This paper presents Video Città, a system that provides services for users to share and comment videos about a physical location. The system lets users upload their own videos, annotate and hyperlink content that was previously uploaded. It provides different modes of access including the Web and PDAs for mobile access. The system explores several new research directions, namely the generation of hypermedia structures from the existing materials as a way to provide personalized movie structures, the collaborative annotation of multimedia materials and the mobile access to geo-referenced contextual information.
Video; Personalization; Mobile Access; Collaborative Work
VideoCittà is a collaborative environment created to promote
interaction between users engaged in tourism or other cultural activities, wanting
to share their materials and their experiences, when visiting or studying a
city or region.
One of the key aspects in the design of Video Città was the support for
heterogeneous client applications, and the ability to deal with those differences
in terms of client technology. A user may be connected to Video Città
trough a mobile handheld computer, a networked PDA equipped with GPS or trough
the web browser of a personal computer. The adoption of an http based client-server
architecture, was the basic integrating element, and made possible the adoption
of a simple web browser as the simplest and weakest form of client platform.
Personalization, in its different forms [2] plays also an important in this
type of system as a way to adapt the information to be delivered to the user
considering its main interests and location.
VideoCittà is structured according to the geographical reality of the
locations it represents, using a map as the main navigation mechanism. Behind
the simple representation of the geographic elements that build the city, VideoCittà
is, as its name implies a view of the city, trough the videos available for
viewing and commenting. These videos can be institutional, commercial or, more
generally, personal videos uploaded to the system by all the users who want
to share their audiovisual materials to the community.
For those preparing a visit, or studying the cultural or historical elements
of a region, the videos may be an important reference, but more important it
is possible to complement the images with small bits of information or annotations.
Taking notes is a natural way of expressing impressions or feelings about an
external source of information.
Paper is undoubtedly the traditional support for annotated materials and has
to be thought as a reference concerning the basic needs and mechanisms of annotation.
In VideoCittà users are invited to share not only their videos, but to
share their views, their impressions, or the knowledge they have about cultural
or historical aspects of the location, by annotating the videos. Video Città
brings collaborative video annotation to the historical and tourism domains,
enabling people to share their videos and to interactively provide additional
information annotating the existing ones. Every member of the community may
annotate their videos and those provided by other members, with no thematic
restriction to the annotations a user can provide. Figure 1 depicts the process
of adding one annotation using the Web interface.
An important element in Video Città is related with the ability of accessing
the system from mobile devices, as is related with the concept of geo-thematic
relevance. The geo-thematic relevance is an element of personalization that
tries to capture the geographical proximity of the user with an entity that
is strongly connected with his main interests.
For those performing scientific work these features could be very helpful, when
integrated with the communication mechanisms for user and for group interaction.
Personalization plays an important role in the management of group interaction
simplifying interaction between explicit groups (teams) and suggesting or promoting
interaction between implicit groups, or members with similar interests.
For the tourist user, we believe that the capacity to collect remotely placed
messages could be fun, and motivating. In respect to group interactions, mechanisms
exists that inform the user of the presence on site, of other members with interesting
points in common. The location based features of the system contain in itself
mechanisms of adaptation that give the user, information about the place where
it is and the events taking place in its vicinity.
The Video Città system also contains mechanisms for the automatic generation
of a movie that illustrates a tour around the city. This movie, made from the
materials that exist in the system is personalized, meaning that the movie tries
to capture the most relevant elements of the city according to the user's main
interests. The current prototype already contains mechanisms for the movie tour
generation, and in a near future these features will be extended in order to
incorporate the generation of narrative content based on the user experience
visiting the city.
Personalization plays an important role in the usability of the system. The
annotations provided by a community of users, make the viewing of an annotated
video, difficult and time consuming. Some sort of user adaptation is necessary
in order to deliver to the user just the information relevant to its needs or
interests. The characterization of the personalization features available in
the system is made according to several aspects [2]:
§ Content and presentation management
§ User interaction
§ Group Personalization
The content of the video and the additional information provided by annotations
can be of an overwhelming diversity. The same historical monument may be relevant
for different users by their cultural importance or simply by some personal
reason. The first personalization approach in VideoCittà is made by simply
trying to collect information about the user and his main interests, in order
to suggest the type of information that may be relevant to that user.
In the context of VideoCittà the user expresses his main general interests
and the particular interest points in the city he is visiting or planning to
visit. The interest points may be not only monuments or locations but also events
or ceremonies that take place in the region and that may be held in different
locations. Besides the content adaptation, also the layout and presentation
of the information needs to be adapted. The multimodal nature of the system
implies the existence of mechanisms that provide content adaptation based on
the characteristics of the user's accessing device. The user may be accessing
the system trough small mobile devices as phones or PDAs, or trough a web browser.
The type and volume of information that must be delivered to the same user,
is constrained by the modality of access he is using. The existence of location
based services is also a form of personalization: the content delivered to the
user may be adapted according to its location. The annotations previously made
by the user may be available when he is near the point of interest, information
about events, related spots may be adapted to its current location.
The role played by the user in the construction of its profile defines two
types of user interaction: implicit and explicit. In the context of Video Città
explicit personalization plays an important role in the definition of the user
profile and in their particular interests about cultural or historical events
in the city or region. If the user is accessing the system to make plans for
a visit or accessing it when touring the town, clear information about the user's
interests are the key to provide him with information that is relevant to the
objectives of his visit. The system also uses implicit interaction, when making
suggestions about information that may be interesting to the user. This implicit
form of user interaction is also present when location based recommending is
enabled. The information delivered to the user is constrained by his location,
the user is implicitly defining its patterns of interest trough is geographic
location. A non-interactive approach is also present and uses the information
available about the user actions and navigation pattern to infer and make suggestions
about items that could be relevant to the user.
The group personalization is extremely important in the context of this project and we plan to address it in two different perspectives: information and communication. The information perspective deals with information targeted to specific groups. Communication in this context is related with group awareness mechanisms, enabling an augmented support for interaction within explicit and implicit defined groups. Explicitly defined groups, can be traveling companions or friends with common interests in a city visit. Common interests may also be found in groups of users implicitly considered, due to strong similarities in their profiles or characteristics. People from the same country probably share a common interest in information that may be related with its national country. Groups of fans of a music artist or adepts of a sport probably have common information interests regarding events related with that artist or that particular sport The main body text should use a 10-point Times Roman font. Please follow the style defined in this document for the title, authors and section headings. There is NO need to leave a blank text box at the lower left corner of the first page as theprevious version of this template suggested.
Most of the components of the system, mainly the server and the Web client were
built and the tests with users will start now. Future work will concentrate
on interfacing profile information for personalization and providing mobile
access to the system. We hope that the current functionality and the features
that are being developed will allow VideoCittà to be used by different
types of user communities, sharing their videos and their interests about the
cities and regions that are more relevant to them.
Our thanks to Pedro Assunção, Luís Soares e Gualter Garcia, from the MaDLabs, for their contribution on the interface design, of the video cittá prototype.