Interaction Design Patterns: Generative Design Supporting Intercultural Collaboration
School of Design, The Hong Kong Polytechnic University, Hong Kong, SAR China.
Since Alexander first introduced design patterns to architectural design, various researchers have built on this approach to create and communicate reusable design knowledge. Interaction design patterns for remote collaboration represent a new and constantly evolving facet of the use of patterns in design.
So far, most collaboration patterns address arrangements of physical objects and spaces rather than the design of processes. In collaborative work, different teams may use different means of communication for different purposes but there are many similarities and commonalities among the processes employed. Patterns involved in work and social interaction can be seen to reoccur in other modalities of collaboration as well.
The author proposes that the design of interfaces to support intercultural remote collaboration can be informed by those more generalized interaction patterns. These interaction patterns would capture and document good solutions used to structure the collaboration process. This knowledge of interaction patterns can be reused to generate and custom-make interface designs for various collaborative modes and purposes.
This paper describes multiple methods - including observations, interviews and case studies in co-located and remote collaboration - which were used by the author to collect interaction patterns in collaboration. It presents preliminary results in the form of evolving interaction design patterns for intercultural collaborative learning and design, and gives examples of how such patterns could be used to generate designs.
Design patterns and pattern languages structure design knowledge into a format, which can be used to generate infinite numbers of good design solutions. A pattern, which is by definition  a good design solution to a problem in a certain context, evolves slowly over generations. Patterns are written guidelines collected into languages. There are numerous approaches to writing, sharing and using design patterns and pattern languages in the field of interaction design, very abstract human-computer-interaction design patterns , and more contextualized interaction design  and website design patterns  to collaborative interaction  and groupware patterns .
This paper explores the question, which pattern and pattern language format is capable to generate good interaction design solutions in the context of intercultural collaborative learning. It presents an interaction design pattern approach and pattern language framework to intercultural collaborative learning. The paper argues, that the collaboration goal and process are most relevant to structure a pattern language for a specific context of use, but ethnographic techniques like observations and interviews capture reoccurring good design solutions in order to write a single pattern.
Research has been carried out to share, compare and evaluate different approaches to pattern writing. Researchers generally agree that patterns have to evolve slowly in order to find a commonly accepted format. But so far researcher did not agree on a common degree of abstraction or focus of interaction design patterns and pattern languages. However, many patterns in the area of interactive and collaborative design remain using the original “Alexandrian” format .
In 1977, Alexander at al.  proposed a pattern language for architecture. They argued that architecture is not created but generated by events that reoccur, including the architecture for us experiencing it. Those events are living patterns indirectly generated by ordinary action of people. [2 pp.xi] Good patterns evolved slowly over generations. In order to document good design knowledge and make it available for reuse, Alexander at al. [1,2] collected patterns in architecture in order to abstract a shared pattern language. This language gives each person who uses it the power to create an infinite variety of new and unique designs. Patterns are described in “three-part rules, which establish a relationship between a context, a system of forces, which arises in that context [repeatedly], and a configuration which allows these forces to resolve themselves [in a good way] in that context.” [2 pp. 247]
It sounds simple, but pattern writing is challenging. Patterns should not be too wide or abstract, nether too limiting nor narrowing the context of use. Alexander’s method in identifying good patterns is easy, almost profane; You have to feel good about it! His suggestion how to identify and write pattern reads almost like a cookbook recipe. Observe and analyze instances of one pattern. Distinguish what feels good and bad about it. Identify and abstract the properties all good solutions have in common. Define forces, which balance this pattern. Describe the circumstances, which lead to a good solution. Define a range of contexts where the named forces bring a pattern into balance. [2 pp. 247-276] Patterns, identified in this way, slowly improve by evaluating and testing them against our experience.
However, one isolated pattern does not work well. They need semantic connections to other patterns. Alexander compared this phenomenon to natural language. A word needs to be related to another word to express a deeper meaning. Hence, a network of related patterns create a pattern language. Alexander et al.  created an immense architecture pattern language offering a network of large range (cities), middle range (buildings) and small range patterns (single components and details of buildings). But they stress that a living pattern language needs to be shared and needs to evolve like a culture. Different cultures share patterns, like relatives share a common pool of genes or natural languages share a common pool of language components. Pattern languages vary from person to person and culture to culture, but overlaps do occur.
Interaction design patterns cover a diverse area of research into human computer, social, collaborative or mobile interaction. Despite of writing style, naming convention, degree of abstraction or format, different pattern collections and languages show overlaps. Similarities of patterns can be found between such different contexts of interaction as websites , human computer systems  or interactive exhibits . As an example, one pattern, which reoccurs throughout the interaction design domain is called “Go back to a save place”  or “Home Link” . This pattern describes the phenomenon that users want to explore the computational space freely, but in case they get lost, users want to have a quick and secure escape, which takes them back to a known place or position. There are more patterns showing commonalities. Those shared patterns illustrate the emergence of a computational interface culture. Many people share this culture nowadays. Shared patterns show a common pool of widely accepted, good design solutions supporting a human interacting with a computational system.
Similar to Alexander’s approach, interaction design patterns can also be categorized into large, middle and small range patterns. As exemplified before, small range patterns show commonalities among various interaction design pattern collections and languages, but large range patterns differ most among various languages. As an example, patterns like Attraction Space , Commercial Website  or Collaborative Virtual Environment  closely relate to a specific interaction design application domain, context of use and interaction goal.
However, approaches to structure those collections and languages vary greatly. Patterns can be structured into networks, showing relations among them by families , hierarchy [14,15], or in parallel pattern languages, which are connected by a model .
1. Groupware Pattern Families
2. Shopping-Website Pattern Hierarchy
3. Model to use parallel pattern languages
Borchers  structures patterns into three separate domains to describe good solutions for interactive music exhibits. Musical patterns describe the application domain of interaction. A HCI pattern language describes the human-computer interaction structure with the exhibit. Finally, software patterns describe the computational implementation of musical and HCI patterns in the interaction process. This approach allows the integration of patterns into a very specific interaction context. Whereas Welie and v.d. Veer [14,15] established a hierarchical pattern language for website design composed of large (goal and posture), middle (experience) and small range patterns (task and action). This approach structures interaction and interface design patterns by interaction goal and process in a certain context e.g. Online Shopping .
Several forms and modes of collaboration and collaborative leaning are practiced in intercultural settings. Full or partial collaboration projects are carried out in educational contexts, e-learning environments or summer schools, using collocated, remote or mixed modes of collaboration.
On one hand, to define a framework of interaction context and goal is most important to write patterns for intercultural collaborative learning in the first place. On the other hand, doing ethnographic studies, observing actual teamwork gives valuable information about intercultural collocated and remote collaboration. As mentioned in [6-8], a design pattern is a good format to organize, present and represent a growing corpus of ethnographic material to outline reoccurring design solutions. Hence, I simultaneously approach writing interaction design patterns from bottom up - identifying the single pattern, and from top down - structuring an emerging pattern language framework.
I observe reoccurring events and processes in collaborative teamwork, where members do not share the same cultural background. I look for overlapping patterns of intercultural collaborative learning in varying settings. Patterns, which occur most often, all intercultural collaborative learning contexts share. In addition to observing single reoccurring patterns, a framework can give guidance. For this purpose, I adapted Welie’s website design pattern framework  (see figure 4). I found it very helpful to establish a pattern network, which describes intercultural collaborative learning processes as well as artifacts being used. Furthermore, I can relate to and extend from existing patterns in the HCI and groupware domain.
Pattern writing is an ongoing process of observing events that reoccur, abstracting them into design patterns and refining those patterns by testing them against experience. I use a human-centered research approach to identify interaction design patterns in intercultural collaborative learning. I observed design and accomplishment of several intercultural collaborative team projects in various contexts. Up to now I tested a few emergent patterns in case studies by conducting interviews and observing the collaboration process. Additionally to those case studies, I interviewed students and users of collaborative systems, design experts and researcher to gain an overview of practices in related contexts to help me evaluate emergent patterns from the observations and design case studies.
I participated in an international information design summer academy with the topic and title "Remote Relations - Tools for Collaboration", held by IIIDj - the International Institute for Information Design Japan. The invited international audience from the fields of design, business, computation and social science was composed of 20 postgraduate university students and young professionals from Japan, Korea, Hong Kong, India, Italy, Serbia, Germany, Norway and USA. The workshop was held in Ogaki, Gifu Province in Japan for 2 weeks in August and September 2003.
Aim of the workshop was the collaborative development of design ideas and scenarios to support remote relationships of individuals, companies and institutions. Participants had been split up into small teams of four to six people with a leader for each group being assigned previously. Each group applied different methodologies to explore this rather broad topic, such as structured or non-structured brainstorm sessions, change of physical work environment or modification and customization of tools and space. There were frequent short presentations within groups and to other groups. The workshop was structured in morning and afternoon table discussions, attended by all participant, continuous group work and lectures by design professionals. In addition to formal collaboration activities, participants had also been encouraged to share daily practices, activities and functions, such as lunch, dinner and accommodation.
In 2004, I participated in Convivio-3, the third international summer school in user-centred interaction design, which was held in Split, Croatia. Themes of study were “Communities in Transition” and “Sustainable Tourism”. Forty-four students from 14 countries including Bosnia, Croatia, Slovenia, Ukraine, Romania, Sweden, India, China and the United States participated in the 2-week program. In addition to attending daily lecture series, participants were organized into 4 atelier groups and asked to design solutions that addressed the local sustainable tourism needs of this post-war community-in-transition.
Each atelier was appointed a leader, recruited in advance. Participants were pre-assigned to the ateliers in efforts to construct teams that were well balanced along the dimensions of culture, research discipline, and institution of origin. Each atelier had a sub-theme, which was connected to the atelier leaders’ areas of expertise as Mobile Devices, Design Methodologies, Identity in Design and Games. Although the design methodologies used within the ateliers varied, they were ethnographically inspired. Each atelier utilized qualitative fieldwork such as interviews or observation and did prototypes in varying degrees of fidelity to inform their resulting designs.
Over a period of 2 years, I observed a university design studio subject titled "Only Connect", where students were asked to accomplish a design project collaborating remotely and collocated. The School of Design at the Hong Kong Polytechnic University organized and taught in collaboration with Design Schools in Korea, Austria and USA. There were approximately 110 Hong Kong participants and 50 international partners. Participants were undergraduate design and fine art students, from various design disciplines and specializations including industrial, environmental, architectural, interaction and fashion design, visual communication and media art. They collaborated in small teams consisting of 4 - 6 people in a stream specific field over a period of seven weeks, from October until December 2003 and 2004. Though distributed in geographical locations in Hong Kong, Austria, USA, Japan and Korea, the meeting point was a shared virtual project and team space in the Internet: the "Only Connect" Project Website [www.onlyconnect.sd.polyu.edu.hk] in 2003 and a “Weblog” team space [www.blogger.com] in 2004. In addition to the shared virtual space students were free to use email, instant messaging, and video conferencing to communicate remotely. In both years, Korean collaborators came to Hong Kong over a period of about a week to collaborate collocated.
My personal agenda for participation in the summer workshops and university courses was the observation of co-located and remote intercultural collaboration. I looked for practices employed to collaborate in intercultural teams within a given time frame and local constraints. I observed teamwork strategies, and solutions to overcome miscommunication and breakdowns in collaboration. My methods included observation, note-taking, informal, formal and contextual interviews with the participants, multi modal discourse analysis of online communication, and designing prototypes and case studies in different stages of fidelity.
Collaborative learning in intercultural contexts shows similarities but also differences to previously described interaction and collaboration design patterns. Similarities to existing interaction design patterns can be found predominantly in small range patterns of human computer interaction and groupware pattern domains.
Even though there are reports of using groupware applications in collaborative learning environments elsewhere [9-10], this study cannot approve the use of groupware especially in the context of intercultural collaborative learning and design. I believe this finding is important as it points out the need to support a flexible setup, as well as adaptation of work practices to the specific context of interaction in intercultural collaboration. The interaction domain of interaction design patterns for intercultural collaborative learning should not be limited to groupware applications. I found that the interaction and collaboration goal and process suggest means and structures of interaction appropriate in the specific context. In addition to this, my findings correspond to Graveline, Geisler and Danchak’s  report of emergent patterns in media use and media richness in remote collaboration, using predominantly emails and instant messaging as forms of group coordination and communication. However, under certain conditions in collaborative learning additional interaction processes including communication, coordination and awareness mechanism need to be supported.
Adapting a website design pattern classification  I propose to structure patterns by collaboration goal and process (goal, posture, experience, task and action) on the x-axis, and collaboration mechanism (awareness, coordination and conversation) identified in , on the y-axis, into a hierarchical large, middle and small range patterns network. The following graph shows a rough sketch of an evolving network of intercultural collaborative learning design patterns using this framework. It exemplifies a pattern connection from the collaboration goal of PEER-LEARNING to the interaction strategy INTER-LINKED TEAM SPACES and the interface element MAIN NAVIGATION.
This example of a pattern network and one pattern is work in progress.
4. Section of an interaction design patterns network for intercultural collaborative learning.
5. “Only Connect” Collaborative Learning and Design Project, Toolkits03 Team Blog, (http://omega.sd.polyu.edu.hk/~blog/onlyconnect/product9/)
Context and Forces (upper level patterns)
… your collaborative learning environment consists of more than one team. For this project all teams work collaboratively with their respective international team members as partners – COLLABORATORS AS PARTNERS. In this context it is important that all teams and team members are treated equally, and are encouraged to freely exchange any kind of information about the project - PEER-LEARNING. You provide an online community space to share information and comment on the team progresses– BLOG or WIKI utilizing a remote collaboration mode. In collocated collaboration you provide a shared physical space, which is divided into team sections or public accessible rooms - TEAM CORNERS.
Intercultural teams profit from being aware what other teams are doing, aiming at and how they progress, especially when they work on similar projects. However, in intercultural collaboration, teams tend to focus on their potentially unfamiliar and new members within the team, missing the opportunity to share with other teams working on the same or similar topic.
Blogger.com offers the possibility to change the basic blog-page layout within a template structure to adapt to specific needs of blog users. The template format is based on html. Hyperlinks can be positioned at different places on the page, linking different contents and spaces. The “Only Connect” collaborative learning project makes use of this possibility to link different teams in the same university collaboration course. Those hyperlinks are placed in the main navigation space in the upper part of the website. User can browse though other team blogs, compare the information collected and reflect on their respective project progress. Collaboration and Communication experiences with international team members can be viewed or directly and indirectly exchanged in order to learn from each other’s mistakes and solutions.
Summer academies make use of available rooms and common areas in the university environment, which usually hosts this workshop. Different collaborating teams are placed in proximity to each other, whether in a section within a shared room, or in public accessible team rooms. All member of the workshop can walk into all team spaces, and share their views encouraged by the visibility and transparency of the team’s collaboration process.
Make a team space always accessible and easy to approach to other teams participating in the same collaborative leaning context. Provide visibly well placed connections to relate team spaces having a similar or the same topic of collaboration.
References (lower level patterns)
In order to point out the importance of this peer-learning option, links should be placed within an easy accessible part of the team space – MAIN NAVIGATION.
The applicability of those patterns is to generate different instances of intercultural collaborative leaning environments. Environments, which utilize remote, collocated and mixed-mode collaboration, are summer workshops, university collaboration projects, e-learning systems and courses. A collaboration and interaction goal and process oriented pattern language gives the designer the power to not just to generate but custom-make intercultural collaborative learning environments. Within my Ph.D. research, this paper presents a first attempt to define an approach and classify a framework for such a pattern language.
Alexander, C. S., Silverstein, M., Jacobson,M., Fiksdahl-King, I., Angel, S. (1977): A Pattern Language. New York: Oxford University Press
Alexander, C. S., (1979): The timeless way of building. New York: Oxford University Press
Graveline, A., Geisler, C., Dancha, M. (2000): Teaming together apart: emergent patterns of media use in collaboration at a distance. Proceedings of IEEE professional communication society international professional communication conference. Piscataway, NJ, USA
Hughes, J., O'Brien, J., Rodden, T., Rouncefield, M. and Viller, S. (2000), Patterns of home life: informing design for domestic environments, Personal Technologies, 4 (1) : 25-38
Lancaster’s pattern collection:
Martin, D., Rouncefield, M., Rodden, T., Sommerville, I. and Viller, S. (2001), Finding patterns in the fieldwork, In Proceedings of ECSCW'01, Bonn, Germany: Kluwer)
Martin, .D, Sommerville, I. (2003): Patterns Of Cooperative Interaction – A Brief Introduction To The Lancaster Perspective. ECSCW 2003, the 8th European Conference on Computer Supported Cooperative Work in Helsinki, 14. - 18. September 2003. http://www.groupware-patterns.org/ecscw2003
Preece. J.; Rogers, Y.; Sharp, H. (20002): Interaction Design: beyond human computer interaction. New York, Wiley & Son Inc.
Rutkowski, A.-F., Vogel, D., Bemelmans, T.M.A., van Genuchten, M.: (2001): IEEE Transactions on Professional Communication, VOL. 44, NO. 2, June 2001
Rutkowski, A.-F., Vogel, D., Bemelmans, T.M.A., van Genuchten, M.: (2001): E-Collaboration: The Reality of Virtuality. IEEE Transactions on Professional Communication, VOL. 45, NO. 4, December 2002
Schuemer, T. (2003): Evolving a Groupware Pattern Language. ECSCW 2003, the 8th European Conference on Computer Supported Cooperative Work in Helsinki, 14. - 18. September 2003. http://www.groupware-patterns.org/ecscw2003
Tidwell,J, Common Ground HCI pattern collection:
van Welie, M., van der Veer, G.C (2003): Pattern Languages in Interaction Design: Structure and Organization. Published at In Proceedings of Interact 2003 International Human–Computer Conference. Zuerich, Switzerland
van Welie, M.: Interaction Design Patterns.