ArticlePDF Available

Personalization in Cultural Heritage: The Road Travelled and the One Ahead

Authors:

Abstract and Figures

Over the last twenty years, cultural heritage has been a favored domain for personalization research. For years, researchers have experimented with the cutting edge technology of the day; now, with the convergence of internet and wireless technology, and the increasing adoption of the Web as a platform for the publication of information, the visitor is able to exploit cultural heritage material before, during and after the visit, having different goals and requirements in each phase. However, cultural heritage sites have a huge amount of information to present, which must be filtered and personalized in order to enable the individual user to easily access it. Personalization of cultural heritage information requires a system that is able to model the user (e.g., interest, knowledge and other personal characteristics), as well as contextual aspects, select the most appropriate content, and deliver it in the most suitable way. It should be noted that achieving this result is extremely challenging in the case of first-time users, such as tourists who visit a cultural heritage site for the first time (and maybe the only time in their life). In addition, as tourism is a social activity, adapting to the individual is not enough because groups and communities have to be modeled and supported as well, taking into account their mutual interests, previous mutual experience, and requirements.
Content may be subject to copyright.
Personalization in Cultural Heritage
1
Personalization in Cultural Heritage: The
Road Travelled and the One Ahead
Liliana Ardissono1, Tsvi Kuflik2, Daniela Petrelli3
1Università di Torino
2The University of Haifa
3Sheffield Hallam University
1+39 011 6706716
2+972 (0) 4 8288511
3+ 44 (0) 114 225 6946
1liliana@di.unito.it
2tsvikak@is.haifa.ac.il
3d.petrelli@shu.ac.uk
2http://mis.hevra.haifa.ac.il/~tsvikak/Home.htm
3 http://dagda.shef.ac.uk/daniela/Daniela_Petrelli/Welcome.html
Over the last twenty years, cultural heritage has been a favored domain for personalization
research. For years, researchers have experimented with the cutting edge technology of the day;
now, with the convergence of internet and wireless technology, and the increasing adoption of the
Web as a platform for the publication of information, the visitor is able to exploit cultural heritage
material before, during and after the visit, having different goals and requirements in each phase.
However, cultural heritage sites have a huge amount of information to present, which must be
filtered and personalized in order to enable the individual user to easily access it.
Personalization of cultural heritage information requires a system that is able to model the user
(e.g., interest, knowledge and other personal characteristics), as well as contextual aspects, select
the most appropriate content, and deliver it in the most suitable way. It should be noted that
achieving this result is extremely challenging in the case of first-time users, such as tourists who
visit a cultural heritage site for the first time (and maybe the only time in their life). In addition, as
tourism is a social activity, adapting to the individual is not enough because groups and
communities have to be modeled and supported as well, taking into account their mutual interests,
previous mutual experience, and requirements.
How to model and represent the user(s) and the context of the visit and how to reason with regard
to the information that is available are the challenges faced by researchers in personalization of
cultural heritage. Notwithstanding the effort invested so far, a definite solution is far from being
reached, mainly because new technology and new aspects of personalization are constantly being
introduced.
This article surveys the research in this area. Starting from the earlier systems, which presented
cultural heritage information in kiosks, it summarizes the evolution of personalization techniques
Personalization in Cultural Heritage
2
in museum web sites, virtual collections and mobile guides, until recent extension of cultural
heritage toward the semantic and social web. The paper concludes with current challenges and
points out areas where future research is needed.
Keywords: Personalized Access to Cultural Heritage, Personalization, Cultural
Heritage
1. Introduction
The extensive amount of existing cultural heritage material, which far exceeds the
space available in museums, and the increasing interest in making collections
accessible to a large public have motivated cultural heritage institutions to adopt
web-based and mobile information tools as a means for presenting their
collections. Most institutions are developing web sites in order to achieve a low-
cost web presence, and many are offering mobile guides to assist visitors on site.
The convergence of internet and wireless technology has made the exploration
of cultural heritage a continuous process, starting before the visit and ideally
never ending, as the user is able to plan the visit online, visit the site, and then
“revisit” places of interest online again. However, any novel technology
introduces new challenges: firstly, large web-based collections are difficult to
browse and entail the risk of overloading users with information; secondly,
visitors are highly heterogeneous and require different types of information, at
different levels of detail; finally, the target users of cultural heritage sites (and
tourists in general) are often first- and short-time visitors to an unknown place.
This means that, on the one hand, they are in constant need of help in finding
relevant information, while on the other hand, providing them with appropriate
information is challenging because, initially, their interests and needs are
unknown.
Cultural heritage has been a privileged application domain for personalization
for many years and recent museum research also acknowledges the need for
personalized, individual support. Specifically, John Falk (2009) points out that
museum visitors differ and their visit experience is composed of the physical, the
personal, and the socio-cultural context, and identity-related aspects. Hence they
may benefit from individualized support that takes into account contextual and
Personalization in Cultural Heritage
3
personal attributes. Moreover, visitors’ behavior may not remain consistent during
the visit and this may require ongoing adaptation.
While technology-oriented research in personalization of cultural heritage
information started in the early ’90s, when pioneer projects developed techniques
aimed at adapting suggestions and presentation of information to the individual
user, museum research studies started to explore the idea of personalized services
only recently. The evolution of these two research areas occurred asynchronously
in the exact/technological sciences (as technology evolved) and in Humanities
(with the evolution of museums visitors’ studies).
This article surveys the evolution of technology-based personalization in
cultural heritage and points out the current challenges. As the systems developed
over the years differ in several aspects, a comparative table has been used to
facilitate the analysis; the criteria for comparison and the reviewed systems are
reported in Section 2. The article then uses an historical perspective and, starting
from the first systems, which presented cultural heritage information in kiosks, it
summarizes the evolution of personalization techniques in museum web sites,
virtual collections, and mobile guides (Section 3); it then discusses methods and
techniques for personalization (Section 4), and concludes by pointing to current
challenges and future research directions (Section 5). A short summary concludes
the paper (Section 6).
2. A Comparative Analysis
During the past 20 years, the research on technology-based personalization of
cultural heritage experimented with different settings and proposed various
solutions. These were developed following the evolution of information and
communication technologies (ICT), specifically concerning: communication
networks (from wired to broadband wireless connections), end-user devices (from
desktop computers to mobile smart phones) and User Interfaces (from textual to
multimodal ones and virtual reality). Table 1 below provides a comparison of
many of the adaptive systems developed so far. A number of different dimensions
have been used as analytical tools; each dimension is further specified to highlight
particular aspects, as discussed in the following:
Personalization in Cultural Heritage
4
Setting captures the situation where the interaction occurs. Indoor and
outdoor both indicate physical presence at the site, while virtual shows the
visitor is using a virtual gallery or collection, most likely remotely from the
site itself, and sometimes in virtual reality 3D.
Device: different devices afford different interactions and therefore we
distinguish desktop, mobile, and wearable/tangible. Specifically, the
“mobile” column describes the type of end-user device for which the User
Interface of the reviewed systems is developed: tablet (T), laptop (L), PDA
(P), and smart phone (S).
Presentation style (Pres. style) captures the form of delivery and the medium
of the presentation, often suggested by the setting or the device. We
distinguish between: map-based when the metaphor used is that of a map on
which the points of interest are displayed; web pages when the display
follows a traditional Web layout; character when the content is delivered by
an agent (generally a cartoon); and audio-video when the presentation is
different from the text and includes delivery modes that are exclusively audio,
or audio and video (could be animations or clips).
Adaptation type (Adapt.) is used to distinguish which features are taken into
account during the adaptation process: context-aware is used when features
such as opening times or proximity are taken into account; individual denotes
personalization to a single person; and group indicates that the adaptation is
targeted to a group of people.
UM representation (UM rep.) is used to classify systems according to how
they represent the user model. Different representations are considered, such
as describing the user’s interest in the concepts of the domain (Overlay),
specifying particular user features (Feature-based), a “bag of words”
(Content-based), or the list of preferred items (List of Items). See Section
4.2.1 for details.
UM initialization (UM init.) differentiates systems according to the way in
which the user model is initialized. Some systems do not initialize the user
model at all, or they start with a general model, the same for every user
(None); others question the user about her/his Preferences, or they employ a
Personalization in Cultural Heritage
5
Stereotype based on some user information. Yet other systems exploit the
Keywords associated with the items selected by the user, or the information
retrieved from other systems (Mediation) to initialize the user model. Finally,
some systems initialize the user models by means of a case-based approach.
See Section 4.2.2 for details.
UM updating captures the different dynamics in the evolution of the user
model during the interaction with the user. Feedback type is used to represent
the intentionality of the interaction: E stands for explicit interaction (e.g.,
clicking on a link), I for implicit interaction (e.g., walking toward an exhibit),
and B for both. The other facets of this dimension denote the particular
inference techniques applied to update the user model (Heuristic inference,
Activation/Inhibition networks, Collaborative filtering, Content-based,
Semantic reasoning). See Section 4.2.3.
Matching user and content (Matching) is aimed at selecting the most
appropriate content to be presented for a given user or users group. Several
techniques have been developed for this purpose: earlier proposals used
condition-action rules (C/A rules), or activation/inhibition networks
(Act/inhib network) for the selection of the most relevant content to be
presented. Recently, users have been enabled to annotate items with rates,
tags and comments (Social), and such information is employed to browse
content on the basis of folksonomies, and to select content by Ranking it.
Items can be ranked by applying various techniques, such as Content Based
filtering, Collaborative filtering, Vector Space, possibly enriched with
Semantic reasoning techniques. See Section 4.3.
The next two sections present cultural heritage systems according to these
dimensions and aspects. As space limits prevent us from describing each system
in depth, Table 1 reports every feature of every system, helping the reader get a
sense of the whole research area at a glance. For each system reported in the table,
but not reviewed in this paper, a footnote shows a relevant bibliography reference.
Appendix A provides a larger font version of the table.
Personalization in Cultural Heritage
6
Table 1. An analytical comparison of personalized CH systems.
System
Setting
Device
Pres. style
Adapt.
UM init.
UM updating
Matching
Name
Indoors
Outdoors
Virtual
Desktop
Mobile
Wearable/tangible
Map-based
Web pages
Character
Audio-Video
Context-Aware
Individual
Group
Overlay
Feature-based
Content-based
List of items
Preferences
Stereotypes
None
Mediation
Keywords selection
Case Based
Feedback type
Heuristic inference
Act/inhib network
Collaborative filtering
Content Based
Semantic reasoning
Social
C/A Rules
Ranking
Semantic reasoning
Act/inhib network
Collaborative Filtering
Content-Based filtering
Vector Space
AlFresco
ILEX
Hyperaudio
P
B
AVANTI
E
Marble
Museum
HIPPIE
P
B
HIPS
P
I
GUIDE
T
B
CRUMPET
P
B
Active
WebMuseum
M-PIRO
I
AmbieSense1
S
B
Tiddler
AgentSalon
E
Archeoguide2
LT
P
I
The Museum
Wearable
I
Gulliver’s
Genie
P
I
INTRIGUE
P
LISTEN
I
COMPASS3
PS
E(c)ho
UbiquiTO
PL
S
B
UbiCicero4
P
I
PEACH
P
B
PeVEP
3D
B
Kubadji
B
iCITY
PS
B
CHIP
E
CHAT
E
PIL
P
*
E
*
ARCHIE
P
Tate Online
Edinburgh
Ad. Gallery
3D
Smartmuseum
PS
E
Kurio5
MUSE
E
DeepMap
I
1 Göker, A. and Mirhaug, H.: 2008, Evaluation of a mobile system in context. Information processing and management 44,
39-65.
2 Vlahakis V., Ioannidis N., Karigiannis J.: 2002, ARCHEOGUIDE: Challenges and Solutions of a Personalised
Augmented Reality Guide for Archaeological sites. Computer Graphics in Art, History and Archaeology, IEEE Computer
Graphics and Applications, 22(5), 52-60.
3 van Setten, M., Pokraev, S., Koolwaaij J.: 2004, Context-Aware Recommendations in the Mobile Tourist Application
COMPASS. In W. Nejdl, W. and P. De Bra, (eds.). Adaptive Hypermedia: Proceedings of the Third International
Conference, Eindhoven, the Netherlands, 235-244.
4 Ghiani, G., Paternò, F., Santoro, C. and Spano L.D.: 2009, UbiCicero: a location-aware, multi-device museum guide.
Interacting with computers 21(4), 288-303.
5 Wakkary, R., Hatala, M., Muise, K., Tanenbaum, K., Budd J.: 2009, Kurio: A Museum Guide for Families. In: N. Villar,
S. Izadi, M. Fraser, S. Benford, D. Kern and A. Sahami (eds.): Tangible and Embedded Interaction: Proceedings of 3rd
International Conference. Cambridge, UK, 215-222.
Personalization in Cultural Heritage
7
3. Historical Overview
Personalization was first introduced on desktop systems but was quickly applied
to mobile technology to exploit its potential in delivering context-aware and
personalized information “on the go”. This section retraces the main milestones in
the management of the user interaction, which has focused on the following
issues:
Improving the selection of the information to be presented, given the users
interests, attitudes and surrounding context, including the social context in
which the user is immersed.
Improving the modality adopted for delivering information to the user, in a
desktop setting and/or “on the go”, and the organization of the presentations,
for better comprehension and navigation of large amounts of information.
Enriching the media used to interact with the user, from text only to audio
and video content, as well as enhanced reality.
We distinguish between two scenarios desktop and mobile due to the
fundamental differences between them that still exist, even though the
personalization technologies used in the two settings are converging.
3.1 On the Desktop: from Hypertext to the Web and Virtual Reality
The personalization of cultural heritage can be traced back to the early ’90s, when
some of the adaptive hypermedia systems looked at museum content (AlFresco
(Stock et al., 1993), ILEX (Oberlander et al., 1998)) and tourism (AVANTI (Fink
et al., 1998)) as possible application domains. AlFresco provided personalized
content as part of a human-machine dialogue on Italian art and combined natural
language processing with a hypermedia system connected to a videodisc.
Personalized content was also the goal of ILEX, which automatically generated
hypertext pages with text and images taken from material harvested from existing
catalogues and transcriptions of conversations with the curator. Natural language
generation techniques maintained a coherent narrative throughout the pages that
were generated. A contemporary of the AlFresco system, AVANTI was a desktop
application running on kiosks and the first to explore the creation of personalized
Web pages of tours in cities, taking into account the user’s interest, knowledge,
and physical abilities.
Personalization in Cultural Heritage
8
With the advent of the web, museums started creating information-rich web
sites, and personalization was considered a good way to support users in finding
information in unfamiliar collections. The online Marble Museum of Carrara
(Paternò and Mancini, 1999) offered three choices to users: 1) a static stereotype
of either ‘tourist’, ‘student’, or ‘expert’ which gave predefined views on the
information; 2) a user profile that should be set manually (e.g., selecting which
information should be displayed); and 3) an adaptive virtual guide that took into
account what had been seen and suggested additional relevant information in a
separate window. The Active WebMuseum (Kohrs and Merialdo, 2001) used the
metaphor of ‘corridors’, represented as lists of artworks, to organize sets of related
paintings dynamically, e.g., same artist, same art movement, same time frame.
The artworks were selected on the basis of collaborative filtering techniques
(Konstan and Riedl, this issue2012) and the users could move along corridors and
change path at any time, exploring the content at their leisure.
Quite naturally, when web technology started to be available on WAP phones6,
cultural heritage applications followed. The new opportunity was explored by
INTRIGUE (Ardissono et al., 2003), which supported browsing cultural heritage
information and planning tours that met the requirements of heterogeneous user
groups, such as families with elderly members and children. The content (text and
images) was dynamically selected, arranged and displayed on web pages tailored
for either a desktop browser or a WAP mini-browser.
The idea of adapting content from the PC to different devices was promoted by
MUSE (Garzotto et al., 2003), which used multichannel Web application
technology to provide visitors with information both on- and off-site: standard
web pages could be accessed on a PC at home or via a hand-held device on site.
The device controlled large displays for multimedia information (e.g., 3D models
or video clips) that could not be presented by the mobile technology of that time.
During the visit, the user could bookmark interesting items or content and this
personal collection was put on a CD as a memento to be revisited at home.
The Adaptive Hypermedia and Virtual Reality threads evolved in parallel. As
far as the former is concerned, the iCITY tourist guide (Carmagnola et al., 2008)
Personalization in Cultural Heritage
9
provided recommendations about cultural events and resources to visitors of the
city of Torino, either in textual format or over a city map. The system offered
multi-device adaptation using XML-based standards for the development of User
Interfaces. At about the same time, CHIP (Wang et al., 2008) experimented with
semantic web technologies to enrich the presentation of the Rijskmuseum
collection with information retrieved from public ontologies.
Experiments with virtual reality started with M-PIRO (Calder, 2005), a follow-
up of ILEX and of the HIPS multimedia tourist guide (Petrelli and Not, 2005)7,
which used simple virtual objects as exhibits. Natural language generation was
used to create labels in multiple languages and, in the next evolution, to provide
the speech of a robot-guide introduced in the Virtual Adaptive Gallery of the
University of Edinburgh (Oberlander et al., 2008). The robot displayed
personalized textual descriptions that took into account what the user had seen to
avoid repetition or to provide background information when introducing new
concepts. Even though the Adaptive Gallery was developed within the Second
Life social virtual space, it created an isolated experience where the visitor was
separated from the group. Conversely PeVEP (Bonis et al., 2009), a platform for
creating 3D virtual museums with personalized content, supported both a
personalized experience and the social context. The visitors could enter non-
personalized thematic rooms, but had a personalized one that contained objects
selected on the basis of their individual interests and behavior. Users could also
see other visitors and chat with them, enter the rooms of others with similar
interests, rate objects, and leave comments or read those that others had left.
3.2 On the Go: Tablet, PDA and Phones, Wearable and Tangible
As soon as mobile technology appeared, it was adopted for delivering context-
aware cultural heritage information both indoors and outdoors. Generally
speaking, indoors there is a limited space, which is extremely rich in content but
offers few services over and above information delivery. In contrast, outdoor
mobile guides covered relatively large areas – a city or part of it – offering a wide
6 WAP phones were the ancestors of current smart phones. They supported the access to Web sites and Web-
based applications using the Wireless Application Protocol.
7 HIPS is described in Section 3.2.
Personalization in Cultural Heritage
10
range of integrated services, such as those providing hotel and travel information.
A few projects tried to cover both cases, e.g., MUSE and Smartmuseum (Ruotsalo
et al., 2009), but most were specialized in one or the other.
Supporting users “on the go” requires knowing the user’s location. Therefore,
context awareness became dominant in mobile guides, using different location-
aware services for outdoors (GPS based) and indoors (where no prevailing
positioning technology yet exists).
The accurate determination of the visitor’s position indoors led to the idea of
interaction through physical movement. This type of interaction was first
introduced by Hyperaudio (Petrelli and Not, 2005): standing for a long time in
front of an exhibit indicated interest while leaving before the audio presentation
was over was considered as displaying the opposite attitude. Hyperaudio also
pioneered mobile guides in the late '90s, using the leading technology of the time:
an Apple Newton modified to detect infrared signals. The system monitored the
visitors’ movement and dynamically composed audio snippets on the basis of
their current position, the chosen perspective (e.g., entertaining vs. informative)
and the interaction history (exhibits seen; for how long; content listened to).
Outdoor personalization was pioneered by GUIDE (Cheverst et al., 2000), which
used a Tablet PC to deliver information on points of interest in the city of
Lancaster, UK, using cellular WiFi technology for positioning. The proximity of
the user to points of interest and the opening times of attractions were used to
select suitable information for the current user’s context and profile.8 Information
was displayed as images and text in web pages; additional services allowed
members of the same group to receive and send messages to each other, and to
leave and read comments.
Smaller than a tablet, but with a larger screen and with more colors than
phones, the appearance of the PDA paved the way for a large number of
prototypes, both indoor and outdoor. The idea of “interacting with the space”, first
introduced by Hyperaudio, was extended in HIPS, which classified visitors on the
basis of their visiting pattern behavior (Bianchi and Zancanaro, 1999): the
position, captured on the PDA, was sent for processing and a dynamically
8 The profile used in GUIDE included the list of preferred attractions, user’s interest, the current position, and
the visited places.
Personalization in Cultural Heritage
11
composed presentation was sent back for the visitor to play. HIPPIE, a spin-off of
the HIPS project, experimented with a different approach, using a dynamic user
model based on the user’s inferred interest and knowledge for personalization
purposes (Opperman and Specht, 2000). While the main delivery medium for
HIPS was audio, HIPPIE produced dynamic text and image hypertext. Both
systems explored Infra Red (IR) technology for indoor positioning. CRUMPET
(Schmidt-Belz et al., 2003) used PDAs for providing dynamic and interactive
maps that showed the current position, recommendations, information about
attractions, and visiting tips. In this system, positioning was based on GPS data.
As mobile and communication technology evolved, so did mobile guides:
PEACH (Stock et al., 2007) dynamically created animation-based video clips
taking into account the user’s focus of attention. A virtual character migrated to
the PDA from a large screen at the beginning of the visit and accompanied the
visitor, presenting information, throughout the visit. At exit time, a personalized
summary was generated, hinting at what seemed to be most interesting content for
the visitor and suggesting future activities. A more affective interaction, though
not personalized, was sought by Damiano et al. (2008), where “Carletto” (a spider
character) dramatized the presentations for visitors, moving away from the idea of
‘guiding’ toward a more engaging ‘storytelling’ approach.
A different kind of service was provided by the iCITY tour guide, which
offered personalized notification of events to visitors using their mobile phones.
The interaction with space introduced by HyperAudio was further explored in
LISTEN (Zimmermann and Lorenz, 2008): the user wore a pair of motion-tracked
wireless head-phones and walked around the exhibition space enjoying
dynamically composed 3D audio presentations. “Speech, music and sound effects
were arranged to form an individualized and location-aware soundscape offering
information related to visual objects as well as creating context-specific
atmospheres”. This type of interaction was further investigated in mobile guides
for visually impaired visitors, triggered by RFID and tilt-based sensors, available
in PDAs (Ghiani et al., 2008).
The ec(h)o project (Hatala and Wakkary, 2005) took an entirely different
approach, experimenting with tangible and embodied interaction, applying vision-
based localization and combining it with RFID technology: a wooden cube
Personalization in Cultural Heritage
12
enriched with orientation sensors was used to interact with the system, and
headphones were used to receive audio content. The interaction metaphor was that
of a two-way conversation, with the user rotating the cube in different directions
to select one of the options proposed by the system.
Localization of visitors in the space was also used to experiment with wearable
computers. In the Museum Wearable project, video clips relevant to the object in
view were played on a single lens goggle (Sparacino, 2002).
After several years of experiments that focused on exploring how the
technology can support the visitor, the experience of mobile personalized cultural
heritage was very rich, though not complete. The social experience of a group
visit was hampered by the individualistic approach. In this direction, addressing
this shortcoming, AgentSalon (Sumi and Mase, 2001) used virtual characters that
accompanied visitors in public places to foster face-to-face interaction between
people. When the visitors approached large screens, the virtual characters
migrated from their mobile device to the screen (together with their user’s profile)
and started interacting in order to trigger face-to-face communication among the
humans on issues of mutual interest. Opening up to the social context for indoor
visits was also the focus of PIL (Kuflik et al., 2011): the group members were
modeled individually, but an effort was made to pull them together in sharing
interesting exhibits (e.g., by messaging each other or by suggesting different
objects to see), fostering follow-up discussions after the visit. Another step
forward was taken by Stock and Callaway (2009) who introduced adaptive,
dramatized information presentations into a museum visitors' guide running on
smart phones. Visitors were provided with different dramatic presentations of the
exhibit, in order to encourage discussion about the individual experience later on.
4. Methods and techniques for personalization in
cultural heritage
Digitization of cultural heritage items is not sufficient to support their retrieval in
very large collections, a fact which challenges the user when navigating through
them (Schmitz and Black, 2008). Concerning this issue, personalization
techniques can be employed to steer the selection of information on the basis of
the user / group interests and context, thus protecting them from data overload.
Personalization in Cultural Heritage
13
Moreover, personalization can be used to adapt the presentation of information to
the user’s device, thus facilitating its exploration.
However, for these purposes, cultural heritage information has to be
represented in a machine-readable format which can be matched with the
interests/preferences/context of the recipient(s). This section reviews the
approaches adopted in cultural heritage personalization for representing
information about items (4.1) and users (4.2), and to perform the informed
matching between them (4.3).
4.1. Information Representation
Some early systems (e.g., AVANTI, INTRIGUE, ILEX, AlFresco) stored the
information about cultural heritage items in structured knowledge bases that
provided rich details about item properties and features but were particularly hard
to develop and maintain. In order to overcome such limitations, some systems
employed natural language generation techniques to create the presentations
directly from the archival descriptions of items; see, e.g., Tiddler (Paris et al.,
2001). Conversely, in other projects, lightweight representations and techniques
were explored to minimize the amount of information about items used by the
systems, or to feed system repositories with information automatically retrieved
from existing archives. Explored solutions include:
A simple list of objects representing the exhibition as “visit paths” (Kubadji
(Bohnert et al., 2008));
Text descriptions and “bag of words” representations of the exhibits on
display9 (Kubadji and PIL);
“Bag of concepts” representations generated by natural language processing
techniques to support a concept-based item classification (CHAT (de Gemmis
et al., 2008));
Collection-specific ontologies for the multi-classification of artworks, such as
location and culture, and multi-faceted search (Delphi toolkit (Schmitz and
Black, 2008)).
9 The “bag of words” is obtained by counting the frequency of words in texts as a representation of the
descriptions (Hanani et al., 2001).
Personalization in Cultural Heritage
14
Alternative approaches looked at improving the quality and the amount of
information available in metadata descriptors by integrating them with additional
semantic information. Astudillo et al. (2008) used ontologies to describe the
“circumstances” of artifacts (e.g., historical, geographical, contexts); CHIP
mapped metadata to concepts defined in different vocabularies in order to enrich
the information repository with geographical, temporal and art history
information.
Currently, user-generated content is seen as a useful source of information to
complement or extend metadata. Ahiara et al. (2008) enrich metadata with
descriptions taken from blogs and Wikis; iCITY employs user tags and
annotations about cultural events to extend the cultural heritage repository with
new knowledge and firsthand experience of items.
4.2. Modeling the User
Personalization implies the modeling of the user. This is an important distinctive
feature, as systems such as Cyberguide (Abowd et al., 1997), mobiDENK
(Krösche et al., 2004) and Cicero (Ciavarella et al., 2004), which select content on
the exclusive base of the context (e.g., current position or time), are unable to
adapt it to the user’s features. In the following, we briefly describe how user
models are represented, initialized and dynamically updated in personalized
cultural heritage systems.
4.2.1. User Model Representation
Needless to say, the user models are closely related to the representation of the
domain information, as outlined below:
Overlay, widely adopted in systems such as PEACH and e(c)ho, describes
user interests, knowledge, and other aspects in terms of the concepts
represented in the domain ontology. This approach was taken also by CHIP,
which overlaid a user model on art ontology, and PeVEP, where a semantic
graph composed by domain items modeled the user.
Feature-based, used in AVANTI, GUIDE, HyperAudio, Hippie, Gulliver’s
Genie (O'Hare and O'Grady, 2003), LISTEN, UbiquiTO (Cena et al., 2006)
and INTRIGUE. This approach represents user characteristics (including
personal, socio-demographical, and interests/ preferences) as a list of feature-
Personalization in Cultural Heritage
15
value pairs. In iCITY, the feature-based user model is multidimensional
(describing the user’s interests for each category of cultural heritage
information), to support content selection on multiple criteria, e.g., proximity
of an event or recentness.
Content-based, where user interests are represented as sets of words occurring
in the textual descriptions of items relevant for the user.10 This approach was
taken by PIL and Kubadji. CHAT goes a step further, by applying a layer of
semantic analysis over the simple “bag of words”, achieving a better
representation of documents and their semantic similarity, than in other
methods (e.g., Feature-based and Overlay).
List of items: in some systems, such as Kubadji, user models only store the
user’s ratings of items (“Collaborative Filtering” approach). Moreover, in
other systems, the user model stored “cases”, which are also lists of items
selected by the users.
4.2.2. Initialization
In many systems (e.g., AVANTI, HIPPIE, UbiquiTO, Gulliver’s Genie, the
Marble Museum and GUIDE), visitors were asked to fill in questionnaires in
order to determine their background, interests and preferences and to set up the
user model. However, it was soon clear that many of the features in the
questionnaires did not affect the visiting style (Petrelli et al., 1999) and that form-
filling was not what visitors wanted to do first (Filippini Fantoni, 2003). This
fostered different approaches: setting up a few simple features (four in
HyperAudio), using stereotypes as a quick means to infer information about the
user (AVANTI, PEACH, Marble Museum, PeVEP and INTRIGUE), specifying
interesting keywords (AgentSalon), or rating artifacts (CHAT), sometimes in
combination with a few questions (CHIP).
If a system offers both online and onsite services, as do MUSE and CHIP, then
core information about the user, collected during an online interaction, can be
used later to bootstrap the onsite guide (e.g., INTRIGUE, UbiquiTO). Conversely,
10 This is the typical user model representation adopted in content-based filtering recommenders based on
unstructured information about items (Billsus and Pazzani, 2007).
Personalization in Cultural Heritage
16
the DeepMap mobile tourist guide (Fink and Kobsa, 2002) explored the idea of
user modeling servers as a source for personal information. PIL explored the use
of mediation (Berkovsky, 2008), incorporating user modeling information
available from other systems to (a) retrieve attractions selected by the user at tour
preparation time from external systems, and (b) combine such case-based
information into a content-based user model steering the presentation of items in
the museum guide. On the other hand, iCITY exploited a common user identity
across services in order to reuse the user information collected by each of them
(by applying mediation techniques in order to address data interoperability
issues).
4.2.3. Dynamic user modeling
A dynamic user model is the core functionality of an adaptive system (Opperman,
1994). The model can be updated on the basis of two possible types of users’
actions: explicit, triggered by direct user interaction, such as rating the
information delivered (e.g., in PIL) and, implicit, such as walking speed or stop
(e.g., in LISTEN, HIPS), or navigation and annotation behavior (iCITY). Implicit
input is less accurate, but also less intrusive than explicit feedback. The two
approaches have also been used together (as in HyperAudio, UbiquiTO and
PEACH).
Various techniques have been used to adapt the user model dynamically,
including:
Heuristic inference, based on weighting mechanisms that combine the
evidence about user behavior to infer the user’s features. For instance,
UbiquiTO and iCITY increment/decrement the user’s interest level depending
on the number and type of actions performed by the user (e.g., clicking on an
item, saving it, rating it, etc.). In other systems, probabilistic methods are
applied for this purpose (e.g., Museum Wearable and AVANTI).
Activation/inhibition networks, where a user model is overlaid on a domain
knowledge base; the user feedback is interpreted as direct interest in domain
concepts and propagated over the links to related concepts (AlFresco,
PEACH);
Personalization in Cultural Heritage
17
Collaborative filtering, where the user model is updated with the items
selected by the user (Kubadji);
Content-based filtering, where the user model evolves on the basis of features
of the selected items: e.g., PIL adapted the “bag of words” representation of
the user model applying the Rocchio algorithm (Rocchio, 1971);
Semantic reasoning, based on an ontological representation of the domain,
which is used to reason about the user’s features, interests, etc., by traversing
semantic relations among concepts (CHIP).
In many systems, a user model is initialized and then continuously updated,
based on visitors’ feedback (e.g., AVANTI, GUIDE, Hippie, Gulliver’s Genie,
PIL, CHIP). In PeVEP and PEACH, the stereotype associated with an avatar,
selected when entering the museum, is updated on the basis of the viewing time,
the manipulation of objects, and the rating and comments provided by the user.
In a few cases, such as HIPS and LISTEN, there is no initialization phase and
the user model is exclusively dynamic. In particular, HIPS’s user model was
based only on the visitors’ walking pace and stops that were classified according
to Véron and Levasseur’s (1983) four animal types. Differently, LISTEN started
from a generic user model and used implicit feedback, movements and time spent
in front of artworks for inferring the visitors’ level of interest in concepts,
combined with a stereotypical representation of motion styles describing different
ways of looking at exhibits (e.g., Sauntering and Goal-driven).
4.3. Matching users and content
Personalization occurs dynamically during user interaction. Researchers have
experimented with different techniques to match the current user profile with the
right content at the right time.
Early systems, such as AVANTI and ILEX, used Condition-Action rules to
map user features with content and presentation style (more or less technical,
verbose, etc.), as well as user interface layouts (e.g., font style).
Ranking has also been explored. The user model is used to rate/rank content
to be considered for selection. For example, INTRIGUE, PIL and iCITY
ranked items with respect to the user’s interests.
Personalization in Cultural Heritage
18
Recently, semantic reasoning has been exploited in combination with ranking
in order to broaden the system’s suggestions. For instance, CHIP integrates
the user’s interest ratings with ontological reasoning (by traversing relations),
e.g., to suggest artworks not only by the same artists, but also by related ones,
such as their students and teachers (Wang et al., 2009).
The systems that employed Activation/Inhibition networks for learning the
user model also applied them for selecting the content to be presented, given
the user’s interest levels concerning the concepts of the knowledge base
(AlFresco and PEACH).
Collaborative Filtering was applied in Kubadji to select the next object to
visit, based on the cumulative visitors’ history: i) using past visitors’ paths to
select, out of a set of known paths, the one to be assigned to the current
visitor(s); ii) using the interest of a visitor in an object (represented by the
time spent on it), compared to other visitor’s interest, in order to predict the
next exhibit of interest; and iii) a combination of the two.
Content-based Filtering has been used to select the most suitable items to
recommend, given the user’s ratings. While in Kubadji this approach is
applied to the “bag of words” representation of the items selected by the
visitor, in CHAT it is applied to the “bag of concepts” representation of
items, identified by lexically analyzing their descriptions.
Social recommendation techniques, based on user-generated content (e.g.,
tags and comments) and folksonomies, are being integrated to enrich further
the systems’ capabilities in selecting the most appropriate content for the
user. For instance, iCITY and CHAT exploit user tagging behavior for
recommendation purposes; see Section 5.3.
PIL used the Vector Space model to calculate the cosine similarity between a
vector representing the user’s interest and a vector representing the content to
be presented. The similarity between the user model and the content was used
to rank the presentations for the individual visitor.
Hybrid approaches: many systems employ more than one basic technique to
improve their own matching capabilities. For instance, CRUMPET combined
collaborative filtering and content-based filtering for recommendation
purposes. PEACH applied several modeling approaches: in one, it inferred
Personalization in Cultural Heritage
19
the relevancy of content, based on the visitor’s level of interest in these
concepts (the same domain knowledge base was applied for both user
modeling and natural language generation of presentation). In another case,
potential new interests were implicitly identified by propagating the user’s
interest, based on feedback, over an activation/inhibition network of
presentations (pre-defined presentations were linked with semantic links).
5. Discussion, Trends and Challenges
Historically, personalization of cultural heritage information has stemmed from
natural language generation research on the one hand and adaptive hypermedia on
the other. Earlier linguistic techniques were replaced with “light weight”
techniques, due to their complexity and computational demand, in favor of less
knowledge-intensive solutions supporting the adaptation to the user’s context, and
of the adoption of lightweight natural language generation techniques based on
templates, canned text, and other similar solutions.
The web soon became an experimental setting for personalization technology.
The Web 2.0, the semantic and social web, and more powerful hardware and
network infrastructures made online and onsite services converge into a set of
integrated services to be used before, during and after the visit.
Despite this progress and interesting results, the cultural heritage industry has
yet to adopt personalization in the sense discussed so far. While mobile guides are
now a common sight in cultural heritage settings and social web technology is
spreading fast, personalized services are not. Recently, the cultural heritage sector
has recognized the value of delivering different content and presentation style to
different types of people (Falk, 2009). Moreover, some institutions already
differentiate their offer; e.g., the Tate Gallery offers customized services
respectively tailored to the needs and preferences of kids, parents and teachers
(Jackson and Adamson, 2009). All in all, cultural heritage institutions now seem
willing and ready to adopt personalization and not only to support explorations
and experiments. However, to meet the needs of the cultural heritage industry,
personalization has to address a number of challenges, described in the following.
Personalization in Cultural Heritage
20
5.1 Standards to Reach Critical Mass
Over the years, much effort has been devoted in the cultural heritage domain to
create and conform to standards that would facilitate the storage and exchange of
information, assuring its evolution, extension and preservation for the future. For
instance, the Europeana project11 is aimed at standardizing access to European
online cultural heritage resources. Wikis and blogs are entering the picture,
providing content that can be constantly revised without requiring technical skills
(Dicker, 2010; Stein and Bachta, 2010). This has so far been an untapped resource
which, if properly exploited, could facilitate the adoption of personalization by
cultural heritage institutions, in order to help the visitor navigate through large
amounts of unstructured, dynamic content. However, some work is needed before
these repositories can become the resource for personalization services, e.g., pre-
processing via language technology to create a coherent representation to be used
by mechanisms that generate presentations. One possibility is for personalization
to become a “pathfinder”, e.g., providing personal views or defining individual
navigation paths in the collections, and a “tour provider” tuning the presentation
to the user’s context.
Standardization should extend from content representation to system
architectures and personalization techniques. Up till now, most projects had their
own research agenda, created their own data and developed complete systems that
were later abandoned. A common standard of infrastructure, data structure and
user model modules would ease the implementation and evaluation of novel ideas
and allow researchers to focus on specific research questions while evaluating
them in a common environment. As for architecture, user models should be
separated from the applications, thus enabling external, standard user modeling
components to use and re-use personal data (Kobsa, 2001) or using standard
ontologies such as GUMO (Heckmann et al., 2005) and communicating in
standard protocols such as UserML (Heckmann and Krüger, 2003). Moreover,
standard architectures (such as the multi-agents architecture suggested in PEACH)
may allow cultural heritage application developers to focus on developing their
specific application while re-using existing infrastructure. This would facilitate
11 Europeana the cultural collections of Europe http://www.europeana.eu/portal/index.html
Personalization in Cultural Heritage
21
the development of more effective and robust infrastructures that could ease the
adoption of personalization by cultural institutions.
5.2 Before, During and After the visit – towards a Lifelong Experience
The cultural heritage experience is being viewed as an ongoing lifelong
experience: curators and museum researchers are continuously looking at how
visitors can be captured and retained over time, both online and onsite (Lord,
2007; Wilkening and Chung, 2009; Falk, 2009). As an interaction continuum,
personalization can play a major role by reasoning on past experience and other
daily and contextual characteristics of the visitor, making the current cultural
heritage experience a link in a lifelong chain. This creates a series of challenges
that accompany lifelong user modeling in general: collecting evidence,
remembering and forgetting (as user’s characteristics change), privacy and user
control what to disclose and what to keep private (Kuflik et al., 2010; Toch et
al., 2012).
Furthermore, a persistent user model re-introduces the well known issues of
trust and transparency that have to be seriously addressed and not just merely
acknowledged. While some projects have investigated different aspects of this
issue (e.g., explanation in INTRIGUE and CHIP (Cramer et al., 2008), scrutability
of user models, as suggested by Kay et al. (2005) and implemented in iCITY),
much work has to be carried out in order to reach satisfactory results.
5.3 Not Me but Us: Social Aspects
As reported in many museum research studies, visiting a museum is a social
experience. Individual user models cannot be directly applied to groups because
group members have different interests, constraints and capabilities (including
physical ones). Therefore, the group visiting experience has to be adapted to
facilitate user interaction and socialization during the visit, as well as to satisfy
heterogeneous requirements that might affect suggestions of the artworks to see
and the overall tour organization. However, in only a few cases has the social
aspect of cultural heritage been considered (e.g., INTRIGUE, PIL, Plua and
Jameson (2002)) and it seems that an extensive amount of work is needed to
achieve satisfactory results.
Personalization in Cultural Heritage
22
It should be noted that, in other research areas, such as Computer Supported
Cooperative Work and collaboration support, group modeling has been studied
extensively, in order to understand how user and group interests evolve along
time, whether users are more or less involved in activities, and the like. For
instance, see (Jameson and Smyth, 2007; Vassileva and Sun, 2007; Masthoff,
2004; Masthoff and Gatt, 2006; Ardissono et al., 2010). While group modeling
has substantially different meanings in the management of synchronous activities,
such as the museum visit, or asynchronous ones, such as the participation in a
virtual community, an intersection between these two research fields can be found
as far as modeling group interests and knowledge is concerned. In fact, such
information can be useful to tailor the selection of content in a cultural heritage
site.
On a different perspective, museums and cultural heritage institutions are
actively supporting the sharing of experience among visitors and the publication
of user-generated material in order to enhance their web presence and establish
long-term relationships with people. For instance, the Brooklyn Museum
(Bernstein, 2008) uses Web 2.0 applications, such as Facebook and Twitter, to
support the clustering of user communities; Steve.museum (Trant, 2006) employs
folksonomies and social tagging to support the navigation of vast collections from
a user perspective (instead of the one held by the curator); the New York Public
Library extracts tags from metadata to integrate a traditional search with
folksonomies (Dalton, 2010). However, user-generated content can cause an
explosion of the amount of available information, so that users lose the
“awareness” of what is happening and of the most interesting contributions to
inspect.
Indeed, the increasing presence of institutions on the Web represents both a
challenge and an opportunity for personalization research. Specifically, the active
participation of the general public and user-generated content could be solicited,
as discussed in (Vassileva, 2012), and effectively used. Information garnered as a
result of inviting visitors to cultural heritage web sites to become curators of the
content, to compose and upload their own virtual galleries (e.g., Tate Kids
Collection in Tate Online), and to share their collection, thoughts and views (e.g.,
“remixing exhibits”, in (Fisher et al., 2007)) can be used by personalization
Personalization in Cultural Heritage
23
systems as a source of reliable data. In turn, personalization can dramatically
improve such systems, e.g., by supporting more effective community formation
via personalized match-making and by enhancing the search facilities offered by
virtual collections; e.g., see (Paliouras, 2012) for a discussion about community-
based personalization. CHAT has made a first step in this direction by integrating
metadata information with user tags in order to improve the personalized
suggestion of items (Lops et al., 2009) but there is more space for experimenting
with existing and future personalization techniques.
5.4. Beyond Hypermedia and Multimedia
With a few exceptions (e.g., PeVEP, PEACH, “Carletto” the spider, e(c)ho,
LISTEN), the delivery mode has been based on the metaphor of hypermedia with
content organized in pages and links, followed by multimedia presentations on
mobile devices. As technology evolves, new possibilities emerge. Augmented
reality could improve the visitors’ sense of being present in the museum and their
enjoyment during the visit (Sylaiou et al., 2010), and body interaction can control
the display or non display of artworks (vom Lehn et al., 2007). These new
interaction paradigms have not been personalized so far and open new avenues for
research.
Following the same line of argument, some degree of affective computing has
been employed in cultural heritage applications, for example to engage visitors in
cooperative learning games (Tate Online, ARCHIE (Van Loon et al., 2007);
Yatani et al., 2004). Affective computing and personalization techniques could be
combined to tailor the environments to the user’s visiting style, mood, and goals,
thus creating a unique and engaging experience.
5.4 Evaluation: A Serious Issue
Evaluation of adaptive systems has received much attention in recent years and a
number of issues that could affect the results of the evaluation have been pointed
out, e.g.: lack of objectivity (compare the adaptive system with its static
counterpart not optimized for normal interaction); lack of distinction between
usability and adaptivity issues; incorrect sampling of participants are among
others discussed in (Weibelzahl, 2005), as well as in (Chin and Crosby, 2002).
Personalization in Cultural Heritage
24
Evaluation of adaptive systems in cultural heritage has followed the same path,
and the results reported in the literature, although providing valuable insights,
suffer from the same drawbacks. For example, the results of the evaluation of the
Marble Museum of Carrara do not distinguish between usability problems
(unclear labels for the stereotypes) and adaptivity. Interestingly, participants did
not like the stereotypic representation and were confused by the need to define a
personal profile manually, all usability issues; however, participants liked the
dynamic personalization as it behaved like a human guide. This fact indicates that
hiding features that are difficult to grasp (e.g., personalization) could be a winning
choice. Similarly, PIL user studies showed that users liked the technology, but the
personalization went unnoticed (Kuflik et al., 2011). As pointed out by
Weibelzahl (2005), this should be the desired effect, as a good personalization
should go unnoticed by the user who becomes aware of it only when something
goes wrong. This seems to contrast with the results of the evaluation of CHIP that
extensively investigated transparency and trust by presenting or omitting certainty
values in the interface. Results showed how transparency is appreciated by users
who can understand better what the system is doing for them and are willing to
change their behavior (providing more feedback) to improve the quality of the
recommendations (Cramer et al., 2008).
These evaluations focused on user interaction and perception, neglecting an
essential part, i.e., that of testing and evaluating the underlying technology
independently from the user. A clear inspection of every individual component is
essential (Weibelzahl, 2002). A multi-layer evaluation framework (Karagiannidis
and Sampson, 2000) that clearly separates different phases, from low-level
monitoring of the user behavior (e.g., keystroke input), through the adaptation
decision, to high-level assessment (e.g., the student has not understood the
concept) has been proposed and successfully validated. However, the evaluation
of personalization in cultural heritage presents a further level of complexity if the
system is to be used in a changing context, such as visiting a museum or a city.
Petrelli and Not (2005) discussed this issue and proposed developing an
environment that allows extensive testing of the system before it is taken out of
the lab for a field trial with real visitors in a real museum. One of the core
functions of the suggested development environment is the possibility of
Personalization in Cultural Heritage
25
simulating the different components, e.g., the user model or the user’s
movements. This two-step evaluation, first in the lab and then in the field, was
adopted in HIPS and allowed separating the evaluation of the system’s
effectiveness and efficiency from the quality of the user experience (Marti and
Lanzi, 2001).
In line with this approach, some projects evaluated aspects of the technology
without involving the user, i.e., user behavior was simulated. PEACH (Zancanaro
et al., 2007), Kubadji, and the Museum Wearable projects all used recorded
visiting behavior to evaluate the quality of the result of the adaptive algorithm.
For the same purpose, PIL simulated the user via synthetic models. UbiquiTO is
another example of evaluation without the user, where the impact of
personalization was evaluated by comparing predefined user interests with system
recommendations.
As for all other aspects of personalization research, the system and user
evaluation would benefit from a common effort and the creation of a shared
understanding of which new systems should be evaluated, and how (Weibelzahl,
2002; Paramythis et al., 2010). A shared, structured evaluation framework, as
common in other specialized areas of computer science such as information
retrieval and speech recognition, would support the comparison of different
solutions, the emergence of best practices and the reuse of components, and it
would lead to faster progress in the field.
6. Conclusions
In the early days of personalization in cultural heritage, 20 years ago, novel
technologies were explored. Research focused on information delivery to an
individual visitor, during a single visit online or onsite. Even then, the availability
of content was recognized as a major and unresolved challenge. Recently, the
focus has shifted from individuals engaged in a single visit to: (i) supporting a
more realistic scenario which complies with the fact that most people visit
physical museums in small groups, such as families and classes of schoolmates;
(ii) managing a long lasting interaction between the visitor(s) and the cultural
heritage site by allowing multiple online and onsite visits; (iii) enabling group
collaboration in both physical and online cultural heritage sites, in order to
Personalization in Cultural Heritage
26
support the formation of virtual communities. This shift happened with the
introduction of Web 2.0, the social and semantic web, and the recognition of
personalization as a tool for lifelong modeling of museum visitors.
The evolution and convergence of technologies, together with the needs
expressed by recent museum research, open new opportunities for personalization
research, which has the potential to improve the presentation of information, the
exploration of content interesting for the specific user/group, the collaboration
among users having similar interests, as well as the adaptation to heterogeneous
user contexts and devices. However, such evolution presents new challenges. For
instance, the dramatic increase in available information and material, the growing
interest in supporting socialization and collaboration in small and large-size
communities and the interest in user-generated content, coupled with a need to
guarantee high-quality information standards at low costs, pose somehow
conflicting requirements, which can be hardly met now but could be addressed in
further interdisciplinary research. Concerning personalization, while there is room
for continuing experimentation with new ideas and new technology, the real issue
is to support realistic scenarios – real visitors and users, as individuals and groups
in daily interactions with cultural heritage. It is time to collaborate more closely
with cultural heritage researchers and institutions, putting personalization research
results to work in practice while basing them on and combining them with recent
museum studies. This can be achieved, in part, by applying personalization in
cultural heritage sites, based on the results of museum studies, by approaching the
cultural heritage community at relevant conferences (e.g., Museums and the
Web), as well as by starting mutual projects with realistic practical goals and
long-term trials.
7. References
Abowd, G., Atkeson, C.G., Hong, J., Long, S., Kooper, R. and Pinkerton. M.: 1997, Cyber-guide:
A Context-Aware Tour Guide. Wireless Networks 3(5), 421-433.
Ahiara, K., Yamada, T., Kando, N., Fujisawa, S., Uehara, Y., Baba, T., Nagata, S., Tojo, T. and
Adachi, J.: 2008, Supporting creation and sharing of contents of cultural heritage objects
for educational purposes. In: L. Aroyo, T. Kuflik, O. Stock and M. Zancanaro (eds.):
Personalized Access to Cultural Heritage: Proceedings of AH2008 workshop. Hannover,
Germany, 15-24.
Ardissono, L., Bosio, G., Goy, A., Petrone, G., Segnan, M. and Torretta, F.: 2010, Collaboration
support for activity management in a personal cloud. To appear in International Journal of
Distributed Systems, IGI Global, 2010.
Personalization in Cultural Heritage
27
Ardissono, L., Goy, A., Petrone, G., Segnan, M. and Torasso, P.: 2003, INTRIGUE: Personalized
recommendation of tourist attractions for desktop and handset devices. Applied Artificial
Intelligence 17(8-9), 687-714.
Astudillo, H., Inostroza, P. and Lopez, C.: 2008, Context/SR: a multi-institutional semantic
integration platform. In: J. Trant and D. Bearman (eds.): Museums and the Web:
Proceedings of the International Conference for Culture and Heritage Online 2008.
Montreal, Quebec, Canada Consulted July 20, 2010.
http://www.archimuse.com/mw2008/papers/astudillo/astudillo.html
Bernstein, S.: 2008, Where do we go from here? Continuing with Web 2.0 at the Brooklyn
Museum. In: J. Trant and D. Bearman (eds.): Museums and the Web: Proceedings of the
International Conference for Culture and Heritage Online 2008. Montreal, Quebec, Canada,
Consulted July 20, 2010.
http://www.archimuse.com/mw2008/papers/bernstein/bernstein.html
Berkovsky, S., Kuflik, T. and Ricci, F.: 2008, Mediation of user models for enhanced
personalization in recommender systems. User Modeling and User-Adapted Interaction
18(3), 245-286.
Bianchi, A. and Zancanaro, M.: 1999, Tracking Users' Movements in an Artistic Physical Space.
In: M. Caenepeel, D. Benyon and D. Smoth (eds.): Community of the Future: Proceedings
of the i3 Annual Conference. Siena, Italy, 103-106.
Bohnert, F., Zukerman, I., Berkovsky, S., Baldwin, T. and Sonenberg, L.: Using Interest and
Transition Models to Predict Visitor Locations in Museums. AI Communications 21(2-3),
195202, 2008.
Bohnert, F. and Zukerman, I.: 2009, Using Keyword-Based Approaches to Adaptively Predict
Interest in Museum Exhibits. In: A. Nicholson and X. Li (eds.): Artificial Intelligence:
Proceedings of the 22nd Australasian Joint Conference, Melbourne, VIC, Australia, 656
665.
Billsus, D. and Pazzani, M.: 2007, Content-based recommendation systems: In: P. Brusilovsky and
A. Kobsa and W. Nejdl (eds.): The Adaptive Web. Methods and strategies of web
personalization. Berlin: Springer, 325-341.
Bonis, B., Stamos, J., Vosinakis, S., Andreou, I. and Panayiotopoulod, T.: 2009, A platform for
virtual museums with personalized content. Multimedia tools and applications 42(2), 139-
159.
Calder, J., Melengoglou, A.C., Callaway, C., Not, E., Pianesi, F., Androutsopoulos, I.,
Spyropoulos, C.D., Xydas, G., Kouroupetroglou, G. and Roussou, M.: 2005, Multilingual
Personalized Information Objects. In: O. Stock and M. Zancanaro (eds.): Multimodal
Intelligent Information Presentation. Netherlands. Springer, 177-201.
Callaway, C., Not, E., Novello, A., Rocchi, C., Stock, O. and Zancanaro, M.: 2005, Automatic
cinematography and multilingual NLG for generating video documentaries. Artificial
Intelligence 165(1), 57-89.
Carmagnola, F., Cena, F., Console, L., Cortassa, O., Gena, C., Goy, A., Torre, I., Toso, A. and
Vernero, F.: 2008, Tag-based user modeling for social multi-device adaptive guides. User
Modeling and User-Adapted Interaction 18(5), 497-538.
Cena, F., Console, L., Gena, C., Goy, A., Levi, G., Modeo, S. and Torre, I.: 2006, Integrating
heterogeneous adaptation techniques to build a flexible and usable mobile tourist guide. AI
Communications 19(4), 369-384.
Cheverst, K., Davies, N., Mitchell K. and Smith P.: 2000, Providing Tailored (Context-Aware)
Information to City Visitors. In: P. Brusilovsky, O. Stock and C. Strapparava (eds.):
Adaptive Hypermedia and Adaptive Web-based Systems: Proceedings of the International
Conference, AH 2000, Trento, Italy, 73-85.
Chin, D. and Crosby, M.E.: 2002. Special Issue on Empirical Evaluation of User Models and User
Modeling Systems, User Modeling and User-Adapted Interaction 12.
Ciavarella, C. and Paternò, F.: 2004, The design of a handheld, location-aware guide for indoor
environments. Personal and Ubiquitous Computing 8(2), 82-91.
Cramer, H., Evers, V., Ramlal, S., van Someren, M., Rutledge, L., Stash, N., Aroyo, L. and
Wielinga, B.: 2008, The effects of transparency on trust in and acceptance of a content-
based art recommender. User Modeling and User-Adapted Interaction 18(5), 455-496.
Dalton, J.B.: 2010, Can structured metadata play nice with tagging systems? Parsing new
meanings from classification-based descriptions on Flickr Commons. In: J. Trant and D.
Bearman (eds.): Museums and the Web: Proceedings of the International Conference for
Personalization in Cultural Heritage
28
Culture and Heritage Online 2010. Denver, CO, USA, Consulted June 27, 2010.
http://www.archimuse.com/mw2010/papers/dalton/dalton.html
Damiano, R., Gena, C., Lombardo, V., Nunnari, F. and Pizzo, A.: 2008, A stroll with Carletto:
adaptation in drama-based tours with virtual characters. User Modeling and User-Adapted
Interaction 18(5), 417453.
de Gemmis, M., Lops, P., Semeraro, G., and Basile, P.: 2008, Integrating tags in a semantic
content-based recommender. In: P. Pu, D. Bridge, B. Mobasher, and F. Ricci (eds.):
Recommender Systems: Proceedings of the 2008 ACM Conference. Lausanne,
Switzerland, 163-170.
Dicker, E.: 2010, The impact of blogs and other social media in the life of a curator. In: J. Trant
and D. Bearman (eds.): Museums and the Web: Proceedings of the International
Conference for Culture and Heritage Online 2010. Denver, CO, USA, Consulted June 27,
2010. http://www.archimuse.com/mw2010/papers/dicker/dicker.html
Falk, J: 2009, Identity and the museum visit experience. Walnut Creek, CA, Left Coast Press.
Filippini Fantoni, S.: 2003. Personalization through IT in Museums. Does it really work? The case
of the Marble Museum website. In: Cultural institutions and digital technology:
Proceedings of ICHIM03: Paris. Consulted June 27, 2010.
http://www.archimuse.com/publishing/ichim03/070C.pdf
Fink, J., Kobsa, A. and Nill, A.: 1998, Adaptable and adaptive information provision for all users,
including disabled and elderly people. The New Review of Hypermedia and Multimedia 4,
163-188.
Fink, J. and Kobsa, A.: 2002, User Modeling for personalized city tours. Artificial Intelligence
Review 18, 33-74.
Fisher, M., Kitchen Interactive, N. and Twiss-Garrity, B.A.: 2007, Remixing exhibits: constructing
participatory narratives with on-line tools to augment museum experiences. In: J. Trant and
D. Bearman (eds.): Museums and the Web: Proceedings of the International Conference for
Culture and Heritage Online 2007. San Francisco, CA, Consulted June 27, 2010.
http://www.archimuse.com/mw2007/papers/fisher/fisher.html
Garzotto F., Cinotti T. and Pigozzi M.: 2003, Designing multi-channel web frameworks for
cultural tourism applications: the MUSE case study. In: J. Trant and D. Bearman (eds.):
Museums and the Web: Proceedings of the International Conference for Culture and
Heritage Online 2003. Charlotte, NC, USA. Consulted June 27, 2010.
http://www.archimuse.com/mw2003/papers/garzotto/garzotto.html
Ghiani, G., Leporino, B., Paternò, F. and Santoro, S.: 2008, Exploiting RFIDs and tilt-based
interaction for mobile museum guides accessible to vision-impaired users. In K.
Miesenberger, J. Klaus, W. Zagler and A. Karshmer (eds.): Computers helping people with
special needs: eleventh International Conference, ICCHP 2008. Linz, Austria, 1070-1077.
Hanani, U., Shapira, B. and Shoval, P.: 2001, Information Filtering: Overview of Issues, Research
and Systems. User Modeling and User-Adapted Interaction 11(3), 203-259.
Hatala, M. and Wakkary, R.: 2005, Ontology-based user modeling in an augmented audio reality
system for museums. User Modeling and User-Adapted Interaction 15(3-4), 339-380.
Heckmann, D., Schwartz, T., Brandherm, B., Schmitz, M. and von Wilamowitz-Moellendorff, M.:
2005, GUMO - the General User Model Ontology. In: Proceedings of the 10th International
Conference on User Modeling (UM'2005), Edinburgh, UK, 2005, 428-432.
Heckmann, D. and Krüger, A.: 2003, A User Modeling Markup Language (UserML) for
Ubiquitous Computing. In: P. Brusilovsky, A. Corbett, F. de Rosis (eds.): Proceedings of
the Ninth International Conference on User Modeling, Johnstown, PA, USA. 393-397.
Jackson, S. and Adamson, R.: 2009, Doing it for the kids: Tate online on engaging, entertaining
and (stealthily) educating six to 12-year-olds. In: J. Trant and D. Bearman (eds.): Museums
and the Web 2009: Proceedings of the International Conference for Culture and Heritage
Online 2009. Indianapolis, IA, USA. Consulted June 27, 2010.
http://www.archimuse.com/mw2009/papers/jackson/jackson.html
Jameson, A. and Smyth, B.: 2007, Recommendation to groups. In: P. Brusilovsky and A. Kobsa
and W. Nejdl (eds.): The Adaptive Web. Methods and strategies of web personalization.
Berlin: Springer, 596-627.
Karagiannidis, C. and Sampos, D.: 2000, Layered Evaluation of Adaptive Applications and
Services. In: P. Brusilovsky, O. Stock and C. Strapparava (eds.): Adaptive Hypertext AH
2000: Proceedings, Springer LNCS 1892, 343-346.
Personalization in Cultural Heritage
29
Kay, J., Lum, A., Niu, W.: 2005, A scrutable museum tour guide system. In: A. Butz, C. Kray and
A. Krüger (eds.): Multi-User and Ubiquitous User Interfaces: 2nd Workshop, Proceedings
Jan. 9, 2005 , San Diego, USA 19-20.
Kobsa, A.: 2001, Generic User Modeling Systems. User Modeling and User-Adapted 11(1), 49-63.
Kohrs, A. and Merialdo, B.: 2001, Creating user-adapted Websites by the use of collaborative
filtering. Interacting with Computers 13, 695-716.
Konstan, J.A. and Riedl, J.: Recommender systems: from algorithms to user experience. User
Modeling and User-Adapted Interaction 22 (this issue), 2012.
Krösche, J., Baldzer, J. and Boll, S.: 2004, MobiDENK - Mobile Multimedia in Monument
Conservation. IEEE MultiMedia, 11(2), 72-77.
Kuflik, T., Kay J. and Kummerfeld, B.: 2010, Lifelong Personalized Museum Experiences. In S.
Berkovsky, F. Bohnert, F. Carmagnola, D. Cheng, D. Heckmann, T. Kuflik, P. Nurmi, and
K. Partridge (eds.): Proceedings of Workshop on Pervasive User Modeling and
Personalization (PUMP'10). June 2010, Big Island Hawaii, 9-16.
Kuflik, T., Stock, O., Zancanaro, M., Gorfinkel, A., Jbara, S., Sheidin, J. and Kashtan, N.: 2011, A
Visitor’s Guide in an “Active Museum”: Presentations, Communications, and Reflection.
ACM Journal of Computers and Cultural Heritage 3(3).
Lops, P., de Gemmis, M., Semeraro, G., Musto, C., Narducci, F., Bux, M.: 2009, A Semantic
Content-Based Recommender System Integrating Folksonomies for Personalized Access.
In: G. Castellano, L.C. Jain and A.M. Fanelli (eds.): Web Personalization in Intelligent
Environments, vol. SCI 229: Springer, 27-47.
Lord, B.: 2007, The manual of museum learning. Lanham, MD: AltaMira Press.
Marti, P. and Lanzi, P.: 2001, I enjoyed that this much! Techniques for measuring usability in
leisure-oriented applications. In: J. Bawa, P. Dorazio and L. Trenner (eds.): The Usability
Business: Making the Web Work. Springer, 49-59.
Masthoff, J.: 2004, Group modeling: Selecting a sequence of television items to suit a group of
viewers. User Modeling and User Adapted Interaction, 14(1), 37-85.
Masthoff, J. and Gatt, A.: 2006, In pursuit of satisfaction and the prevention of embarrassment:
affective state in group recommender systems. User Modeling and User-Adapted
Interaction 16(3-4), 281-319.
Oberlander, J., O'Donnell, M., Mellish, C. and Knott, A.: 1998, Conversation in the museum:
experiments in dynamic hypermedia with the intelligent labeling explorer. New Review of
Multimedia and Hypermedia 4, 11-32.
Oberlander, J., Karakatsiotis, G., Isard, A., and Androusotpoulos, I.: 2008, Building an adaptive
museum gallery in Second Life. In: J. Trant and D. Bearman (eds.): Museums and the Web:
Proceedings of the International Conference for Culture and Heritage Online 2008.
Montreal, Quebec, Canada. Consulted June 27, 2010.
http://www.archimuse.com/mw2008/papers/oberlander/oberlander.html
O'Hare, G.M.P. and O'Grady, M.J.: 2003, Gulliver's Genie: A Multi-Agent System for Ubiquitous
and Intelligent Content Delivery. Computer Communications 26(11), 1177-1187.!
Opperman, R.: 1994, Adaptive user support Ergonomic design of manually and automatically
adaptable software. Hillsdale N.J.: Lawrence Erlbaum Associates.
Oppermann, R. and Specht, M. 2000, A Context-Sensitive Nomadic Exhibition Guide. In: P.J.
Thomas and H. Gellersen (eds.): Handheld and Ubiquitous Computing: Proceedings of the
2nd international Symposium. Bristol, UK. Lecture Notes in Computer Science, vol. 1927.
Springer-Verlag, London, 127-142.
Paliouras, G.: Discovery of Web user communities and their role in personalization. User
Modeling and User-Adapted Interaction 22 (this issue), 2012.
Paramythis, A., Weibelzahl, S. and Masthoff, J.: 2010, Layered evaluation of interactive adaptive
systems: framework and formative methods. User Modeling and User-Adapted Interaction
20(5), 383-453.
Paris, C., Wan, S., Wilkinson, R. and Wu, M.: 2001, Generating Personal Travel Guides - And
Who Wants Them?. In: M. Bauer, P.J. Gmytrasiewicz and J. Vassileva (eds.): User
Modeling: Proceedings of the eight conference. Sonthofen, Germany, 251-253.
Paternò, F. and Mancini, C.: 1999, Designing Web User Interfaces Adaptable to Different Types
of Use. In: J. Trant and D. Bearman (eds.): Museums and the Web 1999: Proceedings of the
International Conference for Culture and Heritage Online 1999, Pittsburgh, USA.
Consulted June 27, 2010. http://www.archimuse.com/mw99/papers/paterno/paterno.html
Personalization in Cultural Heritage
30
Petrelli, D., De Angeli, A. and Convertino, G.: 1999, A User-Centered Approach to User
Modeling. In: J. Kay (ed.): User Modeling: Proceedings of the 7th International Conference
UM’99, 255-264.
Petrelli, D. and Not, E.: 2005, User-Centred Design of Flexible Hypermedia for a Mobile Guide:
Reflections on the HyperAudio Experience. User Modeling and User-Adapted Interaction
15(3-4), 303-338.
Plua, C. and Jameson, A.: 2002, Collaborative preference elicitation in a group travel
recommender system In: F. Ricci and B. Smyth (eds.): Recommendation and
personalization in eCommerce: Proceedings of AH 2002 workshop. Malaga, Spain, 148-
154.
Rocchio, J.J.: 1971, Performance indices for document retrieval. In G. Salton (ed.): The SMART
retrieval system––experiments in automatic documents processing 5767. NJ: Englewood.
Ruotsalo, T., Mälekä, E., Kauppinen, T., Hyvönen, E., Haav, K., Rantale, V., Frosterus, M.,
Dokoohaki, N. and Matskin, M.: 2009, Smartmuseum: Personalized Context-aware Access
to Digital Cultural Heritage. In: Advances in Semantic Processing: Third International
Conference Proceedings of International Conference on Digital Libraries and the Semantic
Web 2009 (ICSD2009), Trento, Italy.
Schmidt-Belz, B., Laamanen, H., Poslad, S. and Zipf, A.: 2003, Location-based mobile tourist
services - first user experiences. In A.J. Frew, M. Hitz, and P. O'Connor (eds.): Information
Technology and Travel & Tourism. Proceedings of the tenth International Conference,
Helsinki, Finland. 115-123.
Schmitz, P.L. and Black, M.T.: 2008, The Delphi toolkit: enabling semantic search for museum
collections. In: J. Trant and D. Bearman (eds.): Museums and the Web 2008: Proceedings
of the International Conference for Culture and Heritage Online 2008. Montreal, Quebec,
Canada, Consulted June 27, 2010.
http://www.archimuse.com/mw2008/papers/schmitz/schmitz.html.
Sparacino, F.: 2002, The museum wearable: Real-time sensor-driven understanding of visitors'
interests for personalized visually-augmented museum experiences. In: J. Trant and D.
Bearman (eds.): Museums and the Web 2002: Proceedings of the International Conference
for Culture and Heritage Online 2002, Boston, MA. Consulted June 27, 2010.
http://www.archimuse.com/mw2002/papers/sparacino/sparacino.html
Stein, R., Bachta, E.: 2010, Breaking the bottleneck: using pseudo-wikis to enable rich web
authoring for non-technical staff members. In: J. Trant and D. Bearman (eds.): Museums
and the Web 2010: Proceedings of the International Conference for Culture and Heritage
Online 2010. Denver, Colorado. Consulted June 27, 2010.
http://www.archimuse.com/mw2010/papers/stein-bachta/stein-bachta.html.
Stock, O., Carenini, G., Cecconi, F., Franconi, E., Lavelli, A., Magnini, B., Pianesi, F., Ponzi, M.,
Samek-Lodovici, V. and Strapparava, C.: 1993, Alfresco: enjoying the combination of NLP
and hypermedia for information exploration. In: Maybury, M.T. (ed.): Intelligent
Multimedia Interfaces, 197224. Menlo Park, CA. AAAI Press.
Stock, O., Zancanaro, M., Busetta, P., Callaway, C., Krüger, A., Kruppa, M., Kuflik, T., Not, E.
and Rocchi, C.: 2007, Adaptive, Intelligent Presentation of Information for the Museum
Visitor in PEACH. User modeling and User-Adapted Interaction 17(3), 257-304.
Stock, O. and Callaway, C.: 2009, Multiple Coordinated Mobile Narratives as a Catalyst for Face-
to-Face Group Conversation. In I.A. Iurgel, N. Zagalo and P. Petta (eds.): Interactive
Digital Storytelling: Proceedings of the 2nd Joint international Conference. Guimarães,
Portugal, 209-220.
Sumi, Y. and Mase, K.: 2001, AgentSalon: facilitating face-to-face knowledge exchange through
conversations among personal agents. In: E. André, S. Sen, C. Frasson and J. Müller (eds.):
Autonomous Agents. Proceedings of the fifth international Conference. Montreal, Quebec,
Canada, 393-400.
Sylaiou, S., Mania, K., Karoulis, A. and White., M.: 2010, Exploring the relationship between
presence and enjoyment in a virtual museum. International Journal of Human-Computer
Studies 68, 243-253.
Tallon, L. and Walker, K. (eds.): 2008, Digital Technologies and the Museum Experience:
Handheld Guides and Other Media. Lanham, MD: AltaMira Press.
Toch, E., Wang, Y. and Cranor, L.F.: Personalization and privacy: a survey of privacy risks and
remedies in personalization-based systems. User Modeling and User-Adapted Interaction
22 (this issue), 2012.
Personalization in Cultural Heritage
31
Trant, J.: 2006, Exploring the potential for social tagging and folksonomy in art museums: proof of
concept. New Review of Hypermedia and Multimedia 12(1), 83-105.
Van Loon, H., Gabriels, K., Luyten, K., Teunkens, D., Roberto, K., Coninx, K. and Manshoven,
E.: 2007, Supporting social interaction: a collaborative trading game on PDA. In: J. Trant
and D. Bearman (eds.): Museums and the Web 2007: Proceedings of the International
Conference for Culture and Heritage Online 2007. San Francisco, CA, Consulted June 27,
2010. http://www.archimuse.com/mw2007/papers/vanLoon/vanLoon.html
Vassileva, J.: Motivating participation in social computing applications: a user modeling
perspective. User Modeling and User-Adapted Interaction 22 (this issue), 2012.
Vassileva, J. and Sun, L.: 2007, Using community visualization to stimulate participation in online
communities. e-Service Journal 6(1), 3-39.
Véron, E. and Levasseur, M.: 1983, Ethnographie de l’exposition: l’espace, le corps, le sens, Paris,
Bibliothèque publique d’Information, Centre Georges Pompidou.
Vom Lehn, D., Hindmarsh, J., Luff, P. and Heath, C.: 2007, Engaging Constable: Revealing Art
with New Technology, Proceedings of CHI 2007, San Jose, California, USA, 1485-1494.
Wang, Y., Stash, N., Aroyo, L., Gorgels, P., Rutledge, L. and Schreiber, G.: 2008,
Recommendations based on semantically enriched museum collections. Journal of Web
Semantics, 6(4), 283-290.
Wang, Y., Stash, N., Aroyo, L., Hollink, L. and Schreiber, G.: 2009, Semantic relations in content-
based recommender systems. In: Y. Gil and N. Fridman Noy (eds.): Knowledge Capture:
Proceedings of the fifth International Conference. Redondo Beach, CA, 209-210.
Weibelzahl, S.: 2002, Evaluation of Adaptive Systems, Ph.D Thesis, University of Freiburg.
Weibelzahl, S.: 2005, Problems and Pitfalls in Evaluating Adaptive Systems. In: S. Y. Chen and
G. D. Magoulas (eds.): Adaptable and Adaptive Hypermedia Systems. Hershey, PA: IRM
Press, 285-299.
Wilkening, S. and Chung, J.: 2009, Life stages of the museum visitor: Building engagement over a
lifetime. Washington DC: American Association of Museums Press.
Yatani, K., Onuma, M., Sugimoto, M. and Kusunoki, F.: 2004, Musex: A system for supporting
children's collaborative learning in a museum with PDAs. Systems and Computers in Japan
35(14), 54-63.
Zancanaro, M., Kuflik, T., Boger, Z., Goren-Bar, D., Goldwasser, D.: 2007, Analyzing museum
visitors’ behavior patterns. In: C. Conati, K.F. McCoy, and G. Paliouras (eds.): User
Modeling. Lecture Notes in Computer Science, vol. 4511, 238246.
Zimmermann, A. and Lorenz, A.: 2008, LISTEN: a user-adaptive audio-augmented museum
guide. User Modeling and User-Adapted Interaction 18(5), 389-416.
Author biographies
Liliana Ardissono is an Associate Professor at the Dipartimento di Informatica of
the Università di Torino, where she obtained her university degree and her Ph. D.
in Computer Science. Her research interests include user modeling, adaptive
hypermedia and cloud computing. She has investigated the development and
exploitation of personalization techniques for various domains, such as
recommendation, awareness support and presentation of Cultural Heritage
information. She is the Secretary of the Board of Directors of User Modeling Inc.
and she is a member of the Editorial Board of User Modeling and User-Adapted
Interaction – The Journal of Personalization Research.
Tsvika Kuflik is a senior lecturer of Information Systems at the University of
Haifa, where he leads a group that focuses on research on Ubiquitous User
Modeling and on personalization and Intelligent User Interface for Cultural
Heritage. Tsvi received his B.Sc. and M.Sc. in Computer Science and Ph. D. in
Information Systems from Ben-Gurion University of the Negev. He has worked
Personalization in Cultural Heritage
32
on personalization and intelligent user interfaces for cultural heritage over the past
ten years. During this period he spent a year in FBK/irst in Trento and half a year
at the University of Sydney working on these areas. In addition he worked on
Information Filtering and Decision Support Systems applications for Agriculture
and Occupational Therapy. He has authored over a hundred technical papers and
has edited several books.
Daniela Petrelli
... The desire to offer differentiated experiences to diverse audiences (Falk, 2009;Falk and Dierking, 2012) has fueled a growth of interest in personalization (Stock, 1993;Oberlander et al., 1998;Paterno, 1999;Bowen and Filippini Fantoni, 2004), including projects that have explored how to design personalized experiences and exhibitions in museum contexts (Kuflik et al., 2011;van Tuijn et al., 2016;Lee and Paddon, 2017;Kontiza et al., 2018). Much research has focused on user modeling and recommender systems which draw on data about users' backgrounds, interests, or behaviors (Ardissono et al., 2012;Fishwick, 2016;De Angelis et al., 2017;Deladiennee and Naudet, 2017;Almeshari et al., 2019;Castagnos et al., 2019;Katifori et al., 2019;Mauro, 2019;Mokatren et al., 2019;Dahroug et al., 2021). While personalization tends to be approached in terms of matching users with relevant content, it can also focus on making an experience feel more personal by developing a personal connection between the visitor and the museum. ...
Article
Full-text available
The capture and analysis of diverse data is widely recognized as being vital to the design of new products and services across the digital economy. We focus on its use to inspire the co-design of visitor experiences in museums as a distinctive case that reveals opportunities and challenges for the use of personal data. We present a portfolio of data-inspired visiting experiences that emerged from a 3-year Research Through Design process. These include the overlay of virtual models on physical exhibits, a smartphone app for creating personalized tours as gifts, visualizations of emotional responses to exhibits, and the data-driven use of ideation cards. We reflect across our portfolio to articulate the diverse ways in which data can inspire design through the use of ambiguity, visualization, and inter-personalization; how data inspire co-design through the process of co-ideation, co-creation, and co-interpretation; and how its use must negotiate the challenges of privacy, ownership, and transparency. By adopting a human perspective on data, we are able to chart out the complex and rich information that can inform design activities and contribute to datasets that can drive creativity support systems.
... The results showed that the visitors performed significantly better on knowledge acquisition and retention tests related to augmented exhibits than to non-augmented exhibits and that they perceived AR as a valuable and desirable add-on for museum exhibitions. [28] A survey of studies carried out on the personalization of cultural heritage information is carried out, including the evolution of personalization techniques on websites, virtual collections and mobile guides, to the recent extension of cultural heritage to the semantic and social web. [29] A study on science learning in informal environments and through the use of digital technologies is presented. ...
Chapter
Full-text available
This article starts with a short review of recent studies and reports on the adaptation of museums toward digital transformation. This first part aims at giving a contextualization of the current importance of the use of technologies, and an understanding of which are the current strategies and struggles museums face regarding their use. In correlation to this context, it becomes relevant to analyze the scientific literature and understand how it reflects the evolution of the adoption of technologies, such as the themes under investigation, inferring trends, among others. Therefore, the second part explores the scientific production related to the use of technologies in museums, in the last decade, by using a bibliometric analysis approach. This study allows us to understand the evolution of the scientific literature on the adoption of technologies in museums, and proposes an agenda for future research that offers the potential to advance research on the use of new digital technologies in museums.
... Despite their pioneering and remarkable work in autonomous navigation and human-robot interaction, these approach do not realize a real Personalization [9] of the user experience. Recent works [4,5,77] try to implement a much more effective and personalized user experience by designing robots that are able to exploit their perceptive and decision-making skills in order to establish a much more deep interactions with visitors However, these approaches do not take into account the mental states of the interacting user; therefore, they do not achieve a real personalization [78] based on complex models. ...
Article
Full-text available
A big challenge in human–robot interaction (HRI) is the design of autonomous robots that collaborate effectively with humans, exposing behaviors similar to those exhibited by humans when they interact with each other. Indeed, robots are part of daily life in multiple environments (i.e., cultural heritage sites, hospitals, offices, touristic scenarios and so on). In these contexts, robots have to coexist and interact with a wide spectrum of users not necessarily able or willing to adapt their interaction level to the kind requested by a machine: the users need to deal with artificial systems whose behaviors must be adapted as much as possible to the goals/needs of the users themselves, or more in general, to their mental states (beliefs, goals, plans and so on). In this paper, we introduce a cognitive architecture for adaptive and transparent human–robot interaction. The architecture allows a social robot to dynamically adjust its level of collaborative autonomy by restricting or expanding a delegated task on the basis of several context factors such as the mental states attributed to the human users involved in the interaction. This collaboration has to be based on different cognitive capabilities of the robot, i.e., the ability to build a user’s profile, to have a Theory of Mind of the user in terms of mental states attribution, to build a complex model of the context, intended both as a set of physical constraints and constraints due to the presence of other agents, with their own mental states. Based on the defined cognitive architecture and on the model of task delegation theorized by Castelfranchi and Falcone, the robot’s behavior is explainable by considering the abilities to attribute specific mental states to the user, the context in which it operates and its attitudes in adapting the level of autonomy to the user’s mental states and the context itself. The architecture has been implemented by exploiting the well known agent-oriented programming framework Jason. We provide the results of an HRI pilot study in which we recruited 26 real participants that have interacted with the humanoid robot Nao, widely used in HRI scenarios. The robot played the role of a museum assistant with the main goal to provide the user the most suitable museum exhibition to visit.
... Personalization in cultural heritage has been a research field for decades [39]. In [40] present an interactive VR web application that allows users to create their own custom virtual art galleries. ...
... As a result, only tourism products that place a strong emphasis on cultural originality may really satisfy the psychological and emotional needs of contemporary travellers. Ardissono et al. adopted the virtual reality development technology Unity3D and produced the Nanyin cultural interactive tourism system, which enhanced the viewer's sense of immersion and experience [12]. Wang et al. used Arc GIS Engine technology to realize the cultural tourism system of colleges and universities, showing the cultural characteristics of the campus [13]. ...
Article
Full-text available
Tourism is evolving constantly as a result of the growth of the creative economy. Currently, ethnic cultural tourism is the focus of cultural tourism’s overall development, while cultural tourism’s development is primarily driven by considerations of social economy, cultural protection, and cultural inheritance. The first section of the article defines cultural and creative tourism. Second, this essay examines the evolution of creative and cultural tourism in general. In the end, cultural tourism’s overall development is unquestionably superior to other development, and the other two commonly receive poor ratings. Its overall development score ranges from 4.4 to 4.9, and its average value without overall development is 4.42. The highest score is 4.89, with the average for cultural tourism with limited development being 4.49 and the average for cultural tourism with full development being 4.71. In a nutshell, the paper’s proposal for evaluating the entire development of cultural tourism is doable.
Thesis
Full-text available
Article
Full-text available
The study of museums is one of the most important qualitative studies dealt with by many researchers in societies. It aims at measuring the service quality of virtual museums from the viewpoint of visitors to museums and the impact on visitor's satisfaction. The study uses a set of dimensions to measure the quality of museums by default: website, interaction, reliability, privacy, safety, effectiveness, empathy, response. The study used a questionnaire, which was distributed to (375) visitors, the study concluded that the quality of museum services affects visitor satisfaction, and study showed the importance of visitors' trust in museum technology, and its role in encouraging visitors to visit the virtual key words: museums, E-services quality, trust, visitor's satisfaction, Sharjah.
Article
Full-text available
تمثل دراسة المتاحف واحدة من اهم الدراسات النوعية التي يتناولها العديد من الباحثين في المجتمعات، ولقد هدفت دراستنا الى قياس جودة الخدمات المتحفية الافتراضية من وجهة النظر الزوار للمتاحف افتراضيا ، وبيان مدى تاثير ذلك على الرضا العام لهم، ولهذا الغرض تبنت الدراسة مجموعه من الابعاد لقياس جودة المتاحف افتراضيا وهي: واجهة الموقع الإلكتروني، التفاعل، الاعتمادية، الخصوصية، الأمان، الفاعلية، ﺍﻟﺘﻌﺎﻁﻑ، الاستجابة، واستخدمت الدراسة لقياس ذلك اسلوب الاستمارة ، والتي وزعت على زوار المتحف افتراضيا عبر تحويل الاستمارة الى رابط لهذا الغرض، وبلغ مجموع العينة الداخلة في الدراسة (375) زائر لمجموعه من المتاحف العاملة في امارة الشارقة ، وتوصلت الدراسة الى ان جودة الخدمات المتحفية المطبقة الكترونيا تؤثر بشكل متفاوت في رضا الزوار، فضلا عن الدور الذي تلعبه ثقة الزوار بالتكنولوجيا المطبقة في المتاحف والخاصة بالمنصات الافتراضية المرفوع عليها برامج ومحتويات المتحف.
Chapter
Full-text available
Empirical studies with adaptive systems offer many advantages and opportunities. Nevertheless, there is still a lack of evaluation studies. This chapter lists several problems and pitfalls that arise when evaluating an adaptive system, and provides guidelines and recommendations for workarounds or even avoidance of these problems. Among other things the following issues are covered: relating evaluation studies to the development cycle; saving resources; specifying control conditions, sample, and criteria; asking users for adaptivity effects; reporting results. An overview of existing evaluation frameworks shows which of these problems have been addressed and in which way.
Article
Full-text available
In the classroom, textbooks and some supplemental materi-als are usually used for studying culture and history as well as language arts and science. Even though such materials are well-organized accord-ing to their corresponding curricula, they lack enough information for pupils to appreciate cultural objects and get more familiar with them. We have been carried out a research project that we call CEAX, which aims to reveal a methodology for establishing a framework for managing various kinds of information on cultural heritage objects and how to use it for educational purposes. This paper describes a system for supporting the creation and sharing of adaptive descriptions about cultural heritage objects, which is a part of the CEAX framework. The primary target users of this system include both experts, such as curators, and educators, such as school teachers. This paper also discusses the features of our framework for personalized access.
Article
Full-text available
Museums offer vast amounts of information, but a visitor's receptivity and time are typically limited, providing the visitor with the challenge of selecting the (subjectively) interesting exhibits to view within the available time. Mobile, electronic handheld guides offer the opportunity to improve a visitor's experience by recommending exhibits of interest, and adapting the delivered content. The first step in this personalisation process is the prediction of a visitor's activities and interests. In this paper we study non-intrusive, adaptive user modelling techniques that take into account the physical constraints of the exhibition layout. We present two collaborative models for predicting a visitor's next locations in a museum, and an ensemble model that combines the predictions of these models. The three models were trained and tested on a small dataset of museum visits. Our results are encouraging, with the ensemble model yielding the best performance overall.
Chapter
Usability evaluation traditionally places a special emphasis on the need to develop software that is usable, Le. used by “specified users to achieve specified goals with effectiveness, efficiency and satisfaction in a specified context of use” (ISO 9241, part 11). But what exactly does it mean to “achieve a goal” in the context of software applications for art, entertainment and leisure? Most activities in these contexts are not guided by a predefined objective based on user intention, instead, the leisure experience is a case in which individuals frequently “adjust” the way they interact with the environment, depending on many different and context-dependent factors. This chapter describes an approach to leisure-oriented usability evaluation developed around a handheld electronic museum tour guide, part of the Hyper Interaction within Physical Space (HIPS) project. It suggests that emotion, culture and context are important influences on perceived usability, and suggests how to apply these to usability away from the work environment.
Conference Paper
Ubiquitous computing offers new chances and challenges to the field of user modeling. With the markup language UserML, we try to contribute a platform for the communication about partial user models in a ubiquitous computing environment, where all different kinds of systems work together to satisfy the user’s needs. We also present an implementation architecture of a general user model editor which is based on UserML. The keywords are ubiquitous computing, distributed user modeling and markup languages.
Chapter
This chapter discusses content-based recommendation systems, i.e., systems that recommend an item to a user based upon a description of the item and a profile of the user's interests. Content-based recommendation systems may be used in a variety of domains ranging from recommending web pages, news articles, restaurants, television programs, and items for sale. Although the details of various systems differ, content-based recommendation systems share in common a means for describing the items that may be recommended, a means for creating a profile of the user that describes the types of items the user likes, and a means of comparing items to the user profile to determine what to re commend. The profile is often created and updated automatically in response to feedback on the desirability of items that have been presented to the user.
Article
ARCHIE is a research project in which the educational staff of the Gallo-Roman Museum collaborates with the Human-Computer Interaction research group of the Expertise Centre for Digital Media (Hasselt University) in the context of expansion of the museum. Starting point of this interdisciplinary collaboration is our strong belief that handheld guides are a promising medium to enhance the visitor's learning experience in a museum. Recently, mobile devices are becoming a common aid to support a museum visit, though the first PDA-based applications also revealed some important shortcomings: the device demands lots of attention, tends to displace the surrounding objects and generates the unintended side effect that it's a quite individual, isolated experience. However, many studies have pointed out that interactions with the exhibit, as well as communication and social interaction between visitors are key points of a successful learning environment. Designing for interaction asks for a mental switch. Main reason why current solutions fail is that the hardware and content are designed and structured for retrieval by one person rather than by multiple persons. We want to deal with the above mentioned unintended side effects and explore the possibilities an interactive, mobile and networked museum guide can offer to (groups of) visitors: a greater versatility for visitors to tailor information to their needs and interests (personalization), to discover the exhibits at their own pace (localization) and to communicate and interact with family or group members (social interaction). Three services make up the core of the system we are developing: a group communication facility, automatic personalization and WiFi-based localization. Collaborative context-aware applications can be built using these services of the framework. A collaborative trading game for (school) groups of children (aged 10-14 years) is the first application we build using these services to verify the usefulness of the services and test whether it supports our goals. The trading game concerns the introduction of social differentiation in West European society (round 825 BC). There are 4 players in the game who participate in the collaborative application through a PDA and who have different roles, property, exchangeable goods and goals. We designed the museum game in this way that every player is dependent on the concrete actions of other players; only through social interaction and cooperation can they come to a good end. Every player has his own character (avatar) and all participants can communicate directly to each other using their PDAs and Voice-over-IP technology. They can use this to negotiate on trade proposals and to help each other solving questions. Indirect communication between participants is carried out by the trade actions, which are necessary to accomplish the game. In this paper we present the ARCHIE project, its objectives and framework overview, make clear why we want to stress on social interaction and give some important design sensitivities that should be taken into account while designing for interaction. We present a first application, a collaborative trading game for (school)groups of children, from conceptual stage towards final implementation and conclude with user test results.
Article
nd different media, and on the high-level integration of the different communication modalities. The role of language seems to be central even here but in a novel way (see also [Stock, 1992]). All these different motivations share the view that the so called teletype approach is too narrow for NLP applications and envision a larger bandwidth of communication. In recent years there have been various projects sharing all or part of the above motivations and the chapters in the present volume give an excellent overview of the problems in designing a multimedial integrated system and of the proposed solutions. A central problem concerns the integration, in the system output, of text generation with forms of graphics. For instance the combination of analogical images and text has been developed in several systems, notably in Comet [Feiner and McKeown, this volume], where the role of a coordinator that has to negotiate with the different mode specialists is emphasized. Particularly interesti
Conference Paper
This paper presents a system called AgentSalon, which facilitates face-to-face knowledge exchange and discussion by people having shared interests, in museums, schools, offices, academic conferences, etc. This system was designed as a sub-system of our ongoing project to construct a personal agent system for tour guidance and knowledge sharing among users. AgentSalon has a big screen for two to five users. The screen shows conversations among animated agents belonging to the users. The personal agent usually runs on our mobile guidance system, called PalmGuide, and guides its user according to his/her individual interests and touring records. When users connect their PalmGuides with AgentSalon by infrared, their personal agents with their personal information migrate to AgentSalon and engage in automated conversation. Contents of the conversation include opinion exchange about tour experiences, mutual recommendations of exhibits on behalf of the users, etc. By observing a chat of the agents, users can effectively obtain appropriate topics: that is, tempts them to follow the chat. This paper shows the first prototype of AgentSalon provided to participants in an academic conference.