ABSTRACT
In this paper we present the project Campiello, which aims to build a community network like system for cultural towns. Thereby Campiello takes new approaches in availability and modality of access to the network. The access to the system will be extended into the real places through new user interface metaphors, and the main functionality for accessing information will not be browsing but recommendation.
Given these objectives, the availability and modality of access can be considered a major issue, because only a broad participation to the community network activities can sustain their growth and wealth. However the experience so far demonstrate that the common user base is mainly composed by computer literate, accessing the network because the already have a PC at home or at the workplace. In fact from the technology point of view community networks are based on large bulletin board and the main user interface is usually PC-based (proprietary application), a Web browser. In some cases the access is possible from public places like Internet cafes or libraries.
In this paper we want to present the EU funded project Campiello which follows a slightly different approach for setting up a community network like system. One main difference is, that in Campiello the community network access will be broadly extended into the real places through new interface metaphors (paper, large screen displays) and not only be accessible from Home PCs. This greater penetration and connection with real life should support a user base truly representative of the community and possibly open the community to more interactions with external members, like in our case the tourists of cultural towns.
Even though the project is not finished right now and we cannot present a usable system or experiences with the usage of a finished system, we think that the approach Campiello is taking might be interesting for existing community networks.
In the rest of the paper we first briefly present the Campiello project objectives (Chapter 2) and then concentrate on two of the new user interfaces that will be used in Campiello to connect the electronic system with the real physical places (Chapters 3 and 4).
So the aim is not only to build a network for empowering an existing community, but to build a network that tries to bring different communities together (tourists and locals) or at least to let them profit from each other.
The system will be installed in two sites, the city of Venice in Italy and the city of Chania in Crete.
In Campiello we try to avoid this weakness by applying an intersection of user driven and design driven approaches. This means, that in addition to the users, we involve a third party, the 'interaction designers' who focus in the design of innovative interaction patterns and not in the development of innovative technologies (as for system engineers). See [Agostini98] for a more detailed description of our methodology.
According to this methodology the project team consists of system engineers, representatives of the users, and a design school that is working in the area of interaction design:
The project started in September 1997 and during the first year we have been working on a set of user scenarios describing which interaction they envision for the system, and on the design of the system that will implement these scenarios. The user scenarios are integrated representation and simulations of the system, including not only the technical solutions in place, but showing also the physical context of use, the roles of the users/actors, the content that circulates within the system, as well as the functionality of the system (see Pictures 1 and 2 in this paper for examples from the scenarios).
For structuring the information space and designing the access to it, we were faced with the well-known tradeoff among structured information system and unstructured 'communication area'. If the system is designed in a structured way it is easy for users to find information. On the other hand if you give users a lot of freedom in creating new information they will more likely contribute, but this usually leads to a loss in structure and so can frustrate people when they try to find something in the network.
We have chosen not to use a hierarchical structure of discussion places, which the user can browse, but to model the main user interface around recommendation and search. The information space consists of a set of items. The major types of items are physical places (e.g. (cultural) buildings, restaurants, and museums), events (e.g. concerts, festivals), and more abstract topics of interest like 'food' or 'parenting'. For search and recommendation issues the items can be related to one or more categories. Additionally, the descriptions or comments of the items can contain links to other items.
The system collects explicit and implicit feedback of the user and combines this with a given profile of interests to select information items that might be interesting for the user. This works in a pro-active way (i.e. without action by the user) and in a reactive way (i.e. the user tells the system to give him items that match some attributes). More information about the idea of collaborative filtering or recommendation can be found in [Shardanand95], KnowledgePump, an existing system from Xerox is described in [Glance98].
On top of the information services devoted to improve the flow of information inside and among the different communities, Campiello offers a layer of functionality to support improved communication and contact facilitation. This layer includes awareness services and services for searching people with similar attributes (e.g. interests) that one could contact for preparing a visit to the town or for whatever.
Finally we provide a personal space (the personal diary) where the users can store information items and partly publish them.
For accessing the information space and the collaborative features we will not only use Web- or PC-based user interfaces, but try to integrate the user interface with the physical places themselves. The idea is to support ubiquitous computing and to provide the possibility to meet people face-to-face at the places themselves. Two of the user interfaces we have in mind for this will be described into more detail in the following chapters.

The main paper item allowing users to enter comments and to request personalized information are the ActiveNewscards. These are pages that contain data from the information space (i.e. parts of the descriptions and comments) and possibilities to express a request for more information or to give feedback.
One type of ActiveNewscards is edited by cultural managers, pre-printed in large numbers, and distributed at several places in the cities. Information cards like this (without the possibility to express feedback) are already available in several cities today. The content of the forms might but does not have to be connected to the place where they are available. Users are free to take these cards, just read the information or also give some feedback and comments and return them at a Campiello service point (a physical place which provides scanning (and printing) equipment). The card can be returned anonymously or be equipped with a personal id that connects the data and the request to a person. In addition to just providing their comments to the system the users can request further information or add information to their personal spaces by returning an ActiveNewscard. Instead of returning the ActiveNewscards at a service point, the users also can fax the pages to a Campiello server. They might also be collected somehow and processed in batch.
The second type of ActiveNewscards are those printed as result of processing an annotated card. They are constructed from information items explicitly selected for the requesting user, taking into account the feedback from the input card and all other information stored in the user and community profiles (using collaborative filtering techniques). In addition to the personalized information selection these types of cards can once again contain means for expressing further feedback.
In addition to ActiveNewscards we will provide paper forms for inserting new items into the information space and for explicitly inserting data into diary and user profile.

The NewsWall is a large screen display, which enables groups of people to browse comments submitted to the Campiello system. Along the top of the NewsWall are a series of subject headings. The space under each heading contains comments relevant to that heading. However the NewsWall display is not static, it is dynamic in two ways. As comments age, they drift off the screen and new comments are added as they are created. This ensures that the comments on display are always recent and timely, unlike the comments that may be returned as the result of a search or recommendation by the other interfaces.
Another aspect of dynamicity which we investigate is that the NewsWall could adapt its display to user interest. If a number of users cluster around a particular subject area, then that area is given more space on the display and less popular subjects are given a smaller amount of screen space.
[Casapulla95] G. Casapulla, F. De Cindio, O. Gentile: 'The Milan Civic Network Experience and its Roots in the Town'. In: Proc. 2nd International Workshop on Community Networking, IEEE Comm. Soc., ACM SIGCOMM, Princeton (NJ), Jun. 1995
[Glance98] N. Glance, D. Arregui, M. Dardenne: 'Knowledge Pump: Supporting the Flow and Use of Knowledge in Networked Organizations'. In: U. Borghoff, R. Pareschi (eds), Information Technology for Knowledge Management, Springer Verlag, Berlin, 1998
[Johnson93] W. Johnson, S.K. Card, H. Jellinek, L.Klotz, R. Rao: 'Bridging the paper and electronic worlds: The paper user interface'. In: Proc. INTERCHI, ACM Press, April 1993
[Schuler93] D. Schuler, A. Namioka (eds.): 'Participatory Design - Principles and Practices'. Lawrence Erlbaum Associates, Hillsdale, NJ, 1993
[Schuler94] D. Schuler: 'Community Networks: Building a New Participatory Medium', Communications of the ACM, Vol. 37, No. 1, Jan. 1994
[Shardanand95] U. Shardanand, P. Maes: 'Social information filtering: Algorithms for automating word of mouth'. In: Proc. CHI'95, Denver CO, May 1995, ACM Press, pp. 210-217.
[i3] i3 web site, http://www.i3net.org/
[Campiello] Campiello web site, http://www.dsi.unimi.it/~campiello Campiello (Esprit Long Term Research Project 25572) started in September 1997 and will last until August 2000.