3D Digital Dossiers -- a new way of presenting cultural heritage on the Web


Anton Eliëns, Yiwen Wang, Chris van Riel, Tatja Scholte

abstract

In this paper we give a comprehensive overview of our work on digital dossiers for the presentation of cultural heritage, in particular contemporary art, on the web using standard 3D technology. Digital dossiers allow for navigation using concept-graphs, and use 3D in an essential manner to present artwork installations, as 3D models, as well as all the relevant information needed for understanding the artwork, and, for curators, for the preservation and possible re-installation of the artwork(s). Our discussion encompasses requirements, implementation issues, and the realization of guided tours in digital dossiers, that provide a narrative facility as well as tools to experiment with exhibition parameters in virtual space.

Introduction

Digital archives for contemporary art that mimic a real-life museum in 3D space lack the ability for effective information retrieval when dealing with a large amount of highly inter-related information. From the perspective of entertainment these virtual museums can certainly amuse but the concept of copying a real life museum in a vir-tual environment is not very original anymore.

To cope with highly inter-related information structures, in such a way that is supports effective information retrieval and relationship detection in 3D space, we created a special kind of digital archive, further called digital dossier, for which we developed a new interface paradigm. XXXX Abramovic dossier.

The case study is a project of a multimedia course at the Vrije Universiteit Amsterdam. The project's aim was the creation of a digital dossier for the Netherlands Institute for Cultural Heritage in the domain of contemporary art. The digital dossier was realized by a team of nine students with different educational backgrounds and skills. The digital dossier presents itself as a digital archive in 3D space, containing information about the artworks of the performance artist Marina Abramovic by present-ing media content and relational structures. The idea of the digital dossier originates from the traditional medical dossier, which contains patient's information of relevance for diagnosis and treatment. In this particular case, the digital dossier presents the artist Marina Abramovic s artworks, serving as an information source for the museum's curators to conserve and install the artworks. Our digital dossier introduces some innovative features with respect to navigation and presentation in 3D environments. For navigation, we designed a concept graph that links multimedia elements in a structured hierarchy. The hierarchal structure is dynamic i.e. the selected information determines the presented hierarchy and visualizes parent-child relationships between information nodes. For the presentation of media content, we designed a content gadget consisting of three windows positioned in a way that allows to deal with multiple media simultane-ously.

The structure of this paper is as follows: In section 3, we give an example of using the digital dossier and in section 4 we explain our approach and implementation. In section 5, we discuss extensions to the digital dossier with respect to data representa-tion and content management and in section 6 we discuss initial user experiences. Fi-nally, in section 7, we draw our conclusions.

Background and requirements

The digital dossier described in this case study was developed in close collaborations with the Netherlands Institute for Cultural Heritage (ICN). ICN is a leading, inde-pendent knowledge institute for the preservation and management of moveable cul-tural heritage. ICN is coordinator of International Network for the Conservation of Contemporary Art (INCCA).
XXXXXX After a first round of the multimedia casus, in which the students produced an application giving an overview of the INCCA information archive, the participants, but only incidental information about the artists and their artworks, we decided to focus on case studies of individual artists, and we introduced the notion of digital dossier:

digital dossier

Create a VR that realizes a digital dossier for a work of a particular artist. A digital dossier represents the information that is available for a particular work of art, or a collection of works, of a particular artist. The digital dossier should be multimedia-enhanced, that is include photographs, audio and other multimedia material in a compelling manner.

Like a medical dossier, the digital dossier was meant to give the information about the artist and the works of art readily at hand, so that it could effectively be used for the task of conservation and the re-installation of the artworks.

Since we were in doubt whether the phrase dossier actually existed in the English language, we looked it up in a dictionary:

Webster New World Dictionary


  • dossier (dos-si-er) [ Fr < dos (back); so named because labeled on the back ] a collection of documents concerning a particular person or matter
  • archive -- 1) a place where public records are kept ... 2) the records, material itself ...
We chose for the phrase digital dossier, and not for archive or library, to stress that our focus lies on presentational aspects. Although issues of data representation and content management are clearly important, our primary interest was with issues of presentation and navigation.

...


main nodeinterviews

1

the abramovic dossier

For the 2004 autumn group, we decided to take the work of Marina Abramovic, a serbian-dutch artist who became wellknown in the seventies with performances with her partner Ulay, and has since then produced numerous installations, videos and performances with what I would like to call 'high existential impact'. The directive with which the students where set to work was, quoting Ted Nelson:

everything must be highly intertwinkled

Since virtual museums are by now a common phenomenon, and the virtual atelier for Marinus Boezem may be considered to be just a variant of this, the 2004 autumn group decided to explore alternative ways of presentation and navigation.

As material for the abramovic dossier there was an interview with Marina Abramovic from ICN, made in cooperation with the Dutch Foundation for the Visual Arts, and a great collection of videos from Montevideo. In addition, a transcription of the contents of the interview made by Michela Negrini, a student of media art at the University of Amsterdam, who also provided an interpretation as well as a categorization of the works of art. Given the material and the categories along which this material was classified, the students decided to explore the use of concept graphs as an instrument for navigating the information space.

navigation -- concept graphs

The reader has already encountered concept graphs in chapter 1, when the notions of multimedia, medium, television and communication were explained by indicating their relations to other concepts.

Concept-relation graphs are a familiar tool in linguistics and have also been used for a long time in Artificial Intelligence to describe the semantic relationships in complex domains. As a navigation instrument it is, to my knowledge only used in a kanji learning tool and the Visual Thesaurus. The Visual Thesaurus allows also for invoking Google image or document search from any of the elements of the concept graph.

...


presentation of video clips from Marina Abramovic

2

After the initial idea was there, one of the students of the group, Olaf van Zon, an AI student, managed to get a first version of a 3D concept graph working in VRML. This prototype implementation demonstrated the potential of the concept graph as a navigation instrument in the abramovic dossier.

presentation -- gadgets

The original idea of presenting information, that is the actual interview, the videos and images of the works of art, as well as the textual information, was to use rooms, where the information could be projected on the walls. The room metaphor, which obviously stems from the virtual museum approach, did however not seem appropriate since it conflicted with the concept graph used for navigation. After some discussion, information rooms were abandoned in favor of information gadgets, that could be expanded from and collapsed into the concept graph.

In the original abramovic dossier, the presentation gadget consists of three panes that can simultaneously show a video of the work, the interview, that is the fragment in which Abramovic speaks about that particular work, and the textual information related to the work and the interview. However, it appeared that in some cases there was not enough information, because the work was not spoken about in the interview, and in other cases there was too much information, for example multiple recordings or text documents. It was then decided to extend the presentation gadget with lists of alternative material that the user could select from and direct to one of the panes for further inspection.

To enable the user to focus on one of the panes, for example to get a better view of the video material a zoom in/out button was provided. All these enhancements, however, did complicate the interaction, as became clear when the abramovic dossier was presented at Montevideo.

In the course of the project, another interesting presentation feature was added, namely the reconstruction of one of the video installations in 3D, incidentally demonstrating the advantages of using 3D.

reconstruction -- recreating the installation

In discussing the abramovic dossier with Bart Rutten from Montevideo, who provided us with all the video material, another project was mentioned which was concerned with 3D-recordings/models of existing installations. Having full confidence in the technical capabilities of my students, I promised to show that such a reconstruction of an installation would naturally fit within our approach.

...



Reconstruction of Terra della Dea Madre in VRML.

3

The installation for which the reconstruction was made is Terra dea degli madre, and installation with two chairs and a television, which was exhibited in the Stedelijk Museum of Amsterdam, in 1986. As a starting point, we took a video produced at the time of the exhibition, which shows the installation in an exposition room in the Stedelijk Museum, and which contains, apart from comments from Abramovic, also the video shown on the televison in the installation.

At this point, we can only speculate how useful such a reconstruction can be as a tool for the conservator responsible for the re-installation, to play around with the presentation parameters, the positioning in space, the overall size, light and ambient effects.
XXXXXX The digital dossier for the artist Marina Abramovic had to satisfy the following requirements:

  • It must serve as an information source for conservators and curators of contemporary art,
  • It must present rich media recordings of all artworks, and,
  • In addition, it must provide background information for the general public (non-expert users).

A scenario of using the digital dossier

In this section, we give an example of using the digital dossier illustrating how to find information related to the artwork 'China Ring'. When starting the dossier, it loads the concept graph that is used to navigate through the available information (see figure 2). In the center of the concept graph, a shining star is shown to illustrate the root of the information hierarchy, and is used as the start object. When clicked, a star structure spreads and child objects appear sur-rounding the center star object (see figure 3). Clicking on the information node Artworks and then on China Ring will bring the node for China Ring into focus. Alternatively, using the keyword search function, the artwork China Ring can be easily found without concept graph navigation. The tree depicted in figure 4, step 4, shows China Ring' as the center node surrounded by all its children nodes which present information in text, picture and video format. When clicking on the current center node China Ring , the content presentation environment appears. It functions as a content gadget with three windows to present different types of information which are grouped into categories (text, pictures and video) listed below the windows. The information items can be dragged-and-dropped to any window for display. As we can see from figure 5, step 5, the picture of China Ring is presented in the left window, video in the middle window and text description in the right window. If desired, the user can focus on any window by using a zoom function (see figure 5, step 6). When the presentation of media content is finished, clicking on the close button will result in going back to the concept graph. Alternatively, the home function of the tool bar may be used to return directly to where we started: the original shining star.

The implementation

In this section we will indicate what technology we used, and we will describe the re-alization of the concept graph in somewhat more detail. We will then discuss the way media content is presented and how 3D models of artworks may be incorporated.

The digital dossier was created with Virtual Reality Modeling Language (VRML). It allows creating virtual worlds connected via the Internet and hyperlinked with the World Wide Web to expose the digital dossier to a broader audience.

As a user interface for navigating the digital dossier, we created a concept graph that represents hierarchical information structures. The concept graph allows the user to detect relations and search for information. Unlike the 3D cone tree, where the com-plete hierarchical structure is presented, only a subset of the hierarchy is shown - three levels deep. The concept graph is implemented as a starstructured-hierarchy diagram represent-ing related information objects. By star-structured we mean that relations between in-formation objects are visualized by lines, getting a parent-child relationships structure by showing a centered information object surrounded by related information objects. This structure, originated by the parentchild relationships, is dynamically generated when selecting an information object. The selected object will be translated to the center of the screen, involving movement in the X and Y direction. It then becomes a parent node showing its children around it. So, the presentation is dynamic and actu-ally determined by the user's choice. To compensate for the lack of an instant over-view, where all information is shown at once, the user can, as already indicated in the previous section, also use keyword search instead of navigation.

In summary, there are two ways of finding information in the digital dossier:

  • concept navigation through information object selection, center nodes and their relation to other nodes are visualized.
  • keyword search the search fucntionality enables the user to located information based on textual input.
Information objects shown in the concept graph are represented by 3D icons. These 3D icons isualize a certain type of information. The icons tell he user what informa-tion s/he can expect when licking on it. We distinguish between two nformation types:
  • Conceptual information types 3d models that represent categories
  • Content information types 3d modles that represent types of content
The content information type itself consists of different media types. These are:
  • Text content type 3d models that represent textual information
  • Picture content type 3d models that represent static visual information
  • Video content type 3d models that represent motion based information

Presentation of media content

Presentation of media content is supported by different visualization facilities. Presen-tation is an essential part of the digital dossier but is separated from navigation. The presentation facilities are deployed when media content is selected for view. The digi-tal dossier contains different presentation facilities for 2D and 3D content. For 2D media content a visualization facility is needed that is able to present video, images or textual information. This facility is implemented as a content gadget with three windows. In each of the three windows the user can view 2D media content. These windows are positioned in such a way that the user can inspect the information simultaneously (see figure 6). In our experience, three images can be presented at the same time without much visual distortion.

fig 6 XXX

Below the three windows a list of all content related to the selected information object is displayed. The content is categorized lists for each content type. The user can control on which of the three windows content is displayed. By using drag-anddrop the user can view content on a window of choice. This functionality gives the user some freedom for customization instead of being bounded to a fixed display. If necessary, the user can focus on a particular window with a zoom option, to avoid distraction from the other windows.

Incorporating 3D models of artwork installations

Since we adopted 3D technology, we could easily accommodate a 3D model for one of the installation art works by Marina Abramovic. We implemented a plain exhibition room, providing a 3D perspective of the instal-lation Terra Degli Dea Madre (see figure 7) that allows the user to manipulate the po-sition of the objects by a click-and-drag function. The 3D environment demonstrates the interactive exploration of the installation of an artwork. By manipulating position and/or angle of objects, museum curators can get in-sight into how the artwork could be exhibited.
XXXXX Drawing the graph This section describes how the concept graph is drawn. When a new node has been chosen to become the center of the concept graph, we could say that the state of the graph changes. This is where the computation of the new positions of nodes in the concept graph starts. The positions of the nodes are computed as positions on a 2D plane. The process can be described in 5 steps:
  • Step 1: First the node that is selected is placed in the center of the space. We call this node the center node. The node represents the information the user is focusing on.
  • Step 2: Next, the available radius for the center node is set. In our application this value was kept constant for each state. The radius available must be larger than the space taken in by the center itself otherwise there is no room for drawing other nodes.
  • Step 3: In this step the radius that is available for each surrounding node of the center node is computed. The surrounding nodes of the center node will be called the child nodes. The children of the center node are the nodes with information related to the information of the center node. The circle around the center node is divided into sectors. The number of sectors is equal to the amount of children of the center node. The space available for a child is the circle that fits inside a sector. However, the circle around the child may not intersect with the center node, this puts a minimum on the distance from the center to the child. If this is the case, the circle available for the child node is decreased to not intersect with the center node (i.e. the radius is derived from the minimum distance and the available space/radius of the center node). Figure B shows how the radius of the area around the child can be derived. Figure C shows the minimum distance from the child to the center node.
  • Step 4: Now the actual radius that will be used by a child node can be computed as follows: First for each node surrounding the child node (which can be regarded as grandchild nodes relative to the center node) we compute the distance it will be placed from the child node. We must make sure that the nodes do not intersect each other, so we arrange them around the child node, as depicted in figure D. The distance to the child node is equal for each node, where each surrounding node has the same sector size to be drawn in. The distance between a surrounding node and the child node is taken as small as possible such that the surrounding node just hits the borders of the sector. But in the case the center node and surrounding node would intersect, the center node and surrounding node are put next to each other, see figure D. The required radius of a child node is the radius of the circle with the same center as the child node in which all its surrounding nodes fit. However in case this radius exceeds the available space of the child node, the available radius is taken as the required radius. Also the child node and its surrounding nodes will be scaled down to fit inside its available radius.
  • Step 5: In this last step the position of all the nodes in the concept graph are computed. The child nodes are arranged around the center node like in step 4: The distance between the child node and the center node is minimized such that the their required space of the child nodes fit in the sectors without hitting the center node. The positions of the child nodes can now easily be derived from the distance and the angle each sector is taking.

    In case the available radius is smaller than the required radius, the child node will be drawn smaller and the distance of the surrounding nodes to the child node are scaled down to fit the available radius. If the available radius for a child node is smaller than a predefined minimum, the children will not be drawn. The positions of the surrounding nodes relative to their child node are computed in the same way as the child nodes. In our application the concept graph could fill a rectangle space. The locations of the nodes are scaled/modified to make optimal use of the entire available screen space.

Resolving conflicts

Because of the (inter) relations of the data, sometimes multiple locations are possible for data items to be placed. (for example both as center node and as surrounding node of a child node). We have chosen to draw only one node for a data item: If a node is already present for an information item, the other possible locations will be left open. Since the center node will be drawn first, and subsequently the children nodes and its surrounding nodes, as a result, most open locations are around the child nodes. The child node is always related to the center node because all relations are symmetric. Therefore there is always one gap in the surrounding nodes of a child node. The locations of the surrounding nodes are rotated such that this gap is in the same direction as the center node from the child node. To indicate relationships, lines are drawn between related nodes. If a node has a relation to a node that is among its children, a colored line is drawn. If the node is not among its children, because the node with the related information was already elsewhere, a visually less apparent line is drawn. The surrounding nodes of the child nodes are drawn slightly transparent to make the center node and the child nodes visually more apparent

Animations

When the concept goes to a different state (because the focus has changed) the concept graph is animated to show the transition of the old state to the new. Nodes that represent information that is also needed in the new state, move to their new locations. Nodes that represent information that is no longer needed disappear from the screen with a fade out effect and new nodes that are needed appear with a fade in effect.

Extensions

The algorithm described here draws nodes that are 2 relations away from the center node. Although it is possible to adapt the algorithm described here to support an arbitrary distance from the center node, for the digital dossier it was not necessary to draw more than 3 levels: it would have resulted in cluttering the screen with more items. Other future work on the concept graph could be to use weighted connections. This can be useful for limiting the amount of nodes on the screen by showing only items that are the most related to the information that is focused upon.

Data representation issues

To preserve the information stored in the digital dossier for future use, we must con-sider the data representation i.e. how the information is stored and structured. To use the information presented by the digital dossier, taking into account future develop-ments in 3D technology or other application contexts, it has to be independent of for-matting information. This means that the same information instance can be used for other presentations in a relatively easy way. In case of the digital dossier the presented information has to be VRML independent.

As an extension to the digital dossier we created a web-based content-management tool that generates XML (eXtendible Markup Language) structured data output from textual information input. XML is independent of formatting information and therefore suited for multiple presentation forms. The tool has initially been created for non-expert VRML users that want to create a 3D digital dossier in a relatively quick and easy way, without programming or adjusting exist-ing code. By using style sheets, the generated XML output can be presented in various ways by giving it formatting information. In case of the 3D digital dossier a style sheet conversion is needed from XML to VRML format.

Fig 8

In re-thinking the abramovic dossier, we first needed to re-establish what are our goals in developing this application and what are our primary data sources. The goal, first of all, is to support conservators in their task of preserving contemporary art, and to assist them with the re-installation of such artworks.

...



Our primary data sources are meta-information, coming from the INCCA initiative, and video-recorded artist interviews, which were initiated by ICN as a means to record information about contemporary art that would otherwise be lost. In addition we media-material, including images and video, that may be regarded as recordings of the works of art, as well as the textual interpretations and classifications that exist, or may be constructed from this material.

At this point, I may remark that one of the pitfalls in creating a dossier is to get trapped in the visually salient features of the dossier, the presentation of the artworks themselves, and forget about the primary focus of the dossier, to make all information accessible in an adequate manner. For many cultural heritage applications, which aim to present art to the layman, presenting the artwork is the primary focus, and giving access to the information context generally comes second.

For our next generation of digital dossiers, we decided to take the following steps:

next generation dossier(s)


  1. adaptation of representation to Dublin Core (+ annotation needed for presentation)
  2. XML-based content management, with php forms (extending what we have now)
  3. there should also be a possibility to present the information and material in a 'plain' web format
  4. as well as in (a new version of) 3D dossiers
  5. we should think about the proper presentation parameters.
Dublin Core is the standard used in the INCCA initiative, to record meta-information about existing information sources. See section 3.3 for a description of the Dublin Core element set and the Resource Description Framework (RDF) on which it is based.

For the abramovic dossier, a collection of record-like structures was developed, together with a simple content-management tool, written in PHP. This content-management system must be adapted to be compatible with the Dublin Core-based resource descriptions.

Further, we decided that, along with the 3D presentation of the dossier, it would be worthwhile to develop a conversion tool that produces standard web-technology based presentations as well. This approach allows us to assess the usability merits of the 3D dossiers in a comparative way.

Finally, as I indicated before, an important issue that must be resolved concerns the proper presentation parameters. What do we present to the user? And how do we allow the user to interact with the material presented?

content management and data representation

For developing the abramovic dossier, we have a fixed number of record-like structures:

structures


  • Video -- to display video fragment, including interviews
  • Picture -- to present pictures of the artwork
  • Artwork -- contains all information connected to a work of art
  • TextItem -- to present text, from the interview or any other source
  • MaterialItem -- to present information about material(s) used
  • GroupNode -- to combine nodes in the concpet graph
  • Information -- acts as the outer container for all nodes
All these structures support a set of common attributes, including shortName, longName, ID, connectedNodesIDs, and description. In addition the Video, Picture and Picture have fields allowing to show a preview image. And the Video, Picture and TextItem, also have a url field giving access to the actual information item.

The Information and GroupNode structures are used for creating the top-levels of the concept graph, whereas the other structures, such as the Video and TextItem give access to for example a fragment of an interview and its transcription.

Below an example is given of the data underlying the concept graph of the abramovic dossier:

concept graph



  Information {
   informationNodes [
    GroupNode {
  	ID "MAIN"
  	shortName "Main"
  	longName "Main"
  	urlModel "models/conceptGraph/main/modelMain.wrl"
  	description [ "Central information node" ]
  	connectedNodesIDs [ "ARTWORKS", "KEYWORDS", 
  	                         "INTERVIEWS", "REPORT" ]
  	}
    GroupNode {
  	shortName "Artworks"
  	longName "Artworks"
  	description [ "Node that connects to all the artworks" ]
  	ID "ARTWORKS"
  	connectedNodesIDs [ "MAIN", "TRANSITORY", 
  	              "ULAY", "VIDEOINSTALLATION", "ABRAMOVIC" ]
  	urlModel "models/conceptGraph/artworks/artworksGroup.wrl"
  	}
  	## ...

   ]
  }
  
The Information node collects all available nodes, and takes care of connecting the individual nodes, based on the information specified for each node.

As an example of an Artwork node, that is an element of the list of nodes in the Information node, look at:


  Artwork {
    shortName "Terra degla Dea Madre"
    longName "Terra degla Dea Madre"
    description ["15:40 min, colour, sound."]
    ID "AV24"
    connectedNodesIDs ["VIDEOINSTALLATION", "DTV24", 
               "TTV24", "PV24", "CV24", "VV24", "G0"]
    urlPreviewImage "images/previewImages/AV24.jpg"
    widthPreviewImage 479
    heightPreviewImage 349
  }
  
This node is connected to many other nodes, giving access to the information items that belong to it, such as the video clips of the interview, shown below.

  Video {
    ID "CV24"
    shortName "Interview clip Terra degla Dea Madre"
    longName "Interview clip showing Terra degla Dea Madre"
    url "interviewclips/interview_terra_degla.avi"
    width 320
    height 360
    urlPreviewImage "images/previewImages/interview_terra_degla.jpg"
    widthPreviewImage 320
    heightPreviewImage 240
    description [""]
    connectedNodesIDs ["CLIP", "AV24"]
  }
  
In the url field of this declaration, the actual video file is indicated, which should be displayed at a resolution of 320x360, as specified in the width and height fields.

And finally, as an example of a TextItem, consider:


  TextItem {
    shortName "Instruction"
    longName "Green Dragon Lying instructions for the public."
    description ["Text explaining the way the public has to interact with the artwork."]
    ID "ITO05"
    connectedNodesIDs ["AO05", "INTERACTION"]
    url "text/AO05_instruction.txt"
  }
  
For constructing the abramovic dossier, Tim Verweij developed the content management tool, that allows the user to browse and edit existing nodes, and to insert new nodes into the graph.

integration with the Dublin Core

The Dublin Core is a general resource description formalism, that allows for specifying resources in a variety of domains. See section 3.3. For INCCA the Dublin Core was chosen, not because it is the most suitable formalism, but because it may serve as the least common denominator, and agreement on anything else simply seemed to be impossible. As a reminder, the Dublin Core provides the following elements:

Dublin Core


  • title -- name given to the resource
  • creator -- entity primarily responsible for making the content of the resource
  • subject -- topic of the content of the resource
  • description -- an account of the content of the resource
  • publisher -- entity responsible for making the resource available
  • contributor -- entity responsible for making contributions to the content of the resource
  • date -- date of an event in the lifecycle of the resource
  • type -- nature or genre of the content of the resource
  • format -- physical or digital manifestation of the resource
  • identifier -- unambiguous reference to the resource within a given context
  • source -- reference to a resource from which the present resource is derived
  • language -- language of the intellectual content of the resource
  • relation -- reference to a related resource
  • coverage -- extent or scope of the content of the resource
  • rights -- information about rights held in and over the resource
Descriptions of items in the digital dossier should incorporate these elements, together with the attributes needed for the insertion of items in the concept graph and the presentation parameters, that are necessary for displaying the (media) material. Technically, the namespaces supported by RDF does allow for merging these different types of annotations. However, the challenge here is to derive the presentation attributes automatically, and to come up with a reasonable default for inserting these items in the concept graph.

Usability evaluation

As a first review of the digital dossier we conducted a cooperative evaluation [1] with potential end-users. The cooperative evaluation is a variation of a think-aloud evalua-tion and has the following advantages:
  • It's easy to conduct, and involves no extra costs
  • It delivers test results in a relatively short time
  • It encourages the user to criticize the application and the style of interaction
The actual evaluation sessions of the digital dossier were recorded on video.

The evaluation was designed to assess domain related tasks where the user uses all functionality related available for navigation and presentation. It focused on the fol-lowing aspects of the digital dossier:

  • Navigational actions
  • Interpretation of navigational results
  • Interpretation and detection of relations between information
  • Presentation of content
  • Functionality of presentation gadget
In general, we were interested in explorative tasks, where the evaluator has a passive role.

The test results give a first indication of the usability of the digital dossier.
Positive results:

  • The concept graph makes it easy to detect relations between information
  • Using a concept graph for navigation appeared to be intuitive for all users
  • The close relation between the concept graph and presented media reduces dis-orientation

Negative results:
  • The meaning of 3D icons was not well understood
  • The users expressed the wish to customize the visual appearance of the concept graph and the icons used.
In general, we conclude that the concept graph supports both intuitive navigation and relationship detection. However improvement of the visual appearance of the digital dossier is definitely possible and desired.
XXXXXX

style issues -- how to improve the dossier

The abramovic dossier does also provide a facility for search, as well as online help. However, as already mentioned, when demonstrating the application to the interested parties, that is ICN and Montevideo, a number of issues came along, that I will here summarize as a list of questions:

style issues


  • what icons should be used to identify the elements of the concept graph?
  • what categories and relationships are most appropriate?
  • how should the information be displayed, simultaneously or more focussed?
  • how do we allow the user to choose between multiple information items?
  • how do we avoid visually disturbing elements?
Obviously, although the abramovic dossier was very positively received, these issues must be dealt with to make it a success. Having a first prototype, we need to rethink our application, not only with regard to its style of presentation, but as we will discuss in section 10.3, also in terms of its underlying data representation.

Related work

Concept graphs or concept maps ... see concept mapping portal A common environment used to present art related information in 3D space is the virtual museum. A virtual museum is a navigation orientated environment where navigation and presentation of content are overlapping [2]. While navigating for con-tent, representation of the artworks are constantly visible. For quick navigation or searching in large information collections, this could lead to confusion and disorientation.

Many different visualizations have already been proposed to navigate hierarchical information structures [4]. These visualizations can be in 2D or 3D and are needed to interpret the intended hierarchy of a particular informa-tion collection. Such diagrams are aimed at reducing the cognitive load for the user.

A well-known example of a 3D information visualization is the 3D cone tree [3] (see figure 1). The 3D cone tree visualizes hierarchical structures and consists of cone objects. The motivation of sing 3D over 2D cone trees is that 3D visualizations make optimal use of screen space and provides the opportunity to visualize larger hierarchical structures. With respect to usability, we observe that the cone tree presents all information at once. In case of a large amount of highly inter-related information structures this could lead to an informationoverflow.

Conclusions

We have argued that a concept graph implemented as a star-structured diagram, where the presentation of the structure is dynamic, as used in the digital dossier, may pro-vide intuitive navigation when dealing with highly inter-related information structures in 3D space. Instead of presenting a complete view of the hierarchy, the concept graph shows only a subset of the information. Presentation of content is separated from navigation but the digital dossier indicates a strong relation between them.

So far, the results of the initial evaluation look very promising for using the concept graph as a navigation paradigm. Evaluation indicates that it is relatively easy to use and that it supports exploratory tasks rather well. As such, the digital dossier can be a solution for dealing with presenting highly interrelated information structures in 3D space. How-ever, to get a more accurate view of the usability of the digital dossier, evaluations that are geared to obtain quantitative results need to be conducted. Also deploying the digital dos-sier in more real world applications may reveal new issues for its further development.

Acknowledgements

We gratefully acknowledge the contribution of the following people to the develop-ment of the digital dossier:
  • Vrije Universiteit Amsterdam : Tim Verweij, Olav van Zon, Eric de Boer, Pascal Snijders Blok, Olaf van Zon
  • Universiteit van Amsterdam : Abeer Mahdi, Jina Menke, Michela Negrini
  • Technische Universiteit Delft: Jurgen Koster
  • The Netherlands Institute for Cultural Heritage : Tatja Scholte, Ijsbrand Hummelen
  • Montevideo : Bart Rutten, Gaby Wijers

References

[Remediation] {\rm\sc Bolter J.D and Grusin R. 2000},
Remediation -- Understanding New Media, The MIT Press
[Dix] {\rm\sc Dix A, Finaly J, Abowd G, Beale R, 1998},
Human Computer Interaction, Prentice Hall 2nd edn.
[Navigate] {\rm\sc Eliens A., van Riel C., Wang Y. 2006},
Navigating media-rich information spaces using concept graphs -- the abramovic dossier, In Proc. InSciT2006, 25-28 Oct. 2006, Merida, Spain
[VirtualArt] {\rm\sc Grau O. 2003},
Virtual Art -- From Illusion to Immersion, The MIT Press
[Graphs] {\rm\sc Ivan H., Melancon G. and Marshall M.S. 2000},
Graph visualization and navigation in information visualization: A survey, IEEE Transactions on Visualization and Computer Graphics, 6(1):24-43
[Agents] {\rm\sc Hoorn J., Eliens A., Huang Z., van Vugt H.C.,},
{\rm\sc Konijn E.A., Visser C.T. 2004,} Agents with character: Evaluation of empathic agents in digital dossiers, Emphatic Agents, AAMAS 2004, New York 19 July - 23 July, 2004
[ConceptMaps] {\rm\sc Riel C. van, Wang Y. and Eliens A. 2006a},
Concept map as visual interface in 3D Digital Dossiers: implementation and realization of the Music Dossier, In Proc. CMC2006, Costa Rica, Sept 5-8 2006
[Guide] {\rm\sc Riel C. van, Eliens A., Wang Y. 2006b},
Exploration and guidance in media-rich information spaces: the implementation and realization of guided tours in digital dossiers, In Proc. InSciT2006, 25-28 Oct. 2006, Merida, Spain
[Cones] {\rm\sc Robertson G.G. and MacKinlay J.D. 1991},
Cone trees: animated 3D visualizations of hierarchical information, In Proc. of the SIGCHI Conference on Human factors in computing systems: Reaching through technology, pp. 189-194, New Orleans, Louisiana, USA
[Gadgets] {\rm\sc Schonhage, B., van Ballegooij, A., Eliens, A. 2000},
3D gadgets for business process visualizations, In Proc. of the fifth symposium on Virtual reality modeling language (Web3D-VRML), pp. 131-138, Monterey, California, USA
[Present] {\rm\sc Wang Y., Eliens A., van Riel C. 2006},
Content-oriented presentation and personalized interface of cultural heritage in digital dossiers, In Proc. InSciT2006, 25-28 Oct. 2006, Merida, Spain