October 24-26, 2007
Toronto, Ontario, Canada

Paper: Integrating Data Management and Panoramic Images for an Effective Representation of Cultural Heritage Information

Stefano Valtolina, Dipartimento di Informatica e Comunicazione, Università degli Studi di Milano, Italy; and Elisa Bertino, CERIAS and CS Department, Purdue University, USA


The effective presentation of cultural heritage (CH) information requires the application of sophisticated techniques from different areas; namely, human computer interaction (HCI) and knowledge management. CH is no longer perceived as just a set of isolated objects, stored in museums or collections, without any links to the specific historical and social contexts of reference. Rather, objects have to be connected to additional information concerning those reference contexts. Since such information evolves over time, based on the research carried out by archaeologists, historical experts, and sociologists, the knowledge base needs to be updated to reflect new discoveries. Moreover, information should be presented at different levels of detail and should support an exploratory approach to information discovery by users. This paper presents a system addressing such requirements. The system uses 360 panoramic images to disseminate CH information retrieved from an open knowledge network, defined as a set of interrelated knowledge sources. In particular the paper explores the integration of panoramic images with ontologies in the context of an application in the area of archaeology (specifically, the Etruscan civilization). The paper also discusses the relevant technical issues for the design of the panoramic image interactive system. Those issues concern on one hand the possible organizations for the domain knowledge and on the other hand the architecture of the system and its interaction patterns. Such interactions are crucial in order to facilitate the development of applications that are easily customizable users and are characterized by a high level of interactivity.

Keywords: cultural knowledge base, panoramic image, data integration, interactive museum tools, narrations


The rapid development of the Information and Communication Technologies (ICT) is deeply influencing today’s society. In particular, the use such new technologies is of increasing interest for the field of cultural heritage (CH) dissemination and valorization. CH is a knowledge intensive domain that requires information interaction techniques in order to address the complexity of the domain’s knowledge base. In real cultural exhibitions, one of the most difficult goals that every curator has to achieve is to make the visitor fully grasp the role of an exhibit in the context of the culture the exhibit belongs to. Achieving such goal would make a visit to the exhibition much more satisfying from a cultural point of view. However, several organizational obstacles must be dealt with. For example in an archaeological context, while it is simple to present the finding’s analytical data, it is much more complex to communicate the functions of the finding in its original context of use. Communicating such contextual information requires to reconstruct the relations between the finding itself and all the elements characterizing its original environment. However, a museum exhibition usually has a static structure, which can be hardly modified after it has been set up. The structure is dictated by the information that the organizer wants to convey, that is, the main theme under which the displayed collection should be read by the viewer. That means that the theme of the exhibition deeply affects the arrangements of the collections. Therefore, shifting the focus of an exhibition requires re-arranging the exhibits, even though the involved findings are the same. Moreover the flexibility need of an exhibition is dictated by the fact that the CH is not a static subject, but evolves over the time through the discovery of new documents and the development of new theses by CH scholars.

For these reasons CH cannot be perceived anymore as a set of single objects, storable in museums, without any link to their specific historical and social context. . Information technology’s instruments and techniques can prove useful in helping cultural institution in disseminating their content. At the present, the most common ICT applications to CH are Web sites allowing users to obtain information about museums and their collections or, in more sophisticate applications, to browse among the museum exhibits through a virtual environment. However, in our opinion, the real added value of a CH application is represented by the enhanced flexibility making possible for users to customize their visits to the exhibitions, according to their own interest. We thus need innovative applications, tools and approaches to the management and presentation of CH information.

We present such an approach. The approach has been developed as part of the T.Arc.H.N.A. project (Towards Archaeological Heritage New Accessibility). T.Arc.H.N.A. is a European project in which Archaeologists and Computer Science experts collaborated in order to develop an innovative system for enhancing accessibility to the CH both in the field of education and dissemination. By closely following the discovery process carried out by the archaeology experts, the system is able to offer a rich contextualization of the collections in order to perceive each finding no more as a single object missing its links with the specific historical, anthropological and social context that produced it. Therefore, CH value widens in space and time and it is under this particular point of view that it must be disseminated to the great public.

The need to provide a customized visit able to cover the whole aspects of a CH of interest creates an opportunity for developing products and services with a new interaction paradigm. In particular, we propose an interaction method based on the use of panoramic images, with the intent of integrating, enhancing and visualizing information in a form that responds to the fruition context of the final user. The result is the development of a “virtual museum wing” displayed through Web applications or totems in real museums that complements the display of the artefacts, physically present in the museums, by presenting additional knowledge and “virtual artefacts”.

We describe in detail a virtual wing aimed at the dissemination of the archaeological information concerning the collections owned by the Etruscan National Museum of Tarquinia. The use of panoramic images appears to be a promising technology to integrate and relate artefacts placed in the Tarquinia museum with other information coming from databases belonging to several Etruscan cultural institutions. The integration among the various information sources is achieved by means of an ontology describing the relevant concepts of the archaeological domain.

The rest of the paper is organized as follows: Section 2 overviews the state of the arts in the field of the cultural dissemination. Section 3 presents our approach to an integrated access to the CH. Section 4 presents a network of interactive museums semantically interconnected by means of panoramic images. Section 5 provides some implementation details concerning the developed application. Finally Section 6 outlines some conclusions and future work.

Cultural Dissemination

In the cultural field it is common practice to build applications expressly developed for a museum or a gallery, which are often little more than an on-line brochure with the addition of a browseable collection (Van Welie & Klaassen, 2001). These applications often exist in isolation, even when they share common heritages, for example paints, sculptures or other assets made by the same artists or belonging to the same artistic movement. 

Panoramas are used in many domains such as entertainment, e-commerce and e-science, to show events, estates, furniture, cars, jewelry, astronomic images, and so forth.  However, the domain in which panoramic images have been mostly used is by far is CH (Zheng et al, 2003)(Pan et al, 2004; Horry et al, 1997; Nakano et al, 2004): city tours, relevant architectures, museums, caves, and beautiful landscapes are available on the Web as panoramic images. To give you an idea of the phenomenon, consider that there are portals such as http://www.fullscreenqtvr.com/, http://www.vrway.com/, and http://www.arounder.com/ dedicated to such technology from which users can access thousands of panoramic images from all around the world. Among other technologies, an important class is represented by Virtual Reality (VR) (Mosaker, 2001)(Papagiannakis et al, 2002). VR techniques allow one to reconstruct and render complex three-dimensional models; however, VR applications require in most cases specialized hardware or accessories such as high-end graphic workstations, stereo displays, gloves or 3D goggles. Compared with such VR techniques, panoramic images only require a browser and an Internet connection. Besides, through the use of 360 panoramas it is possible to offer interactions based on high-quality photographic representations, and not only on computer rendered ones.

Besides passive visualization, it is possible to experiment new active forms of interaction (Hoeben & Stappers, 2006) through the use of use hotspots. When a user clicks on an active area of the panorama, referred to as hotspot, a wide variety of actions can be triggered, including the presentation of a photograph or text, an audio recording, or another panoramic scene. Unfortunately, the way panoramas are supported suffers from the limitation of being simple views of a landscape. In our opinion, panoramas could become a new interaction medium. The idea is to follow the approach of Web sites such as Flickr (http://www.flickr.com), and Bubbleshare (http://www.bubbleshare.com/upload) that offer users the possibility to store and share their own photos, possibly enriched with additional multimedia content. For example Bubbleshare allows its users to publish photographic albums integrated with textual captions, audios comments, and notes of other visitors. By adding such content, the user is able to narrate the story behind each image, sharing not only a simple piece of data, but also his experiences and ideas.

Panoramic Images For An Integrated Access To Cultural Heritage

Panoramic images have not been widely adopted as a technology for information dissemination.  In our opinion, this is due to the fact that panoramic images are not integrated with the rest of the information typically available at Web sites; panoramas are often only an eye-caching Web component representing some object, such as a monument or a landscape, usually described by a short text. When interaction is supported, it is limited to the hotspots which connect another panorama or link another page with some additional information of the selected region. In order for panoramic images to become an appealing technology for CH information dissemination on Web, we believe that one of the main steps is its integration with database (DB) technologies (Valtolina et al, 2006). Such integration would allow the integration of dynamic elements in a panoramic image. To understand the advantages of such integration, one should consider the evolution Web technologies have undergone since scripting languages (like CGI) were integrated with DB technologies through the use of the OBDC protocol.  In addition, in the case of CH solutions - but this applies to many other application domains - museums already have databases containing digital information about the elements represented by the panoramic images. In this context, panoramic imaging technology should be used to create environments that, by hiding the complexity of the underlying DB's and archives, are more natural to users and easier to use when interacting with the CH information represented in the panoramas. To address such integration, we have introduced a new notion of hotspots that we refer to as semantic hotspots. Such hotspots make use of a semantic model to dynamically link relevant regions of a panorama with the information stored in one (or several) data sources.

Another major limitation of panoramas today accessible from Web is the fact that it is not possible to customize the visit according to the user’s interests and to contextualize the panorama’s elements according to their historical, artistic, or anthropological meanings. In particular, domain experts should be able to enhance the information presented in a panorama by guiding the attention of the end-users to the elements which are more relevant to a specific theme. This is what happens in a real visit: the expert guide does not wait for the visitor to ask additional information about an area (e.g., a fountain or a church), but points out to the visitor the elements which she/he things are more relevant to a specific context or better suited to a specific target of visitors.

To address this issue, we provide a methodology, based on the concept of narration (which is conceived as description of a story, an interpretation of the information presented) through which domain experts can contextualize and connect different semantic hotspots specified in a panoramic image (or in different panoramas) in order to generate customized presentations of the contents. The link between narrations and semantic hotspots is maintained through a semantic model describing the heritage in the panoramas.

With the introduction of semantic hotspots and narrations, panoramic images go from a set of digital pictures stitched by a software and included in a Web page towards a new paradigm of interaction according to which users not only can specify the places to visit, but they can also have an entry-point into the digital archives relevant to the elements in the panoramas as well as enjoy customize tours with the help of experts guides.

Knowledge Base For Etruscan Heritage

The approach used for describing the information domain highly influences the way the panoramas system is organized. Adopting a richer model allows one to provide a semantically richer representation of the domain of interest. In order to achieve this goal it is necessary to define a knowledge representation model able to describe the cultural aspects of domain in order to meet users’ expectations and needs. In this paper we focus on a solution based on the identification, creation, representation, and distribution of information through a knowledge base (Thomas et al, 2001)(Wexler, 2001).  A knowledge base provides the media for the computerized collection, organization, and retrieval of knowledge. It is commonly used to capture explicit concepts and relations at the base of the information domain by means of an ontology.

In the case study presented in this paper, computer scientists and archaeologists of the University of Milano have collaborated to create an interactive panoramas system for the dissemination of Etruscan heritage spread in several European museums. A fist step of the project was the definition of a shared knowledge which provides formal definitions for the relevant Etruscan domain’s concepts. The final result is an “architecture of concepts” representing a huge quantity of objects and architectural monuments and a range of relevant scientific themes regarding the development of the Etruscan culture. In order to connect these concepts, an Etruscan ontology had to be designed and developed.

This semantic model for the Etruscan ontology has been derived from a well-known CH ontology, that is,  the CIDOC Conceptual Reference Model (CRM). CIDOC/CRM is the culmination of over 10 years work by the CIDOC Documentation Standards Working Group  and CIDOC CRM SIG  and it is intended to promote a shared understanding of CH information by providing a common and extensible semantic framework that any CH information can be mapped to (Croft et al 2006). It can be described as a property (relationship) driven model, the definition of which is exhibited by the 132 unique properties that associate the 81 discrete concepts.  This ontology is suitable to create a common language and it is usable as the basis for a good practice conceptual modeling in each cultural contexts. For these reasons it defines concepts and relationships at an abstract level, thus hiding the peculiarities of a specific language or the intrinsic features of a well determinate cultural context. However, a shared knowledge base describing the Etruscan culture has to be able to reflect the specific definitions of the relevant domain’s concepts in a rigorous and detailed way. Therefore we specialized CIDOC in order to develop an ontology specific to the Etruscan culture

The development of our ontology has been accomplished through an iterative processs. The first step was to determine domain and scope of the ontology. This means to clearly establish which are the real-world entities described by the ontology, and what kind of informative requirements the ontology must fulfill. By adopting such perspective, the initial step in the ontology development required a major involvement of the Archaeologists in that they had to identify the most important entities, either physical or conceptual, of the domain, like findings, monuments, finding representation relevant from an artistic point of view, and so forth. The second step was to label each class of elements using named entities, like Finding, Monument, Decoration, etc.  The third step was the definition, for each entity, of a set of attributes, which expressed some of their relevant characteristics, such as the raw material composition of the finding, its chronology, its style and so on. If needed, new entities were introduced in order to model the attributes, for example FindingMaterial, FindingCronology, FindingProducer, etc. In this way it was possible to conceptually organize a set of classes able to describe the architecture of concepts that our ontology had to represent. The outcome of this activity was a set of entities, referring to the entire set of possible information items useful in our archaeological context. Next, we created the proper Etruscan ontology. Instead of defining a new hierarchy of classes, we took advantage of the CIDOC/CRM, adapting its conceptual reference model to our archaeological heritage domain. During such step, computer scientists and archaeologists collaborated in order to choose proper CIDOC/CRM concepts and relations according to the entities defined in the previous step. For example, the entity “Etruscan Hero” in the system is represented by the class TC1.Tarchna_Person. This class is a simple mapping of the E21.Person CIDOC/CRM class, which comprises people who lived or are assumed to have lived (http://cidoc.ics.forth.gr/). By exploiting the CIDOC/RM hierarchical structure, we defined a semantic model customized for our archaeological context, readable and understandable by the Archaeologists, which could be used as a media through which Computer Scientists and Archaeologists can share a common understanding and therefore communicate unambiguously.

Besides representing the knowledge base structure, the ontology is also used to integrate and support access to information from heterogeneous data sources. The integration is carried out by expressing the elements defined in each logical schema in terms of the classes forming the ontology and of a proper a set of mapping information. This means that each entity defined in the database must be associated with an Etruscan entity class expressing the same semantics. Once such a relation has been established, all the entity’s attributes must be mapped as well; therefore an entity, expressing the same semantics of the attribute meant to be mapped, must be found. Since the Etruscan entities have been associated with the CIDOC/CRM classes, the data source entities are automatically mapped on the ontology. This means that it is possible to take advantage of the relations infrastructure set up while defining the ontology. Moreover, there are no specific requirements for a data source in order to participate in the system. The only requirement concerns the database schema that must be able to specify the mapping information. In this context, no modifications are required on the existing archives and the participation of a data source in the network can start/terminate at anytime.

Enriched Knowledge Base By Means Of Narrations

The ontology described in the previous section represents the knowledge base structure of the Etruscan domain and is used to integrate and to access information of heterogeneous data sources. The result is an architecture of concepts offering a uniform description of the CH which is independent from the specificities of each DB schema. Exploiting such an ontology, it is possible to address the shortcomings in the use of panoramic images that we discussed in Section 3, that is, to contextualize the panorama’s elements according to their historical, artistic and anthropological meaning, and to meet the end-users’ wishes and interests. 

In order to address the above issues, we have defined a method for the semantic enrichment of the available information according to the directions indicated by the domain expert community. Such method enriches the ontology schema by means of a new class called “Narration”. A narration allows one to contextualize the data in a novel way by relating different types of artifacts. It is composed by a text and a context. Our approach supports two different types of narration: direct and indirect. In a direct narration, the text is a detailed domain expert’s interpretation, such as an explanation of a given finding or monument, or a given group of invariable and limited objects and its context is defined by means of an explicit description of the relations between the text and the objects involved in it. For instance, a possible direct narration can concern the function of a specific trumpet as musical instrument, or the role of a given object as a votive offering. Instead, in an indirect narration, the text is a more general document that deals with arguments involving a (possibly large) group of artefacts. This narration typically focuses deeply on a given topic, such as a description of iconography and of its formal origin, a class of monuments, the role of a given group of objects and so on. In this case the context expresses the network of relations that connect the text with the set of artefacts related to it. For instance, a possible indirect narration can deal with the theme of devotional practices or the music in ancient Etruria.

It is important to note that each archive in the federation of the databases connected by means of the ontology can be updated, for example because the institution maintaining it acquires some new exhibit. Moreover the federation can be extended by new entries of institutions, which bring to the federation new databases. In all these cases the set of artefacts collected as the result of the interpretation of a context in an indirect narration is different from the previous ones. Hence the context and the narration itself are systems of documents varying in time. 

Therefore the narration has a double function. On one hand it provides to the domain expert a medium for creating an interpretation of the information content, and on the other it offers to the final user a medium by means of which to query the knowledge base according to a specific thematic interest.

Panoramas System

In order to give to the large public an easier accessibility to the bulk of information defined through the “architecture of concepts” represented by the Etruscan ontology according to specific interest of the user, and to give the cultural heritage a contextualized perspective, our solution offers a new model of museum visit that exploits the concepts of narration.

Through the use of the narrations, it is possible to connect panoramic images by means of which different European museums can build a shared knowledge network and promote integrated access the Etruscan heritage using seamless environments. With respect to existing solutions, our solution provides contents the organization of which is not  predefined but can be customized according to the user needs and interests. A user, reading the narration’s text, can browse through the collection of artifacts constituting the narration’s context. She/he can thus experiment a new kind of virtual visit enriched with the knowledge from the domain experts who wrote the narration.

In this perspective, our notion of semantic hotspot has a key role in order to integrate panoramas with narrations. Such hotspots make use of narrations to dynamically link relevant regions of a panorama with the information stored in one (or several) data sources. When the visitor clicks on a region associated with a semantic hotspot, she/he can access additional information about the artifact. As a consequence, using Web services exposed by the engine developed for T.Arc.H.N.A (called T-engine), the application retrieves all the narrations connected to the selected artifact. When the visitor selects a narration among the ones prepared, the system extracts and highlights the semantic hotspots in the panorama which are related to the selected narration. If a narration is associated with hotspots belonging to multiple panoramas, the user has the opportunity to preview them and eventually to jump into the new panorama where the selected artifact is placed. The T-engine takes care of translating the requests expressed in terms of the concepts defined in the semantic model into SQL statements understandable by the integrated data sources.

Figure 1 shows a narration about Etruscan Music as it appears on a PC screen to a visitor. The text (frame A) describes the social role of music and features, roles and functions of musical instruments in the Etruscan age. The context (frame B) is a set of four documents displayed on the right side of the text: the top document shows and describes a lute, a musical instrument, the second and third displays two frescos from the Leopard Tomb, displaying musicians playing musical instruments and the fourth a vase on which a playing musician is represented. The four documents have been retrieved from different databases, and illustrate artifacts (instrument, frescos and vase) which can be seen in several institutions, geographically dispersed in Europe.

Figure 1
Figure 1: The materialization of narration as seen by the end-user: frame A is the representation of the text created by the archaeologist and arguing Etruscan musical instruments whereas frame B is the representation of the archeological context, that is a set of documents involved in the content of the narration

The National Museum Of Tarquinia: A Virtual Tour

The described approach has been applied to develop a virtual tour available from the National Museum of Tarquinia Web site. The application is accessible both through the Web and by the multimedia installations placed inside the museum.

The panorama Web framework has been developed using pages dynamically generated according the tour chosen by the visitor. Combining JSP (java server page) and JavaScript, our solution generates a semantic query based on the content of the selected sensible area. The query is then forwarded to Sesame (Aduna & Sirma Ltd, 2006), an open source java framework for storing, querying and reasoning with an ontology; this query is expressed in SeRQL, a RDF/RDFS query language combining traditional features of classic query language with the possibility of going through a semantic net. Sesame accesses the ontology representing the knowledge domain in order to enrich the panorama with mutable and flexible information according to the number and types of the databases from which to retrieve data.

 figure 2Figure 2: A screenshot of the panoramic Web application (Italian release)

Figure 2 shows an example of our final application. In this screenshot it is possible to view a panorama regarding an area of the National Museum of Tarquinia. The scenario is the following: a user browsing the panorama has clicked on the finding representing a bronze statuette. As a consequence, below the panorama, the user can view as much tabs as the narrations connected to the selected finding. Selecting the narration titled “I bronzetti votivi” (the votive bronze artefacts), on the right, it is possible to view all the artifacts referred in the narration and contextually to highlight them in the panorama.

Conclusion And Future Work

In this paper we have presented a system that aims at including knowledge in the process of organizing and disseminating heritage using 360 panoramic images. With our solution, digital archives are integrated with the panoramas and domain experts can actively participate, specifying narrations, in the process of disseminating information to the public.

Future work is oriented towards providing visitors with tools for annotating the panoramas in order to share comments with others using a sort of collaborative pattern. We are also interested in exploring recent research in the area of social networks in order to support the formation of cultural communities centered on our notion of knowledge base.  Another direction of our work is related to the creation of to an Integrated Development Environment (IDE) able to offer to any user (museum curators, researchers, students…) the possibility to create its own panoramic images (according to our guidelines) that can be automatically integrated in a network of museum panoramas.


This work has been partially founded by the European Community under the T.Arc.H.N.A. project (Towards Archaeological Heritage New Accessibility) under the EC-grant No 2004-1488/001-001, CLT-CA22 (CULTURE2000 2004-2007).


Aduna B.V., Sirma AI Ltd. User Guide for Sesame. Updated for Sesame release 1.2.6., last updated 2006, consulted June 2007. available http://www.openrdf.org/doc/sesame/users/index.html

Crofts, N., Doerr, M., Gill, T., Stead, S. and Stiff, M. (2006). Current Official Version of the CIDOC CRM version 4.2.1 of the reference document. Definition of the CIDOC Conceptual Reference Model,  last updated October 2006, consulted June 2007. available http://cidoc.ics.forth.gr/docs/cidoc_crm_version_4.2.1.pdf

Hoeben, A. & Stappers, J. (2006). Taking clues from the world outside: navigating interactive panoramas. Springer-Verlag. (Eds.) Personal and Ubiquitous Computing. London, UK: January 2006, Vol. 10, N. 2, 122-127.

Horry, Y., Anjyo, K. and Arai, K. (1997). Tour into the picture: Using a spidery mesh interface to make animation from a single image. ACM Press/Addison-Wesley Publishing Co. (Eds.) SIGGRAPH’97. Los Angeles: August 1997, 225-232.

Mosaker, L. (2001). Visualising historical knowledge using virtual reality technology. Routledge, part of the Taylor & Francis Group (Eds.) Digital Creativity, Vol. 12, N. 1, 15-25.

Nakano, A., Rai, C. and Hoshino, J. (2004). Panorama-based immersive story environment. ACM Press (Eds.) ACM SIGCHI International Conference on Advances in computer entertainment technology ACE ’04. Singapore: September 2004, 354-354.

Pan, Z., Fang, X., Shi, J. and Xu, D. (2004). Easy tour: a new image-based virtual tour system. ACM Press (Eds.) ACM SIGGRAPH international conference on Virtual Reality continuum and its applications in industry. Singapore: June 2004,  467-471.

Papagiannakis, G., Ponder, M., Molet, T., Kshirsagar, S., Cordier, F., Magnenat-Thalmann, N. and Thalmann, D. (2002) LIFEPLUS: Revival of life in ancient Pompeii, Virtual Systems and Multimedia. IOS Press and OHMSHA publishers (Eds.) VSMM 2002 Virtual. Gyeongju, Korea: October 2002.

Thomas, J. C., Kellogg, W.A. and Erickson, T. (2001). The Knowledge Management puzzle: Human and social factors in knowledge management. IBM Systems Journal, Vol. 40, N. 4, 863-884.

Valtolina, S., Mazzoleni, P., Franzoni, S. and Bertino, E. (2006) Semantic Approach to Build Personalized Interfaces in the Cultural Heritage Domain.  ACM Press (Eds.) 8th International Working Conference on Advanced Visual Interfaces 2006. Venezia, Italy:  May 23-26, 2006, 306-309.

Van Welie, M. & Klaassen, B. (2001). Evaluating Museum Websites using Design Patterns, Technical report number: IR-IMSE-001, December, 2004, 76 pages.

Wexler, M.N. (2001). The who, what and why of knowledge mapping. Journal of Knowledge Management, Vol. 5, N. 3, 249-263.

Zheng, J. Y., Shi, M., and Kato M. (2003). Route panoramas for city navigation. ACM Press (Eds.) Eleventh ACM international conference on Multimedia. Berkelet, CA, USA: November 2003, 108-109.

Cite as:

Valtonia, S., and E. Bertino, Integrating Data Management and Panoramic Images for an Effective Representation of Cultural Heritage Information , in International Cultural Heritage Informatics Meeting (ICHIM07): Proceedings, J. Trant and D. Bearman (eds). Toronto: Archives & Museum Informatics. 2007. Published October 24, 2007 at http://www.archimuse.com/ichim07/papers/valtonia/valtonia.html