IA Summit 2006: Main Conference details

Saturday, March 25

Sunday March 26 Monday March 27

Opening Keynote

Saturday, March 25, 8:30 - 10:00
Dr. David Weinberger

Dr. Weinberger began his "career" in the late '70s teaching philosophy at New Jersey's Stockton State College for five years. (He has a Ph.D. in philosophy from the University of Toronto.) During this time he maintained his steady freelance writing of humor, reviews and intellectual and academic articles, publishing in places as diverse as The New York Times, Harvard Business Review, Smithsonian, Alfred Hitchcock's Mystery Magazine and TV Guide. In 1985, after being denied tenure because the tenure quota was filled, and after an enthusiastic but well-mannered student demonstration in his support, he became a junior marketing guy at Interleaf, an innovative start-up with new ideas on how to create and structure documents. At Interleaf he helped launch the industry's first document management system and its first electronic document publishing system, years ahead of the Web. He left Interleaf after 8 years, as VP of Strategic Marketing.

He founded the one-person strategic marketing company, Evident Marketing, in 1994 and within two years counted among his clients a wide variety of companies, including RR Donnelley, Intuit, Sun Microsystems, Esther Dyson's Release 1.0 and CSC Index.

In late 1995, he joined Open Text as VP of Strategic Marketing because he saw an opportunity to help shape the way intranets are used. As part of the senior management team, Dr. Weinberger helped Open Text move from one of the first Web search engine companies (the engine behind Yahoo!) to market- and thought-leadership in Web-based collaborative software.

After helping to take Open Text public in 1996, Dr. Weinberger returned to consulting, writing and speaking, helping to found a couple of dot-coms, and serving on industry and company boards. In 2000, Perseus published The Cluetrain Manifesto, of which is is a co-author. It became a national best-seller.

In 2002, Perseus published Small Pieces Loosely Joined to enthusiastic reviews.

Dr. Weinberger currently writes too much, including 3 weblogs, articles for Wired, Salon, USAToday, Esther Dyson's Release 1.0, and many more. During the 2004 presidential campaign, he was Senior Internet Advisor to the Howard Dean campaign, consulting on Internet policy. In 2004 he was made a Fellow at Harvard's prestigious Berkman Institute for Internet & Society. http://www.hyperorg.com/speaker/bio.html

Download slides from the opening keynote (22MB, PPT).

Setting the Agenda for IA Research

Saturday, March 25, 10:30 - 12:15, Track 2
Don Turnbull, Peter Morville, Jamie Bluestein, Keith Instone

IA practitioners often say that too much research related to IA is often abstract or not useful. When then asked about the kinds of difficult problems they are facing that some empirical or theoretic research and innovation could help with, they are often at a loss.
With this in mind, this panel will help to frame issues in IA research to synchronize ideas with those in the academy and the field. A set of upcoming research agendas that show both the promises and the perils inherent in Information Architecture as a whole. The overall goal of the panel is to allow for the panel and the audience to clearly communicate their wants and needs to devise some immediate scenarios to begin building a useful body of IA research. This panel will have a substantial Q&A session to share ideas and build working relationships.
Each panel member will discuss his take on IA research, along with their own core areas of research and suggestions for cooperation between researchers and practitioners:

• Peter Morville - findability. The future of this area of research (and how attendees can help by providing sites for study or work processes to observe).
• Keith Instone - faceted browsing. What's next in this exciting area, how tools could support it and what can be learned from their use or applied from other research areas (cognition, task analysis, etc.)
• Jamie Bluestein- how hypertext research can inform IA (and what we should look forward to when merging these two areas).
• Don Turnbull - a behavioral model of information seeking (user search and navigation). A review of this area of research and a call for participants to help update the model previously published with user data collection and site log analysis.

Download slides for 'Setting the Agenda for IA Research' (143KB, PPT).

IA: Not Just for the Web Anymore

Saturday, March 25, 10:30 - 12:15, Track 3
Dan Brown, Lou Rosenfeld, James Melzer, James Robertson, Seth Earley

Overview: More and more information architects are discovering that their services are employed outside the scope of their organization's or client's web site. Organizations face numerous information management problems: from archiving old data and making it accessible to managing the document creation process and capturing informal knowledge. When information architecture must influence an enterprise's overall information management -- process, policy, and implementation -- the IA becomes Enterprise IA. This panel will shed light on these challenges and discuss strategies for addressing them, leaving attendees with some ideas on how to they can bring enterprise thinking to their own day-to-day work. The panel will explore how to jump-start enterprise-wide IA projects and give attendees some ideas for how to put together a long-term strategy.

Description: Some large organizations are taking a more holistic approach to their information management strategy, merging practices from disparate fields such as data standardization, web content management, and records management. This approach is known in the IA community as enterprise information architecture. Enterprise IA (EIA) directly addresses organizational outcomes, business culture and change management. Its domain is more strategic than web-based information architecture, which focuses on tactical design decisions about navigation and user experience. EIA is a nascent area, and we're just beginning to map it and develop language to describe its basic concepts and challenges (much as we did with "traditional" IA a decade ago).

Though enterprise information architecture is relatively new, a handful of practitioners have been developing theory and practice around it. This panel will explore the work done in this niche to-date. Attendees will hear how the panelists have engaged clients in enterprise-level discussions, collaborated with professionals in related fields. The panel will solicit stories and case studies from the attendees, to highlight techniques used throughout the community. By the end of the session, the panel and attendees together will have identified EIA challenges and shared tools for addressing them. The tone of the panel is conversational: attendees should feel that they can participate in an industry-wide dialog on how to do enterprise information architecture.

We Are Not Alone: IA's Role in the Optimal Design Team

Saturday, March 25, 10:30 - 11:15, Track 1
Jared Spool

In most organizations, design is not a solo activity. Different individuals, each with their own set of skills, come together to create an ideal user experience. Yet, we know very little about how to form these groups, and, more specifically, how to best do our jobs within them.

By studying several dozen design teams, the UIE research team has spent the last few years looking at how different compositions lead to the most effective results. How do you build a team to produce the best possible designs?

In this presentation, Jared will discuss the three models of design team organization: Consulting, Review &Approve, and Educate &Administrate. He'll talk about the different approaches that teams take, where they succeed, and where they run into trouble.

He'll also discuss the differences UIE has found between the approach of a centralized user experience team and distributed approaches that put skilled individuals within each of the development teams. He'll talk about the issues of scaling design to meet the needs of the enterprise once its value becomes realized and the pitfalls of specialization.

Taxonomy Testing and Usability

Saturday, March 25, 11:30 - 12:15, Track 1
Joseph Busch, Ron Daniel, Jr

Effective taxonomies are not carved in stone. They must be modified based on changes in the content, user needs, and budgeting realities. Developing taxonomies in an incremental fashion requires that we be able to measure how well it is working in order to plan its next version. How is that done? This talk describes qualitative and quantitative taxonomy testing methods. It answers questions such as:

• Tagging just enough content to see if it works is a good first step to test taxonomies, but how much content is enough for validation?
• What other approaches have been shown to be effective such as open and closed card sorting, use-based scenario testing, and focus groups?
• How should test results be evaluated, what do they mean, and what corrective steps can be taken?

Download slides for 'Taxonomy Testing and Usability' (1.6MB, PPT).

Exploring patterns in website content structure

Saturday, March 25, 1:45 : 2:30, Track 1
Svetlana Symonenko

This paper describes a pilot study for ongoing dissertation research on the indications of conventionalization in the observable structure of website content. Until now, research has mainly focused on interface-related aspects of Web-based information (Research-Based Web Design & Usability Guidelines, 2002; Ivory and Medgraw, 2005; Morville and Rosenfeld, 2002) or on content-related aspects at the page level (Bates and Lu, 1997; Dillon and Gushrowski, 2000; Haas and Grams, 2000; Ihlstrum and Ekesson, 2004; Rehm, 2002), with insufficient attention to patterns in the organization of content of entire websites. To fill this gap, the dissertation will explore regularities in websites' content structure and in users expectations of such structure.

The study reported here piloted the phase of the dissertation, which will assess the extent of regularities in the observable content structure of websites of two types by means of qualitative content analysis. For the purposes of the study, websites are categorized by the type of their originator (Shneiderman, 1997), and, thus, are contextualized in particular discourse communities, which include site sponsor and target audience. To control for cultural differences, the scope of the study is limited to sites in English and by U.S.-based entities.

Download paper for 'Exploring patterns in website content structure' (295KB, PDF).

Download slides for 'Exploring patterns in website content structure' (881KB, PPT).

Building bridges between information behaviour research and information architecture

Saturday, March 25, 1:45 - 2:30, Track 2
David G. Hendry

Information Architecture is unquestionably a design profession. As such, its practitioners seek to change a current situation into a preferred one (Simon, 1996). Projects in Information Architecture, thus, restructure or expand the built world (Schön, 1991), even if it is computer mediated and often intangible. Because the outcomes of Information Architecture lie in a built world, and a large and diffuse world it is, it can be very difficult to set useful boundaries on the subject. Consider that to some degree the traditions of social science, information technology, and design all influence the practice of Information Architecture. Indeed, like the field of Human-Computer Interaction, Information Architecture is a boundless domain (Barnard, May, Duke, & Duce, 2000). How do we flourish in a boundless domain?

Donald Schön answers this question by proposing an epistemology for practice, where professionals learn to Reflect-in-Action (Schön, 1991). Under his view, knowledge is generated in the process of doing. The reflective practitioner applies a discipline and he "strives to make the situation confirm to his view of it while remaining open to evidence" (Schön, 1991, p. 74). In a virtual world (e.g., a sketchpad), free of the constraints of a built world, the professional applies exploratory experiments to a representation only for the sake of seeing what happens or tries move-testing experiments to examine specific implications. By holding a conversation with his materials and listening to the backtalk the designer, according to Schön, is able to change a problematic situation for the better. In this way, the problem is defined only with its solution is fully expressed (Rittle & Webber, 1974).

On the other hand, when conducting research on information behavior, the social scientist normally embraces a different ethos. He or she values fidelity of description of observed phenomena and the invention of models that hold explanatory or predictive power. Objectivity is assumed. Abstraction, within appropriate bounds, is the gold standard. Design and research, in short, entail different standards for creating and valuing knowledge.

One might reasonably expect that research findings about the current conditions of the built world can significantly inform design. In other words, descriptive studies, models or theories of information behavior should be important to information architects who seek to change patches of the built world by enabling new information experiences. Experience, however, shows that this is not always the case; indeed, the knowledge generated in the act of design is often ahead of what knowledge exists in a research base. But, even more critically, research that is intended to be applied in design is often ineffective. Indeed, much theory and many research approaches in Human-Computer Interaction have not been useful to practitioners (Rogers, 2004). Fisher, writing as Pettigrew, Fidel, and Bruce (2001), furthermore, conclude that research studies of information behavior have had little significant impact on information system design.

In this essay, we examine the issues that surround the application of Information Behavior Research in Information Architecture and seek to understand how the relationship between these two areas of inquiry can be enhanced. The argument we shall make is that these two fields are ideally positioned to benefit for each others' unique perspectives. In fact, they are likely better positioned to mutually benefit than the fields of psychology and computer science were at the beginning of the 1980s when interest in Human-Computer Interaction began to grow rapidly (Carroll, 1997). To enjoy the benefit, however, dialog is needed so that the relationship between the fields becomes one of "mutual understanding" rather than ignorance. We seek to facilitate this dialog.

Download paper for 'Building bridges between information behaviour research and information architecture' (27KB, PDF).

Game Changing: How You Can Transform Client Mindsets Through Play

Saturday, March 25, 1:45 - 2:30, Track 3
Jess McMullin

A significant barrier to success on many consulting projects is lack of buy-in from business stakeholders. Sometimes this shortfall of support simply requires IAs to be fluent in the language of business and communicate the business benefits of a project. However, the challenge is often deeper when business stakeholders hold a mindset that doesn't appreciate the approach or objectives of a project. It's these mindsets that are often dismissed with a rejoinder that they just don't get it'. Instead of this abdication, IAs can actually change mindsets to increase project success.

Changing mindsets takes more than talking with business people in their own language, it takes transformative experiences that make perspectives flexible and open to new views. These reframing experiences most often come through participation in successful projects, but this leads to a chicken-and-egg problem - how do we provide success with IA projects for an audience of people that are causing failure? The common answer is to run a pilot or prototype to tackle low-hanging fruit and build buy-in with a track record of success - but what happens when you don't have time, budget, or support for a pilot?

One avenue for accelerating mindset change is playing design games. Games create low-risk situations that let stakeholders experience new perspectives, approaches, and methods without the time, expense, and commitment of a pilot. By scaling human-centered design principles and methods down to the level of games, we can rapidly expose entire teams to new ways of thinking, opening up their world views, and offering valuable perspectives that can bring new appreciation for IA to their mindset.

This session will mix principles with examples to help IAs use games that can change business mindsets and develop increased buy-in for human-centered design and information architecture. Fundamental principles will lead to a discussion of various game options and outline how IAs can pitch games as a serious tool to facilitate project progress. Attendees will have the opportunity to participate in some gameplay during the session.

Download slides for 'Game Changing: How You Can Transform Client Mindsets Through Play' (1.4MB, PDF).

The International Information Architecture Slam

Saturday, March 25, 2:45 - 5:45
Eric L. Reiss, Matthew Fetchko, Chris Chandler, Lynn Boyden

The rules of engagement for the IA Slam call for the brief presentation of a consulting scenario by the Instigators (that's us), followed by the completely random division of participants (that's you) into teams who work furiously to develop a response to the problem in less than an hour, using only the classic IA tools of paper, post-its, and markers. At the end of the breakout session, teams present their solutions to the Instigators (and the other participants as well). The Slam requires that participants employ not only the core tenets of IA theory, but also those that inform the realpolitik of IA practice: managing team politics, identifying the problem, considering the customer experience, brainstorming and designing solutions, and communicating that solution.

Solutions are judged on a wide variety of criteria, including plausibility, feasibility, teamwork, leadership, project management, client relationship management, presentation skills and format, and problem identification and solution. Other criteria, too arcane to elaborate in this forum, are also evaluated to arrive at a clear winner of the annual Information Architecture Slam medal. The judges' decisions are often arbitrary but always final.

Re-invoking Culture and Context in Digital Libraries and Museums

Saturday, March 25, 2:45 - 3:30, Track 2
Ramesh Srinivasan

I will argue for the creation of digital library, museum, and other collections with an acknowledgment of the local context and culture of objects. This can be achieved without sacrificing interoperability issues. The approach in this panel will describe how to begin to achieve such a harmony between local knowledge of an object and standards, information sharing, and information access.

Selling IA - Getting Execs to say Yes

Saturday, March 25, 2:45 - 3:30, Track 3
Samantha Starmer

The purpose of this session is to provide some specific recommendations on how to sell IA to executives. Whether you want to begin a particular IA project, increase IA staff or get more budget for IA work, it is likely that at some point you will need to submit a proposal to some level of executive or to those who control resource allocation. I will review my top recommended strategies from my experience of getting IA projects resourced and completed with executive approval and support.

1. Show the problem and how you can fix it
2. Indicate benefits to the bottom line
3. Manage the politics
4. Don't sell a silver bullet
5. Pay attention to style

Selling IA should be looked at as a full project itself, and one where understanding your users and their environment is just as important as with any other IA or user experience task. It can also be one of the most fun parts of IA because when you are successful the benefits to your own career and to IA in your organization can be great and long lasting.

Download slides for 'Selling IA - Getting Execs to say Yes' (4.34MB, PPT).

Architecting self-organizing learning communities

Saturday, March 25, 2:45 - 3:30, Track 1
Faison "Bud" Gibson

Higher education frequently boils down to one-way learning communities built around professors, where a lot of information other participants could provide lays fallow. We attempt to crack this problem by creating an online, self-organizing knowledge sharing space for participants in a course at Michigan's Ross School of Business. Our prototype illustrates how information architects can harness simple technologies coupled with social processes to dramatically increase participation and information in knowledge communities.

Substantive Issues Addressed: A perennial issue in knowledge and learning communities is exposing the full range of information available from participants. Limited time and social dynamics frequently dictate that only a few will participate. Then there is the question of how to organize the information that is contributed. This proposal outlines a web-based learning remix project at Michigan's Ross School of Business that is designed to remove limits on participation and simultaneously highlight participants' locus of attention by topic. The system is based on three principles:

1. Everyone participates.
2. Almost all participation is online.
3. Online participation is self-organizing.

Our system achieves the first and second goals and makes good progress on the third. On the input side, each participant makes fifteen microcontributions per week by bookmarking web pages in del.icio.us and posting to their personal blog. Multiple times each hour, our publishing platform gathers, remixes, and presents these contributions based on the tags participants themselves have supplied.

One of the things that most excites us about the remix is that it provides an almost real-time indication how participants are attending to external items found on the web and internal items generated by their peers while simultaneously creating a structured, searchable long-term archive. Tags indicate how participants want their and others' contributions organized in the overall record of the conversation. The frequency of linking to given items and the frequency with which tags are used point to the central topics.

We are using simple behavioral approaches in combination with technology to bring coherence to the remixing process. In class, we come to consensus on key tags. Participants also consult the class tag cloud and use the same tags as the items their contributions follow. Thus, we have an evolving and self-organizing archive of class attention located at: http://thecommunityengine.com/bit320/remix/

Download slides for 'Architecting self-organizing learning communities' (608KB, PDF).

From task to activity: A case study of developing for innovation

Saturday, March 25, 4:00 - 4:45, Track 2
Nancy Kaplan

Donald Norman conjured up a storm of commentary when he published his essay, "User Centered Design Considered Harmful." The ensuing discussions on the SIG-IA were voluminous and occasionally heated but shed little light on important pragmatic questions:

• What is activity centered design anyway?
• What does it look like in practice?
• How do its outcomes differ from those produced by UCD methodologies?

In this research paper and presentation, I will address a number of those questions through the lens of a case study, a development project in which UCD practices gave way to ACD practices and in which ACD methodologies yielded design directions that differed from those that were emerging from UCD investigations. The presentation will highlight important methodological differences between UCD and ACD and will report on three studies of the web-based software that we developed as a result of ACD methods.

In the course of the presentation and the full paper, I will provide definitions of "tasks" and "activities" to show how the differences between these two analytic tools produce strikingly different outcomes. I will also discuss ways that academic research can support and augment more pragmatic and results-driven projects in business settings. I aim to help information architects understand more fully how methodologies that have become professional best practices shape the work that they do and how careful re-examination of accepted practices can inform and even transform our business.

In short, the presentation will offer some answers to questions like these:

• What is Activity Centered Design?
• How do we know that ACD methodologies are valuable?
• What methodologies for doing ACD are useful?
• How can I use these methodologies in my IA and XD work?

Download slides for 'From task to activity: A case study of developing for innovation' (3.3MB, PPT).

A Room of Our Own: Starting IA Locals and Bringing IA to Work

Saturday, March 25, 4:00 - 4:45, Track 3
Stacy Merrill Surla

Defining, inventing, and teaching IA practices is very important work. Someone's got to do it! But creating environments in which IAs can be IAs is just as important. The most fertile IA seeds can't bear fruit if they have only stony ground to support their growth. Fortunately, creating IA-friendly environments is not magic. It's a skill that can be learned and practiced by anyone willing to summon the nerve and put in the effort.

This session will describe strategies for creating and sustaining two kinds of IA environment - local IA groups and IA culture at work. This session is for you if:

• IA is a total mystery to your boss, or work provides little scope for IA practice
• You can't talk with friends about your exciting IA projects without first explaining and justifying yourself into a state of exhaustion
• You have created supportive IA environments in the past and now want some new ideas

You will learn:

• The different challenges and opportunities inherent in organizing for local groups and for work culture
• The role of the local group as the "IA Spa"
• Using IA to contribute project value and career currency to your organizational unit
• How-to skills for building local groups in two critical areas: volunteer organization and party planning
• Furthering your career, one way or the other, through bringing IA to work

Download slides for 'A Room of Our Own: Starting IA Locals and Bringing IA to Work' (61KB, PPT).

Research BOF

Saturday, March 25, 5:00 - 5:45, Track 1

BOF - Tagging

Saturday, March 25, 5:00 - 5:45, Track 2

Business & Design BOF

Saturday, March 25, 5:00 - 5:45, Track 3
Jeff Lash

This Birds of a Feature session will allow for attendees interested in the intersection of business and design to "flock together" for a moderated discussion.

As Selling IA becomes more important, the distance between the often juxtaposed areas of business and design is decreased. In 2005, we saw a dramatic increase in awareness in the business community about the power of design. With more design concepts filtering into traditional business disciplines, there are opportunities and expectations for information architects to understand and influence the business environments.

The session will begin with a brief presentation about the current state of business + design to set the stage for the discussion. More experienced practitioners may have more to contribute to the discussion, but attendees at all levels should benefit from listening.

Sunday, March 26

Monday March 27 Saturday March 25

Wireframes: A comparison of purposes, process, and products

Sunday, March 26, 8:30 - 9:15, Track 1
Anders Ramsay, Dave Heller, Jeff Lash, Laurie Gray, Todd Warfel

Few areas in information architecture generate such heated debate as wireframes. A staple of the information architect toolkit, wireframes were originally very simple and somewhat-standardized documents created by a limited number of people.

As the discipline has progressed and as web development methodologies have grown up, the context in which wireframes are created and used has changed. Recent discussions in the IA community show not only the differing philosophies behind the purpose of wireframes but also the varying processes and tools used to create wireframes.

This session will present a wide variety of different approaches to wireframing. Each presenter will discuss a different approach to wireframing, describing the purpose of wireframes, showing how they are used in the development process, and detailing the tools and techniques used, with a fair and honest discussion of the benefits and drawbacks of the approach.

Panelists will address questions such as:

• How would you define a wireframe?
• What purpose do wireframes serve?
• Who are the audiences for wireframes?
• How do wireframes fit in with other documentation produced - both by IAs and by others - for the project?
• How does wireframing fit in with other activities, such as prototyping and usability testing?
• How does wireframing change based on the context (innie vs. outie, timelines, budget, size of design team)?
• Where in the process do wireframes fit, and when does wireframing start/end?

The goal of the session is not to agree on one preferred approach, but instead focus on the context in which wireframes are used, and present different options for different situations. Audience participation and interaction with the presenters will be encouraged. Attendees should obtain a better sense of the various options for wireframing, knowledge of when each method is appropriate and what its benefits and detriments are, and new ideas for approaching the wireframing process.

Download slides for 'Wireframes: A comparison of purposes, process, and products' (1.2MB, PPT).

Indexing consistency & its implications for IA

Sunday, March 26, 8:30 - 9:15, Track 2
Hope Olson, Dieter Wolfram

Consistency in indexing has long been considered essential for effective information retrieval. In the process of indexing, indexers choose what topics to represent and what to call those topics. The goal is to select and name topics consistently so that all of the material about any given topic will be found together. Decades of research on consistency between indexers and by the same indexer at different times has documented medium to high levels of inconsistency. People simply do not choose the same concepts or the same words for those concepts. This body of research has identified various factors that contribute to consistency, but it has not addressed the underlying nature of consistency. Can consistency be achieved or is inconsistency inevitable? Does it follow predictable patterns? Consistency is also essential to good information architecture (IA). When developing access points to content, reliance on commonly used terminology is needed for effective access.

One thing we know about consistency is that indexers usually agree on a core of topics, but vary in representing peripheral ones. This suggests that the choices might adhere to a power law that can predict the distribution of topics. If the distribution of topics indexed is predictable it might be possible to minimize the variation or to develop interfaces for searching databases that would take that distribution into account. It could also be that the variation is a positive factor. This study explores these possibilities.

Previous studies that have explored consistency generally have data from only a handful of indexers. That is where the current study differs. It gathers data from a much larger group of people than previous studies. The data for this pilot study were derived from responses by MLIS students in an indexing exercise. Data were collected from two sections (one online, one onsite) of a course on Information Organization. Students were asked to provide five terms that describe what an assigned conference paper is about. Students were asked to submit their terms without identifying themselves. Aggregate data were tabulated based on the frequency of occurrence of terminology.

Initial findings indicate that over a large number of people, consistency in identifying key concepts is small. In fact, the frequency distribution of common topics follows a strong inverse pattern, with few terms being agreed upon by many participants, and many uniquely identified terms. The relationship, although not completely Zipfian (i.e., following a classic power law), demonstrates the wide degree of scatter in terminology used by participants in the study. It follows that information searchers who try to identify query terms for searching for information will also rely on different terminology to describe a similar concept of interest.

Recognition of these differences has implications for the design of systems. Multiple access points that accommodate the different ways that users interpret content are needed so that users may be guided to relevant content despite using different terminology.

This project addresses the conference themes of "learning" and "doing." By studying patterns in user behavior regarding the assignment of topicality of system content, systems developers may incorporate an understanding of these patterns into information system design.

Download paper for 'Indexing consistency & its implications for IA' (94KB, PDF).

Ambient Findability

Sunday, March 26, 8:30 - 9:15, Track 3
Peter Morville

At the crossroads of ubiquitous computing and the Internet, the user experience is out of control, and findability is the real story. Access changes the game. We can select our sources and choose our news. We can find who and what we need, when and where we want. As society shifts from push to pull, findability shapes who we trust, how we learn, and where we go. In this cyberspace safari, Peter Morville explores the future present in search algorithms, embedded metadata, ontologies, folksonomies, mobile devices, findable objects, evolutionary psychology, and the long tail of the sociosemantic web.

Download slides for 'Ambient Findability' (26.9MB, PPT).

Wireframing Challenges in Modern Web Development

Sunday, March 26, 9:30 - 10:15, Track 1
Nathan Curtis, Bill Scott, Livia Labate, Thomas Vander Wal, Todd Warfel

As technology and customer expectations continue to evolve, information architects are faced with a growing list of challenges. Our industry has been plagued with a poor toolset for years, but we've managed to make ends meet and create workarounds. Now with things like AJAX and Agile Development gaining press and coming into the spotlight, we're faced with the challenges of documenting RIA interactions, transparent transitions, and non-page based metaphors with a static page-based toolset. Something has to change.

Metadata games: cutting the metacrap

Sunday, March 26, 9:30 - 10:15, Track 2
Karen Loasby

Cory Doctorow may have been right when he said there is no point in getting authors to apply metadata to their content. Not to help us search the whole internet, at least. For him, people are lazy and stupid when it comes to metadata. Getting every kid in their bedroom and every proud grandmother posting pictures of her grandchildren to tag well enough to fix websearch... well that is indeed probably a pipe dream. But does that mean the same is true for you and your company and your company's employees.

Journalists are wilful, opinionated and rebellious (that's why they're good at their jobs) but they are definitely not lazy or stupid. It's more that when it comes to metadata they are lacking in motivation. This isn't just true of journalists. People make choices about what matters to them and what they will take time to do well. Our job is to make metadata matter to them.

In this session Karen will demonstrate some of the games she and the BBC IA team use to make metadata matter to their journalists and web producers, hopefully amusing them along the way. There's not much presentation involved as you will be expected to join in with the games for most of the session.

Download slides for 'Metadata games: cutting the metacrap' (1.3MB, PPT).

Design Patterns in the Real World

Sunday, March 26, 9:30 - 10:15, Track 3
James Reffell

Design patterns are a nice concept, but how do they fare in the real world? This session will draw lessons and pose questions based on the experience of that eBay's User Experience & Design organization. Large UI and IA groups face enormous challenges in the creation and dissemination of design patterns, IA guidelines, style guides, and similar approaches. These challenges include:

• Scarcity of time & people to document & disseminate the design patterns
• Difficulty of establishing guidelines across different and widely distributed business units
• Pace of change - technology, understanding of users, business requirements are all changing at an exceedingly rapid rate
• Balancing the need for extremely exact guidelines (e.g. consistent visual specifications) versus flexible design patterns that are useful for new situations

At eBay, we've implemented an extremely lightweight, flexible approach to these challenges. This approach was born of necessity (due to limited resources) but has been surprisingly effective. The approach is based on a few core principles:

• Move quickly and don't think too hard
• Incent mass documentation
• Allow and acknowledge messiness
• Allow no exceptions (subvert them!)
• Document the pattern and the specification
• Push patterns into code whenever possible

This session will illustrate these principles through real-world examples, including the successful and the unsuccessful. Following the presentation, lively discussion will be encouraged around the key topics.

A Process By Any Other Name...

Sunday, March 26, 10:45 - 11:30, Track 1
Adam Polansky

• Why do a wireframe when a cocktail napkin will work?
• But, don't use a cocktail napkin if a wireframe is needed.
• How does an IA decide what tools to use?
• Where can they get examples, stencils or templates?

At its highest level, the purpose of an IA is to close gaps between groups. Designers need to understand the functions of an interface. Developers need to understand what to code. Business owners need to understand how their requirements are being accounted for. And, visitors to a site need to understand how to find what they are seeking.

The IA exists at all of those junctures. Their talent lies in finding a way to communicate information from one group to the other in ways they can understand.

The IA confronts the challenges that come with being a liaison, referee, sage, counselor and translator. One of those challenges has to do with "Process". Rational, Agile, Waterfall, De-centralized, Collocated - These are all terms that invoke specific ideas about how a product is delivered.

One area that is becoming more prescriptive surrounds the development of the deliverables. Sitemaps, Wireframes, Personae and Functional Specs are some of the most commonly expected artifacts that are associated with Information Architecture.

The purpose of this presentation is to back away from what have become both an exhaustive approach to artifact development and the overarching processes that govern their creation. We can re-cast the deliverables as elements in a tool-kit and put the emphasis on doing what is necessary to close the gaps - focusing on tool selection based on context.

The first part of the presentation will discuss process in terms of needs that the IA can fulfill. While remaining outside of any particular methodology, we'll chart the logical points of impact along a project life-cycle where an IA can contribute most.

The second part will go into how to determine the need for particular artifacts to create, by describing some of the variables such as the project constraints of time, quality and cost. We'll also consider how personalities and relative expertise of other project team members may affect your approach. We'll further discuss some of the challenges associated with the decisions about the choice of artifacts with examples and resources.

As a take-away, there will be Visio Stencils for the creation of deliverables made available to attendees.

Download slides for 'A Process By Any Other Name...' (5.2MB, PPT).

Lakoff's 'Women, Fire & Dangerous Things' - What every IA should know

Sunday, March 26, 10:45 - 11:30, Track 2
Donna Maurer

George Lakoff's book 'Women, Fire and Dangerous Things' is a fundamental work on categorization theory, explaining categorization from a linguistic and cognitive perspective. Many IA's (myself included) have had a paradigm shifting moment on reading it.

But it is 583 pages long, weighs 2lbs, and is a very, very hard read. Let's take a short cut - let me do the hard work.

In this presentation, I'll examine the fundamentals of Lakoff's theories and those scholars from which his theories draw. I'll explain prototype theory and basic level categories and will discuss classical categorisation theory and how it fails to describe the real world we live in.

More importantly, I'll discuss how these relate to everyday IA - particularly how we can use basic level categories and prototype theory to create more intuitive structures. I'll even explain how folksonomies/tagging are a natural outcome of the failure of classical categorisation theory.

Download slides for 'Lakoff's 'Women, Fire & Dangerous Things' - What every IA should know' (328KB, PDF).

Emotion, Arousal, Attention and Flow: Chaining Emotional States to Improve Human-Computer Interaction

Sunday, March 26, 10:45 - 11:30, Track 3
Trevor Van Gorp

Other authors: Dr. Ron Wardell, University of Calgary; Edie Adams, Microsoft

Over the last few years, concern for the emotions elicited by products has become more prevalent. Designing for people's emotions represents an acknowledgement of the increasing awareness and importance of people's affective responses in both the selection and use of products. The focus of this paper is on presenting an understanding of emotions as they relate specifically to the design of interfaces for electronic products, including computers and mobile devices.

Although some current models utilized for designing to elicit emotional states describe how emotions are processed or how they are elicited, several important questions remain:

• Which emotions we should be attempting to elicit?
• In what sequence should be they elicited?
• What is the overall goal of eliciting emotions?

This paper collects existing research on emotional design and synthesizes it with related research not normally considered to fall under that category (i.e., persuasion, social interaction and state chaining).

The role of emotions in determining the focus of our attention and the level of our motivation will be explored. This includes the role that emotional states play in affecting how we learn, process and use information. Emotional responses and other affective states ultimately influence what we pay attention to in our lives by affecting how we appraise situations and persuading us to change some aspect of our attitudes or behaviour. On a day to day basis, the things that we pay attention to may seem to be a small concern, but over the course of months and years, the direction and focus of our attention influences and builds the self of each individual.

Download slides for 'Emotion, Arousal, Attention and Flow: Chaining Emotional States to Improve Human-Computer Interaction' (12.4MB, PPT).

Communicating Concepts through Comics

Sunday, March 26, 11:45 - 12:30, Track 1
Kevin Cheng Jane Jao

"How will people use this feature?"

"How does it integrate with their lives?"

At Yahoo!, we've used a number of tools such as requirements documents, personas, user scenarios and storyboards to assist in answering questions like these. However, these tools have yielded mixed and often suboptimal results. For example, requirements and personas were rarely consumed or were interpreted differently between individuals. Traditional storyboards detailing screen by screen progressions created a focus on the interface, rather than the concept.

In one of our upcoming products, we explored a new method of testing and communicating product concepts through the use of comics. Comics are a unique medium between video and static sketches and provide versatility beyond either of the mediums if applied correctly.

During our presentation, we will provide an in depth walkthrough of our process and the advantages that we found comics to have over other methods. In particular, we will discuss how comics can be used to convey subtleties of timing and emphasis and at the same time abstract the details using iconography and art to help the reader relate to the characters and focus on the concepts instead of minute details. Example comics to illustrate each of these areas will be included in the presentation and proceedings.

The flexibility and portability of comics - how we used them to convey multiple paths in scenarios and used Flash to create a portable, distributable version which could be easily consumed online by remote stakeholders, will also be discussed.

To contrast, we will conclude with some of the limitations and challenges we found when using comics, such as the delicate balance of presenting sufficient interface elements to instill discussion but not so much as to distract from the overall concept.

Read more about 'Communicating Concepts through Comics' (slideshow is available there)

New Approaches to Managing Content

Sunday, March 26, 11:45 - 12:30, Track2
Dan Brown

Most content management systems depend on the same tightly woven set of assumptions, positing content as the product of a discrete and rigid process, framing people as automatons with singular responsibilities. Many content management systems end up being used in very limited capacities - or worse, go unused entirely - because these underlying assumptions do not hold true in every situation. This session explores some different approaches to managing content, starting afresh with a new set of assumptions. It introduces analytical methods for digging beneath the surface and builds on these methods to create new theoretical frameworks for content management. We will explore what content management systems might arise when we throw out the assumptions entirely and when we change certain aspects. The session will conclude with a participatory brainstorming session to imagine what systems would like in these new frameworks. To be clear, there are no proven strategies here, only the possibility for radical new ideas and a means for changing your thinking on web content management.

The purpose of this presentation is not to solve these problems, which would require a much larger ego than mine. Instead, it's to demonstrate that there is value in questioning assumptions which can lead to innovative solutions to common, seemingly intractable problems. No doubt there is lots of work left to do on re-thinking content management. I would like this presentation to leave attendees with inspiration to think about these and other issues facing our field.

Read more about 'New Approaches to Managing Content' (slideshow is available there)

Stone Age Information Architecture

Sunday, March 26, 11:45 - 12:30, Track 3
Alex Wright

What can pre-literate societies teach us about information architecture? While we tend to think of complex information systems as the province of technological societies, in fact tribal societies employ remarkably refined systems for managing their collective intellectual capital. This presentation will explore the little-studied world of tribal information architecture, including folk taxonomies (not to be confused with "folksonomies"), kinship and social networks, and stigmergy (methods of projecting information into the outside world). Surveying the anthropological literature and related fields, we will explore the heritage of these ancient information architectures in search of patterns that may shed light on our own approaches to organizing information. Ultimately, we will look for clues to the future of information architecture through the filter of our distant cultural past.

This session will explore the little-studied world of tribal information systems, a topic that has gained a small foothold in the anthropological literature but gone all but unnoticed in the literature of both information architecture and traditional library and information science. In this session, I will survey major findings in this under-studied field that I believe hold important lessons for the practice of information architecture.

My central thesis is that tribal societies engage in practices of information gathering, organization and management that closely resemble what we think of today as "information architecture." In contrast to the popular stereotype of tribal societies as predominantly oral cultures lacking the facility for systematic knowledge, in fact many tribal societies have devised surprisingly complex, durable systems for collecting, organizing and distributing large bodies of shared information.

Topics covered will include:

• Folk Taxonomy
• Kinship and Social Networks
• Stigmergy

Ultimately, I will argue that the study of tribal information architecture reveals deep-seated impulses that still influence our relationship to externalized information systems. If, as Walter Ong argued, the electronic word is propelling us towards a new kind of tribalism, then perhaps our deep cultural past holds critical clues for the future of information architecture in our modern tribal world.

Download slides for 'Stone Age Information Architecture' (2.7MB, PPT).

Web Applications and Real-World IA: Five Techniques for Making Methodology Deliver

Sunday, March 26, 2:00 - 2:45, Track 1
Hala Heymassi, Robert Moll, Elyse Sanchez, Charles Field

Methodologies are employed to guide us through a project so we can develop an effective, creative IA solution within time and budget constraints. Ironically, many methodologies are too complex or high-level to apply to real-life projects. We will present five practical and proven techniques for applying a deliverables-driven methodology to your project or organization.

You'll learn how to be a more effective and innovative IA. Our presentation will enable you to plan more successfully; utilize our requirements visualization process to facilitate communication and a common understanding of project goals among team members; research complex domains without breaking the budget; and ensure your ideas are on target and in line with client deadlines. We'll provide practical tips, techniques and tools that will enable you to "sell" IA expertise internally and to clients; better incorporate IA into the project plan and implement quicker, more creative solutions.

Download slides for 'Web Applications and Real-World IA' (5.1MB, PPT).

IA for Efficient Use and Reuse of Information

Sunday, March 26, 2:00 - 2:45, Track 2
Thomas Vander Wal

The sites we work on have traditionally focussed on single use for the information, they are meant to be read. We know that people have frustrations actually using information in their lives and across their devices. The content needs to be structured in a way that eases the use of the information in formats that can easily be used in the appropriate applications and/or devices. It is in the interest of the content providers to enable ease of information use and reuse.

Information Architecture is one of the best roles in the design and development workflow where this framing of the information to meet needs should take place. This requires us to modify how we approach the information and to build this into our deliverables.

Thomas will describe how to perform granular content inventories. He will show how to use these to help identify prospects for implementing technologies that can ease putting information in a format that people can drop into their applications of choice and move between their devices. He will provide an examination of RSS, microformats, standard data formats, podcasting, etc. that currently exist and how to understand when to best use them, as well as demonstrate their use.

Mind-shift: is IA equipped for Web 2.0?

Sunday, March 26, 2:00 - 2:45, Track 3
Michael Arrington, Dan Brown, Kevin Lynch, Brandon Schauer, Gene Smith

The Web 2.0 meme is popping up everywhere. If you haven't yet been exposed to the "dot-oh" hype, then you must have been hiding under a very reclusive rock somewhere.

The assumption behind all of this Web 2.0 chatter is that some fundamental capabilities of the Web have emerged to create an exciting new class of experiences. Flickr, Wikipedia, Google AdSense, and blogs are some of the recurring poster children chosen to exemplify these new Web 2.0 services and platforms. The participatory nature, scalability, and collective intelligence of these exemplars are just some of the compelling characteristics of this new breed of web experiences.

Just as Information Architects adopted and applied appropriate skills and rigor to address the problems of Web 1.0, the challenge will be to define the competencies necessary to design for Web 2.0 capabilities. The current IA toolbelt might not be ready to design for remixability, user collaboration, and emergent systems of organization. Some say the social scientist or the interaction designer may displace the Information Architect within the context of Web 2.0. So what will IA bring to the table, and what's the skill gap for IAs working in the Web 2.0 world?

IAs have always played a pivotal role for organizations by aligning the needs of users with the capabilities of technology and the objectives of the business. This panel seeks to expose these relationships within the context of Web 2.0, and offer some new mindsets for IAs to prepare them to introduce and harness the value of Web 2.0 within their organizations.

On the panel sits some of the major proponents of Web 2.0 selected for the diverse perspectives they can provide on IA's potential in the context of Web 2.0.

Download slides for 'Mind-shift: is IA equipped for Web 2.0?' (4.9MB, PPT).

In search of common grounds: Introducing Grounded Theory to IA

Sunday, March 26, 3:00 - 3:45, Track 1
Lada Gorlenko

Grounded Theory is a methodology used in the social sciences to build systematically an integrated set of concepts from qualitative empirical data. The method is used to generate concepts in areas where little is already known, as well as offer a fresh take on the existing body of knowledge.

Grounded Theory may potentially add value to IA research by providing a unified research framework. Some techniques, such as contextual enquiry or card sorting, are already indispensable to the IA toolbox. However, Grounded Theory puts the familiar techniques in a new perspective. It makes them integral to the whole research process, rather than treating them as stand-alone procedures. Such integration and the routines that come with it may become a useful skill in IA research, especially when investigating and building concepts and patterns ‘from scratch’.

Grounded Theory has not been directly applied in the IA practice yet. The session explains the foundations of Grounded Theory and invites participants to evaluate whether the method is applicable and practical in the IA context. The session is a theoretical introduction to Grounded Theory, aimed predominantly at advanced IA practitioners and instructors. However, anyone who wishes to take a break from hands-on work and push the envelope instead might find the session stimulating.

Download slides for 'In search of common grounds: Introducing Grounded Theory to IA' (374KB, PDF).

Object-oriented design

Sunday, March 26, 2:00 - 2:45, Track 2
Ann Rockley

Object-oriented structured content is fast gaining popularity with the increasing prevalence of XML-based systems, desire for content reuse and multi-channel content delivery. Effective object-oriented design is supported by content models. Organizations can either create custom content models, adopt an existing standard like DITA, or start with a standard model and customize/specialize it.

This session covers:

• Pros and cons of object-oriented design
• Analysis of content for appropriateness of structure and reuse
• Content modeling o Information product models o Element models o Structural and content reuse o Identification of granularity (object size)
• Assessment of content standards o DITA o Other standards (e.g., XBRL, S1000D, SPL) o The pros and cons of standards o Determing if standards will meet your requirements
• The role of content management in supporting object-oriented content o XML-based content management o Support of granularity o Support of reuse

Download slides for 'Object-oriented design' (406KB, PPT).

Clues to the Future: What the users of tomorrow are teaching us today

Sunday, March 26, 4:15 - 5:00, Track 1
Andrew Hinton

According recent academic and business research, there is an enormous wave of people on its way to adulthood that may very well take us by surprise. And while many designers may be aware of this, we still face the challenge of making it clear to our clients and stake-holders.

Beyond the hype and more obvious implications of the "net generation" are key questions that affect how business and design plan for the future. For example: the shift from hierarchical to nodal paradigms; the rise of new kinds of literacy (and authority); "virtual" vs. "real" money; the splintering of identity (the multiple user-id paradigm); and users who, frankly, expect your web environment to be as well designed as the best games on their X-Boxes.

It's important not to focus on the surface gadgetry, but to understand what is different about how these users think, how they solve problems and manage resources, how they socialize and organize, and how vastly different it may be from the assumed conventions of most business and design decision-makers (i.e. people born before 1985).

This presentation will:

1. Survey some of the current research and insights on the issue;
2. Explore some of the more challenging theoretical questions raised;
3. Discuss the practical business and design implications of those questions; and
4. Suggest how those implications might help make stronger cases for innovative design.

Download slides for 'Clues to the Future: What the users of tomorrow are teaching us today' (13.1MB, PDF).

Content Analysis: Methods and Mentoring

Sunday, March 26, 4:15 - 5:00, Track 2
Chiara Fox

A good information architecture requires a thorough understanding and analysis of the content to be architected. Content analysis is the process where relationships, interdependences, patterns, and content genres are identified. However, more and more clients are performing this time consuming step on their own, looking to information architects for guidance and mentoring. How can you teach a non-IA how to do a content analysis, and still end up with the high quality conclusions and artifacts that you need to base your architecture upon? This session will look at the methods and tools needed so clients can build skills and knowledge in-house, and IAs end up with the data they need to inform the rest of the project.

Download slides for 'Content Analysis: Methods and Mentoring' (1.6MB, PPT).

What do AJAX, RIAs and Web 2.0 Really Mean for IAs?

Sunday, March 26, 4:15 - 5:00, Track 3
Dave Heller

In an attempt to break through the hype and anti-hype surrounding these buzzwords, this presentation will attempt to explore patterns within existing applications and explain how these can provide real-world applicability to solutions from informational marketing sites, to highly transactional applications presented through a web browser. With a concentration on AJAX and Web 2.0 this session will give attendees skills to help them apply these emerging paradigms to their own solutions.

First the presentation will explore breaking down the buzzwords into meaningful and useful pattern descriptions. Then I will explore these patterns and try to work on existing sites to find ways where solutions from this new harvest of ideas can provide improvements. Then finally I will look at issues especially grounded in problems of Rich Internet Applications and how to make decisions about choosing a technology and being sure that the usability of a solution. The end-user will leave this workshop with the following:

• understanding the concepts behind collaborative participation in information spaces
• get a handle on buzzwords and contextualize them within the history of the web
• ability to deconstruct patterns out of looking at existing applications
• how to take those patterns and use them creatively

Download slides for 'What do AJAX, RIAs and Web 2.0 Really Mean for IAs?' (7.7MB, PPT).

Bringing More Science to Persona Creation

Sunday, March 26, 5:15 - 6:00, Track 1
Steve Mulder, Ziv Yaar

Too often, personas require project teams to make a leap of faith: We talk to a few users and create personas based on limited qualitative research, hoping that this cast of characters accurately represents the breadth of real users. As a result, market researchers and stakeholders sometimes roll their eyes at the lack of rigor in our persona creation process, and they have an easier time dismissing personas as merely a cutesy design tool.

It's time to bring more science to persona creation. When personas are derived from statistically significant quantitative research, their usefulness and credibility increase greatly. Statistical analysis techniques such as emergent cluster analysis can be applied to survey results, behavioral data, and financial data to ensure that the segmentation we use to create personas is valid and actionable. The result is a set of personas that is defendable and leads to better decision-making throughout a project.

This session will describe Molecular's methodology for persona creation. We will cover how to get the most out of user interviews, surveys, and log file analysis for generating personas, as well as which statistical analysis techniques are the most useful for segmentation. We will also present examples and lessons learned using this methodology, and how to better integrate art and science in customer research.

Download slides for 'Bringing More Science to Persona Creation' (5MB, PPT).

Measurement of Semantic Distances: an Introduction to IEML

Sunday, March 26, 5:15 - 6:00, Track 2
Yves Marleau

With the rapid evolution of computational systems and semantic web inspired technologies, the need to explore new ways to enhance Information Architecture practice is becoming a necessity.

The Information Economy Metalanguage (IEML) offers a novel way to map documents and ideas, which will allow to measure and analyze semantic distances. Furthermore, IEML would be able to bridge and provide interoperability with other architecture languages, and greatly improve information searching and classification; all essential elements of the IA field.

The Impact of RIA on Design Processes

Sunday, March 26, 5:15 - 6:00, Track 3
Matthew Moroz, Jeanine Harriman, Jenica Rangos, Christopher Follett

The increased ubiquity of Rich Internet Applications [RIA] contributes to the evolution of user experience design goals. AJAX, Flex and other analogous tools can create experiences that are more highly dynamic and user-data driven. As a result, user experience professionals are increasingly breaking away from designing page-based, linear navigation to create more immersive and interactive experiences.

This opportunity challenges many conventional UX design methods and deliverables. With increased interactivity comes an increase in potential options available to a user. A challenge emerges. Due to the limit of the number of permutations that can be represented using conventional methods, such as page-based schematics, UX professionals are tasked with modifying how design is represented to developers. In short, how do we convey the fullness of a proposed solution without mocking up every pixel?

At Avenue A| Razorfish, our experience design practice and its tools continue to evolve as we begin to leverage RIA. Our presentation will describe 3 cases of RIA work with: a media conglomerate, a major automotive manufacturer, and a shoe manufacturer. We will describe how our deliverables changed to best convey design and its challenges, as well as our lessons learned along the way and the best practices that we are working with now.

Questions that the Presentation Asks and Answers:

• UX Design Challenges: How will we adapt UI and Interaction Design Principles for RIAs?
• Selecting RIAs: RIAs are not a one-size-fits-all solution. What user tasks and goals do RIAs best support?
• Supporting RIA Development: Do traditional deliverables succeed in informing RIA web development? How must they be modified?

Download slides for 'The Impact of RIA on Design Processes' (320KB, PPT).

Monday, March 27

Saturday March 25 Sunday March 26

Facets are fundamental: Rethinking information architecture frameworks

Monday, March 27, 8:30 - 9:15, Track 1
Abe Crystal

Commonly-used IA frameworks (Brinck, Gergle, & Wood, 2002; Duyne, Landay, & Hong, 2003; Rosenfeld & Morville, 2002) have three serious flaws:

1. They minimize or discount non-topical characteristics of information. An extensive body of research has investigated the criteria users rely on to assess whether a document or Web page will be relevant and pertinent to their needs (Borlund, 2003; Crystal & Greenberg, in press; Mizzaro, 1997; Tombros, Ruthven, & Jose, 2005). The key insight of this work is that users identify many aspects of documents as important - not just what the document is about, but also its structure, level of difficulty or technicality, practical implications, approach or methods discussed, and so forth. IA frameworks and practices should incorporate these wider conceptions of relevance, and faceted classification provides a way to do so.


2. They treat facets as supplemental, rather than fundamental. As Bates (2002) argues, faceted classification should be the foundation of Web-based information retrieval systems. Facets can be incorporated directly into information retrieval interfaces, and user studies have shown this approach be effective (Yee, Swearingen, Li, & Hearst, 2003). As evidenced by the numerous presentations on facet-based approaches at the 2005 IA Summit, facets are quickly becoming part of mainstream IA practice. However, Bates, as well as Rosenfeld and Morville, set up an unnecessary and confusing dichotomy between hierarchical and faceted classification. As described by Yee et al (2003) and Kwasnik (1999), a hierarchical structure can be applied to a particular facet. Using this framework, facets can be seen as fundamental to IA, not supplemental. Rather than thinking of facets as a way to improve IA, we should think of facets as the foundation of IA. Making facets fundamental also requires clearer definitions of basic terms: what facets are, how they are chosen and defined, and how they are used.
3. They conflate organization and representation. Rosenfeld and Morville, and Brinck et al both distinguish between organization and navigation systems. These distinctions simply aren't clear. In some examples, a particular set of links is called "navigation" while another set is called "organization." A better distinction is between the organization system created by the IA to "make order" (Levy, 1995) within the information space, and the representations that build on the organization system to make specific information objects accessible to users. A framework based on this distinction also provides an elegant way to model "hybrid" organization approaches.

It's time to develop frameworks that put IA practice on a sounder footing, and help us to identify challenges that can push the field forward. In this presentation, I will explain the problems with existing frameworks and outline an alternative framework based on faceted classification. I will ground this framework in information science theory, illustrate its application to typical IA problems, and outline the possibilities it offers for IA practice and research.

Download paper for 'Facets are fundamental: Rethinking information architecture frameworks' (878KB, PDF).

Global Taxonomies meet Interface Design: Challenges and Best Practice

Monday, March 27, 8:30 - 9:15, Track 2
James Kalbach

This case study will draw from the presenter's experience developing multilingual search systems. The focus will be on interface design and how to best surface taxonomies within the search experience. Challenges in creating taxonomies in different languages will also be discussed, as well as organizational hurdles in the development process.

A fundamental problem facing searchers is selecting appropriate terms. In many cases they are not aware of powerful indexing and metadata in information products such as LexisNexis. Solutions to this problem were derived from user research. From novice searchers to information professionals, users' information seeking behavior directly impacted the system design.

One strategy is to let users explore controlled vocabularies while constructing a search. Another strategy is to narrow searches through clustering after a search has been conducted. Finally, while viewing individual documents, users can base new searches on the concepts found in those documents. Strengths and weaknesses of each approach will be provided with examples.

This presentation is intended for all audiences. Participants can expect to learn about interface design of taxonomies and the challenges of multilingual search systems through practical examples. Additionally, tips on selling ideas internally will be communicated.

Download paper for 'Global Taxonomies meet Interface Design: Challenges and Best Practice' (200KB, DOC).

Download slides for 'Global Taxonomies meet Interface Design: Challenges and Best Practice' (1.5MB, PPT).

Tagging and Beyond: Personal, Social and Collaborative Information Architecture

Monday, March 27, 8:30 - 9:15, Track 3
Gene Smith, Danah Boyd, Scott Golder, Jane Murison, Rashmi Sinha, Mimi Yin

The first wave of tagging applications made it easy for people to save and share metadata with others. Whether people used tags primarily for personal findability (as in del.icio.us) or social sharing (as in Flickr) a kind of massively multiplayer information architecture emerged - collaborative categorizations that support personal, local and global views of the information space.

Tagging is being adopted for personal information management (Gmail was first, with Chandler following) to support more flexibility in organizing and finding information and dealing with information overload. In the social context, it is being used to share information, make recommendations, construct identity and have fun. As tagging matures, it is becoming a better discovery tool (e.g. Flickr's clustering algorithms) and is beginning to augment other discovery technologies like search (e.g. Yahoo's My Web 2.0).

Tagging is the best known example of what we can call social information architecture (collaborative filtering and Amazon's Listmania are two others). This panel will discuss these first-generation social IA applications and key concepts - like collaboration, glocalization, identity and trust - that connect them. We will explore the relationship between personal and social IA (is it a virtuous cycle? Or is it self-limiting?) And we will examine the possibilities and pitfalls of personal, social and collaborative information architecture with real-world examples. Examples will include:

• Ownership and moderation of tags on message board conversations (the BBC)
• Tagging and personal information management in Chandler (OSA Foundation)
• Tagging and group identity in a corporate staff directory (nForm User Experience)

The strict faceted classification model: an effective alternative to free-form tagging

Monday, March 27, 9:30 - 10:15, Track 1
Travis Wilson

If you're using faceted classification to organize your data, then perhaps you're familiar with this situation: you want to assign something two different headings from the same facet. For example, a "Flavor" facet contains "chocolate" and "raspberry" headings, and you want to assign both those headings to a bowl of chocolate-raspberry ice cream.

This innocent-looking scenario cuts open a surprisingly wide swath of information theory. It turns out that a strict faceted classification model forbids you to assign both headings, and with good reason. This is counterintuitive, controversial, and if you subscribe to S.R. Ranganathan's original facet theory, heretical.

Yet a faceted classification is most effective when built upon that restriction. This session explores how to use the restriction constructively, whether the faceted classification model really fits the user's intent, and what the alternatives are. We'll critically examine facet structures, especially in light of such new metadata schemes as tagging. Along the way we'll discover:

• the real nature of the faceted classification model
• the similarities to and differences from free-form tagging
• orthogonality
• binary facets
• recent suggested enhancements to the faceted classification model
• new technology that leverages the strict model to sort tags into facets
• how to make your classification as usable as possible

Over the last couple of years, tagging has emerged as a flexible method of assigning headings to resources, regardless of other headings on that resource. Its applications sharply illuminate its differences from faceted classification; in fact tagging is the closer of the two to Ranganathan's canons of facet theory.

Although tagging allows the information architect a more intuitive free association between headings and resources, it comes at a cost. Part of that cost is simply a limitation of underlying technology (properly designed database systems can handle far more headings in faceted classification schemes than in tagging schemes). But another cost is the clarity of the information itself. Since "chocolate" and "raspberry" have not been separated into different facets, they bypass the intent of faceted classification and therefore deny benefits to the less conventional users of the data. In the Web 2.0, these users include data-mining software that relies on orthogonality between facets, and software that looks to the faceted data for definitive answers.

In many cases, IAs who adopt the strict faceted classification model are compelled to define their facets more clearly, yielding better long-term results. This 45-minute tour of the issue provides the principles necessary to do that.

Download paper for 'The strict faceted classification model: an effective alternative to free-form tagging' (182KB, PDF).

Ich Bin Ein Website!The Impact of Language and Culture on Internationalization and Localization

Monday, March 27, 9:30 - 10:15, Track 2
Evan Gerber

On June 26, 1963, John F. Kennedy famously declared "Ich bin ein Berliner." To some, this sounded like "I come from Berlin." Others, however, interpreted it as "I'm a jelly donut." Everyone perceives information through individual filters of culture and language. Accounting for these filters in a globalization or localization project is challenging and complex. Content management technology helps surmount these obstacles, but only after careful preparation and execution. This presentation explores the steps involved in developing global websites for both a manufacturing and insurance corporation. Attendees will garner practical insights on challenges, strategies, and the lessons learned which are critical to success on an enterprise wide internationalization project.

The presentation starts with a review of work required in the analysis and design phases, to ensure that the project incorporates linguistic and cultural constraints from the beginning. A review of best practices explores options for building teams that extend across boundaries of nationality and language, while educating clients about culture and idiom. The discussion then concerns critical tasks and strategic requirements for conducting research on all of the separate cultural and linguistic groups. The dialogue describes the composition of survey audiences, and discusses tools used to assess the current website. After this, the topic covers processes to synthesize data from different cultural and linguistic entities into meaningful solutions. To wrap up the analysis and design segment, the exploration will turn to documenting culture-specific requirements, and obtaining buy in from different groups.

The next part of the discussion addresses additional work added to the design phase by internationalization and localization. The impacts of culture and multiple languages on taxonomy structure are addressed. The material covers different ways to optimize the taxonomy for cultural expectations and linguistic differences, while allowing for future growth. Best practices in visual design will be shared to ensure that cultural and linguistic concerns are addressed while designing and testing for multiple languages. The presentation explores how the front-end meshes with the administrative site to optimize the CMS for different regions and parlances. Tools such as entry templates, directory structure, and naming conventions are discussed, with an emphasis on promoting maintainability and adoption upon rollout.

The presentation reviews the challenges encountered in developing internationalized and localized content. Impacts of cultural issues on translations will be covered, and the processes to predict, detect, and manage these problems. Finally, the presentation will cover the best ways to roll out a website to global audiences. Myriad small problems which crop up during rollout, and the conversation explores different means to manage them through client contact and training.

Internationalization and localization are fraught with pitfalls, many of which are unexpected the first time around. Throughout the project lifecycle, teams must understand constraints imposed by language and culture, and their impact. This presentation teaches attendees about those pitfalls, and explores ways to ensure success. While the process isn't brain surgery, it's harder than eating a jelly donut.

Tags and facets, tags and languages: a case study.

Monday, March 27, 10:45 - 11:30, Track 1
Peter Van Dijck

Peter Van Dijck will discuss a project (a public website) that attempts 2 things: to combine a folksonomy with a faceted approach, and to localize a folksonomy.

Tag clouds are notably hard to navigate - using a faceted navigation system might help. But how can you combine facets with tags? Who decides what the facets are? Who assigns tags to facets? And do users really get this?

Tag clouds are also notably hard to localize. Most tag clouds currently simply present a combination of languages to the user.

What approaches are possible to fix that, and what works?
Can algorithms be used efficiently?
Should tags be translated?
Manually, or by the computer?

We considered and tried various approaches. The audience will learn what worked in this particular project, what we learnt from our research on other projects, and where we failed.

Download slides for 'Tags and facets, tags and languages: a case study' (2.8MB, PDF).

Montreal, Paris, Dakar: Conducting an International Intranet Needs Analysis

Monday, March 27, 10:45 - 11:30, Track 2
Isabelle Peyrichoux

In 2002, I met an important challenge when I was nominated as Information Architect for the French-Speaking University Agency (Agence universitaire de la Francophonie), an international and non-profit organization located in 35 french-speaking countries around the world. The challenge was to design, from scratch, a highly efficient and practical intranet, which would facilitate the everyday work of approximately 400 employees across 40 different locations, on five continents.

To achieve that, I performed an extensive needs and tasks analysis, conducting one on one interviews with 50 employees in three locations: Paris (France, Europe), Dakar (Senegal, Africa) and Montreal (Canada, North America). From the moment I chose the method to the moment I went back from my survey travel, I faced many challenges and obstacles - some expected, others not. These challenges had to do mainly with interpersonal, organizational and cultural issues.

This presentation is the journal of my 2 month-journey and describes the different obstacles I faced and how I dealt with them. Below are the main topics I will address:

• Choosing a method to collect useful information about users' actual needs, in a limited time.
• Dealing with cultural andorganizational differences or issues.
• Dealing with people who are really reluctant to what you are doing.
• Making the most of one-hour individual face-to-face interviews with people you don't know and who don't know you.
• Making people speak the most freely as possible of the tasks there are doing and the problems they are facing in their everyday work.

Through very concrete examples taken from my personal experience in an international and multi-cultural organization, I intend to share some of my learnings with the audience.

Download slides for 'Montreal, Paris, Dakar: Conducting an International Intranet Needs Analysis' (443KB, PDF).

From Pace Layering to Resilience Theory: the Complex Implications of Tagging for Information Architecture

Monday, March 27, 10:45 - 11:30, Track 3
D. Grant Campbell, Karl V. Fast

This paper will present a framework for adapting theories of complexity, pace layering and resilience to the question of tagging and folksonomies, and their influence on the practice of information architecture. In so doing, it represents a step towards the development of a body of theory that can serve the purpose of information architecture as both a professional practice and a field of intellectual study. The tagging phenomenon represents the greatest challenge we have yet faced in our field's working relationship with complex systems. Instead of replacing "mess" with controlled vocabularies and sound hierarchies, we must now accept the mess of folksonomies into a permanent relationship with our ordered structures. We need theoretical and practical insights from other complex systems to ensure that this relationship remains productive, rather than destructive.

Information architecture has a history of incorporating approaches and theories from other fields: graphic design, library science, cognitive science, usability and human-computer interaction. To date, however, these cross-disciplinary influences have been used directly to improve practice. We hope, through this initial framework, to show how research can be used to enrich this process. We aim to show how other fields can used to create a body of theory that belongs uniquely to information architecture, and which, once created, can affect IA practice on a much wider scale. At the very least, such efforts will enrich our understanding of our field; at best, they may ensure the survival and the healthy evolution of our best work.

Download paper for 'From Pace Layering to Resilience Theory: the Complex Implications of Tagging for Information Architecture' (182KB, PDF).

Download slides for 'From Pace Layering to Resilience Theory: the Complex Implications of Tagging for Information Architecture' (2.5MB, PPT).

Exploring the context of user, creator and intermediate tagging

Monday, March 27, 11:45 - 12:30, Track 1
Margaret E. Kipp

One of the important tasks of information architecture is to reduce the difficulty inherent in searching large document spaces for information. A solidly designed taxonomy using terms and keywords appropriate to the context of the intended user can help make the difference between a usable document space and a space which is difficult to navigate and find the information sought.

While the creation of generic hierarchical classification systems or subject specific taxonomies has a long history, the design of these systems has largely been left to professional intermediaries such as librarians and information architects. Because of the increasing amount and specialisation of information being collected and user requests for more fine grained access, these systems tend to be too generic for user needs. While full text search can provide fine grained access, this access tends to be at the expense of precision due to the use of differing terminology. Increasingly, user tagging and folksonomies created in a distributed fashion through social bookmarking sites are being touted as a potential solution to these problems. (Mathes 2004; Hammond et al 2005) This use of user tags combined with topic maps and tag clusters may have the potential to provide the benefits of a controlled vocabulary which controls for terminological differences while still allowing the use of natural language vocabulary. (Shirky 2005)

This paper examines the differences in the context of user, author and intermediary assigned keywords or tags using the social bookmarking sites citeulike (citeulike.org) and connotea (connotea.org), which are specialised for academic articles. A sample of journal articles tagged in citeulike and connotea by at least 2 people was selected. The chosen articles were manually restricted to a set of journals known to include author assigned keywords and to journals indexed in INSPEC, a database which provides an intermediary assigned controlled vocabulary for searchers, so that each article selected would have 3 sets of keywords assigned by three different classes of metadata creators. Co-word analysis, concept clustering via the INSPEC thesaurus, and descriptive statistics were used to examine differences in context and term usage between the three classes of metadata creators.

This study has implications for the design of systems for accessing, indexing and searching document spaces. The popularity of Google has demonstrated that users prefer to be able to search for items in a more natural way using one interface to locate items of a varied nature, but controlled vocabulary usage can be expensive. (Campbell and Fast 2004) User tagging, with its lower apparent cost of production, could provide the additional access points with less cost, but only if user tagging provides a similar or better search context.

Download paper for 'Exploring the context of user, creator and intermediate tagging' (46KB, PDF).

Download slides for 'Exploring the context of user, creator and intermediate tagging' (244KB, PPT).

How can information architecture address challenges to the Web in third world and developing contexts?

Monday, March 27, 11:45 - 12:30, Track 2
Jason Hobbs

Third world contexts hold some of the greatest challenges to the use of the Web: low bandwidth; limited and costly access; and small user bases. I would like to share my experiences working in Africa where information architecture processes and practise promise to solve many of these problems.

Presentation content:

1. Anecdotal accounts and insights into the use of the Internet and telecommunications in Africa
2. Applied IA: a needs-based approach, CRM models as interaction models and user journeys (case study: Volkswagen website, South African)
3. The strategic implementation plan to localise the UNESCO Digi-Arts global network for African

Out take: People are using the Web in the third world but a very different style of usage exists in developing contexts, one which is based around core needs that people have. Given the high costs of access, low bandwidth and the fact that these conditions are not taken to heart in the design of websites (or more fundamentally in the strategic approach to projects) adds up to poor user experiences that limit the potential of projects and businesses.

Global initiatives need to take these conditions to heart if they wish to meaningfully and successfully address the people and industries in third world localities. Information architecture based on needs based approaches and user journeys, are the starting point for creating user experiences with these conditions in mind promising to build trust (in the use of digital channels) and thereby increase use. These context sensitive approaches offer far stronger return on investment arguments for operating in the third world and developing contexts.

Download slides for 'How can information architecture address challenges to the Web in third world and developing contexts?' (5.8MB, PPT).

The life of tags

Monday, March 27, 11:45 - 12:30, Track 3
Anthony Charles, Jason Toal

Our paper will pursue the potential opportunities for adding a usable layer of rules to the process of tagging. We will discuss the potential of adding metadata to the metadata' through weighting of tags or sequencing of tags through more formal structures. Where possible, we will propose experiments and usability studies that could test the validity of our theories.

Both Jason and Anthony are graduate students in the School of Interactive Arts at Simon Fraser University, Surrey BC, Canada. They have collaborated together on many course assignments and are nearly completed their work. The Information Architecture Summit 2006 represents their first joint publication and fulfills a basic presentation requirement for the master certification. They are very pleased to have this opportunity.

Download paper for 'The life of tags' (576KB, DOC).

Innovation vs. Best Practice conflict or opportunity?

Monday, March 27, 2:00 - 2:45, Track 1
Eric Reiss

"Best practice" implies doing things in the best possible manner, based on past experience. But we like to think of ourselves as innovators in a dynamic industry - we want to go where no one has gone before. Thus, "best practice" and "innovation" are like oil and water - they don't easily mix.

How can we, as UX professionals, balance the need for consistency that "best practice" provides, with our on-going mission to improve the quality of our products? How can we create genuine improvements - and when have we been seduced by the evil twins, Fad and Fashion?

"Innovation vs. Best Practice" is a highly interactive 45-minute exploration of the elements that make up these two ends of the UX spectrum. It's a combination of demonstration and common-sense review, in a single, high-powered, bullet-point-free session.

During the presentation, we'll take a closer look at the popular definitions of both innovation and best practice - and discover why these are frequently inadequate, misleading, or both. Why is a "standard" not always a "best practice"? And if "invention" can be spontaneous, why is "innovation" always planned?

We'll also take a closer look at some of the worst reasons to innovate, which are also some of the most common, plus the Japanese concept of "chindogu" - "useless innovation." Perhaps most important of all, we'll see how User Driven Design helps us avoid harmful innovation in comparison to the more common User Centered Design methodology.

Want to learn the four Laws of Innovation? Want to be able to recognize the three danger signals of Fashion and Fad? Want to know why Jakob Nielsen may be right 37% of the time?

More info available at www.e-reiss.com

Download slides for 'Innovation vs. Best Practice conflict or opportunity?' (2.6MB, PDF).

Information Architecture for the Spatial Web

Monday, March 27, 2:00 - 2:45, Track 2
Matthew Milan, Michael MacLennan

The use of spatial data in web applications and websites has become widespread over the last year, and this increased access to spatial data has brought its application out of the realm of specialized uses into the mainstream. As spatial data becomes a more common element of web sites and web applications, effective design practices are needed to help users access, browse and search spatial data in ways that help them find appropriate information. The application of information architecture concepts and practices can help to improve the usability and findability of spatial data and improve the user experience of sites and web applications that use mapping technologies and spatial data.

This presentation will focus on providing practitioners with a solid understanding of geospatial information design issues for the web, emphasizing on how to effectively structure web navigation and interaction that includes spatial data elements.

Download slides for 'Information Architecture for the Spatial Web' (19.4MB, PPT).

Sorting in an age of tagging: How Information Architects can use sorting to address just about any research question

Monday, March 27, 2:00 - 2:45, Track 3
Rashmi Sinha

The information age needs new and different type of research methods. Instead of focusing on what people want, we need to go deeper - to understand how people think, their subjective understanding of a domain. An easy but powerful way to understand how people think is sorting. Whether in academic fields like psychology / cognitive anthropology, or applied fields like market research / mind-mapping / creativity techniques, sorting comes up again and again.

Sorting (open & closed) are dominant research methods in Information Architecture, used as a starting point for creating a new IA or to refine an existing one.

I will start with a cognitive analysis of sorting and examine why it is a powerful method to understand people's thinking. An interesting contrast to sorting is tagging (recently made popular on sites such as flickr and del.icio.us), and it is illuminating to contrast the cognitive processes in the two. Both sorting and tagging get at the same underlying mental models, but can be used in different ways. Tagging is suited to an emergent collaborative information architecture: patterns evolve naturally over time in web spaces. In contrast, sorting as a research method is fast and efficient, suited to exploring user thinking, and crafting aggregate mental models, with a relatively small sample. From an IA perspective, methods like Tag Sorting allow both to be used together, leveraging the natural strengths of each method.

Any discussion of sorting is incomplete without an examination of sorting on computer systems, and the recent disenchantment with multi-level hierarchies. Some of this disenchantment can be traced to the cognitive complexity of sorting/finding in multi-level hierarchies. But it also has to do with interface problems with categorization systems. I will examine these problems and describe alternative interfaces, such as Microsoft's ribbon UI, Gmail, treemap interfaces, and a game-like sorting interface that I worked on recently.

Sorting has applications beyond IA - it is used in the corporate domain for understanding mental models for positioning / branding, and as a creativity technique. As a "positioning research" tool, it is used to understand how to position a company / product, and to explore brand conceptions (what qualities brands are associated with; how products / brands are grouped in users minds). A second usage is "KJ" / "snowball" techniques used for consensus building or as creativity techniques. The third example, "mindmapping" used to understand how groups view a domain - applications range from stakeholder analysis to problem solving.

Information Architects currently use sorting for designing architectures. However, there are other questions that a company faces that could benefit from understanding user mental models. As practicing information architects expand their role in the company into other domains, practices and positions, sorting is one technique they can carry with them - a focus on understanding how people think, and ways to get at that understanding.

Download slides for 'Sorting in an age of tagging: How Information Architects can use sorting to address just about any research question' (1.3MB, PPT).

Closing Plenary

Monday, March 27, 3:00 -400
Peter Merholz

Download slides for the closing plenary (6.5MB, PDF).