Satisfaction through Empathy and Orientation in 3D Worlds ALEJANDRO C. FRERY, Ph.D.,1 JUDITH KELNER, Ph.D.,1, JOÃO MOREIRA, Ph.D.,2 VERONICA 1 TEICHRIEB, M.Sc.,
ABSTRACT This paper describes a methodology for navigation and exploration assistance intended to enhance user satisfaction when exploring three-dimensional virtual environments. The complexity of such environments often make navigation and information retrieval difficult, making it necessary to add assistance components to the world in order to turn it easier to manipulate. This methodology uses three-dimensional “intelligent” avatars as interactive guides, along with information based navigation strategies. The intelligence of the avatars is represented through physical features, behaviors and knowledge about the user and the environment. These components establish the avatar’s architecture. Content personalization according to the user’s interest, navigation assistance according to the desired content, and avatar guides that make the virtual place more realistic and pleasant are proposed in order to involve users. A three-dimensional model of the Guggenheim Museum Bilbao, in Spain, is presented as a prototype for the validation of this methodology. In the museum the guide is represented as a fish that, according to the user’s preferences, assumes gender and age. The avatar swims through the museum, following navigation routes that lead through exhibitions previously chosen by the visitor. exploration possibilities. It is, however, important to ensure that the development INTRODUCTION of virtual environments is carefully done, This paper describes a methodology that in order to provide the user with as many intends to enhance user satisfaction when as possible friendly tools, permitting exploring three-dimensional (3D) virtual him/her to enjoy the advantages of using environments. This methodology was such advanced interfaces. desktop developed using virtual reality The proposed methodology considers (VR) technology, but can easily be three basic matters: extended to other platforms. Despite the fact that 3D paradigm is actually largely used for information representation, there are many aspects related to the intuitive representation of 3D information objects and mainly. One of the crucial issues is the pleasant and interactive navigation and exploration of these environments. VR interfaces are one of the most advanced types of interface available, offering a lot of interaction and
•
Information representation through the use of rich semantic 3D icons;
•
The use of “intelligent” avatars as interactive guides for navigation and information recovery assistance;
•
The use of spatial navigation strategies to make the exploration of the environment easy and pleasant.
1
Universidade Federal de Pernambuco, Centro de Informática, CP 7851, 50732-970, Recife, PE – Brazil.
2
Aero-Sensing Radarsysteme GmbH c/o DLR, Oberpfaffenhofen, 82234, Weßling, Munich – Germany.
1
aimed at facilitating the interpretation by the user, is achieved using cognitive categories represented by a room2. Furthermore, in order to decrease the site design complexity, a mapping of abstract data into a meaningful representation from which one can get instantaneous understanding was made.
A short description of each these topics will be made in the following sections. Finally, a case study will be presented.
METHODOLOGY The methodology proposes some guidelines to solve the problems of navigation, exploration and information recovery in three-dimensional virtual environments. Content personalization according to the user’s interest, navigation assistance according to the content that the user wants to explore through the construction of navigation routes, and avatar guides that make the virtual place more realistic and pleasant are proposed in order to make users feel more involved with the 3D world.
Navigation Navigating is one of the most common difficulties faced up by 3D environments users. This may be due to the fact that they do not have experience in using the navigation tools offered by plug-ins and browsers used for the visualization of the scenes. But, in most cases, the problem relies on the fact that three-dimensional environments become large and do not offer proper navigation assistance. The lost user is unassisted by the environment. This aspect should seriously be taken into account during the construction of 3D worlds.
Information Representation The 3D paradigm is useful mainly because it offers the possibility of representing information in a realistic way, while organizing content in a spatial manner. Virtual reality interfaces allow the use of multimedia components and 3D objects to enhance the intuition of information visualization, and also permit the user to explore the information in an interactive manner.
A great deal of research3,4 has been done proposing solutions for these navigation and disorientation problems. Most of such solutions are the adaptation of 3D models according to architectonic metaphors used in the real world5. It is important that the virtual worlds offer spatial navigation strategies, helping the user to explore and to accomplish efficiently his/her objectives, even if he/she does not have previous knowledge about how to move and locate in the environment.
In order to construct these information objects, rich semantic metaphors for the modeling process should be used. According to the type of information to be represented, geographical (the location of the atrium in the Guggenheim Museum, illustrated in FIG. 2), architectural and/or conceptual metaphors (see FIG. 7) may be used.
An advantage of 3D worlds is that they allow visualization from different viewpoints and, through the definition of relevant points of view the user may grab a general idea of the world as well as a detailed one, depending on his/her objectives. Some strategies6 for navigation suggest the use of routes for the visualization of the whole world.
A new methodology for mapping Web site structures into architectural metaphors1 is an example of the use of an architectural metaphor for the design of personal home pages. The main page was represented as a hall, while topics form an arbitrary number of rooms. This would make the visitor feel “at home” while navigating through the house. The metaphorical representation,
Navigation routes are a simple but efficient method for animation creation7. 2
as representatives for the user that visits the place. On the one hand for example, the three-dimensional representation of a museum contains guides that can help the visitor to visualize and appreciate pictures, sculptures etc., as well as offering information about the museum’s content. On the other hand, the visitor that enters the virtual museum could also have an avatar representing him/her in the world. Avatars, as representatives of the user in a virtual textual or 3D world have proven being an attractive tool to enhance user’s level of involvement with the environment, because they make his/her presence more realistic. This is mainly true in threedimensional worlds, since they can be rendered similarly to real characters.
They are composed by a collection of points that indicate by which positions an object will pass during the navigation process, establishing also the orientation in which this object will visualize the environment during the animation, through the specification of another set of coordinates. A route based navigation strategy proposed by our methodology offers a socalled “God’s Eye View”. This viewpoint gives a general overview of the virtual environment, being very efficient in maintaining the user oriented in the virtual space without overwhelming him/her with details. Another aspect related to the building of routes is that they have to show the environment from different perspectives. This implies that it is interesting that the user visualizes the environment from the perspective that he/she would do it in the real world, according to its height (tall, short) or age (child, adult). This strategy allows a better visualization of the environment and its content, and may bring greater empathy with the virtual world.
Another possibility is adding some levels of intelligence to an avatar, giving the avatar behavior and personality. Features such as the ability to acquire information about user’s preferences and interests, about other avatars that, for example live in the environment, or about the virtual environment itself allow dealing with aspects such as user satisfaction, dynamic adaptability of the 3D environment to the users needs. These features also tend to make the environment more pleasant and adequate, since the user will encounter characters that communicate with him/her in a similar fashion to what he/she encounters in the real world.
Routes that allow approximation to selected objects during the navigation, for more detailed visualization of specific parts of the content, such as to a picture in a museum room, should also be offered by the navigation support system. Readers interested in a survey of navigation issues in virtual reality environments are referred to8.
The object of this work is the use of an intermediate approach between avatars as mere graphical entities and intelligent avatars: avatars as interactive guides. This approach is illustrated in the section Case Study, where a family of fish is presented as the guides from the virtual Guggenheim Museum. Next section describes the architecture of the proposed avatars.
Avatars Avatars are an important mechanism when working with Web based environments, in both two and threedimensional worlds.
Avatar Architecture
An avatar, in the context of this paper, is a representation of a user in a synthetic world, becoming a character of the same.
Avatars need to have some functionality to interact with the user and with the virtual environment, in order to be interactive guides.
These avatars could be used as characters that inhabit the environment or 3
They need to be informed about what the user would like to do in the environment and, based on this information, they have to perform relevant activities. Having information about the environment, avatars should readily start the exploration of the world, knowing where to find each category of information needed by the user and knowing how to recover it.
Guggenheim Museum in Bilbao, Spain. This museum stands out by its architectural design, composed by a collection of interconnected curved blocks. Its creator, Frank O. Gehry9, found inspiration in the curve shapes of fish, snakes etc., and projected the museum so that it resembles organic shapes, both in format and appearance. The Guggenheim Museum Bilbao is one of the five contemporary art museums maintained by the Guggenheim 10 Foundation . Three floors, accessible through stairs and elevators with several places and passageways that intersect, compose the museum. FIG. 2 shows external and internal views of its structure.
FIG. 1 shows how virtual environment, user and avatar should interact. The virtual environment presents the user the profile interface. Once this textual interface is completed, the environment stores the information entered by the user. According to the user’s preferences, the environment will adapt its content, and the requested avatar will infer which tours it has to make with the user.
The User Profile Interface One of the most important modules of the application is the user profile module. Its objective is to collect information about the users preferences, in order to generate profiles that allow the creation of adaptive functionalities in the environment. These functionalities increase users satisfaction during the exploration of the threedimensional environment, since they have personalized content, navigation routes that take them to the places where this content is exhibited and interactive guides that will go along with them during the visit.
After this adaptation process in which avatars acquire knowledge, they become intelligent (from the user’s point of view), being able to help the user to navigate and to explore the environment. According to the content to be visualized, avatars look for additional information that could be made available if needed. An example of this is the background preparation of additional content in the form of navigation strategies or as extra (unrequested though possibly desired) content. The physical features of the avatar are also adapted to user desires (profile characteristics), for example to be assisted by an elder or young guide or a male/female one. Clearly, the behavior features pursued by the avatars enhance its realism.
Through the interface the user chooses the kind of information to be visualized and a guide to orient and accompany him/her during the visit. If the user is not currently looking for something specific, he/she can choose to visit the whole museum. With this option selected, all expositions available in the virtual museum will be presented, but it is also possible for the user to stop the visit at any time, if the content does not appear interesting or for some other reason.
CASE STUDY: THE VIRTUAL GUGGENHEIM MUSEUM BILBAO This case study aims at validating the representation of information metaphors, navigation strategies and the use of interactive guides. It is related to a very large and complex museum: the
Some users may not want to be guided during the visit, so it is also possible to select an unguided tour and start the visit alone. 4
The use of 3D icons to present this semantic content has proven to be very efficient, since the user interprets the underlying information easily. The use of icons based on real world metaphors, presented in1, is an efficient tool for information recovery. This occurs mainly because the user can easily recognize the information associated to the icon.
FIG. 3 shows the user profile interface filled in with information from a user who wants to visit all expositions presented by the virtual Guggenheim museum, as well as to be guided by the father-fish avatar. The user has also to inform his/her name and e-mail address. This information will be used to recover the user’s profile next time he/she revisits the virtual museum. It is not, then, necessary to complete the user profile interface over and over.
One important aspect of the application is the need of furnishing the museum with pictures, sculptures etc., as it is the case in the real world. So, beyond the external structure, fictitious exhibitions were built. They were arranged in rooms, and users can visit them and recover additional information as two-dimensional homepages. This additional material is presented using HTML11 and Macromedia Flash12, and can be accessed through a standard Web browser, by activating a mouse click based event embedded in the icon. One of the exhibitions was composed by several pieces exposed in the Virtual Museum of the Northeastern Brazilian Man13, a real anthropological museum in Recife, PE, Brazil.
Clicking the link on the bottom of the interface, the visit to the virtual Guggenheim Museum will be start, and the avatar will be positioned in the entry of the museum. The selected avatar will recover the information entered in the user profile interface before starting the tour, knowing exactly what the user wants to see. Only the expositions chosen by the user will be loaded, enhancing considerably the navigation performance inside the environment. The user profile interface is responsible for content, as well as navigation personalization. These features allow an inexperienced user to easily navigate through a virtual environment and save time.
Furthermore, an exhibition of all Guggenheim Museums (FIG. 6) around the world was built, where museums are shown as 3D maquettes icons (architectural metaphors).
The Museum The Guggenheim Museum was modeled according to its real counterpart, allowing virtual visitors to visualize it as if they where in the true museum.
Other types of icons shown in the environment represent information about other Guggenheim Museums. These icons, built as words, represent their semantic content in an easy understandable way for the user. One of these icons, representing news about the Museums, is shown in FIG. 7.
The blueprint of the three floors was used in order to make a precise modeling. FIG. 4 shows the God’s Eye View of the 3D model.
The use of 3D icons presents not only the advantage of being an intuitive way to represent information but also makes it possible to develop reusable content for use in a variety of applications.
Content Icons 3D virtual environments representing real or imaginary places should exhibit information in an organized and comprehensible way. Objects that will represent or contain this information ought to possess a rich semantic content.
Navigation in the Museum Since the virtual Guggenheim Museum 5
is a 3D representation of a real world place some restrictions are advisable in order to maintain the reliability of the model. For instance, it is not adequate to add extraneous signs or landmarks to help the user during navigation within the environment. But this museum, in particular, is a very large place and it is difficult to remain oriented during the navigation. It was therefore necessary to add some navigation strategies to facilitate user’s exploration.
Another navigation-related aspect considered here is that the routes followed by the avatars are generated according to the avatar’s age. This kind of navigation strategy enhances considerably the realism of the navigation, because it uses a real world metaphor. In the next section its implementation will be described.
Avatars as Interactive Guides Avatars representing fishes were embedded in the virtual environment to help the visitors during the exploration of the museum. Fishes were used because the architectonic proposal of the Guggenheim Museum was inspired by organic shapes that give the impression of mobility as fishes also do.
According to the navigation strategies described before, predefined navigation routes were developed, in order to promote tours through the museum. The viewpoints that compose each tour were defined through the analysis of the blueprints of the three floors of the museum. After the definition of the viewpoints, these were grouped in sequences, creating routes that take to every place in the virtual world. Once defined, these routes were tested in order to verify the best orientation for each viewpoint of the route, in order to enable the user to visualize the relevant content available at each position.
Four avatars were created, representing a “family of fishes”: the father, the mother, the son and the daughter (see FIG. 8). According to the architecture presented in section Avatars, they need to have physical features, specific behaviors and knowledge about the environment where they live and about the user, in order to give the illusion of being intelligent characters. Moreover, they need to have physical features and behaviors that make them attractive and intuitively useable, enhancing user’s empathy with the virtual world.
Navigation routes that allow the visualization of the whole environment (overview) were created using viewpoints from the external part of the world. Different angles were used to offer a complete visual access for the user.
The physical features are especially important because they give the avatars realism. In this work two main characteristics were exploited:
It is also possible for the user to freely navigate through the museum, without using predefined routes. This is useful when the user already knows the environment and does not have orientation difficulties, or if he/she prefers to randomly explore the museum until finding something of interest. In this case, the user has to practice the navigation functions available on the visualization plug-in associated with the Web browser.
Age: avatars are modeled as elder or younger fish (father and son, mother and daughter); Gender: avatars are modeled as male or female guides (father and mother, son and daughter). This characterization was obtained by the use of “accessories” added to the basic default geometric model of the fish: the father wears a hat and a bow tie (FIG. 8a), the mother uses red lipstick (FIG. 8b), the
The navigation strategies adopted allow the generation of personalized tours for the users, according to the choices made before entering the virtual environment. 6
son wears a cap (FIG. 8c) and the daughter wears a pink tie (FIG. 8d). These features help transmitting the semantic meaning of the models to the user.
the offering of guided tours has proven to be a good solution for the exploration of unknown worlds. In such worlds, it is very difficult for the user to know what information exactly is presented and where to find it.
Beyond their physical features, these avatars possess some innate behaviors that enhance their realism. The fishes in the Virtual Guggenheim Museum Bilbao swim through the environment during the tour, simulating the behavior of live fishes in the water. They perform a tail movement, through animation embedded in the geometric model.
Moreover, guided tours performed by interactive avatars give a pleasant illusion of realism to the environment, helping the user to explore more intuitively the information presented, until he/she is oriented without further help. These functionalities allow the user to adapt to the environment, while navigating and interacting with it.
Another behavior modeled is that the user, guided by one of the fish, visualizes the content in the museum according to its age. If the user selects one of the child guides, he/she will see the content pieces exposed in the rooms from the perspective of a child, corresponding to the height of a young person. This kind of behavior is also based on real world metaphors and enhances the realism of the world.
The results presented by this work do not take into account the effect that such type of assistance could have in users that need to participate in activities such as model-based simulations. It is possible that these activities are impaired when performed in real environments without the assistance of guides.
Each guide knows the museum, the rooms that will be visited by the user and which exhibitions are offered in each room. The avatars also infer information about the user that requires their help, knowing what are his/her interests in the museum.
Representing information in the environment using real world metaphors proved being efficient. The functionality of known objects is immediate, facilitating the adaptation of the user to the 3D environment and its virtual tools. It was observed that it is very important that objects modeled according to real world metaphors exhibit a similar behavior of their counterparts in the real world.
Although avatars could explore more knowledge becoming smarter, experiences made with these avatars show considerable enhancement of the user’s empathy and involvement with the environment.
Future and ongoing works include the use of the presented techniques for other domains of application, such as remote sensing14. This work will allow the analysis of the necessities of another group of users, with a high level of knowledge about the content presented by the 3D world. The environment should present a great degree of realism and interactivity, in order to offer an efficient tool for data manipulation and visualization.
CONCLUSIONS AND FUTURE WORK People exhibit different behaviors when they are exposed to the same situation, and also have different desires and needs to be solved. Therefore, the modeling of a threedimensional virtual environment in such a way that a visitor feels satisfied is a difficult task.
REFERENCES
The adaptation of the environment, through the personalization of content and
1. Frery, A. C., Groth, B., Jähnichen, S., Kelner, J., Pessoa, B. D. S., Teichrieb, V. (1999).
7
Organizing Web Sites with Architectural Metaphors. In Proceedings of the 3rd World Multiconference on Systemics, Cybernetics and Informatics and 5th International Conference on Information Systems, Analysis and Synthesis 99. (pp. 43-49). Orlando, USA.
Ambientes Virtuais Tridimensionais. Unpublished Master of Science Dissertation, Universidade Federal de Pernambuco, Recife. 9. toMeta (2001). Frank O. Gehry: The Guggenheim Museum in Bilbao. Online document http://ihong.hypermart.net/project/gehry.htm
2. Drogoul, A., Gortais, B., Hutzler, G. (1998). Grounding Virtual Worlds in Reality. In Proceedings of the 1st Virtual Worlds 98. Paris, France.
10. The Solomon R. Guggenheim Foundation (2001). Guggenheim Museum. Online document http://www.guggenheim.org/ 11. Stanek, W. R. (1996). HTML, Java, CGI, VRML, SGML Web Publishing Unleashed. New York: Sams.
3. Dieberger, A. (1994). Navigation in Textual Virtual Environments using a City Metaphor. Unpublished Doctoral Dissertation, Vienna University of Technology, Vienna.
12. Macromedia (2001). Macromedia. document http://www.macromedia.com
4. Pausch, R., Burnette, T., Brockway, D., Weiblen, M. E. (1995). Navigation and Locomotion in Virtual Worlds via Flight into Hand-Held Miniatures. In Proceedings of the 15th ACM SIGGRAPH 95. Los Angeles, USA.
Online
13. Frery, A. C., Kelner, J. (2001). Museu Virtual – Museu do Homem do Nordeste. Online document http://www.cin.ufpe.br/~if291/galeria/museuhomem /museu-virtual.htm
5. Darken, R. P. (1996). Wayfinding in LargeScale Virtual Worlds. Unpublished Doctoral Dissertation, The George Washington University, Washington.
14. Frery, A. C., Kelner, J., Moreira, J. R., Pessoa, B. D. S., Alheiros, D. M., Araújo Filho, M. S. C., Teichrieb, V. (2000). Desktop Virtual Reality in the Assessment of Tidal Effect. In Proceedings of the 20th International Geoscience and Remote Sensing Symposium 2000. (pp. 1-3). Hawaii, USA.
6. Frery, A. C., Kelner, J., Teichrieb, V., Pessoa, B. D. S., Araujo, M. S. C. (2000). Avatars as Interactive Guides for Navigation Assistance in Three-dimensional Virtual Environments. In Annals of the 3rd Workshop on Virtual Reality 00. Gramado, Brazil.
Address reprint requests to: Alejandro C. Frery Universidade Federal de Pernambuco
7. Pöhlker, B. (1999). Entwurf und Prototypische Integration eines Verteilten, Objektorientierten Werkzeuges zur Interaktiven Definition von Pfaden in die 3D-Echtzeitvisualisierungsumgebung SCORE. Unpublished Degree Dissertation, Technische Universität Berlin, Berlin.
Centro de Informática CP 7851, 50732-970 Recife, Pernambuco – Brazil E-mail:
[email protected]
8. Teichrieb, V. (1999). Avatares como Guias Interativos para Auxílio na Navegação em
8
Avatar physical features
inference module
knowledge
behavior
physical features
knowledge
user profile
behavior
Virtual Environment
User
FIG. 1. Avatar Architecture
FIG. 2. Guggenheim Museum Bilbao
9
FIG. 3. The User Profile Interface
FIG. 4. 3D Model of the Guggenheim Museum Bilbao
10
IG. 5. 3D Icon Associated with a Description Homepage
a
b
c
d
FIG. 6. Icons Representing Architectonic Information: a) German Guggenheim Berlin; b) Peggy Guggenheim Collection; c) Solomon R. Guggenheim Museum; d) Guggenheim SoHo
FIG. 7. Icon Representing Conceptual Information
a
b
c
FIG. 8. Guide Avatars: a) Father; b) Mother; c) Son; d) Daughter
11
d
