«1 Introduction and previous work By the term Interactive Digital Artwork (IDA, for short) we mean any artwork where digital technology is an ...»

Another real IDA is described in [1]: an installation receiving texts, images and sound files from its audience through their Bluetooth enabled handheld devices and converting them into sound compositions played through the seven loudspeakers located in seven arms of the sculpture (see figure 1 right). 3D sound effects are thus obtained in the space defined by the sculpture itself. The globe of the sculpture contains a lighting system changing light colors according to the different sounds it plays. The classification label for such an IDA is [content provider:audience, processing dynamics:static, processing contributors:artist].

Our classification overcomes the limitations of previously presented ones and explicitly targets IDAs by means of an approach that it is rooted on the standard input-processoutput view used for discussing Information Systems. Hence our proposal considers interactivity just as one of the components of the classification and is therefore more balanced.

We have not considered in our classification scheme issues related to hardware and software, either in terms of IDA development environment or in terms of the environment where the work is viewed, since both these issues are too much dependent on the current state of development of technology. Hence the characterization of these aspects, while useful from an historical point of view, does not make much sense for the intended use of IDA classification.

The classification might anyhow be refined by taking into account also the sensory channels by means of which interaction between IDA and its users happens, but this will be subject of further work.

3 Comparison with previous frameworks In this section we provide a comparison between our classification framework and the previous ones. The comparison is not easy, since previous classification frameworks 6 E.Nardelli were based on the various kinds of interaction, while we have taken the more general approach of the Information Systems view.

To make the reader able to better understand the comparison we first recall here below the definition of previous classes provided by the previous works in the literature.

Sommerer and Mignonneau [8] discuss two types of interaction:

Pre-Designed: the viewer can choose her path of interaction among a set of limited and pre-defined possibilities, Evolutionary: the artwork’s processes are linked to interaction and interaction is evolving continuously.

Hannington and Reed [4] distinguish three types of interaction:

Passive: the content has a linear presentation and the user interacts by only starting and stopping the presentation, Interactive: the user is allowed to choose her personal path through the content, Adaptive: the user is able to enter her own content and control how it is used.

Edmonds, Turner, and Candy [3] discuss four categories of "relationship between the

artwork, artist, viewer and environment":

Static: there is no interaction, Dynamic-Passive: the artwork response is triggered by environmental factors, Dynamic-Interactive: the human presence and/or actions (purposeful or not) are used as parameters for changing the artwork, whose processing rules are static, Dynamic-Interactive (varying): the processing rules used by artwork to change its output are modified by an agent (the artwork software or a human).

Trifonova, Jaccheri, and Bergaust [9] consider three categories, with various subdivisions:

Interaction Rules: whether the rules controlling the interaction are static (i.e., they never change during artwork’s life) or dynamic (i.e., they may change), Triggering Parameters: whether the interaction rules depend just on the human presence, or it is required some form of human action, or it is the environment that controls them, Content Origin: whether what the artwork shows is predefined by the artist, or provided by the user, or generated by the software, possibly through some evolutionary algorithm.

4 Validation of the proposed framework In this section we discuss how the IDAs analysed in [2] can be classified on the basis of the proposed framework.

In [2] a complete analysis of interactive art installations in Italy is provided, focusing on technological tools used by artist in the field. The analysis focused on the 54 installations for which there was enough available information, and they were classi��ed according to the framework of [9]. Many of them received the same classification, hence we considered for our validation only one instance of each class identified during this classification process.

It is interesting to note that in [2] it is explicitly recognized a weakness of the framework of [9], since it is reported that all the 54 installations were classified as static under the dimension "interaction rules" of the framework of [9]. This means that such a dimension was useless to classify the installations since its values were not able to differentiate among installations.

8 E.Nardelli We report in table 2 the most important descriptive data of one installation for each of the classes we have identified according to our framework. The number in the first column is the number by which installations are referred to in [2]. When the cell for the column "Physical Interface for direct manipulation by the user" is empty it means that the audience has no direct means for interacting with the IDA. Under the columns "Input Device(s)" and "Output Device(s)" we list the actual devices used by the IDA to, respectively, obtain data from the Audience or Environment and to produce data towards them. In the table 3 we show how the installations listed in table 2 are classified according to our framework. When more than one value appears under some dimension this means the IDA has received more than one label. The value PD/CC change is used to represent both the labels pre-defined change and casual change, since [2] does not distinguish between them. As you can see, our classification does not have the above cited weakness of the classification defined in [9], since in all dimensions more than one value is used. We use all the labels, but for evolutionary in the dimension "Processing Dynamics". This was expected, since this kind of processing dynamics is

very sophisticate in mathematical terms and also in [2] it is noted that no installations of this kind was found in Italy.

The last column lists the numbers of all other installations classified with the same set of labels. One class (the one in the second row) covers almost one half of all analyzed

installations, and two other classes (first and fourth rows) cover about one third of them:

this coarseness is shared with the classification in [9] and suggests it is of real practical importance to extend this work by distinguishing also among the various sensory channels by means of which interaction between IDA and its users happens. Moreover, a further empirical validation of our classification needs to be done by directly interviewing the artists and getting their direct feedback on the new classification. Finally, the issues recalled in Section 1 in the research theme Computer Art and the Commercial Perspective are worth further investigations.

In this paper we have presented and discussed a novel framework for classification of interactive digital artworks (i.e., artworks based on ICT and where the user is directly involved in the production of the artistic output and called, for short, IDAs). The need 10 E.Nardelli for such a classification derives from the needs of relating and comparing homogeneous IDAs, of having a common description framework for researching, discussing and teaching about IDAs, and of definining "how-to" procedures for IDAs production and management.

We have built our classification framework on the basis of a critical revision and refinement of previous work. Its novelty lies in its being directly based on the inputprocess-output view considered for discussing Information Systems. Hence it allows to overcome weaknesses and limitations of the previous proposed ones. Our classification framework is validated by applying it to a set of 54 real-life examples of IDAs in Italy.

Acknowledgments. We would like to thank Letizia Jaccheri for useful and interesting discussions during the development of the work here described. Comments from referees helped in improving the presentation.

References

1. Salah Uddin Ahmed, Letizia Jaccheri, and Samir M’kadmi. Sonic onyx: Case study of an interactive artwork. In Fay Huang and Reen-Cheng Wang, editors, Revised Selected Papers of the 1st International Conference on Arts and Technology (ArtsIT’09), pages 40–47, YiLan, Taiwan, Sep 2009. Springer Verlag, Lecture Notes of the Institute for Computer Sciences, Social Informatics and Telecommunications Engineering (LNCSIST) vol.30. http:// www.springerlink.com/content/nx243t25h1448750/fulltext.pdf.

2. Leonora Cappellini. Interactive installation art in Italy: an analytical survey. Master Degree Thesis in Informatics for Humanities, Supervisors: Luciana Vassallo and Letizia Jaccheri, University of Pisa, 2009. http://www.idi.ntnu.no/ ∼letizia/tesi_cappellini.pdf.

3. Ernest Edmonds, Greg Turner, and Linda Candy. Approaches to interactive art systems. In 2nd International Conference on Computer Graphics and Interactive Techniques in Australasia and South East Asia (GRAPHITE’04), Singapore, 113–117 2004. ACM.

4. Anne Hannington and Karl Reed. Towards a taxonomy for guiding multimedia application development. In 9th Asia-Pacific Software Engineering Conference (APSEC’02), Gold Coast, Queensland, Australia, December 2002.

5. Enrico Nardelli. A software based installation to assist self-reflection. In 11th Consciousness Reframed International Research Conference (CR-11), Trondheim, Norway, November 2010.

TEKS - Trondheim Electronic Arts Centre.

6. Briony J. Oates. New frontiers for information systems research: Computer art as an information system. European Journal of Information Systems, 15:617–626, 2006.

7. Franc Solina, Peter Peer, Borut Batagelj, and Samo Juvan. 15 seconds of fame - an interactive computer-vision based art installation. In 7th International Conference on Control, Automation, Robotics and Vision (ICARCV’02), pages 198–204, Singapore, 2002.

8. Christa Sommerer and Laurent Mignonneau. Art as a living system: Interactive computer artworks. Leonardo, 32:165–173, 2001.

9. Anna Trifonova, Maria Letizia Jaccheri, and Kristin Bergaust. Software engineering issues in interactive installation art. International Journal of Arts and Technology, 1(1), 2008.