Knowledge-assisted Visualization in the Cultural Heritage Domain
Case Studies, Needs and Reflections
Patricia Martín-Rodilla
Institute of Heritage Sciences (Incipit), Spanish National Research Council (CSIC),
Rúa San Roque, 2, Santiago de Compostela (A Coruña), Spain
Keywords: Knowledge-assisted Visualization, Discourse Analysis, Cultural Heritage, Inferences, Visualization
Requirements.
Abstract: We investigate on how to build software systems that assist cultural heritage researchers to reconstruct past
events on the basis of present data. In this setting, knowledge-assisted visualization can be a useful
mechanism to improve the knowledge generation process and emphasise collaboration. However, a useful
visualization depends on the goals of the user and the specific research problem involved. In this position
paper we present a set of case studies to defend the study of cognitive inferences through discourse analysis
and its typologies as a starting point in the knowledge-assisted elicitation process. Once a complete study of
usual inferences in the cultural heritage domain is done, the visualization needs in this domain will be easier
to determine and apply, attaining our objective of knowledge-assisted visualization.
1 INTRODUCTION
Cultural heritage activities such as archaeological
excavations or anthropological studies manage a
large amount of complex data. In previous empirical
studies, an elicitation process about the interaction
preferences of researchers in cultural heritage
(Martín-Rodilla, 2012) or conceptual modelling
studies in cultural heritage domain (González-Pérez
and Parcero-Oubiña, 2011) were carried out as a
previous work to understand the needs of cultural
heritage researchers. The visualization of this data
and it application to generate knowledge is one of
the problems detected in these studies.
Some literature and applications of visualization
in cultural heritage exist (Vote et al., 2002); (De
Luca et al., 2011). However, these approaches are, in
most cases, ad hoc proposals not aimed at
knowledge-assisted visualization.
Knowledge-assisted visualization involves a
number of areas in visualization theory, including
theory frameworks and taxonomies (Zhou and
Feiner, 1998); (Amar and Stasko, 2005), cognitive
studies about the knowledge generation method and
visualization cognition (Huang et al., 2006), data
characterization (González-Pérez, 2012) or
interaction principles (Van der Vlist el al., 2011).
This paper shows an initial study based on
discourse analysis as technique to characterize the
usual inferences that occur in the cultural heritage
domain and their implications in knowledge-assisted
visualization.
2 METHODOLOGY AND
PROPOSAL
According to the literature (Tufte, 1990); (Chen et
al., 2009) it is possible to assist in the knowledge
generation process of humans by improving the
visualization of the data and the processes
involved.In knowledge-assisted visualization (Chen
et al., 2009), the user’s knowledge is an important
part of the visualization. Characterizing this
knowledge, we can incorporate it as expert
knowledge, building best-fit visualizations in a
specific domain.
The question that we find unanswered, therefore,
is what are the visualization needs in the cultural
heritage domain to design useful knowledge-assisted
visualizations and what is the best method to
characterize this knowledge.
We propose a deep study of the most common
types of inferences, in order to correctly support the
processes of knowledge generation in a specific
546
Martín-Rodilla P..
Knowledge-assisted Visualization in the Cultural Heritage Domain - Case Studies, Needs and Reflections.
DOI: 10.5220/0004281605460549
In Proceedings of the International Conference on Computer Graphics Theory and Applications and International Conference on Information
Visualization Theory and Applications (IVAPP-2013), pages 546-549
ISBN: 978-989-8565-46-4
Copyright
c
2013 SCITEPRESS (Science and Technology Publications, Lda.)
domain. The future users of a visualization tool or
technique in this field are archaeologists,
anthropologist, historians and researchers or
practitioners in similar disciplines. These
professionals aim to reconstruct past events through
present data. For this reason, the best-fit
visualizations to assist the processes that they carry
out are those that can support the most common
inferences that occur in the cultural heritage domain,
which often entail strong spatio-temporal
components.
The case studies presented below were
developed by applying the technique of discourse
analysis (Hobbs, 1985) in order to dissect the
relevant cultural heritage statements and determine
visualization user needs. This method analyses the
syntactic of the discourse looking for elements of
connection such as prepositions or similar particles
and evaluates the knowledge implied in the text.
Hobbs’s analysis is put into practice in other
research areas such as biomedical domain (Lacson et
al., 2006), legal arguments (Moens et al., 2007) or
mathematics research (McGrath and Kuteeva, 2012)
to extract information and requirements. We try to
find common inferences types in archaeological
research as a subfield of cultural heritage. The
differences between archaeology and other cultural
heritage disciplines in this argumentation and the
generalization of the conclusions are future needs in
our research.
Establishing types of inference presented in the
discourse allows us to determine the needs of the
users, introduce it as expert knowledge and elaborate
a set of suggested visualization solutions that could
be used as input for a research-oriented software
system.
3 CASE STUDIES AND NEEDS
The following sections present three case studies.
Each case study includes a literal statement taken
verbatim from a research work in archaeology, an
analysis of the inference according with the
characteristics of the discourse in said statement, and
a reflection about the user needs that this analysis
reveals in terms of visualization requirements and
suggested visualization techniques.
3.1 Case Study One
Cultural Heritage Statement: “There is a group
of pots with no decoration, low quality clay
and hand made. They could have been used as
housewares. This hypothesis is fitting with the
chemical analysis of the pots” (Cobas and
Parcero-Oubiña, 2006).
Inference Analysis: This is a common kind of
hypothesis in archaeology. Based on the
classification of pots or pot fragments found at
a particular site, researchers assess the quality
and analyse the chemical properties of the
pots, and they try to reconstruct the function
they had. The first objective is to find groups
or similarities between pots. The presence or
absence of decoration, the quality of the clay,
and the manufacturing process (such as by
hand or wheel) for each pot are determined
and compared with the outcomes of the
chemical analysis.
User needs: For this kind of inference, the first
process is one of clustering. The second
process is a relational one: the researcher
wants to find a relation between the pot
groups and the characteristics of the pots in
order to infer the function or common use of
the pots. To improve this knowledge-
generation process, the researcher needs to
visualize data in groups. To support the
second process, the researcher may need to
focus on a particular group and visualize the
specific characteristics of its pots and their
chemical properties in order to determine the
function of the pots.
3.2 Case Study Two
Cultural Heritage Statement: “In funerary
contexts in the Bronze Age of Atlantic
Europe, we can find individual tombs with
decorated and high-quality manufactured pots
with handles and brilliant colours” (Prieto-
Martínez, 1998).
Inference Analysis: In this case, the first part
of the statement is a selection of a space-time
slice. The researcher relates the funerary pots
with their characteristics (material, colour,
quality, handles). The main objective is the
characterization of the funerary pots in
Atlantic Europe during the Bronze Age.
User needs: In order to compare the target
pots with pots of similar but different
contexts, the researcher needs to visualize a
map of their spatio-temporal situation plus the
characteristics being analysed (material,
colour, quality, handles) for each region of the
map. In addition, the researcher may want to
visualize the sequence of temporal events for a
Knowledge-assistedVisualizationintheCulturalHeritageDomain-CaseStudies,NeedsandReflections
547
specific area of the map, or for a specific pot,
or other spatio-temporal relationships.
3.3 Case Study Three
Cultural Heritage Statement: “The source of
inorganic matter implies that the scope of
interaction does not exceed 50 km. This
means that animals were required to transport
the wheat to the mills” (Fernandez, 2010).
Inference Analysis: The first part of the
inference involves the grouping of inorganic
matter found, as well as the delimitation of the
area where those inorganic matters were
found. Knowing that traces of wheat farming
were found on the site, the research infers that
animals must have been required to transport
the wheat, because no mills were present
within the studied archaeological area.
User needs: This kind of inference is more
complex than the previous two, because it
involves a concatenation of simpler
inferences. In the first part, the researcher
needs to visualize the area where inorganic
matter and wheat farming traces were found.
In the second part, the researcher needs a
relational visualization to match the distance
and the nearest mills on the map.
3.4 Overall Observations
There are some differences in the inference types
and knowledge-generation methods that are revealed
in the case studies presented. These differences
mean that the knowledge-assisted visualization
requirements are arguably different too.
In addition, we detected that the most common
types of inferences are clustering or relational
mechanisms. This idea is not exclusive in cultural
heritage and it is applicable to other research areas
(Morato et al., 2003); (Fujita and Teplovs, 2009).
However, this characterization allows us to identify
a set of needs in visualization for researchers in
cultural heritage and suggest some association
visualization techniques:
We detect clustering inference as a starting
step in the knowledge-generation process. We
think that cluster visualization and related
techniques (Van Wijk and Van de Wetering
,
1999); (Stasko and Zang, 2000) are good
choices for the first step in the knowledge-
assisted process.
We detect strong spatio-temporal components
in the following steps, e.g. in cases two and
three. It is arguable that techniques related to
spatio-temporal visualizations (Andrienko et
al., 2003); (Aigner et al., 2011) are good
candidates for the next steps in the
knowledge-assisted process.
We also detect needs in terms of varying
granularity of the information that it showed
in the visualization. In all three case studies,
statements involve large amounts of data. In
this situations we argue that the application of
well-known patterns such as
Abstract/Elaborate (Yi et al., 2007), Filter (Yi
et al., 2007), Snap-Together (North and
Schneiderman, 2000), or patterns at the
implementation level (Heer and Maneesh,
2006) would work well to solve these needs.
This list of needs is not complete. The main
purpose of this paper is to start a discussion about
the analysis of discourse and the typology of
inferences as a method to understand a domain and,
in particular, the cultural heritage domain, and
improve knowledge-assisted visualization.
4 REFLECTIONS AND
DISCUSSION
This study is an initial conceptualization about the
main features in the knowledge-generation process
in cultural heritage and it needs about visualizations.
Our final objective it is to introduce this domain
between the topics in visualization, to discuss
scientifically the needs extracted in this study and
compare with other domains with a successfully
application of knowledge-assisted visualization
techniques.
In this paper we defend the need of a deep study
about the knowledge-generation process in each
domain in terms of inferences and discourse
analysis, as a part of the process of elicitation
requirements in the knowledge-assisted
visualization.
This proposal is an appropriate initial-point in
the requirements visualization process, because
allow us to understand the mind process involved in
each domain and the final objectives of each
research analysed. With this information, we could
choose better visualization principles and techniques
for each objective. Following these ideas, our future
work is focused on the conceptualization of the
method and in the elaboration of test-bed and
empirical studies with users.
IVAPP2013-InternationalConferenceonInformationVisualizationTheoryandApplications
548
REFERENCES
Aigner, W.; Miksch, S.; Schumann, H. and Tominski, C.
(2011).Visualization of Time-Oriented Data, Springer.
Amar, Robert A.; Stasko, John T. (2005). Knowledge
Precepts for Design and Evaluation of Information
Visualizations. In IEEE Transactions on Visualization
and Computer Graphics 11(4): 432 – 442.
Andrienko, N., Andrienko, G. et al. (2003). "Exploratory
spatio-temporal visualization: an analytical review." In
Journal of Visual Languages & Computing 14(6):
503-541.
Chen, M.; Ebert, D.; Hagen, H. et al. (2009). Data,
Information, and Knowledge in Visualization. In IEEE
Computer Graphics and Applications 29(1): 12-19.
Cobas, I.; Parcero-Oubiña, C. (2006). Alto Do Castro
(Cuntis, Pontevedra). Síntesis De Resultados Y
Estudios De Materiales, Campaña 1993. In Tapa
(Traballos De Arqueoloxía E Patrimonio), Spanish
National Research Council. Santiago de Compostela.
De Luca, L., Busayarat, C., Stefani, C., Véron, P.,
Florenzano, M. (2011). A Semantic-Based Platform
for the Digital Analysis of Architectural Heritage.
Computers and Graphics 35(2): 227-41.
Fernández Martín, S. (2010). Los complejos cerámicos del
yacimiento arqueológico de La Motilla del Azuer, PhD
Thesis. University of Granada.
Fujita, N.; Teplovs, C. (2009). Automating analysis of
collaborative discourse: Identifying idea clusters. In
C. O'Malley, D. Suthers, P. Reimann & A.
Dimitracopoulou (Eds.), Computer Supported
Collaborative Learning Practices. International
Society of the Learning Sciences, Inc. Rhodes, Greece:
162-164.
González-Pérez, C. (2012). A conceptual modelling
language for the humanities and social sciences. In
RCIS'12: Sixth International Conference on Research
Challenges in Information Science, Valencia, Spain.
González-Pérez, C., Parcero-Oubiña, C. (2011). A
Conceptual Model for Cultural Heritage Definition
and Motivation. In CAA'11: 39th Annual Conference
on Computer Applications and Quantitative Methods
in Archaeology. Beijing, China.
Heer, J. A., Maneesh, A. (2006). Software Design
Patterns for Information Visualization. In IEEE Trans.
Vis. Comput. Graph. 12(5): 853-860.
Hobbs, Jerry R. (1985). On the coherence and structure of
discourse. In Technical Report, Center for the Study of
Language and Information (CSLI), Stanford, CA.
Huang, Z., Chen, H. et al. (2006). Expertise visualization:
An implementation and study based on cognitive fit
theory. In Decision Support Systems 42(3): 1539-
1557.
Lacson, R.C.; Barzilay, R.; Long, W.J. (2006). Automatic
analysis of medical dialogue in the home
hemodialysisi domain: Structure induction and
summarization. In Journal of Biomedical Informatics
39(5): 541-555.
Martín-Rodilla, P. (2012). The role of software in cultural
heritage issues: Types, user needs and design
guidelines based on principles of interaction. In
RCIS'12: Sixth International Conference on Research
Challenges in Information Science, Valencia, Spain.
McGrath, L.; Kuteeva, M. (2012). Stance and engagement
in pure mathematics research articles. Linking
discourse features to disciplinary practices. In English
for Specific Purposes 31(3): 161-173.
Moens, M.F.; Boiy, E.; Mochales Palau, R.; Reed, C.
(2007). Automatic Detection of Arguments in Legal
Texts. In ICAIL '07 Proceedings of the 11th
International Conference on Artificial Intelligence and
Law. California, USA.
Morato, J; Llorens, J.; Genova, G.; Moreiro, J.A. (2003).
Experiments in Discourse Analysis Impact on
Information Classification and Retrieval Algorithms.
In Information Processing and Management 39(6):
825-851.
North, C.; Schneiderman, B. (2000). Snap-Together
Visualization: A User Interface for Coordinating
Visualizations via Relational Schemata. In AVI '00
Proceedings of the working conference on Advanced
Visual Interfaces. Palermo, Italy.
Prieto-Martínez, M. P. (1998). Forma, estilo y contexto en
la cultura material de la Edad del Bronce gallega:
cerámica campaniforme y cerámica no decorada, PhD
Thesis. University of Santiago de Compostela.
Stasko, J., Zang, E. (2000). Focus+context display and
navigation techniques for enhancing radial, space-
filling hierarchy visualizations. In InfoVis'00: IEEE
Symposium on Information Visualization, Salt Lake
City, UT.
Tufte, E. R. (1990).Envisioning Information. Cheshire,
CT: Graphics Press.
Van der Vlist, B.; Bram, J.J.; Niezen, Gerrit; Hu, Jun et al.
(2011). Interaction Primitives: Describing Interaction
Capabilities of Smart Objects in Ubiquitous
Computing Environments. In IEEE AFRICON 2011,
Livingstone, Zambia, IEEE.
Van Wijk, J. J.; Van de Wetering, H. (1999). Cushion
treemaps: visualization of hierarchical information. In
InfoVis'99: IEEE Symposium on Information
Visualization, San Francisco, CA.
Vote, E., Acevedo Feliz, D., Laidlaw, D H.; Sharp
Joukowsky, M. (2002). Discovering Petra:
Archaeological Analysis in Vr. In IEEE Computer
Graphics and Applications 22(5): 38-50.
Yi, J.S., Kang Y., Stasko, J.T., Jacko, J.A. (2007).
Toward a Deeper Understanding of the Role of
Interaction in Information Visualization. In IEEE
Transactions on Visualization and Computer Graphics
13(6): 1224-1231.
Zhou X., M.; Feiner, S. K. (1998). Visual task
characterization for automated visual discourse
synthesis. In CHI '98 Proceedings of the SIGCHI
conference on Human factors in computing systems.
ACM Press/Addison-Wesley. Los Angeles, CA.
Knowledge-assistedVisualizationintheCulturalHeritageDomain-CaseStudies,NeedsandReflections
549
