Achieving a State of Trans

Architects discovered long ago that buildings were really about the space they define - its dimensions, certainly, but also its affordances for the people and activities that can gracefully occur therein. Today's information architecture is more like city planning, in that we are dealing with extended and related spaces of activity. Today, information architecture spreads out across PCs, phones, wireless devices, Wi-fi, GPS and RFID tags, exploding the "space" of computational interaction. Likewise, through any of these platforms, a variety of media experiences are enabled, from text to voice to image, video, motion graphics, 3-D models and VR. The inhabitants of our architectures blur across the landscape, slipping from keyboard to earbud to real spaces with sensors and effectors. A truism in that discipline that we used to call "interface design" was that the computer should be seen as an environment for humans. But looking at today's information landscape, we can see that a miraculous inversion has occurred: humans are the environment for information architecture. Our design criteria are becoming transpersonal, transmedia, and transportable. This talk will focus on achieving a state of "trans" through a process involving both research and revisioning.

Back to Top

Taxonomies, Controlled Vocabularies, and Ontologies

The Panel will explore the evolving nature of Taxonomies, Controlled Vocabularies, and Ontologies and discuss how these fields are emerging as extensions of Information Architecture. Topics such as the resurgence of interest in taxonomies, how assets are being used to generate product revenue, and a new defintion for Ontology that meets business needs will also be addressed.

Back to Top

Emerging Content Requirements for News Products

This presentation identifies trends in online news delivery and presentation of news content. Trends are driven in part by the Internet and by knowledge-based capabilities of content management and information retrieval systems, and demonstrate how "traditional" text-based news production has responded to new media opportunities. One of these trends is personalization of news content and its presentation. Another trend is represented by the use of summaries as a format for customized presentation. Summaries are also important because of the need to accommodate new modes of delivery where screen size is limited. Issues involved with presentation of online news content, particularly with respect to usability, will in turn drive a new set of content requirements that will provide opportunities for news providers to add value to their products. Of special interest are value-added features that support the journalistic aim of providing explanations for news events. We view explanation as an important product differentiator, and we explore how explanations require supporting presentation capabilities in order to meet the personalized requirements of consumers. New content features, derived through sophisticated processing of news content, may be represented as extended XML-based enriched text elements. XML-based specifications also support the objective of separating the abstract representation of content from its presentation, a prerequisite in order to support the multi-purpose nature of news content. While these content requirements and associated technologies might also be applied to other industries or solutions, the focus here is on opportunities for creation of news products, in part because news production presents a number of unique and interesting challenges, not the least being its role in helping the news consumer learn about events in the world. Research underlying this presentation explores the intersection of capabilities emerging from within the news production environment with capabilities derived from leading edge knowledge-based technologies.

Back to Top

The 2004 Information Architecture Slam: The Workshop with a Winner

Results from the IA Slam Preconference Workshop will be presented. The Friday Workshop gave participants a chance to share their techniques, ideas, and points of view, and allowed people with a passion to "put their money where their mouth is."

Back to Top

XIA @ UT: An Extreme Makeover

The University of Texas at Austin (UT) library system, the General Libraries, encompasses over seventeen different physical facilities and 8 million volumes. In addition to maintaining the physical collection, the General Libraries embraces the virtual world through its web sites. An enormous undertaking, the library web presence attempts to meet the needs of many, if not all, potential library users.

For a class project, a group of graduate students in the university's School of Information re-engineered portions of the General Libraries web site to create a jump-start portal for researchers. The intent was to design a web site that would help new or experienced library visitors to maximize the research potential of the library, according to their unique research needs. The graduate student team used a new set of information architecture (IA) methods, XIA (extreme information architecture - based on the extreme programming methodology), to organize and frame their development work as they cooperated to redesign elements of the UT library web site as a research portal. The XIA team sought to advance the state of the art in conceptualizing IA by adopting a more modular approach to development and conceptualizing the library services. New and cutting-edge technologies, including Cascading Style Sheets, XML and a web services architecture were implemented in the design, transforming the library web site to be more compliant with web standards, and easily upgradeable as technologies change.

This presentation will discuss the process of designing the research portal of the web site, with emphasis on the lessons learned by the XIA development team.

Back to Top

A Case Study of Redesigning a Digital Video Digital Library

The Open Video Project is a digital library of contributor-donated digital video and associated metadata, used by a diverse set of users with wide variety of interests and goals. Based at the University of North Carolina at Chapel Hill, the project began in 1999 with a very basic Web site and 195 videos. Several iterations later, the Web site currently contains about 2000 digital videos in a variety of genres, lengths, and formats, and is accessed by thousands of users a month.

The current iteration of the site (http://www.open-video.org) is the result of an effort to use guidelines and principles from the information architecture community, along with findings from our own video browsing user studies, to produce a Web site that can better meet the needs of our users. The particular challenge we faced was that Open Video users vary widely in their individual characteristics (e.g., demographics, technology experience, video experience), their goals (e.g., learning, work, research, entertainment), and their specific tasks (e.g., finding a known video, finding a specific type of scene or visual style, browsing for something fun to watch), while the characteristics of our content--digital video--also vary greatly in terms of genre, topic, and style.

By providing users with different ways to search, browse, and evaluate video, we believe the redesigned Open Video Web site enables users to perform their tasks in ways that better support their particular characteristics. As publicly-accessible collections of digital video become increasingly common, we also believe that the issues we faced in the redesign process and the ways in which we resolved them will become relevant to more and more information architects and Web developers, and we propose this case study as a way to share our experiences and lessons learned with the IA community.

Back to Top

A Critical Review of Enterprise Content Management (ECM)

"Enterprise" Content Management has emerged as a favorite buzz-phrase for analysts and software vendors alike. In theory it holds the alluring promise of connecting, consolidating, and aligning the disparate information management efforts across a major enterprise into a simpler and more effective whole.

In practice, ECM means many different things to different organizations, and early returns from the ECM front suggest that attempts to implement content management on a truly enterprise scale have been fraught with great difficulty. In some cases effective business processes are being sacrificed at the altar of technical consolidation. In other cases, large enterprise-level projects are collapsing under their own weight. And in still other cases, rationalization of vendors and suppliers is forcing substandard technical solutions on individual business units.

CMSWatch founder Tony Byrne will argue that "ECM" does indeed hold some promise, and that IA specialists will play a critical role in making ECM a truly useful practice. The two keys to successful ECM are: 1) successful Enterprise Information Architecture, without which cross-divisional content sharing and consolidation are nearly impossible, and 2) successful recognition and application or emerging ECM design and implementation patterns. Byrne will share some preliminary thinking on some initial ECM patterns and invite the IA community to critique and extend those patterns based on concrete project experiences of their own.

Back to Top

Making Personas More Powerful

Personas are supposed to be tools to aid in making design trade-offs. But too often they've become more of a check-off item. Arguably this is because personas as commonly done lack enough actionable detail to help make hard decisions that occurs as strategy resolves into interface.

Consequently, I've been developing a persona template that goes into great detail, including looking at the aspects of the interaction (for example, how frequently and regularly is a task performed, what's the intensity when it's performed); information characteristics (for example, what's the volume of information being processed, what modes are involved); sensory/immersive characteristics (for example, what mood is to be evoked, what styles/genres best evoke it); as well as emotional characteristics (e.g. what personality should the produce convey, what sort of brand relationship is involved); as well as questions around the context of users, accessibility and usability. The dimensions involved in the template are based on research from not only the user experience community, but software engineering, film-making, marketing and branding.

This new approach to personas is in line with the conference theme of breaking new ground. Participants will not only be able to take away an extensive set of questions to pose when thinking about the users they're designing for, but hopefully spur them to think about additional questions of their own that will allow them to make more powerful usage of personas.

Back to Top

XIA: Xtreme Information Architecture

This presentation overviews a new set of methods for Information Architecture based on the Xtreme Programming (XP) development methodology. These new XIA methods include: making the customer an intimate team member, user stories that develop into test cases, modular testing for rolling development, using code as product specification, adaptive release planning and developer-controlled velocity.

Information Architects are the focal point of XIA involving managing the development process including tracking requirements, organizing the design decisions, and letting customers and developers do what they should: set project priorities and choose technical tasks (respectively).

This new, adapted methodology has great potential to transform both small group IA projects and large Web development efforts. By focusing on building small, functional elements of an overall Web site with iterative methods, XIA can help to avoid design paralysis and functionality creep among other pitfalls in traditional IA design and development.

Back to Top

Fun with Faceted Browsing

This presentation will define faceted browsing and related information architecture concepts. Faceted browsing is defined as an interaction style where users can filter a set of items by selecting from a set of values to get a smaller set of items. Facet values can be selected in any order. Only valid values that lead to non-zero results are listed.

Many examples of faceted browsing will be presented and analyzed so that attendees can understand the best practices. Various ways of providing facet selection and facet history are being used, for example.

Faceted browsing applications are starting to become common on the web because they enable users to filter a set of products in a flexible manner and based on their needs, not on a pre-set hierarchy. Faceted browsing application are breaking new ground because of their innovative merging of user interface and information architecture.

In addition to examples and definitions to help IAs communicate faceted browsing concepts, a review of the research literature and a set of open research questions will be presented in order to establish a roadmap for future work on faceted browsing.

The handout includes an overview of the IA aspects of faceted browsing and a few examples. The presentation will include more of both and a section on research.

Back to Top

IA of Content Management Based Reuse

The design and reuse of content has been receiving greater acceptance in all industries in paper-based, web-based and other media based delivery. The addition of a content management system and XML can greatly improve the functionality of reuse by enabling systematic (automatic) reuse, a full audit trail for every piece of reusable content including where it is reused, and reuse at a very granular (small content object) level. However, content management based reuse brings a whole new level of information architecture to the task of designing, managing, and delivering content. This session will provide guidelines on information architecture of content management based reuse including:

• Identification of appropriate forms of reuse (opportunistic, systematic, derivative, nested, locked)
• Identification of opportunities for reuse (across media, between information types, within documents)
• Design of:
  
  
• Information product models
• Element models
• Structural reuse maps
• Content reuse maps
• Element level metadata
• XML ready models
  
  
• Development of business rules for determining what content is the approved/authorized version to reuse (e.g., how does an author determine which is the source when a derivative version of an element has been created or what happens if two authors edit/update two or more derivative content chunks (siblings from the same parent) simultaneously?)
• Determination of how to most effectively support the models (authoring tool, CMS, delivery tool, or across all tools, and the role of XML)

Examples and real case studies will be used throughout.

Participants will be introduced to the steps involved in modeling effective reusable content, how to develop business rules for reuse, and how to support their models with appropriate technology.

Back to Top

Incorporating Research on Navigation into a Design Method

At the 2003 IA Summit the divisive panel on navigation and wayfinding spurred a flurry of discussion at the conference and afterward on the SIGIA mailing list. It reminded us that there are few hard laws that govern our navigation design decisions: ours is a craft of trying and testing. And yet our process as design practitioners can be made more systematic if we incorporate findings from the research literature. This presentation will survey and summarize major research on navigation such as the Shape of Information (Dillon et al), Information Scent (Pirolli et al), Effective View Navigation (Furnas), and Mental Models (Young et al). Options for incorporating these findings into the common user-centered design process will be offered. The goal is to increase our ability to create successful designs through use of a more systematic design method that incorporates navigation research.

Back to Top

Selling User Experience Through Value-Centered Design

The effectiveness of information architecture and user experience work is directly linked to buy-in from business decision makers. Without management on board, these efforts have limited impact. However, attempts to build support for user experience often fail to connect with decision makers using their own language and have limited success.

Value-centered design includes user experience while focusing on a business perspective. This model provides a platform for bringing business decision makers on board with user experience efforts. The basic premise of value-centered design (VCD) is that shared value is the critical objective for business initiatives. This value comes from the intersection of business goals and context, individual goals and context, a product offering, and a delivery channel. By placing individual goals on the same plane as business goals, VCD dramatically changes the conversation with business decision makers. VCD makes it clear that sustainable projects require more than Return on Investment - projects and products must also provide individuals with "Return on Experience", where users gain some benefit for the attention and effort invested in the experience.

This session will explain the basic value-centered design model, and then expand on it with practical tools for everyday use. These resources include a basic methodology, a framework for choosing project activities, setting project priorities, and developing team competencies. These methods will be explained with project examples, sample deliverables, and informative illustrations.

Value-centered design reframes current thinking, offering a fresh approach to user experience evangelism that builds business buy-in by buying into business. This balanced perspective lets user experience practitioners start real conversations with business decision makers - conversations that provide a foundation for long term impact and adoption of user experience within the project and across the organization.

Back to Top

Using Facet Analysis for Improving Information Access to Marginalized Communities

Information architecture is concerned with connecting people with content, using classification and vocabulary techniques developed in library and information science. This presentation will look at the specific challenge of connecting people who belong to marginalized communities: particularly members of gay and lesbian communities. People in such communities often have difficulties connecting with important and needed information, because formal systems in libraries, bookstores, newspapers and web information systems lack meaningful categories. Many gay men and lesbians have an adversarial relationship to "official" categorization, arguing that their concerns are all too often classed as "illness," "deviance," or simply "departures from the norm." In addition, many users have ambivalent feelings about their membership in that minority. They would like to have the option of "turning off" their orientation when it suits them, so that they can prevent their sexuality from being an issue, and to turn it "on" when it suits them, so that they can filter information sources for relevance to themselves.

This presentation will offer preliminary findings from a study in progress: a series of qualitative interviews studying the information searching and classification techniques of fifteen lesbians and gay men. Early findings suggest that the practice of facet analysis, as developed in information architecture and demonstrated in experimental browsing systems such as Epicurious and Flamenco, provides a very effective way of connecting marginalized groups with information content. Sites which use faceted metadata give the user control over facet sequence; in so doing, they enable the user to decide when, if at all, to introduce a "gay" facet into the browsing or searching process. This suggests that information architecture could be a very useful basis on which to enhance the quality of life for members of marginalized communities.

Back to Top

When Ninety-One Years of Content Goes Digital

Girl Scouts of the USA has been in operation since 1912; however that legacy includes a culture that is slow to change. The organization operates similarly to a nonprofit franchise, with 316 independent charters referred to as councils. Combined, there are 8,000 staff members, mostly women, who run the day-to-day operations of Girl Scouting.

The project was to develop a new extranet for the councils with a small team that included one information architect. The project sponsor, acting CEO and the councils themselves, supplied our goals:

• Reduce GSUSA production costs associated with intellectual capital.
• Provide a digital space that supplements off-line knowledge sharing among the Councils.
• Leverage the features of the recently licensed content management system, Documentum 4i.
• Develop a model solution for the nonprofit community that rivals anything found in the commercial
sector.

I established my IA goals as maximizing usability and ensuring scalability. To meet the usability challenge, I conducted 40 interviews, facilitated card-sorting exercises, and developed user personas. To support scalability, I adopted modularity as a solution by developing nine types of content with corresponding entry screens and seven groupings of functionality for landing screens. I also leveraged standards such as the Dublin Core Metadata. The supporting documentation includes a range of site maps and wireframes in which all functionality is fully annotated.

Despite the project's challenges, we still have a high ROI estimated at $250,000 a year, plus the intangible benefits of a stronger community to fulfill the Girl Scout mission.

Back to Top

Creating a Consistent Enterprise Web Navigation Solution

The Problem

To accommodate multiple user types with multiple tasks Vanguard produced a family of sites. While these sites are mostly self-contained there are times when users may cross over between sites.

These sites have been built over the past 6 years by various teams and there is little consistency between them - either in architecture or design.

Navigation Systems are of particular concern since inconsistencies exist not only between sites but also within sites. These inconsistencies have caused usability problems and have affected our productivity as we keep "re-inventing the wheel" for every navigation problem.

The Solution

To solve this problem we launched a project to create a set of navigation systems to be used by the design teams for all Vanguard sites.

After consulting the published work of many experts and analyzing our situation, we found new patterns and new definitions that were not reflected in the current literature on web navigation. By breaking the subject of navigation down to its most fundamental building blocks (semantic relationships, content types, patterns of navigation, and the basic interplay of navigation and content), then recombining them we created a new taxonomy and a new classification scheme for navigation systems. This new taxonomy and classification scheme gave us a powerful tool to create the systems and understand how they would work together with content to create a consistent and beneficial user experience.

Though we have not tested it far beyond the context of our own family of sites, we feel that this new approach could be of value to any designer or architect who is confronted with creating navigation systems for their sites.

Back to Top

Rebuilding Trust in the User Centered Design process through the Redesign of the Wachovia.com Investing Center

It's a distressingly common scenario: In a complex and politically charged environment, work has devolved from design to negotiation and from negotiation to capitulation. Exhausted and time-pressed, the user-centered design team settles and produces a less- than-ideal design. The design is tested and the usability results are unequivocal: the design doesn't meet users' needs. The client is unhappy because they want solid usability results-and from their perspective they've participated in a user-centered design process (perhaps forgetting, by now, the caveats and concerns of the design team).

Now what?

This is a first hand account of how our team went through this cycle and how, after two failed attempts to "get it right," finally succeeded. This case study will review the redesign of the Investing Center on the Wachovia.com website. Investing and brokerage are complex topics for users and a complex part of the way we do business. We worked with our internal business partners to fully understand the issues facing us and to learn how to produce a truly user-centered design that could be embraced by our business partners.

This presentation will cover: *Managing relationships: repairing the relationship that has become adversarial and moving from relationships that result in design by committee to true partnerships *The changing face of subject matter expertise: How do you figure out what you need to know before it's too late when your area of expertise is classifying the information, not the information itself? *Leveraging usability: Our initial work on this project demonstrated that we were applying usability testing too late in our process for the findings to radically change our designs. After getting the buy-in of our business partners we applied usability early in the process, testing 3 radically different approaches to drive us to the successful design we ultimately proposed.

Back to Top

Metadata Standards & the Enterprise Information Architecture

Outside of web development, information professionals in various settings use metadata standards to help establish information management conventions within their organizations. One such metadata standard is the Dublin Core Metadata Initiative. Originating from the library world, these information professionals looked to establish a simple set of metadata to describe information at a very basic level. With close to a dozen years in use, the corporate world has begun to see the need of such metadata standards within their organizations. Dublin Core has recently become incorporated as part of many metadata standards internationally. It has been adopted by CEN Workshop Agreement CWA 13874, as the NISO Standard Z39.85-2001, and as a draft with ISO as DIS 1583.

This presentation will focus on how corporations are using Dublin Core and other enterprise metadata standards to help manage information on their intranets, in business applications, as well as document management.

We will also describe other areas where metadata can help manage statistical or transactional information produced within an enterprise. We hope to bring insight in how metadata from the data world, web world, and digital assets management world can help us learn from each other and help us establish a more robust understanding of enterprise information architecture. By "Breaking New Ground" in learning what has been done in the Dublin Core and Data Warehousing/Data Modeling community we can expose ourselves to new methods, tools, techniques, and possibly establish best practices together for enterprise information architecture.

Back to Top

Stories from the Field: Never Consider Yourself a Failure, You Can Always Serve as a Bad Example

On my course syllabus I include the line, "never consider yourself a failure, you can serve as a bad example." I believe I received the quote from a fortune cookie.

In this presentation, I'll present three case studies of projects where the results were not as grand as I would have preferred.

What happened? We'll see how people, politics, perceptions, time, and budgets conspired in fascinating ways leaving users with products that could have served them better.

Thom will tell tales using a variety of narrative maps. You should get a chuckle or two, and learn some strategies for avoiding similar fates.

Back to Top

Producing No-Duh Deliverables

Whether you're preaching to the faithful or trying to convert the masses, one of the information architect's biggest challenges is to make necessarily complex solutions easily understood.

There's seldom a direct line from an IA's solution straight through to production. An IA (or their supervisor) first needs to garner support for the idea, usually from people in different departments and with varying backgrounds. Solutions frequently get changed in the process and the less understood the solution is, the greater the risk that its effectiveness will be compromised.

The challenge is to get everyone to talk about the same things in the same ways, but this puts intense pressure on the deliverables that define solutions. Ideally, these deliverables need to get this kind of response: "Yeh, no duh! I get it already!"

This presentation will show participants how to produce No-Duh deliverables. We'll do this by reviewing No-Duh examples from various fields; walking through a couple of No-Duh case studies; and outlining key No-Duh guidelines. Finally, a few brave participants will produce No-Duh solutions on the fly and be critiqued by the group.

The interactive presentation will be energetic and fun and supplemented by some No-Duh PowerPoint slides. Participants will also get handouts for the two case studies.

Back to Top

Recycle, Reuse, and Rebuild: Information Architecture on a Budget

You have seen them in design magazines, or on cable TV - those wacky, award winning homes built from recycled truck tires, large metropolitan area phone books, or carpet samples. While we quickly recognize the components from which they were made, we might never think of building something useful from these cast-offs. Clearly, effective recycling and reuse requires a special vision. This same challenge faces many information architects and developers today: reuse materials from pre-existing designs to build something more dynamic, scalable, and, most of all, customer-friendly. We must provide multiple formats, each designed for an individual audience, each crafted from the same basic pieces.

Our case study presents an approach to the problem of information reuse. We were asked to provide Web access to all product documentation with the following goals: " Make the information easy to locate using familiar navigation methods, such as indexes and tables of contents " Provide points of reference between any hard copy guides and the web-based information " Provide regular updates to the web-based versions " Allow customers to download the updates " Provide a comprehensive search facility that lets customers access not only product documentation but also technical support database solutions, and marketing information such as white papers.

We addressed this problem by reviewing the components of the existing structures, developing standards for components to facilitate reuse, and designing new structures using these recycled items. Our presentation will describe the process of identifying existing structures and will present before and after samples of several designs making use of these recycled and recast building blocks.

Back to Top

Birds of a Feather Discussions

• Controlled Vocabularies & Semantic Web
• Content Management Systems
• IA Tools & Applications
• IA Process & Selling IA

Back to Top

Posters

• Social Virtual Interface (SVI). Baris Aksakal
• Blueprinting: Moving into Precision. Clifton Evans
• Taxonomy and Thesaurus: Analogy with a Play. Mark Geljon, Almar van der Krogt
• 3D images and Metadata: Can We All Just Get Along? Elise Lewis
• Visualizing Site Traffic Alongside Site Hierarchy or Flow. Brett Lider
• Joint Evolution of Web Browsers and Online IA. Jennifer E. Jobst, Don Turnbull
• LUCIA: A Comprehensive IA Process Model. Arno Reichenauer
• Building Information for Computerized Systems: Developing Country Scenario. Daniel N. Ruheni
• URI as Navigation Tool. Manu Sharma
• Metric Aggregation for Social Network Analysis in Blogspheres. Frederic Stutzman, Butch Lazorchak, Jesse Wilbur
• Information Architecture & Social Communication: Twins? Javier Velasco
• The User Experience from Design to Use, and Back: A Causal Model. Javier Velasco and Ricardo Baeza-Yates

Back to Top

Bottom-Up Information Architecture: Re-Designing an Enterprise-Class Web Site

This case study is about a bottom-up information architecture exercise, a site re-design that was primarily a re-architecture of an enterprise-class web site. The company had a acquired a number of companies in a short period of time, and the company's digital asset had grown organically without a centralized web team to shape its direction. Information was hosted at a number of different locations by different groups, updated manually, and there was no sense within the company of a single, unified view of it's digital assets. The case study outlines some of the information organization problems that were defined and walks through how those problems were addressed by re-architecting the site from the bottom-up, improving the site's search capability and implementing a content management system. A metadata schema was implemented that improved search and facilitated findability across the site. Content and audience usage were both analyzed and a competitive analysis of other sites was performed. Business requirements were defined, as were metrics for success. A design system was developed and detailed templates were specified to accomodate all information within the company's taxonomy. The methods used and lessons learned will also be covered in the presentation.

Back to Top

The Use of Narrative in Interaction Design

What role can narrative and the tools of narrative play in interactive design to create more enriching user experiences? The time has come where a discussion of narrative structures in web design can move beyond the conceptual into a practical discourse on how narrative is being used and can evolve in use for conceptualizing and developing focused user experience solutions.

Our presentation will explore how narrative theories and structures can inform the interactive development process and enrich user experiences. Three key points will be covered:

1. A background on narrative in the context of interactive media.
2. Applying narrative tools in the creation of interactive experiences. How can the narrative be used to conceptualize, communicate and create websites that strategically balance creativity with bottom-line business requirements?
3. The evolving role of narrative in web development. How can more advanced narrative structures be utilized in creating rich, immersive experiences? How will narrative inform the evolution of interactive media? What new processes and deliverables will be required to meet this new reality?

The presenters collectively have more than 30 years of experience in interactive design, with backgrounds in User Experience Development, Content Strategy, Information Architecture and Film Production. The presenters have experience designing websites and applications for corporate clients in industries ranging from pharmaceutical, information technology, retail, professional services and entertainment.

Back to Top

14 Things Users Want to Know

The design of information rich web sites is all about helping users gain the information they need. It's about answering questions. How do we know if we've answered all the questions users have? How do we know if we've answered each question well?

The research team at User Interface Engineering spent the last year systematically analyzing thousands of questions that users asked each other online. After studying more than 3,000 posts, they've produced a framework that explains what users want to know. This framework is essential for understanding the key to quality user assistance.

In this presentation, Jared will talk about the 14 different types of questions users ask. You'll find out how you can use this framework to create your own user assistance project plans and ensure that you answer all the questions users have.

You will learn:

• How to identify the motivations behind user's questions
• How to tell when users are satisfied with the content
• New techniques for planning, writing, and validating the user assistance to ensure its completeness

Back to Top

The 2003 Dublin Core Conference On-line Proceedings: Building Metadata-Based Navigation Using Semantic Web Standards

One of the touted benefits of the Semantic Web is that it will make searches more precise and efficient by leveraging metadata about web site content. Faceted metadata retrieval is an approach to providing users access to large collections of semi-structured data and content that promises an improvement in usability over that available using more traditional search methods. In this case study, we describe how the on-line proceedings for the 2003 Dublin Core Conference were implemented by combining these two ideas and generating a faceted metadata retrieval interface from instance metadata, ontologies, and controlled vocabularies expressed in RDF and RDF Schema. We share lessons learned in the design and implementation of the proceedings, and in particular focus on emerging best practices for representing and sharing metadata using Semantic Web standards.

Back to Top

Information Design and Information Architecture - Together Again for the First Time

Although the field of Information Design is recognized as one of the components of Information Architecture, it is one of the more overlooked fields when it comes to our day-to-day work. How can our use of Information Design enhance the practice of Information Architecture, and the quality of our products?

IAs spend a lot of time thinking about how content is organized. We deal with site-wide systems and navigation schemes, and use techniques like card sorting and content inventories to understand how the product's pieces fit together. But if we do not pay attention to the content within the pages themselves, and how that is presented, we may seriously jeopardize the goals of the tools we're building.

This talk will feature work from CTB/McGraw-Hill, a division of the publisher McGraw-Hill that publishes standardized tests. CTB is moving away from traditional paper-based systems and towards giving customers the tools to schedule, manage, deliver and even create their own tests. These tests result in a massive amount of data about the students, which is delivered in the form of reports. The reports were traditionally designed for by Researchers, Statisticians and Psychometricians for an expert audience of Test Coordinators at the State and District levels, but now need to make sense to teachers, parents and sometimes even students.

How did a change in audience affect the page-level design of products? How was Information Design used to address the issues involved? And what impact did it have on other components of the Information Architecture, such as content organization and interaction design?

Back to Top

An Ethnographic Study of How Stockbrokers Use a Web-Based Trading System, and Recommendations Based on Information Learned

This project studied how independent stockbrokers used a web-based trading system to manage their clients' trades. This system is part of a broker workstation that is used daily by thousands of independent financial advisors. The inspiration for this project came from an earlier study in which it appeared that there was considerable inconsistency in the way that the advisors completed online trades. A study team of three primary investigators visited the offices of 18 independent financial advisors in 14 different cities across the country. The team used ethnographic research methods such as participant observation, recording of detailed fieldnotes, and interviews based on open-ended questions. The team also gathered forms and documents that the advisors had designed to help them in their daily work. The team quickly discovered subcategories of advisors (advisor types). These types, which varied in their approach to trade management, interacted with the system in very different ways. The core functionality of the system, trading, supported all advisor types fairly well, but there were many opportunities for improvement in the system that would better support all Advisor types. The team produced a complete set of task flows that delineated the differences between advisor types during the trading process, typical use cases based on advisor types, advisor type personas, and an interaction diagram that outlined recommended additions to the system to support user needs based on advisor type categorization. To summarize, the use of ethnographic research tools and methods proved to be highly effective in the discovery and documentation of different advisor types during trade processing. The thoroughness of the data collected allowed for very detailed documentation of user behavior, and also allowed the team to produce highly targeted recommendations for future development in the system.

Back to Top

Using Machine Learning Techniques to Populate Dynamic Interfaces

In recent work (Marchionini & Brunk, 2004) Marchionini and Brunk argue for the utility of a general class of interfaces called Relation Browsers (RBs) for information architecture. The goal of the RB is to help system users develop an overview of the contents and global structure of a website. The relation browser accomplishes this by allowing users to define context-dependent slices of a data set. This slicing proceeds by articulating necessary conditions for relevant information on orthogonal classificatory facets. For instance, a user of movie database might query the RB to select Dramas, produced between 1945-1950, that are available in DVD format.

This talk pursues the application of machine learning (ML) techniques to the deployment of RB interfaces. The work is important in two respects. First, there is the practical matter of populating the RB interface. How may we assign facet relations to documents? Automatic text classification has proven useful for this task. Second, can we discover facets automatically, without a pre-existing class scheme? Projection pursuit methods show promise for inferring new topical relations from pre-existing structures. Many websites have portions of their content topically classified. Even a partial taxonomy gives us useful evidence as we construct a model of a site's global information space.

The goal of this research is to enable information architects to utilize novel interfaces in the absence of pre-defined topical relations. For illustration, the presentation describes ML's role in an ongoing information architecture project for a large governmental website. The talk uses this example to foreground the successes and to articulate the challenges inherent in automated approaches to inferring inter-document structure in service to the deployment of RB-style interfaces.

Back to Top

Information Visualization: Failed Experiment or Future Revolution?

For more than a decade, information visualization research has been promising vastly superior interfaces for coping with large amounts of complex information. If successful it will revolutionize information architecture. But the revolution is still pending. Why? What happened? Where are these vastly superior visual interfaces? Has information visualization failed, or will there be a future revolution?

This presentation will answer these questions. It will examine the history and guiding principles of information visualization, analyze seminal visualization applications, and summarize current research--all within the context of information architecture theory and practice. It will provide both an introduction to the fundamental concepts of visualization and a discussion of the major challenges that currently exist. Special attention will be given to work in digital libraries, information retrieval, and mapping large document spaces.

The presentation is about both "what might be" and "what currently is." However, it is not intended for the IA who needs immediate solutions to today's problems. The aim is to examine a promised future based on interactive visual representations of information. It will explore why this future has not appeared, if it might yet appear, how we might get there, and what such a future would hold for information architecture.

Back to Top

The Blind Leading the Blind: Theorizing a Web for the Visually Impaired

How do you plan a site for the visually disabled? Section 508 tells us what not to do, but it doesn't present a model for doing things right. This case study shows how one web site approached the problem of planning a site for the visually disabled, and how the lessons we learned can work for you, regardless of your audience. Understand how browsing behaviors are similar for the average user and the visually impaired user. Learn how a linear underlying structure can improve the browsing experiences of both audiences, and simplify the web team's life by making content independent from visual design.

Back to Top

Using Taxonomies to Improve Information Architecture for Enhanced Information Discovery: A Case Study of the Nursing Portal at Singapore General Hospital

Nurses in a major hospital in Singapore realized that they were constantly in need of instant knowledge for their day-to-day work and were also inundated with patient and non-patient related information from multiple sources. Enthusiastic about portal solutions, they hastily developed a nursing portal to capture all related information to derive knowledge from it, but quickly started experiencing difficulties in knowledge discovery. On their request for advice, we set up a multidisciplinary team to help organize knowledge on the portal. Steps to improve information architecture were considered crucial. Among other things, we focused on developing and deploying taxonomies for better content organization and management using an integrated approach. While initial tasks of user requirements, content analysis, and assessment of relevant tools have already been accomplished, work is satisfactorily progressing for implementation of taxonomies and related activities to improve overall information architecture. We expect to complete the remaining work by the end of 2003. A user assessment is planned to be carried out to study the effectiveness of use of taxonomies and impact of improved information architecture on information discovery. The project will be concluded before the conference dates. This case study will highlight how a team with diversified expertise was able to use their combined skills for improving the information architecture and as a result enhancing information discovery. We will also share the problems we faced in gathering data about the users and the contents, building & deploying taxonomies to improve the information architecture. The case study will also demonstrate how existing tools and technologies were integrated with the local system for more cost-effective implementation. We also expect to share information on usability testing to be carried out to study the impact of our proposed end-to-end solution on the improvements in knowledge discovery.

Back to Top

Information Search Experience: Emotions In Information Seeking

Emotions in design is an emerging topic of interest within human factors work. This is a broad concept that does not lend itself readily to clear boundaries and concise definitions. Nonetheless, some significant work has already been done. Though much attention has been given to emotions elicited by product appearance, affective considerations are not limited to visual and aesthetic aspects: user interaction with information systems also involves, evokes and influences our moods, feelings and emotions in many different ways.

Information seeking on the web, in particular, can be seen as an emotional experience. Unfortunately, confusion and uncertainty tend to dominate feelings of enthusiasm and optimism. The joy of discovery and pride of learning can be rare feelings against a backdrop of frustration and a sense of being overwhelmed for many web searchers. Though this situation has been widely commented on and acknowledged, specific design techniques that account for the entire information seeking experience seem lacking.

This presentation focuses on information seeking behavior on the web and the potential role emotions play in the search experience. Participants can expect to be exposed to some relevant literature and concepts from information seeking research. A holistic design framework that accounts for user emotions as they seek information will then be presented. This framework intends to facilitate general design processes by directly mapping user information needs to feature development. Finally, a theoretical model of the entire Information Search Experience (ISX) from a user's perspective is proposed.

Back to Top

Children as Design Partners for Educational Web Sites: Challenges and Lessons from the NIDA for Teens Project

Building a Web site that provides serious educational content for teenagers can be challenging. To understand and meet the needs of users, designers need real users to guide their work. Traditionally, users' perspectives are solicited through techniques for requirements-gathering and through focus-group to gauge the acceptability of proposed solutions. When teenagers are the target audience, however, developers may need to go beyond traditional methods for involving users in order to create effective and engaging materials.

In this case study detailing the development of the NIDA for Teens Web site (http://teens.drugabuse.com), we describe the process through which IQ Solutions, a health information education and dissemination services company, produced a drug education Web site for teenagers sponsored by the National Institute on Drug Abuse (NIDA) - http://www.drugabuse.gov. NIDA, a Federal agency within the National Institutes of Health, conducts 85% of the world's research on drug abuse and addiction. As part of its educational outreach effort, NIDA wanted to develop a Web site to provide students (and their parents and teachers) with the latest science-based information. The challenge was to maintain the serious, science-based focus of the site while making it useful, interesting, and engaging for its intended audience (teens from 11 to 15 years old).

This case study explores differences between focus group inquiry and collaborative design as sources of information for site and software development. Focus group work elicits critique; the collaborative design process goes beyond critique by allowing teen participants to generate, prototype, and test concepts. We describe the collaborative effort between IQ Solutions and the Intergenerational Design Team in the School of Information Arts and Technologies (http://iat.ubalt.edu) to show how this novel approach to developing content for teens produced the NIDA for Teens Web site. The study includes discussion of metrics for success and lessons learned from the collaborative design process.

Back to Top

Rapid User Mental Modeling at eBay.com: A Case Study

This case study describes the user research methodology (called Rapid-User-Mental-Modeling or RUMM) used in preparation for the IA redesign of eBay.com. The eBay website has grown in an organic manner, leading to an inconsistent user experience. This project, the first effort to delineate user mental-models for the entire website, posed many challenges. Ambitious plans to change the site meant that the IA needed to be future-oriented. The site has a large and diverse user population. Senior eBay management required proof of the need for such drastic restructuring.

Stage 1: Explore domain

To begin, we interviewed eBay users and senior management to generate a comprehensive list of tasks. To develop a future-oriented IA, free-listing was used with site-stakeholders to elicit "Horizon-Tasks" (tasks planned for the future). Horizon-tasks constituted 20% of our final, representative 100-item task-list.

Stage 2: Understand user mental-models

Next, the 100-item task-list was used in an open card-sorting study with main eBay user types. Aggregate mental-maps (based on cluster-analysis) were used to generate a preliminary IA, including 5 top-level categories.

Stage 3: Verify & refine mental-model

To verify the proposed 5-category scheme, an online, closed card-sorting study was conducted. 1000 participants placed each task in one of the 5-categories or suggested another category, identifying tasks that required further attention.

The results of the three stages helped us design a site-map showing the proposed structure for the entire eBay website; the site-map is now guiding the IA redesign.

The RUMM methodology helped us understand business context and user mental-models. Results convinced senior management that the current site did not match user mental-models and that an IA redesign was needed. Having access to user mental-models and a comprehensive site-map has given eBay designers a broader perspective on the site's overall design direction, clearing the road towards a more consistent and unified user experience.

Back to Top

Brand-Driven Information Architecture

There's more to branding than logos and packaging. Even if you don't work for an advertising agency, brand strategy may influence your information architecture work in ways you haven't anticipated. In this presentation, Jesse James Garrett examines the often overlooked intersection where brand marketing and information architecture converge. Reflecting the Summit's theme of "Breaking New Ground", we'll look at ways information architects can adapt and evolve their approaches to incorporate brand considerations.

The presentation will explore questions such as: How can an understanding of brand influence the way we create information structures? How can information architecture enhance a customer's impression of a brand -- or undermine it? What insights can IAs gain from the methods marketers use to develop brand identities? How do the impressions your customers have of your brand shape their expectations of the information architectures you provide for them?

Back to Top

J-Flow: From Sitemap to Prototype

Tools for presenting Information Architecture deliverables such as sitemaps, navigation structures, screenflows, wireframes/blueprints and mockups are rare. EzGov's information architecture tool "J-Flow" is used to present the design of interactive systems, from sitemap to individual screens, using one, commonly available delivery format (HTML).

The J-Flow has been developed by information architects of EzGov (e-Government Technology Specialists, http://www.ezgov.com) and applied in the design process of many e-government online presences, ranging from information-oriented portals to transactional web applications.

The J-Flow is based on Microsoft Visio's "Save as Web Page" feature but in-house developed Visual Basic scripting behind well-designed stencil objects allows for features that are useful for information architects.

With the current version of the tool information architects can: - create screenflows, from the top-level conceptual sitemap to low-level alternative flows for error conditions. - link high-level flows to low-level flows to create hierarchies of screenflows. - link wireframes to screens. - link mockups to screens. - link together ("package") sets of screenflow and wireframe documents. - export the resulting package in a demonstration-ready format (HTML + images).

The results have been used: - by IA's to share designs with team members (fellow IA's, interaction designers, visual designers, requirement analysts, software engineers). - by IA's to show designs to clients and other stakeholders. - by requirement analysts to check completeness of business and validation rules. - by software engineers to understand the navigation and interaction design. - by clients' helpdesks employees to answer end-user questions about the application.

The presentation will introduce the J-Flow, its place in the design process, show examples, and discuss future developments.

Attendees will learn how a set of Visio templates and stencils helps information architects to develop deliverables that can be used to present designs.

Back to Top

Architecting Time: Designing Online Events and Other Magic Tricks

What is an online event? How is it different from a run-of-the-mill web site? And why should we care?

As corporate travel and marketing budgets shrink, we’re seeing more companies creating events online in place of staging in-person conferences. If done well, online events can provide valuable content to users while deepening the relationship between enterprises and their customers. If done poorly, these events can be flat, undifferentiated experiences that offer little more than a standard press release.

Our challenge as information architects and experience designers is to trick time so that we create experiences that convey the energy and excitement of an in-person event, without the attendees ever leaving their desks. This presentation will explore these concepts, including rules for creating the illusion of time and how this impacts design.

In the second half of the presentation we look at a case study of an extremely successful online event that was launched in September of 2003. We will walk through the process we used to create the information architecture and the resulting design.

If time allows, the final part of this presentation may include an optional group exercise of designing the ASIA conference as an online-only event.

Back to Top

The Aesthetic Imperative: Four Perspectives on Aesthetics to Impact the User Experience

Emotion and aesthetics are all the rage in the user experience community, with luminaries like Don Norman and Patrick Jordan extolling their virtue in designing products. But what does this all mean for information architects cranking out site wireframes and taxonomic schemes? Should IA's care about this stuff? Absolutely! I believe that aesthetics help ensure that "total integrative experience" inside and outside of a product, beyond the visual veneer-whether an elegant navigational flow or powerful digital library. More than simply slick style, it's about insight into the user's desire for delightful interaction. By having this in mind, the IA's creation of site maps, search UI's, and metadata models can support the beauty and emotional content of a product, at a subtle but influential level.

By connecting to other disciplines' viewpoints on aesthetics, I present a four point model for interpreting and applying beauty to products. This model draws upon writings by Dewey, Cziksentmihalyi, Gelernter, and Gropius (philosophy, psychology, computer science, and architecture, respectively). With this model, I hope to expand upon and enrich the typical IA repertoire, towards an "aesthetic imperative" that IAs can guide in collaboration with other designers, and even programmers! My presentation will feature actual and speculative products from prior work to reinforce the four perspectives: products that invite personal fulfillment, that promote lifestyle enhancement, that convey functional elegance, and that engender spiritual/cultural balance.

Thus, the IA audience should walk away valuing aesthetics as an integrative element, learning how to apply different interpretations in a project, and how to relate aesthetics to traditional IA activities. Finally, the model provides an effective language that can structure a conversation about aesthetics in design.

Back to Top

Then A Miracle Occurs...

One of the most important steps in a user-centered design process is synthesizing user research. How do we take a potentially overwhelming amount of data and be able to generate design concepts and solutions from it?

In this talk, I'll provide a brief history of work modeling, and then present two new modeling methods that have been developed by Adaptive Path. The Mental Model clearly delineates how a user perceives and approaches their tasks and goals, and affords information architects the ability to group material where it best supports the customer. The Thematic Model encourages the emergence of psychological narratives that in turn focuses efforts on solutions that will provide the most value for the user.

The talk concludes with a look at evolving our methods for synthesize user research, at times borrowing from other disciplines: archaeologists employing photographic data, or social network analysts 'socio-grams.'

Throughout the discourse, I'll pay attention to two distinct threads. The first is the content of these models. The meat. How they work, and what they help us find out.

The second will be the visual language employed by these methods of synthesis. Synthesis relies on breaking free of a text-only world, to incorporate shapes, lines, drawings, photos, and even video in our work. What implication does this have on our practice?

Back to Top