IA Summit 2005

Since the advent of the Internet, corporations have been faced with a glut of new and ever-proliferating ways to deliver information online.

However, basic business functions such as decision-making, real time collaboration, and information analysis are still more likely to be done offline. Why?

This poster presents an information model for analyzing the efficacy and appropriateness of various Internet-based information delivery channels given a specific business context.

The poster provides:

• Descriptions of each channel--including mailing lists, forums, instant messaging, web conferencing, news feeds, updates, alerts, commentary, email, briefings, wikis and blogs, with links to authoritative sources for further follow-up
• An attributes table, including synchronous or asynchronous communication, number of participants, criticality and permanence of the message, formality of the format, and level of detail
• Graphs showing which channels are more appropriate given a primary business task (e.g., analysis, collaboration, broadcasting, reporting)
• A timeline placing various channels in a historical context
• Conclusions

top of page

The Web is now very widely used by school students, including those in the elementary grades, as a source of information to support class assignments and projects. Unfortunately, children encounter major problems in finding this information when they need it; research has shown that it is difficult for children to formulate searches and to revise them if they do not work, and that when browsing on the Web they often lose their way. As a consequence, they too often miss relevant information.

One solution to this problem is to develop better ways for children to seek information on the Web. In our earlier research we have been designing portals that try to overcome such problems. In this new research, however, we are experimenting with a new approach that relies upon a virtual rather than a conventional interface for such a web portal. The virtual reality interface will enable children to find information by navigating (browsing) through a virtual library, much as they navigate through games on their computers. The library will have corridors and bookshelves, and information will be stored on the shelves as books. These books can be taken down from the shelf and opened to reveal the actual information. There will also be a library catalog that can be searched to find where books on particular subjects are shelved.

Our earlier research suggests that elementary school students are attracted to the concept of such a virtual reality interface as another way of looking for information. One reason why we think such an interface may prove successful is that children often like to browse rather than search for information; the virtual reality interface should facilitate this activity. The ideal portal, then, may provide a conventional interface for searching and a virtual interface for browsing. Alternatively, we may find that some children like one interface better than another, depending on factors such as their age, gender and spatial abilities. These are the kinds of questions that we intend to explore in the research by asking students to evaluate the virtual interface alongside the conventional interfaces that we already have designed.

In a broader context, we hope to learn more about designing virtual realities and about how children respond to such designs in order to understand better how this approach might be applied to other tasks than web portals, and to determine the impact of such an interface on the information architecture of a site.

top of page

Mark Geljon, Almar van der Krogt

This poster reflects a project we carried out for a worldwide semiconductor producer.

Problem: Most organizations have a lot to communicate with their clients. To facilitate this process they create Information Heavy websites, with predefined (top-down) structures and appealing interfaces that should lead visitors by all the information. It is even expected that visitors engage in an interaction with the organization through the website. Eventually, this should lead to a long term (online) relationship, realizing a true involvement of the client with the organization. However, for visitors in most cases the homepage is not the entry point. And only providing the right information is not enough. Binding visitors is getting more difficult, but only when they have a visit of a certain amount of time and a certain depth the organizational message has the opportunity to get across. Taking into account that the set-up and maintenance of these websites are expensive, finding ways to increase retention becomes a real issue. In addition, due to the enormous amount of informa tion on the web, users have a habit of Googling all the information they want to find (especially from information heavy websites). This is not wrong; the perception of users is that they will reach their goal easier with a search.

Solution: Always provide alternatives according to the tasks at hand from a visitor perspective (the information scent should indicate that staying on the site is more profitable). Through an extensive analysis on the internet related tasks and the tasks that will follow, a mapping of the scenario's that will follow previous scenario's can be made. These will be presented as alternative routes. The internet related tasks are transformed in web visit scenario's. These scenario's are 100% customer oriented, the translation to pages on the organizations website takes place in a matrix.

Implementation: providing alternatives The website becomes independent of the navigation a visitor uses. The various webpages are more or less independent information bites (the so called 'info snacks'). After a page is being found, a list of alternative links is presented, which (with appropriate labelling) will take the visitor to other interesting information on the site. By clicking these alternative links, the length and depth of the sitevisit will be longer.

top of page

Mike Munnelly, Joanne McLernon

The graphic ostensibly defines a process to capture the voice of the customer. The story it tells is why it's so important, from a business perspective, as we're no longer just validating with users during a design process, but reaching out to them earlier- as we plan to create a solution.

As the image shows- an advanced firm will have much information- and many projects. This process is intended to help identify what is most important to specific groups of customers- in order to provide solutions and experiences that have lasting impact.

The image (broken into 2 pages- but will be one for the poster) shows how you can bring the customers perspective into your business cases. By going through an effort like this you can get real customer insight, be able to quantify them, with statistical relevance and use the outcome as a lens to apply to business problems.

It begins with a collection of the extant data- and filtering it with your existing user knowledge- adhering to persona and segmentation best practices. You then undertake a qualitative effort (field research, contextual inquiry, observations, interviews) to gain initial context to then validate with qualitative measure (site usage data, clickstream analysis, user surveys) in order to create the lens, or perspective that is the captured Voice of the Customer.

As firms become more savvy about the user experience, they can take more advantage of efforts like this. The more we understand about a user, they more deeply we can probe. The goal is to being the customer into the strategy planning process- and the output from an effort like this would bring audio, video of customers into the conference rooms.

top of page

Rahul M. Kulkarni

International Children's Digital Library (ICDL) is a research project to develop innovative software and a collection of books that specifically address the needs of new generation children as readers. The ICDL collection mostly consists of historical literature, which is no longer under copyright and a small collection of contemporary literature, whose copyrights are given to ICDL by a few authors. But majority of the current children's books cannot be added in the ICDL due to their length of copyright protection. Thus the goal of this research project is to bridge the divide between what exists and what does not exist in the ICDL and provide access to contemporary books that are under copyright protection, since their presence in the digital library environment in some form is essential.

Digital booklists, which consist of lists of book titles, serve this purpose to some extent since adding book titles to digital lists do not require permission from the authors. However, most of the children's digital booklists are created by adults and do not provide features and functionalities that offer enough information to enrich the digital booklist experience from a social standpoint and motivate children to use them. Most of the research till date has been done on how to enhance DLs and how to design information and social spaces for adults. Not much research is done to explore how digital booklist can be designed for children as an information space that supports sociable environment for peer-to-peer review and knowledge exchange. Thus working with the Intergenerational Design Team of children 9-14 years old and adults together as research partners, this project applies cooperative enquiry methodology, developed at the Human Computer Interaction lab (HCIL) at the Univer sity of Maryland (UM), to investigate and extend prior research on various information and social needs of children.

This project addresses issues such as:

• What are various types of children's information needs and how do information-seeking behaviors of children (users) vary according to their information needs?
• Why is it important to know about these needs and behaviors in cyberspace while designing architecture of information space that support sociable environments for children?
• Why is Information Architecture important while creating children centered designs and how it helps organize their information space in cyberspace.
• What methodologies can be used to create Children centered Information Architecture of an online sociable environment?

By collaborating with children as design partners and implementing principles of user-centered and participatory design, this research project further discusses how ICDLs sociable booklist system is developed by designing the architecture of such a system along with prototype interfaces to provide engaging user experience. Finally, we state that since the children's digital library (ICDL) which is currently static, aims at the international audience, the collaborative and sociable environment of the digital booklist will help bridge the gap between the static and the dynamic environments as well as bridge the gap, to cross social and cultural boundaries across the world.

top of page

Alex Kirtland

This poster describes what a dashboard is (aka Executive Dashboard), what are some key things to keep in mind when developing a dashboard, and what makes for a successful dashboard.

top of page

Jamshid Beheshti, Andrew Large, Charles Cole

An information architecture (IA) matrix was used to evaluate four children's Web portals. The Matrix is developed to assist designers in defining a portal's purpose and to serve as a template for a systematic approach to information architecture (Large, Beheshti & Cole, 2002). The IA Matrix consists of 51 attributes divided into three main components: Model, Interface, and Retrieval.

The IA Matrix is designed so that once the Model is defined, the Interface and the Retrieval components are relatively easily identified.

Four children's Web portals were evaluated using the IA Matrix, and the results were compared to the evaluations conducted by four children focus groups (Large, Beheshti & Rahman, 2002). The portals were chosen based on the intended audience (children between 10 and 13), the purpose (mainly informational), variety of interface design, and relative popularity. Although since the time of evaluation one portal (LycosZone) has significantly shifted its orientation to serve a higher age group, it is included in the results for comparative analysis.

The four portals were evaluated by assigning a score on a five-point scale to each of the 51 attributes in the IA Matrix. Yahooligans! achieved the highest score due to its many navigational aids, interface consistency, its relatively fast response time, and its help facilities. LycosZone (at the time of evaluation) had the best overall portal objective. The portal delivered information, albeit not too effectively, as well as education and entertainment.

KidsClick was designed by librarians, who paid much attention to searching and information retrieval, at the expense of the portal's interface. Ask Jeeves for Kids had the lowest scores of the four portals. In terms of the Model and Interface, its scores were very similar to KidsClick even though its philosophy and approach are different.

By using natural language and a question/answer model, Ask Jeeves for Kids attempts unsuccessfully to mimic a library reference interview situation.

The results of the focus group evaluations of the four portals were mixed. The girls groups unanimously chose LycosZone as the best overall portal, whereas for the boys the top choice was Yahooligans! The unanimity within groups did not extend to choices of least liked site, although Ask Jeeves for Kids tended to fare worse with the boys and KidsClick with the girls.

The results of the evaluations by focus groups, in general, correspond to the scores achieved using the IA Matrix to assess the four portals. Comparatively, the top choices of boys and girls had higher scores in the IA Matrix than the portals deemed to be disliked by children. The results demonstrate that the IA Matrix can be used to evaluate children's Web portals.

Further research is needed to assess the inclusion of a new attribute, gender, in the Matrix.

top of page

Are Halland

WM-data User Experience has evaluated the search experience of 25 norwegian websites. The poster will present the criterias used, results found and insights gained from the evaluation. We sum up a road map for implementation of best pracice search experiences in the form of a Pyramid for the Search Experience (loosely based on Maslow's Pyramid of Needs).

top of page

Joe Lamantia

This poster depicts how a small set of standardized Information Architecture structures and elements was used to create an effective suite of interconnected Executive Dashboards at low cost and without substantial redesign effort.

This suite of dashboards meets the diverse information needs of senior decision makers working within many different business units in a global pharmaceutical company. These dashboards incorporate a wide variety of data types and functionality, but present everything within a consistent and usable User Experience by employing modular tiles and navigation structures.

The poster shows how the basic IA component of a tile or portlet with a standard structure, elements, and labeling, can contain a tremendous variety of content types. The content types include qualitative and quantitative visual and textual data displays, as well as complex functionality syndicated from other enterprise applications. It also shows how tiles are easily combined with other tiles or portlets to create more larger scale structures that are still easy for users to comprehend, allowing them to synthesize and compare formerly siloed information to guide strategic decisions.

The poster shows how simple information architecture components common to all the dashboards allow rapid access to a tremendous amount of information, from many sources. The poster shows how this framework scaled well and responded to changing business needs over time, allowing the addition of large numbers of new tiles, views, and types of information to existing Dashboards without substantial redesign or cost.

The poster demonstrates how a set of IA components allows designers to present critical business information by operating unit, geography, topic, or specific business metric, at varying levels of detail, based on the needs of specific audiences.

The poster shows how this set of IA components allowed numerous design teams to innovate within a framework, thus creating an extensive library of reusable tiles and views available for syndication throughout the suite of Executive Dashboards.

The end result of this approach to solving diverse design problems is a series of well integrated User Experiences, offering substantial business value to a wide audience of users.

top of page

Joan C. Bartlett

The advent of bioinformatics, the application of computer technology to the storage, analysis and representation of biological information, profoundly impacts all areas of biomedical research. However, in spite of the critical role that bioinformatics analysis plays in the advancement of biomedical research, and the need to integrate bioinformatics analysis and laboratory research, there is a lack of research into this challenging problem.

The proposed research presented in this poster is part of an ongoing study addressing the problem of linking bioinformatics analysis to scientific practice. To date, this work has produced a systematic protocol describing and detailing the application of bioinformatics analysis to the scientific problem of predicting gene function from sequence data (the functional analysis of a gene sequence).

The current problem is to develop and evaluate a system to make the bioinformatics analysis protocol accessible and usable to a laboratory scientist. From the perspective of information architecture, the challenge is to develop a system that accesses not one, but many different information resources, and presents them, and the information obtained from them in an integrated, coordinated and cohesive manner.

The system will comprise a hierarchical series of web pages, paralleling the levels of representation of the bioinformatics analysis protocol. At the upper level, the system will present the protocol at a broad level, providing an overview and allowing orientation within the protocol/system. At the most detailed level there will be a series of web pages, each providing description and instruction on the use and pplication of a specific bioinformatics analysis step. Intermediate level information will support navigation among the analytical steps, allowing scientists to determine which steps are most suitable to their particular research scenario.

The second part of this study involves testing the protocol and system. In a within-subjects experimental design, participants will conduct a functional analysis of a test gene sequence under two conditions, both with and without the system. Among the factors to be considered in the evaluation are effectiveness, efficiency, and user satisfaction.

The system being designed in this research will not provide access to a single information system. Instead, it must integrate and coordinate the access and use of information from over seventy individual, pre-existing information systems, each of which has its own unique information architecture. The challenge is to provide a framework that logically connects and integrates the use of a diverse variety of resources, and rationally applies them to the accomplishment of a goal.

Information interaction is intrinsically influenced by the information architecture of the system. In this work, an understanding of the workflow and the desired information interaction is used as the basis for information architecture.

Within the domain of bioinformatics, this research is innovative in that it is aimed at developing a task-oriented information system. The system discussed here, for the first time, takes a holistic, integrated approach to the architecture of a bioinformatics information system, linking multiple analyses and resources to accomplish a goal. back to program

top of page

Vinay Venkatraman, Alejandro Zamudio Sanchez

Improving Accessibility by Mobile Communication is an Applied Dreams project from the masters program at the Interaction Design Institute Ivrea in Italy

OBJECTIVE This project sought to improve accessibility for wheelchair users in urban settings in Italy by providing ubiquitous, location-sensitive information.

USER STUDIES The project started by examining the accessibility issues, real life problems affecting wheelchair users in Milan, Turin and Ivrea. . We observed users, emulated situations and utilized previous experience and knowledge of the subject. We discovered that merely providing physical aids like ramps, elevators, and marked pavements does not create accessibility. The lack of information about such features, facilities and their specific presence in localities was a serious deterrent in the free movement of users with special needs. In addition, users needed much more detailed and verified information, not just on-site data about a particular locality to plan and execute a trip. We developed video scenarios to communicate the realities and limitations of accessibility in public and private spaces.

ANALYSIS Accessibility is just not physical. It is equally mental and cultural. Offering adequate information can increase confidence in mobility.

There exists a need for very location-specific information like finding accessible restrooms, the closest parking space, back door ramps, accessible restaurants and other services. This is often soft information not published anywhere and come only out of experience, trial and error. Traditionally, such information passes by word of mouth. If it could be shared in a more systemic way, the implications would be greater. Enabling accessibility means having good, timely information dissemination and collaborative information gathering.

SOLUTION The final solution emerged as a city-specific, audio-visual database that can be accessed through a mobile phone when on the move and via the Internet when available. The content of the database provides relevant and very location-specific information about accessibility. It provides guidance, ratings and tips in the form of text, images and video, disseminating specific, validated information that assists users in making mobility decisions. The content of the database is largely generated by users themselves, i.e., wheelchair users. This user-generated content is moderated to avoid of conflicts, redundancy and misuse. The users are also encouraged to rate information to make it more reliable. The entire database is also available as a website where users can add further detailed information, and discuss and actively participate in debates.

PROTOTYPE: A basic prototype of the solution was made to show the relevance and value of real-time, location-sensitive, multimedia information. It helped work out the concept's feasibility and assess the technological barriers and limitations, and evolve workarounds. The prototype combined an MMS gateway, a web server and a scripted application on the Symbian OS running on a Nokia 6600 Mobile phone.

top of page

France Bouthillier, Tao Jin

One of the main objectives of Information Architecture is appropriate information delivery to ensure that the information needs of specific users are met. Competitive intelligence (CI) is the process by which an organization systematically and legally collects, organizes, analyzes, and disseminates information about its competitors and the external environment, in order to gain and/or maintain its competitive advantage. There is a wide array of software products available for CI purpose. Some desktop applications address the whole CI process including the identification of CI needs; acquisition of information; the organization, storage and retrieval of information; analysis of information; development of CI products; and distribution of CI products. There is a set of applications that address one particular step in the CI process: the acquisition of information. To search and retrieve information, intelligent agents are particularly useful because they can carry out some tasks on behalf of a user. In addition, given that CI involves a constant examination of business news, RSS tools have also been increasingly useful and used as news feed aggregators. All these tools have some value but it remains unclear to what extent they can contribute to facilitate the CI process. In other words, this proliferation of tools creates some confusion for CI practitioners and IA developers: to what extent intelligent agents and RSS tools should be integrated in a global information architecture addressing CI needs?

To answer this question, this poster proposes an examination and a comparison of the value-added dimensions of popular intelligent agents and RSS applications. Although these applications are very attractive, without a thorough examination of their features, it is impossible to assess their value-added dimensions and make recommendations on their importance in the context of CI. This presentation seeks to generate useful insights regarding issues related to the retrieval and storage of unstructured information (business reports, news, white papers), representing major challenges for the field of Information Architecture. Considering that few attempts have been made to establish conceptual links between Information Architecture and Competitive Intelligence, this presentation will generate new knowledge on the role of specific CI tools in IA strategies for improving access to information. The examination of the value-added dimensions of intelligent agents and RSS tools should be useful for further assessment of these applications for creating competitive intelligence. Although the potential of intelligent agents has been examined by various authors, a systematic assessment of their value has not yet emerged and, given the recent popularization of RSS tools, there is a need to understand their usefulness from a global perspective.

top of page

Andrew Large, Jamshid Beheshti, Valerie Nesset, Leanne Bowler

Children use the Web to find information for both recreational and educational purposes, but in doing so encounter problems in query formulation, search strategy construction and navigational orientation. One remedy is to design portals that will offer better support than either the regular portals that they typically employ (such as Google, MSN and Yahoo) or the specialized portals designed for children but which children rarely use (such as KidsClick and Yahooligans!).

Two intergenerational design teams comprising respectively elementary school students from grade six (11 years' old) and grade three (8 years' old), together with three adult researchers, designed using a Cooperative Inquiry methodology two low-tech web portal prototypes specifically targeted at the students' peers. These portals were intended to support information retrieval in an educational rather than a recreational context. They were subsequently converted into working web portals that can be used to find information in French and English relating to Canadian history and deemed appropriate for an elementary school audience. The portals provide multiple search and browse tools, offer personalization and help facilities, and present the interface in both English and French. A specialized hierarchical indexing structure was created to reflect students' representations of Canadian history. The two portals were evaluated by students in several elementary schools in both experimenta l and operational environments.

The poster presents the design methodology and the two portal architectures as well as briefly summarizing the evaluation data. The overall research objectives were to elaborate design criteria for children's web portals intended to support class-based assignments, to explore the role that children themselves can play in the design process, and specifically to test an intergenerational approach using Cooperative Inquiry.

top of page

Bryan P. Carson, Michael Lynch

Context of the work: The Middlebury College Library recently had to bring its entire web site into a content management server along with the rest of the college. This presented many challenges, but also many opportunities for improvement of the web site from the users' point of view as well as from the content providers point of view.

Many of the problems with the previous web site consisted of duplicated effort among the content providers (mostly librarians), dead links, varying links to the same content, and widely varying presentations of information within analogous documents (library guides for various academic subjects). Furthermore, documents were not very easily searchable because they were large static documents.

These problems were all innovatively solved through the information architecture we implemented fully utilizing the functionality of the Microsoft Content Management Server.

Innovative points: In many parts of our site we had long lists of links which also had accompanying blocks of text which could also include links or images. These static documents consisting of individual content blocks were often duplicated across the site on pages that were created by and maintained by many different content providers (and thus cause people to duplicate effort.) We created a special CMS document template which we called link with blurb. This document template allowed us to pull each link and associated text block apart into an individual document (posting in the CMS lingo). Once these were broken down into individual entities, they only had to be created once and then could be shared. This was tremendously helpful during our migration from the old site to the CMS.

This IA also allowed us to do several other things:

1. Through the use of various rendering scripts, the link blurbs can be mixed and matched into different dynamic documents as needed.
2. Using rendering scripts, we allowed documents built on other document templates to combine with the link blurbs in various manners. (This will be illustrated in the diagrams on the poster.)
3. Our central list of databases became an effective master list for most of the links in the link-blurb type documents.
4. This not only allowed us to easily keep these links up to date across the site, but it also allowed us to mask the actual target URLs of our databases (important when URLs contain usernames and passwords.)

Some of these things are possible with existing applications like Dreamweaver. However, implementing them was extremely cumbersome in a population of users who only infrequently update their content and thus never maintain the required software proficiency level to implement such features.

top of page

John R. Fisher

This poster discusses the design of an experimental API (application program interface) for vocal user interfaces. The emphasis is on component design which supports vocal interaction design issues. Related matters, including interaction diagrams, voice XML and intelligent AI techniques are also briefly discussed. Keywords: vocal user interfaces, vocal interaction design.

top of page

James Melzer

This case study is one in a set of three studies to analyze how the lack of an overarching profit motive skews performance measures used in government projects, and how this bias makes the job of user-centered design teams more difficult. An intranet portal implementation is described as an example of this problem.

In government contracting, the concept definition is often substantially completed before a contract is awarded. This means concept definition is often written by government staff who are not experts in the task being performed, or by contractors who have not been allowed to thoroughly research the task being performed. In both cases, the overarching goals of any government project must be to avoid waste and corruption and to minimize risk. The result is that projects tend to be measured by cost, time, and scope rather than performance metrics that evaluate the effectiveness of the overall purchase in accomplishing business goals.

This example demonstrates how a common public sector management method can result in a project that succeeds at common project management metrics but still fails to accomplish basic organization goals.

The project manager's primary performance metric was delivery of non-prioritized features within time and budget allotted. This was fundamentally incompatible with user centered design (UCD) principles, which stress metrics based on the successful and efficient completion of business goals, the satisfaction of users, and user adoption rates for new systems. Even if UCD practices are followed during design and implementation, this strategy opened the door to features that were usable but potentially not useful. Without an overarching strategy guided by a unified and well planned experience design, the project risked failure despite success in delivering effective applications on time and within budget.

top of page

Brian Manning

The purpose of this poster is to show how the impact of a user experience design can be measured. Modeling before and after usage data of a typical user flow to a funnel structure provides an easy means to visualize metrics and understand the impact of the solution on the user and the business.

top of page

Are Halland

An assessment of the potential to influence and control the user experience of large governmental web information spaces

Can we control how users experience an information space? This thesis seeks to explore relations between theoretical insights in the fields of navigation and rhetorics, to enable us to build a better understanding of the problem space and use to analyse examples. Ultimately, the goal is to form a combined theory of how navigation through a web information space can be intentionally influenced and controlled by a rhetor.

The starting point for this exploration was a failed, personal user experience with the governmental information portal norge.no. When trying to find information about kindergartens in my vicinity, the resulting experience was one of frustration, non-fulfillment of personal goals and certainly no efficiency effect on the side of the governmental bodies responsible for the website(s) implied.

Drawing on insights from classical and modern rhetorics, including the newly formed field of captology, and a large theoretical body in the field of navigation theory, including wayfinding and information scent, I then analyse three governmental websites to identify further problems, solutions and insights.

Finally, I use the combined insights from the theoretical and empirical investigation to form a new theoretical model of rhetorical navigation - the navigation pyramid.

top of page

Svetlana Symonenko

Although the notion of the Internet as an information environment that has difficulty lending itself to organizing efforts still prevails, there appears to be a trend towards conventionalization in website architecture, in terms of both explicit structure and implicit organization of sites' content. This is a bottom-up process, driven by practices rather than by prescribed standards. On the other hand, research on information seeking behavior and HCI offers evidence in support of the existence of certain expectations of sites' external design and internal structure, which users have developed based on accumulated experience in online searching and browsing. To date, major research effort focused on interface-related aspects, with insufficient attention to similarities in the organization of websites' content, similarities both existing and as perceived by users. These two aspects, however, appear mutually informing and critical for ensuring findability of Web-based information.

To address this need, a dissertation study is under way to explore users' expectations of the websites' content organization, and whether such expectations depend on the type of the site. Similarly to the way the results of system usability studies inform the system's design, implications of such study of content usability will contribute to the improved organization of web-based content, as well as to a more user-oriented design of online search interfaces. The study is conceptually related to the notion of information shapes, integrating semantic and spatial characteristics, as well as to digital genre research. The novelty of the study is in taking a holistic approach to a website, as it extends the concept of digital genre (which, to-date, has been largely applied to individual web pages) to cover an entire website as an information entity. To tap into users' expectations, the study employs a theoretical framework of internalized mental representations that people form of empirical phenomena in order to perform various tasks.

top of page

Dan Saffer

Pile Cabinets is a poster on my master's thesis project, examining a new way of managing digital documents: through a metaphor of piling. Rather than file things in folders or have random files on the desktop, users instead put documents into piles on the desktop. This better emulates how people work with physical desktops.

The poster will show research methodology, research models, personas, and the status of the project as of December 2004, which will likely be tested paper prototypes.

top of page

Laurie Gray

This presentation illustrates a method of information-gathering employed in December 2004 to gather initial design feedback for a website redesign. This approach was utilized to gather high level input from 10 users of a commercial website. The content for this approach comes from a large-scale web survey completed in November of 2004. Results from this work are to be turned into tangible designs that will be formally usability-tested in January 2005.

Specifically, this approach combines the interactivity of a paper prototyping session with the ease of administration and data collection of a web-based survey. In this approach, a series of images representing potential screen designs and content were placed into a commercially-available web survey site. Questions scattered through the survey gather immediate user feedback. Appointments are set with the participants, and during these appointments, a researcher calls them and, using a predefined script, walks with them through the web survey. Using this hybrid approach, the researcher is able to gather immediate thoughts and reactions to the information viewed in the survey, answer questions, and ensure that the participant understands the task.

One unique approach attempted in this task is the assignment of content to screen locations as outlined by a grid superimposed on a screenshot of the site. Use of this method is anticipated to provide a feeling for the relative importance and location of content in the final design. At this writing, the testing is not complete and so the accuracy of this hypothesis is not confirmed, but it will be by the time of presentation.

Handouts of the script will be provided for those visiting the presentation.

top of page

Dan Brown

The purpose of this poster is to catalog different ways of representing sample data in wireframes. By exploring these different approaches, the poster will help IAs identify their own strategies for developing their deliverables. The poster will describe five different ways of representing data in wireframes, comparing the pros and cons of each approach. For each approach, the poster will identify the implications for different audiences, like visual designers, business owners, and technologists, and questions each might ask.

top of page

Keith Instone, Dana Chang, Iromie Weeramantry

November 2004 was the first major redesign of ibm.com in 4 years. It included 200 properties, 1.6 million pages, 9 countries and 6 languages. The redesign included a new visual design system and information archtecture. This poster focuses on one part of the new IA: task-based navigation.

Task-based navigation is a new navigation system for country home pages and top level pages. Based on the customer sales cycle (learn, shop, use, manage), user tasks are supported with a variety of technologies, including faceted browsing and targeted content.

The poster will cover:

• Aspects of the previous design that were addressed with task-based navigation
• Determining which tasks to support and how to support them
• Other ideas we experimented with for task-based navigation
• How task-based navigation was applied to 9 different countries based on culture, business model and other factors
• How existing pages and existing applications were folded into the task-based navigation system
• How new pages were developed to support certain tasks
• How applications like faceted browsing were used to address a wide variety of tasks
• How the search engine was used to better target specific results for some tasks
• The results so far and the plans for increased task-based navigation on ibm.com

back to program

top of page

Dan Brown

Using the iPod as a benchmark for findability issues, this poster will walk readers through the key elements of music navigation systems of three other devices. For each system, the poster will describe the navigation system: the structure of the menus, and the controls used to move through and select menu options. The poster will also describe the display: the way the unit shows the current state, and the kinds of information revealed to the user in different modes. By dissecting the different approaches to displaying information about music libraries, the poster will ultimately shed light on the kinds of IA problems facing our industry in the future.

top of page

Dan Saffer

The Sounds of the Amazon is a fun mapping of data from Amazon.com, linking the top 20 albums from April 2, 2004, in a complex web of connections.

top of page

Marcy Jacobs

This case study is one in a set of three studies to analyze how the lack of an overarching profit motive skews performance measures used in government projects, and how this bias makes the job of user-centered design teams more difficult. A web-based geospatial tool is described in this case as an example of this problem.

The concept definition phase of a project is crucial in setting the strategic direction and outcomes (metrics). In government contracting, the concept definition is often substantially completed before a contract is awarded. This means concept definition is often written by government staff who are not experts in the task being performed, or by contractors who have not been allowed to thoroughly research the task being performed. In both cases, the overarching goals of any government project must be to avoid waste and corruption and to minimize risk. The result is that projects tend to be measured by cost, time, and scope rather than performance metrics that evaluate the effectiveness of the overall purchase in accomplishing business goals.

This case study highlights the challenges associated with moving a stove piped, departmental web site to a topically structured enterprise information architecture. Factors including funding, organizational structure and political motivations perpetuate this challenge.

In contrast, in the private industry, an overarching profit motive drives web sites to be managed through a centralized budget, communications and approval process. Although offices may have their own presence, there is more control and regulation toward a unified communication goal. To some degree, this difference is fueled by the motivators of success for government and commercial sites. Where a commercial site must recognize user audience and attend closely to the audience's needs for financial gains, a government site defines success by providing accessible information. In this project, several customers at the department level approached us to assist in an enterprise-wide mandate to topically reorganize sites towards the goal of reducing redundancy across the enterprise and unifying the voice of the organization.

top of page

Edward Bilodeau, Kimiz Dalkir

The objective of this research project is to create a virtual environment to host a community of practice (CoP) through which students and practitioners in the field of knowledge management (KM) can interact and share knowledge for continuous learning. Virtual networks supported by information and communication technology help connect isolated professionals and local pockets of expertise. They facilitate the diagnosing and addressing of recurrent business problems whose root causes tend to cross team boundaries. This process is facilitated when members link and coordinate unconnected activities and initiatives by addressing similar knowledge domains. As members create, collect, package and make available their collective know-how in the form of best practices and lessons learned, they effectively bring the overall performance of their community up to its highest potential. Bringing together students and practitioners in communities of practice appears to enhance student learning.

The virtual community is much like a form of a virtual lab that complements the classroom lectures, discussions and exercises. At present, there are few tools to enable students in the KM course to practice what they learn in class. WebCT has been used to a limited extent for some asynchronous knowledge transfer and sharing between students but this project makes use of WebCT in a novel way: as the host site for a virtual community of practice focusing on the theme of knowledge management. Along with specialists, academics, alumni, and other KM practitioners, students registered in all four courses can participate in the online community as part of their learning environment over the course of multiple semesters. Although there is an open component, in that participants will be free to participate as much as they wish, there is also a concurrent structured program of learning that makes use of the virtual community as its infrastructure. Students can undertake assigned roles w ithin the community on a rotational basis in completing a number of short assignments. These roles typically would include: librarian, journalist, mediator, apprentice, taxonomist, archivist, competitive intelligence and mentor. Results of formative evaluations from the preceding sessions are used to revise the roles, activities and general design of the infrastructure supporting the community of practice. A summative evaluation will take place at the end of the winter 2005 class session in order to assess how the online community contributed to learning success.

top of page

Ashley Cook

This case study is one in a set of three studies to analyze how the lack of an overarching profit motive skews performance measures used in government projects, and how this bias makes the job of user-centered design teams more difficult. A web-based geospatial tool is described in this case as an example of this problem.

The concept definition phase of a project is crucial in setting the strategic direction and outcomes (metrics). In government contracting, the concept definition is often substantially completed before a contract is awarded. This means concept definition is often written by government staff who are not experts in the task being performed, or by contractors who have not been allowed to thoroughly research the task being performed. In both cases, the overarching goals of any government project must be to avoid waste and corruption and to minimize risk. The result is that projects tend to be measured by cost, time, and scope rather than performance metrics that evaluate the effectiveness of the overall purchase in accomplishing business goals.

This example demonstrates how a public sector project approach centered on reuse of technology can result in a project that successfully meets the reuse goal in the interest of waste prevention, yet does not best meet basic organizational goals.

Focusing solely on the delivery of a tool reusing existing functionality is fundamentally incompatible with user centered design (UCD) principles. UCD practices are centered on the successful and efficient completion of business goals and the satisfaction of users. Even though pilot and usability testing are performed on a tool, so long as strategic UCD analysis are not conducted prior to design and implementation, this strategy can result in features that are usable but not useful.

top of page

Maria Lee, Thom Haller

www.plainlanguage.gov serves as a resource to promote the public's right to clear information. Created and maintained by Federal volunteers, the site provides content on what it means to write clearly it provides tools for Government writers and managers (and other audiences).

The original site was compiled by a general adding of links. Content was seldom linked, and the site did not support the purpose: that information can be clear and understandable. Two information architecture classes (one entitled information engineering) tackled the problem of reshaping the site so it better met audience needs.

Following a phased structure we referred to as GECKO (for Gathering, Evaluating, Chunking, Knowing (Testing) and Optimizing, students gathered information on different audience groups, developed comprehensive task lists of what people wanted to accomplish on the site, organized content into groupings based on user needs, tested structure with different audiences (with two test sequences with paper prototypes, testing of graphic mockups, and testing of HTML buildout). The site content was optimized to follow the rules of Plain Language and good web content.

The project included work of 20 class members and approximately 20 Government volunteers. The poster shows the progress through the site development. It includes visuals of the prototypes and volunteer project manager Maria Lee will be available to describe the site building process, challenges, and opportunities.

The poster supports anyone interested in site structure and the necessity to craft usable content.

top of page