreover, the revision of IEEE-Std-1471-2000, “IEEE Recommended Practice for Architectural Description of Software-Intensive Systems” may include explicit representation for design decisions as first-class entities. Despite the current flurry of research activity, it became clear that as researchers, we lacked the instruments for evaluating the effectiveness of various rationale and design decision modeling systems, for comparing these systems to each other, and for using these analyses to drive future research.

The goal of this research project is to develop a better understanding of the documentation needs of maintenance designers for software systems and to build a model of how those maintenance designers interact with knowledge about legacy designs. To accomplish this goal, we propose a course of empirical study that uses surveys, expert interviews, talk-through protocol analysis, and ethnography as the basis of a theory of maintenance design. The theory will be tested using experimentation by protocol analysis and validated through further ethnography in real-world industrial contexts. In particular we are investigating the differences in how initial designers and maintenance designers interact with the design space, and what the role of experience is when interacting with legacy design artifacts.

How to embody digital information into product interfaces, Heekyoung Jung (Indiana University)

Abstract Background: While conventional computers or information appliances have been typically manipulated through screen-based graphic user interface, interfaces of recent interactive devices have become diversified into different forms such as touch screens, smart materials, and physical controllers. Introduction of dynamic relations between information and interface have opened up new design opportunities in terms of product aesthetics and usability. However, at the same time, they have brought problems of cognitive overload or confusion to users from arbitrary couplings between information and interface. As information is embodied into visual or physical interfaces of various products, characteristics of information and user activities should be reflected in interface considering appearance, symbolic meaning and context of product.

Research objectives and questions: This study aims to explore new interaction design factors to represent unique characteristics of information into everyday objects in various contexts of use. The study also includes discussions about how to evaluate their influences on product design and user experience. Followings are specific research questions:

- What types of design factors can we explore for intuitive interaction with digital information?

- Which of them (new design factors) can users control or perceive as product interface?

- How can we match new design factors to appropriate types of information and tasks?

- How can we evaluate influences of those new interaction design factors?

Research plans and ideas: As a systematic approach for exploring new interaction design factors, we suggest an interaction model. Referring to Chris Crawford’s listening-thinking-speaking model, we specified unit modes of interaction between user and interactive product into user’s input (action), data, data processing (or task) and object’s output (response). Then we exemplified some possible design factors for each unit mode. Based on this framework, we will create new types of interactive products by combining different factors of unit modes. Such case studies are expected to provide a common ground for designing new forms of interactive products considering tight couplings between information and interface.

Designing Deeply Engaging Online Communities for Adolescents, David Gurzick (University of Maryland, Baltimore County)

Abstract For many of those who are connected to the Internet, online communities have become a staple of modern life. Though it is only within the last few decades that they have become mainstream, recent years have seen an increase in both the number and type of people who use them. One group that has experienced considerable growth in their use of online communities is American adolescents (which, at the time of this study, are those aged 12-18, and born in the years 1988-1994). They are characterized by a hyper-connected lifestyle and constitute the sub-population with the largest number of members online, using mobile phones, and participating in socially based e-services. However, though members of this group are among the largest consumers of online communities, it is unknown whether the current understanding of online community design applies to the design of communities specific to adolescents.

This study looks to bridge this gap, presenting a research strategy for addressing, 1) In what ways does an online community, designed in accordance with existing best practices, succeed or fail when used by a population of adolescents, and 2) How might best practice be emended by new theories and guidelines for the design of adolescent online communities? This strategy follows a design science approach centered on the detailed investigation of an adolescent –centric online community as it progresses through an iterative design.

Two broad stages are identified for this research, each consisting of a build/evaluate cycle of an online community artifact. In the first stage, an online community for adolescents will be created and evaluated in accordance with existing design guidelines. As a better understanding is constructed on how an adolescent population responds to existing design practices, this research will move to create theory to explain the phenomena that are discovered and to suggest improvements for any uncovered deficiencies. The second stage will allow for the evaluation and refinement of those theories that emerge from the first stage that might offer improvement to the existing design guidelines. Within this stage, these theories will be codified into modifications to the online community artifact. To explore the effects of the new design theories, a study using new participants will compare the redesigned online community (modified according to the refined guidelines) and a replica of the previous existing online community (designed around current best practice). The results of this second stage will be incorporated into refinements to the design theories, with the resultant refined theories organized and presented in terms of design guidelines for online communities for adolescents.

By evaluating, and possibly improving, the guidelines from which this type of online community is created, the future development of such systems in practice can be better informed. Moreover, investigating the design of online communities for a population that is, in many ways, divergent from those populations that underpin the current understanding of design, may lead to new and novel ways of approaching the design of online communities in general. Such conceptual understanding might have far reaching results, leading to the creation of more sustainable and effective online communities.

Facilitating and measuring older people's creative engagement in a user centred design process, Helena Sustar (City University of London)

Abstract Interfaces and products are usually designed for a broader market and general use by younger users. In the classic design process older people are rarely involved. If they are, they are brought in for focus groups or at the end of the design process as participants in usability tests. Despite the fact that there is existing research that (a) looks at the involvement of older people in participatory design activities and (b) supports the creative potential of older people, there is no research that looks at combining the two themes i.e. ways of facilitating and measuring creative involvement of older people in participatory design activities. The goal of my research is to address this limitation in the research of the relationship between creativity and age in the user centred design (UCD) process for user interfaces. As a first step I wanted to check whether classic UCD techniques could initiate creative engagement. To achieve this I used UCD methods to design a simple and intuitive interface with technology and wireless media to the stage of mock-up model. The main interactions happened between the user and device with touching the “plants” – gadgets, and changing their position in the device. This mock-up model has been used in a creative process as a stimulator of new ideas. Three different age groups of users were chosen: students (23+), older people in their sixties, and older people in their eighties. A series of interviews based around interacting with the mock-up model were conducted (see photos). We tried to find out how familiar the participants were with certain parts of the device and how they will use different pieces of the model in a new innovative way. In addition, we asked them to use a model in different situations: having a conversation with a friend, taking medicine, being reminded of daily activities, managing the household (smart house) and thinking how a device could provide different services to support these situations. I was especially interested to find out: (a) what was the role of the model in stimulating new ideas, (b) what kind of ideas the model generated and (c) what were the main difficulties that block creativity. The results of the qualitative research showed that the mock-up model was able to stimulate ideas at all age groups. New ideas were stimulated by the participants’ interaction with it, from the mock-up model as visual presence, the participant’s existed experience and from environment that sounds participant. The only differences in define age groups were in way how participants generated ideas.

Supporting Scenario Generation in Product Design, Irene Anggreeni (University of Twente, Netherlands)

Abstract Designing consumer products has grown more challenging due to the increasingly complex characteristics of these products. Consequently, product design has recently turned to scenario based design to address this challenge. Scenario based design is originally a heuristic method that arises from computer system development. The main idea behind scenario based design is to use concrete descriptions of people using technology, referred to as scenarios, to discuss and analyze how the technology could fit into their activities. Scenarios have promising qualities to help a project team formulate, communicate and evaluate design ideas with stakeholders early on. Despite the potentials, available scenario approaches mostly address the high-level use of scenarios and therefore often miss the low-level guidance for identifying, creating and communicating the scenarios. We believe that a scenario generation support tool could aid designers by providing a framework to create, use and organize their scenarios as an integral part of designing.

There exist scenario-related support tools that provide more prescriptive steps towards the construction of scenarios (e.g. [1, 2]). However, they are often focused in specific activities such as the capture and collection of design requirements. The application of scenarios could be more useful when they are created, maintained and reused throughout a design process. Therefore, this research started broad by studying scenario uses from theoretical as well as practical point of view. Six categories of scenario uses over different design phases have been identified. This categorization was used to identify the difficulties currently experienced by designers in practice. Furthermore, designers will be given an active role in the concept development of the support tool. In the end, the success measure of the project will be whether the designers find the support tool useful in their practice and whether it helps them to design better products.

This research is expected to contribute in an understanding on how product designers can entangle the chaos of designing by structuring their rationales and using scenarios to register them. As an illustration, consider the complexity of nowadays consumer products, in which a single product is often designed to have many functions to accommodate diverse users in different settings. Designing such a product requires the design team to deal with an enormous amount of relevant, yet not uncommonly contradicting design aspects at once. The scenario generation support tool will be developed to give method guidance to the process, specifically focusing on generating ‘useful’ scenarios. The generated scenarios will construct a frame of reference for the designers to deal with all relevant design aspects.

DESRIST 2008 Day 1 – May 8, 2008 (The Westin Buckhead Atlanta)

PE11-2: Keynote: Charles Eastman (Georgia Institute of Technology) (9:30 – 10:30 am)

“Sciences of the Artificial in Architecture and Building Design”

“The Philosophy of Design Science”

DESRIST 2008 seeks to focus on the development of
design science researchers. Our Keynotes, Invited Speakers, and Panels discuss areas of advanced research designs, methodologies, empirical issues, epistemology and the dissemination of knowledge, as specifically embraced by Design
Science Research approaches. In particular, we envision these sessions to constitute a Philosophy of Design Science track that will informally focus the expertise of all DESRIST
2008 attendees on the needs and development of the research students.

Sessions Coded as P (as in P132-A) are components of The Philosophy of Design Science Track.

Sessions Coded as PE (as in PE11-2) are components of The Philosophy of Design Science Track as well as the Exemplars Track.

Sessions Coded as E (as in E14-2B) are components of the Exemplars of Design Science Track.

P12-A: Advancing DS Approaches (11:00 am – 12:30 pm)

Session Chair: Juhani Iivari (University of Oulu)

Theory Development in Design Science Research: Anatomy of a Research Project by Bill Kuechler (University of Nevada, Reno) and Vijay Vaishnavi (Georgia State University) (Nominated for Best Paper Award)

Abstract The common understanding of Design Science Research in Information Systems (DSRIS) continues to evolve. Only in the broadest terms has there been consensus: that DSRIS involves, in some way, learning through the act of building. However, what is to be built – the definition of the DSRIS artifact – and how it is to be built – the methodology of DSRIS – has drawn increasing discussion in recent years. The relationship of DSRIS to theory continues to make up a significant part of the discussion: how theory should inform DSRIS and whether or not DSRIS can or should be instrumental in developing and refining theory. In this paper we present the exegesis of a DSRIS research project in which creating a (prescriptive) design theory through the process of developing and testing an IS artifact is inextricably bound to the testing and refinement of its kernel theory.

The Use of Focus Groups in Design Science Research by Monica Chiarini Tremblay (Florida International University), Alan R. Hevner (University of South Florida) and Donald J. Berndt (University of South Florida)

Abstract Focus groups to investigate new ideas are widely used in many research fields. The use of focus groups in design science research poses interesting opportunities and challenges. Traditional focus group methods must be adapted to meet two specific goals of design research. For the evaluation of an artifact design, exploratory focus groups (EFGs) study the artifact to propose improvements in the design. The results of the evaluation are used to refine the design and the cycle of build and evaluate using EFGs continues until the artifact is released for field test in the application environment. Then, the field test of the design artifact may employ confirmatory focus groups (CFGs) to establish the utility of the artifact in field use. Rigorous investigation of the artifact requires multiple CFGs to be run with opportunities for quantitative and qualitative data collection and analyses across the multiple CFGs. In this paper, we discuss the adaptation of focus groups to design science research projects. We demonstrate the use of both EFGs and CFGs in a design research project from the health care field.

Dissemination and Importance of the "Method" Artifact in the Context of Design Research for Information Systems by Tobias Bucher (University of St. Gallen, Institute of Information Management) and Robert Winter (University of St. Gallen, Institute of Information Management)

Abstract During the last decade, design research (DR) has evolved into a key research paradigm in information systems (IS). DR for IS is aimed at creating solutions – so called "design products" or "artifacts" – to problems of practical relevance. From a product perspective, these artifacts can be distinguished into the five types "constructs", "models", "methods", "instantiations", and "theories". This paper analyzes the proceedings of the First and Second International Conference on Design Science Research in Information Systems and Technology (DESRIST) and categorizes the presented papers in regard to the topics covered. It turns out that the "method" artifact is addressed least frequently. Moreover, almost all DESRIST publications which do address the "method" artifact describe the design of some kind of algorithm or mathematical/statistical technique. The "method engineering" (ME) discipline is concerned exclusively with the processes of constructing, adapting, and implementing generic methods that are aimed at the design of socio-technical information systems. However, explicit references to the ME discipline and its body of literature are used rarely – if at all – in the DESRIST proceedings. To overcome this gap, this paper briefly introduces the principles of the ME approach. Moreover, it discusses potential reasons for the lacking integration of DR for IS on the one hand and the ME approach to IS development on the other hand. The paper argues that the transfer and application of the basic ideas of the ME discipline hold a strong potential of bringing forward the DR paradigm.

Goal-Framing and Breakdowns in the Design of Boundary-Spanning Information Systems by Susan Gasson (Drexel University)

Abstract It is argued that design groups which span organizational boundaries are becoming increasingly common and present a unique problem for information system design. Such groups attempt to integrate design understanding across many different domains of expertise and knowledge. This means that no one design group participant is likely to understand either the problem domain or the solution domain in full. This paper critiques traditional and emerging process models of design, in terms of their applicability for boundary-spanning design groups. Findings from studies of boundary-spanning design groups were synthesized across multiple levels of analysis to present an integrated model of design for such groups. The model builds on the notion of technological frames, to develop a theory of framing at the three levels of project coordination, group consensus-building, and individual learning and dissonance.

E12–B: Invited Papers – NSF Science of Design Projects (11:00 am – 12:30 pm)

Session Chair: Veda Storey (Georgia State University)

Self-Adaptation in Software Agents: An Initial Case Study in Game-Playing Agents by Ashok K. Goel (Georgia Institute of Technology), Joshua Jones (Georgia Institute of Technology), Christopher Parnin (Georgia Institute of Technology), Spencer Rugaber (Georgia Institute of Technology) and Avik Sinharoy (Georgia Institute of Technology)

Abstract As the task environment of a software artifact evolves, so must its design. For example, as the task environment in a computer game evolves, so must the design of the software agent that plays the game (or the agent’s behavior is likely to become more suboptimal than before). We are exploring how a software artifact may adapt itself as its task environment evolves incrementally. In particular, we are investigating how a game-playing agent may adapt itself as the percepts, actions, rules and constraints of its environment evolve from one version of the game to the next. A core research question in our work is what must an agent know about its design so that it can identify and make the right self-modifications to meet the needs of the new task environment? Our hypothesis is that the agent’s self-knowledge of its teleology (i.e., the mechanisms by which its design achieves its functions) may support the process of self-adaptation. In this paper, we describe the preliminary design of an interactive environment called GAIA in which a human game engineer and a game-playing software agent cooperatively adapt the agent’s software design and program code. As the game-playing agent uses its self-knowledge of its teleology to identify modifications to its design and code, the game engineer may (or may not) accept specific modifications and thus guide the process of self-adaptation. We also illustrate a first, simple example from FreeCiv, an interactive turn-based strategy game, at a high-level of specification.

Designs Can Talk: A Case of Feedback for Design Evolution in Assistive Technology by William Robinson (Georgia State University)

Abstract Design science researchers gain insights and make improvements on their designs, as they are applied in natural contexts. Software artifacts are particularly useful design instantiations because feedback can be obtained directly from the artifact. A software design talks to its designers.

We illustrate software design feedback with a case study in assistive technology (AT). A specialized emailing system was designed for cognitively impaired patients in an effort to decrease their social isolation, which often occurs after a brain injury. The patients continue to expand their email system usage, which is remarkable for AT. We attribute this unusual success to the feedback obtained directly from the software, through monitoring user goal models. Such monitoring has allowed the designers to understand and evolve their designs to meet the changing user needs. It illustrates how an operational artifact, like software, can drive design evolution faster than artifacts that are more static.

Adaptation Architectures by Kevin Sullivan (University of Virginia)

Abstract The complexity of emerging software-intensive systems, and the complexity and dynamism of the environments in which they are developed and operated, are growing to unmanageable levels. We now even foresee the emergence of ultra-large-scale (ULS) systems, which are project to be so complex that developing them will be unmanageable with the current scientific, engineering and managerial paradigm of largely centralized control and substantially top-down development [ULS]. Notwithstanding some important progress away from traditional methods, such complex systems will nevertheless still require significant advances beyond the current state of the art in design theory and methods. As the U.S. President’s Council of Advisors on Science and Technology (PCAST) put it, “As software’s complexity continues to rise, today’s... problems will become intractable unless fundamental breakthroughs are made in the science and technology of software design and development [PCAST].”

In this paper we focus on the need for fundamental advances in the particularly important discipline of software and software-intensive systems architecture. While recognizing that many important advances that have been made, we see advantages in exploring a significant change in perspective to support design for adaptation of complex software-intensive ecosystems. In this paper we outline one promising direction and summarize evidence developed to date in support of ongoing work in this direction.

PE-Plenary Speaker: Prasad Joshi (Infosys) (1:30 pm – 2:00 pm)

Enterprise IT – Quo Vadis?

IT is playing an increasingly important role in making business agile and innovative. CIOs are moving away from reactively aligning IT with business and are proactively synchronizing IT with business. The best practices powering agility and faster innovation include globally distributed work models that engage talent across the world, best of breed pre-configured industry vertical and business function specific IT platform solutions, and transaction based pricing models for generic IT capabilities. While most organizations are leveraging some or all of these practices, only a few have realized their full potential. IT organizations today leverage global talent for cost reduction but only a few have harnessed them for innovation. Agility is provided by IT platform based business solutions that make it easy to configure, run, and reconfigure business processes. Software as a service is becoming mainstream even in domains that were once considered “core” to the organization. So how can we create the blueprint for the next generation enterprise IT? Researchers and practitioners will have to co-create the blueprint for next generation enterprise IT using design research. The new paradigms, methodologies and tools that result in the next generation enterprise IT model should become part of the research agenda.

PE13-1: Invited Speaker: Anindya Datta (Walking Stick Solutions) (2:00 pm – 2:25 pm)

From the Research Lab to Enterprise Data Centers - My Experiences

in Commercializing Science

Moving ideas from the research lab to the marketplace is a long journey with few well-trodden paths to show the way. In this talk, Anindya Datta will describe his experiences in commercializing research-based ideas, e.g., radically differing expectations in academia and the marketplace, and applying for grants versus asking for venture funding, and why the experience is worth all the effort and work.

PE13-2: Panel I (2:30 pm – 3:30 pm)

Moderator: William Robinson (Georgia State University)

Panelists: Sham Navathe, (Georgia Institute of Technology); Sandeep Purao (Penn State University); Al Hevner (University of South Florida)

What makes design science research different?

Design science research (DSR) can be defined from a number of perspectives. Recent foundational articles present frameworks for understanding DSR. From inception to experimentation and test, many frameworks aim to be comprehensive rather than focus on a few core DSR concepts. Consequently, DSR is associated with quantitative, qualitative, and action research projects. This panel will discuss core concepts, methods, and technologies that distinguished DSR from more common IS research methodologies.

PE14-1: Invited Speaker: Jay Nunamaker (University of Arizona) (4:00 – 4:25 pm)

Design Science Involves Traveling the Last Research Mile

Design Science is a research methodology for creating new innovative IT artifacts. Traveling the Last Research Mile in IT means going from an idea to acceptance in the marketplace and having achieved wide spread use of the IT artifact. However, it is all too common to generate an idea, a concept, and a theory, develop a model, develop a prototype to test in the lab, write an article, dust off your hands and the task is complete. “The rest is just trivial details.” Yet the best academic adventure lies in traveling “the Last Research Mile.” The moon is 238,000 miles from Earth. Would you go 237,999 miles on a trip to the moon and then abandon the expedition if by going one last mile, you could find out what the moon is really made of? You might prove that Newton was right about a lot of things—a feather really will fall as fast as a cannon ball in a vacuum, and you really can jump six times higher on the moon. If you don’t go that last mile, you leave many key questions unanswered. Sometimes as you contemplate the Last Research Mile, the effort it requires may seem to be all uphill and f