Tangible Interfaces to Digital Connections,
Centralized versus Decentralized
Matthijs Kwak, Gerrit Niezen, Bram van der Vlist, Jun Hu, and Loe Feijs
Designed Intelligence Group, Department of Industrial Design
Eindhoven University of Technology
Den Dolech 2, 5612AZ, Eindhoven, The Netherlands
Abstract. In the era of distributed digital media, technology is moving
to the background and interoperability between devices increases. The
handles for users to explore, make and break connections between devices
seem to disappear in overly complex menu structures displayed on small
screens. Two prototypes have been developed that introduce a tangible
approach towards exploring, making and breaking connections between
devices in a home environment. Findings suggest that users are better
able to project their mental model of how the system works on decen-
tralized representations and that a tangible solution is not necessarily a
better one.
Keywords: Ontology, semantic connections, tangible user interface,
internet of things.
1 Introduction
In the era of distributed digital media, especially in a home environment, devices
are connected to one another to create preferred experiences. A home theatre
system is one example of how multiple devices can create one joint experience
when interoperating [5,6,3,7]. With the introduction of portable media players,
possibilities and needs for content sharing are even bigger. Currently these de-
vices are connected wirelessly or with all kinds of cables, and users are currently
occupied with finding the right cables to connect devices and have to deal with
cables that physically allow for connections that are not possible. Even more,
some possible connections never get explored, simply because physical cables do
not allow for it. Wireless technologies such as Bluetooth solve part of the prob-
lem, but introduce overly complex menu structures and devices without proper
interfaces. A single task like sharing music from the one device to another cur-
rently involves multiple steps on both devices, while one single high-level effort
would be desirable.
In ‘The Internet of Things’ [8] and ‘Shaping Things’ [14] a world is sketched
in which each everyday object has an unique identity and is connected to the
internet. In this world, technology has moved to the background and interoper-
ability between devices has been achieved. Provided that these devices are able
to communicate with each other and with the user, this could mean the end of
Z. Pan et al. (Eds.): Transactions on Edutainment V, LNCS 6530, pp. 132–146, 2011.
c
© Springer-Verlag Berlin Heidelberg 2011

Tangible Interfaces to Digital Connections, Centralized versus Decentralized 133
compatibility problems and the hassle of using cables, and that users will have
less physical and visual handles to make sense of their environments and the
devices therein. Design can play an important role in this sense-making with
paradigms like Tangible User Interfaces [4], that believe that physical handles
for digital information provide users with more freedom and control.
The SOFIA project is a European research project that targets to “make
‘information’ in the physical world available for smart services - connecting the
physical world with the information world” [13]. Within this project a “Semantic
Connections” demonstrator was developed named Interaction Tile. This demon-
strator allows users to tangibly explore, make and break connections between
devices in a smart home environment [11,16,17].
A second demonstrator, named Interaction Tabs, was developed to explore
alternative possibilities of Tunis. Where the Interaction Tile provides users with
a centralized way of exploring, making and breaking connections, the Interaction
Tabs provides users with a decentralized way to perform the same tasks.
In order to see which demonstrator would be the easiest to use and allow for
a better projection of the users’ mental model, a user experiment was set up to
answer the following questions:
– Are the demonstrators a better alternative, compared to the conventional
method?
– Will the users be able to work equally well with both demonstrators?
In the first question, “better” is in the sense that exploring, making and break-
ing connections are easier (more efficient) and more satisfactory (positive user
experience). An important aspect is the mental model that the participants have
and how it compares to the actual architecture of the system.
2 Background
2.1 SOFIA Project and the Interaction Tile
SOFIA (Smart Objects for Intelligent Applications) is a European research
project addressing the challenge of Artemis sub-programme 3 on Smart Environ-
ments. The overall goal of this project is to connect the physical world with the
information world, by enabling and maintaining interoperability between elec-
tronic systems and devices. Our contribution to the project is to develop smart
applications for the smart home environment, and to develop novel ways of user
interaction. For users to truly benefit from smart environments, it is necessary
that users are able to make sense of such an environment. One way of facilitating
this “sense making” is through design. Our contribution to the SOFIA project
aims at developing theories and demonstrators, and investigating novel ways of
user interaction with the smart environment, through interaction with smart
objects in the space.
To illustrate the concepts and ideas developed in the project, a demonstrator
was developed. The demonstrator is a tile-like interactive object that allows for
both exploration of a smart space in terms of connections, and manipulation of

Tangible Interfaces to Digital Connections, Centralized versus Decentralized 135
We want to enable users to explore and manipulate the connections within the
smart space without having to bother with the lower-level complexity of the
architecture. We envision this “user view” to be a simplified view (model) of the
actual architecture of the smart space. Conceptually, the connections are carriers
of information; in this case they carry music. Depending on the devices’ capa-
bilities (e.g. audio/video input and/or output) and their compatibility (input to
output, but no output to output), the Interaction Tile will show the connection
possibilities. In our current demonstrators we do not distinguish between differ-
ent types of data since we are only dealing with audio, but it will be inevitable
in more complex scenarios. The Interaction Tile acts as an independent entity,
inserting events and data into a triple store and querying when it needs infor-
mation. The different types of events and the connections between smart objects
and their related properties are described in an ontology. The ontology with “is-
a” relationships indicated is shown in fig. 2. Fig. 3 shows the architecture of the
current setup.
Owl:Thing
NFCEvent
SmartObject
Event NetworkEvent
MediaPlayerEvent
NFCExitEvent
NFCEnterEvent
DisconnectEvent
ConnectEvent
PlayEvent
CueEvent
StopEvent
is-a
is-a
is-a
is-a
is-a
is-a
is-a
is-a
is-a
is-a
is-a
is-a
Fig. 2. Ontology indicating “is-a” relationships
We implemented the demonstrator using the Jena Semantic Web framework,
the Processing library for Java, and Python for S60. Every interaction with either
the music players (smart phones) or the interaction tile results in an interaction
event. A semantic reasoner (Pellet) is used to reason about these low-level events
in order to infer higher-level results.
When the user shakes the tile to establish a connection, two NFCEnterEvent
events (generated by the RFID reader inside the interaction tile) by two differ-
ent devices that are not currently connected, will result in a new connectedT o
relationship between the two devices. Because connectedT o is a symmetric rela-
tionship, the reasoner will automatically infer that a connection from device A to
device B means that device B is also connected to device A. Since connectedT o
is also an irreflexive property, it is not possible for a device to be connected to it-
self. A generatedBy relationship is also created between the event and the smart
device that generated it, along with a timestamp and other event metadata.

136 M. Kwak et al.
SIB
Windows XP with Jena
KP / Music Player
Python for S60
Nokia 5800 XpressMusic
KP / Music Player
Python for S60
Nokia N95
KP / Interacon Tile
Arduino-based with Processing /
Python (RFID)
TCP Socket over WiFi
TCP Socket over WiFi
Serial over USB
Serial over Bluetooth
KP / Ambient Lighng System
Arduino-based
KP / Surround Sound System
Windows XP with Processing
TCP Socket
Fig. 3. Overview of the demonstrator
(a) Interaction Tile (b) Interaction Tabs
Fig. 4. Two versions of the demonstrator
2.2 Interaction Tabs: Decentralized
A second demonstrator was developed to explore other tangible solutions. The
Interaction Tabs demonstrator was implemented using the same set-up and soft-
ware, but replacing the Interaction Tile (fig. 4(a)) with the Interaction Tabs
(fig. 4(b)).
First the Interaction tile was analyzed using the Frogger Framework (fig. 5)
[18]. The Frogger framework is a design framework that allows for both analyz-
ing and synthesizing interactions. Six relations (couplings) between action and
reaction are described:
Time. The product’s reaction and the user’s action coincide in time.
Location. The reaction of the product and the action of the user occur in the
same location.

Tangible Interfaces to Digital Connections, Centralized versus Decentralized 137
Ac
tio
n
In
he
re
n
t i
nf
o
rm
a
tio
n
Au
gm
e
n
te
d
in
fo
rm
a
tio
n
Fu
n
ct
io
na
l in
fo
rm
a
tio
n(time)
(location)
(direction)
(modality)
(dynamics)
(expression)
(time)
(location)
(direction)
(modality)
(dynamics)
(expression)
(time)
(location)
(direction)
(modality)
(dynamics)
(expression)
(time)
(location)
(direction)
(modality)
(dynamics)
(expression)
(time)
(location)
(direction)
(modality)
(dynamics)
(expression)
(time)
(location)
(direction)
(modality)
(dynamics)
(expression)
Fig. 5. Frogger Framework [18]
Direction. The direction or movement of the product’s reaction (up/down,
clockwise, right/left and towards/away) is coupled to the direction or the
movement of the user’s action.
Dynamics. The dynamics of reaction (position, speed, acceleration, force) is
coupled to the dynamics of the action (position, speed, acceleration force).
Modality. The sensory modalities of the product’s reaction are in harmony
with the sensory modalities of the user’s action.
Expression. The expression of the reaction is a reflection of the expression of
the action.
Furthermore, Wensveen distinguishes between three types of feedback and feed-
forward; functional, augmented and inherent [18]. Feedback is “the return of
information about the result of a process or activity” [2]. Functional feedback is
“the information generated by the system when performing its function”. Aug-
mented feedback is information generated by an additional source, not directly
related to the system and its function. Inherent feedback was defined by Lauril-
lard [9] as “information provided as a natural consequence of making an action.
It is feedback arising from the movement itself.” Feedforward is the information
provided to the user before any action has taken place. Inherent feedforward com-
municates what kind of action is possible and how one is able to carry this action
out. When an additional source communicates what kind of action is possible it
is considered augmented feedforward. Functional feedforward communicates the
more general purpose of a product.
There are many improvements one can consider for the Interaction Tile (fig. 6)
when putting it in the Interaction Frogger framework. We decided, though, to
stay as close to the original design as possible; for research purposes it is best to
change as little as possible in order to be able to clearly identify what exactly
causes change in user behavior (if users’ behavior actually changes).

138 M. Kwak et al.
Ac
tio
n
In
he
re
n
t i
nf
o
rm
a
tio
n
Au
gm
e
n
te
d
in
fo
rm
a
tio
n
Fu
n
ct
io
na
l in
fo
rm
a
tio
n
(time)
(modality)
(time)
(location)
(direction)
(modality)
(dynamics)
(expression)
Fig. 6. Interaction Tile in the Frogger Framework
Ac
tio
n
In
he
re
n
t i
nf
o
rm
a
tio
n
Au
gm
e
n
te
d
in
fo
rm
a
tio
n
Fu
n
ct
io
na
l in
fo
rm
a
tio
n
(time)
(direction)
(modality)
(time)
(location)
(direction)
(modality)
(dynamics)
(expression)
Fig. 7. Interaction Tabs in Frogger Framework
By removing the center tile and moving its functionality into the cubes that
represent the devices being connected/disconnected, the demonstrator would
become simpler as this would allow for a more direct manipulation of the con-
nections. The (digital) states of the connections would be physically represented,
cubes being aligned means that the devices that they represent are connected,
and not being aligned means they are disconnected.
Inspired by Siftables [10] the cubes were transformed to tabs, because tabs
have a clear top and bottom. This does still afford stacking, but hopefully users
would understand that the tabs were to be aligned (fig. 4(b)). An LED at each
side gives the feedback:

Tangible Interfaces to Digital Connections, Centralized versus Decentralized 139
Red. No connection possible. This occurs when no relation is possible between
the two devices of which the tabs are aligned.
Green. This occurs when a relation exists between two devices of which the
tabs are aligned.
To make a connection, the tabs that represent these devices have to be aligned.
To break the connection, the alignment has to be broken.
As a result of removing the center tile, the Interaction Tabs will no longer
allow for the exploration of existing connections and connection possibilities
without immediately manipulating the connections. Moving towards having a
more physical approach might influence the scalability and have some other
practical implications; however, this is beyond the scope of this paper as we try
to focus on the interaction itself.
We expect these differences to also have different influences on the user’s men-
tal model, in the way users conceptualize connections (including properties like:
persistence, transitivity and directionality) and differences in how they imagine
devices to be connected, e.g. devices connected in a networked fashion versus
connecting devices peer-to-peer. When we also analyze the Interaction Tabs us-
ing the Interaction Frogger framework (fig. 7), there are two ways of comparing
the two demonstrators:
First we consider the Interaction Tile and Interaction Tabs as being part of the
same demonstrator set-up as shown in Fig. 3, serving as a device to manipulate
the connections between the various devices in the set-up. The changes might
improve the interaction with regard to:
Direction. With the center tile removed, the direction of making and breaking
connections (although done remotely) corresponds better.
Modality. With the shaking interaction removed, the modality of making and
breaking connections corresponds better.
Secondly, we look at the interaction devices themselves as if they were stand-
alone products. This reveals more improvements: Information about time, loca-
tion, direction and modality are augmented and inherent when the center tile
and shaking interaction are removed. For the Interaction Tile only location is
inherent, and time, direction and modality are augmented (fig. 8 and fig. 9).
3 Experiment
In order to answer the questions raised in the beginning of this paper, an exper-
iment was conducted.
3.1 Participants
12 participants were invited to the experiment, in which 3 were women and 9
were male. The participants were between 21 and 26 years old. All but one had
a BSc. in Industrial Design. One also had a MSc. in Industrial Design.

140 M. Kwak et al.
Ac
tio
n
In
he
re
n
t i
nf
o
rm
a
tio
n
Au
gm
e
n
te
d
in
fo
rm
a
tio
n
Fu
n
ct
io
na
l in
fo
rm
a
tio
n(time)
(location)
(direction)
(modality)
(location)
(time)
(direction)
(modality)
(time)
(location)
(direction)
(modality)
(dynamics)
(expression)
Fig. 8. Interaction Tile in the Frogger Framework (stand-alone)
Ac
tio
n
In
he
re
n
t i
nf
o
rm
a
tio
n
Au
gm
e
n
te
d
in
fo
rm
a
tio
n
Fu
n
ct
io
na
l in
fo
rm
a
tio
n(time)
(location)
(direction)
(modality)
(time)
(location)
(direction)
(modality)
(time)
(location)
(direction)
(modality)
(time)
(location)
(direction)
(modality)
(dynamics)
(expression)
Fig. 9. Interaction Tabs in Frogger Framework (stand-alone)
3.2 Apparatus
The following equipment was used:
– Dell laptop (Windows XP) with Wi-Fi, Bluetooth antenna, audio out and
two USB ports.
– Nokia N95 mobile phone with Python installed, running a script to be able
to play a sample and communicate with the laptop.
– Nokia 5800 XpressMusic mobile phone with Python installed, running a
script to be able to play a sample and communicate with the laptop.

Tangible Interfaces to Digital Connections, Centralized versus Decentralized 141
– Ambient Light lamp: A Bluetooth Arduino-based lamp that renders the
music in colored lighting using RGB LEDs, with code running to be able to
communicate with the laptop.
– Samsung NV8 digital camera mounted on a tripod to record the experiment.
– Philips speaker set with two satellite speakers and a subwoofer, connected
to the Dell laptop.
– Netgear WPN824 wireless router.
– Interaction Tile and Interaction Tabs, including software in Java and Python.
A controlled setting was used to conduct the tests. The study took place in
the ‘Contextlab’ at Eindhoven University of Technology. The lab is furnished to
look like a living room, which is the context in which the demonstrators would
normally be used.
3.3 Measures
We gathered data about the usability of the demonstrators in comparison to
conventional methods of connecting devices, using Bluetooth pairing. Here the
usability is divided in three aspects; efficiency, effectiveness and satisfaction.
The setup of the test was exploratory, but includes two proven methods to gain
insight in the participants’ mental models and have the participant score the
usability, respectively the ‘Teach-Back protocol’ [15] and the ‘System Usability
Scale’ [1]. The action cycle by Norman [12] was also used to gain insight in the
participants’ mental models.
Added to these methods, we also collected data about task completion time,
errors, recovery from errors and participants’ satisfaction with using the method.
A between-subjects design was used.
3.4 Procedure
Participants explored, made and broke connections between two mobile phones
(a Nokia N95 and a Nokia XpressMusic), a sound system and an Ambient Light
lamp. This was done using the Interaction Tile, Interaction Tabs and Bluetooth
pairing.
Every session was recorded and notes were made by the moderator.
In this study, each participant worked through four phases of tasks starting
with one of three methods (Interaction Tile, Interaction Tabs and Bluetooth
pairing). Bluetooth pairing was tested as a comparative conventional method to
measure the usability of the demonstrators.
Briefing. Participants received a brief explanation (5 min.) before the test,
outside the ‘Contextlab’. They were guided through the task path by the mod-
erator. After the explanation, participants filled in a pre-test questionnaire and
signed an informed consent form. The pre-test questionnaire included questions
about: age, gender, occupation, self-report of familiarity with interactive systems
(computers and mobile phones) and the participant’s experience with usability
studies and focus groups.

142 M. Kwak et al.
Tasks. After the briefing, the participants worked on the actual tasks for about
30 minutes (including intermediary discussions). The task path for each method
(Interaction Tile, Interaction Tabs and Bluetooth pairing) is as follows: First,
users were introduced to the method and given three task descriptions. For each
description they were asked to connect the devices or configure the demonstrator
to perform the tasks (9 minutes). Second, users were given a task description
and asked to fill in an Action Cycle diagram (6 minutes). Third, users were
presented with three scenarios. For each scenario they were asked to explain
which connections there were (9 minutes). Fourth, users were asked to explain
what the method was they had used and how it worked using the teach-back
protocol (6 minutes). The order of tasks was random but the same for each
participant.
Debriefing. After the main tasks were performed there was a post-test ques-
tionnaire (5 minutes), where participants filled in the SUS questionnaire to rate
the satisfaction of using the method. The session was concluded with a post-
test discussion (5 minutes), where the moderator followed up on any particular
problem that came up for the participant.
3.5 Moderator Role
The moderator sat in the room with the participant while conducting the ses-
sion. The moderator introduced the session, conducted a short background inter-
view, and then introduced tasks as appropriate. Because this study is somewhat
exploratory, the moderator sometimes asked unscripted follow-up questions to
clarify participants’ behavior and expectations. The moderator also took notes
and recorded the participants’ behavior and comments. The session was digitally
recorded on video using a Samsung NV8 digital camera.
4 Results
Unfortunately the system was not stable enough to accurately measure the per-
formance data that was intended to be measured. The stability program also
influenced the grades given by participants in the SUS questionnaire. Therefore
the SUS scores were not reliable. However, we did gain certain insights from the
observations and other measurements.
4.1 Action Cycle Diagram
The participants clearly had problems with filling in the Action Cycle Diagram.
Only a few descriptions correspond to the predefined description. This can be
explained by the fact that people do not consciously think about the seven steps
as defined by Norman [12] during everyday activity. It is also not uncommon
to go through several cycles before a goal is reached and not all of these cycles

Tangible Interfaces to Digital Connections, Centralized versus Decentralized 143
have to include all seven steps. This would require participants to fill in several
diagrams or include several cycles in one diagram. Because this issue did not
surface during the pilots or the first test, it would have been incorrect to change
the procedure.
All participants followed roughly the same steps in achieving their goal. All
but one participant forgot to mention the breaking of existing connections in
‘Action specifications’. The participants using the Interaction Tabs and Blue-
tooth noticed this during the execution (before they thought they had achieved
the goal) and went through another iteration immediately. Of the participants
using the Interaction Tile, all but one participant noticed this after the execu-
tion (after they thought they had achieved the goal). These participants went
through another iteration at a later stage but were also able to achieve their
goal.
4.2 Teach-Back Protocol
While it is possible to draw conclusions concerning the actual mental models of
participants, the protocol was mainly used to see if there were notable differences
between the methods.
Although there were some differences between the participants individually,
amongst the methods the drawings and explanations were roughly the same.
None of the participants went into details about what happened in the back-
ground, but instead focused on the matters ‘at hand’. Three participants (2
for the Interaction Tile and 1 for Bluetooth pairing) mentioned extending the
current system with more devices (more mobile phones and a TV). One partic-
ipant (Interaction Tile) was able to conclude that the connected devices were
networked; the rest explained the connections in a hierarchical way.
In one of the examples given in [15] the researchers were able to conclude
that participants tend to draw little when the system is transparent. If it is less
transparent they are likely to make more detailed drawings to better support
their story.
In this test the level of detail amongst the methods was roughly the same.
4.3 Observations and Post-Test Discussion
None of the participants that worked with the Interaction Tabs had trouble
working with that method. During the post-test discussion they only wondered
what was happening in the background. This was not because they had been
unable to perform certain tasks, but because they suspected more was going on
than what was visible to the user.
None of the participants working with Bluetooth pairing had trouble working
with that method. They all mentioned that they were familiar with this way of
connecting devices but had never experienced Bluetooth working this well.
The only real trouble for the participants working with the Interaction Tile
was the initial experience with that method. It was not clear what the relation
was between the center tile and the cubes and all four interpreted the pulsing

144 M. Kwak et al.
green LED as a ‘working connection’. One participant initially thought the LEDs
were lasers that could ‘read’ the cubes when placed on top. Another participant
thought it was only necessary to align the ‘main’ device to the center tile and
align the other devices to the ‘main’ device.
During observations and post-test discussions it became clear that all but one
participant were not able to get from the method that the connected devices
were networked. The tasks given and the methods at hand led them to conclude
that the connections were hierarchical and participants mainly followed one of
two modes of arranging connections:
Linear (from one device to the next) - This was seen with the Interaction Tile
and Interaction Tabs.
Centralized (from one device outwards) - This was seen with the Interaction
Tabs and Bluetooth pairing.
Some participants sporadically arranged connections with the Interaction Tile
in a way that indicated they took it for a network, but they explained verbally
that they expected the system to make a hierarchy out of their arrangement.
Some participants also explicitly mentioned that certain connections should not
be possible while in fact they were.
5 Discussion and Conclusion
The most interesting results came from the observations and post-test discussions
with the participants. The fact that all but one thought and worked in hierarchies
is an interesting one. The Interaction Tile was designed to convey a different way
of thinking, but instead participants projected their hierarchical way of thinking
on the method. By making connections between no more than two devices at a
time they did not use the full capacity of the system, took longer to perform the
tasks and were slightly annoyed by the ‘extra’ work. Also, for those who thought
in centralized hierarchies (one device in the center, the others around it), there
was no way of projecting this thought on the Interaction Tile.
This is where the power of the Interaction Tabs showed, because it allows
more ways of thinking (hierarchical, ontological, linear, and centralized). The
participants found meaning in the arrangement of the tabs and the location
of the tabs in relation to each other. For the system this does not matter; a
connection is a connection and if devices are connected, they are networked.
This leads to conclude that the Interaction Tabs are a better fit for this
scenario. Because of the setbacks, it is not possible to say whether the Interaction
Tile and Interaction Tabs are better than Bluetooth pairing, although it appears
that the Interaction Tabs are. It also appeared that participants were better able
to perform the tasks with Bluetooth than with the Interaction Tile but this can
be attributed to the fact that they had experience with Bluetooth pairing and
connecting devices using a GUI.
For further research it would be interesting to see whether the hierarchical
thinking of people can be generalized to scenarios other than the ones used in this
user experiment. This could include other or more devices and media, or even

Tangible Interfaces to Digital Connections, Centralized versus Decentralized 145
completely different contexts. If it can be generalized, an interesting question
would be whether solutions like the demonstrators should allow for hierarchical
thinking while working with ontologies, or not.
Although the described experiment was useful and showed interesting results,
it clearly has some limitations. For such an experiment to be successful, more
participants are required, preferably without a design background. Six partici-
pants for each method is limited, four even more so. It is clearly not possible to
collect reliable quantitative data with this number of participants.
Added to that, not all the methods used in the experiment worked out as
expected. While the ‘speak-out-loud’ step of the Action Cycle diagram is useful
to get insights in what participants think when performing tasks, the other steps
seemed often unclear to them.
The results indicate that it may also be possible to elicit the mental model
of the participants using the Teach-back protocol. It was useful for this test to
see that all methods equally provided the participants with information, but the
full potential of the protocol was not utilized.
If this user experiment were to be repeated at a later stage, the advice would
be to have at least three people to be present during the tests; someone to
manage the software and hardware, someone to guide the participant through
the test and someone to make detailed notes. For a more qualitative approach,
the fourth step of the Action Cycle diagram (think-out-loud) could be consid-
ered for each task. For a more quantitative approach, the Action Cycle diagram
could be removed from the test completely, as well as the Teach-back protocol.
This would allow for more tasks to be performed, which results in more data
to be analyzed. A more elaborate usability questionnaire could be considered,
although one has to take into account that lengthy questionnaires might annoy
participants. This is especially to be considered when questionnaires are com-
bined with performing tasks; this could lead to participants not paying enough
attention when answering the questions.
Acknowledgement
SOFIA is funded by the European Artemis programme under the subprogramme
SP3 Smart environments and scalable digital service. We would also like to thank
the participants involved in the user experiment.
References
1. Brooke, J.: Sus-a quick and dirty usability scale. Usability Evaluation in Industry,
pp. 189–194 (1996)
2. Dictionary: American heritage dictionary. The American Heritage Dictionary of
the English Language (2000)
3. Feijs, L., Hu, J.: Component-wise mapping of media-needs to a distributed pre-
sentation environment. In: The 28th Annual International Computer Software and
Applications Conference (COMPSAC 2004), pp. 250–257. IEEE Computer Society
Press, Hong Kong (2004)

146 M. Kwak et al.
4. Fitzmaurice, G., Ishii, H., Buxton, W.: Bricks: laying the foundations for graspable
user interfaces, pp. 442–449. ACM Press/Addison-Wesley Publishing Co.
5. Hu, J.: Design of a Distributed Architecture for Enriching Media Experience in
Home Theaters. Phd thesis, Department of Industrial Design, Eindhoven Univer-
sity of Technology (2006)
6. Hu, J., Feijs, L.: An adaptive architecture for presenting interactive media onto
distributed interfaces. In: Hamza, M. (ed.) The 21st IASTED International Con-
ference on Applied Informatics (AI 2003), pp. 899–904. ACTA Press, Innsbruck
(2003)
7. Janse, M., van der Stok, P., Hu, J.: Distributing multimedia elements to multiple
networked devices. In: User Experience Design for Pervasive Computing, Pervasive
2005, Munich, Germany (2005), http://www.fluidum.org/events/experience05
8. van Krannenburg, R.: The internet of things: A critique of ambient technology and
the all-seeing network of rfid. Institute of Network Cultures (2008)
9. Laurillard, D.: Rethinking university teaching: A framework for the effective use
of educational technology. Routledge, New York (1993)
10. Merrill, D., Kalanithi, J., Maes, P.: Siftables: towards sensor network user inter-
faces. In: TEI 2007, pp. 75–78. ACM, New York (2007)
11. Niezen, G., van der Vlist, B.J.J., Hu, J., Feijs, L.M.G.: From events to goals:
Supporting semantic interaction in smart environments. In: 2010 IEEE Symposium
on Computers and Communications (ISCC), Riccione, Italy, pp. 1029–1034 (2010)
12. Norman, D.: The design of everyday things. Basic Books, New York (2002)
13. SOFIA: Project website, http://www.sofia-project.eu/
14. Sterling, B.: Shaping things (2005)
15. van der Veer, G., Del Carmen Puerta Melguizo, M.: Mental models. In: The
Human-Computer Interaction Handbook: Fundamentals, Evolving Technologies,
and Emerging Applications, pp. 52–80 (2003)
16. van der Vlist, B., Niezen, G., Hu, J., Feijs, L.: Design semantics of connections
in a smart home environment. In: Design and Semantics of Form and Movement
(DeSForM 2010), Lucerne, Switzerland (2010) (accepted)
17. van der Vlist, B.J.J., Niezen, G., Hu, J., Feijs, L.M.G.: Semantic connections: Ex-
ploring and manipulating connections in smart spaces. In: 2010 IEEE Symposium
on Computers and Communications (ISCC), Riccione, Italy, pp. 1–4 (2010)
18. Wensveen, S., Djajadiningrat, J., Overbeeke, C.: Interaction frogger: a design
framework to couple action and function through feedback and feedforward. In:
DIS 2004, pp. 177–184. ACM, New York (2004)