Speleothem - An Information System for Caves

Based on Semantic Web Technologies

Nikolaos Fanourakis1,2 and Panagiotis Papadakos2

1
Computer Science Department - University of Crete, Greece
2
Institute of Computer Science - FORTH-ICS, Greece
{fanourakis,papadako}@ics.forth.gr

Abstract. Humans have used caves throughout history for a wide va-
riety of needs, showcasing the importance of caves in human evolution.
Nowadays, speleological clubs organize expeditions around and under the
globe in an effort to understand, study and record the still unexplored
and complex network of caves that lies underneath. Unfortunately, a
common vocabulary for recording information related to caves and cav-
ing activities does not exist and most speleologists use adhoc and even
non-digital forms to store it. This demo showcases the Speleothem sys-
tem, an open-source information system for the domain of caves that
exploits semantic web technologies and is built on top of a proposed
vocabulary for caves and caving activities. Speleothem is designed in col-
laboration with a Greek speleological club, where it is planned to become
operational in the near future.

1 Introduction
Caves have been used by humans throughout history for a wide variety of needs
and purposes, and are highly interlinked with human evolution and civilization.
They have been considered sacred places, decorated with parietal cave paintings
by the first human artists (e.g. caves of Lascaux3 and Altamira), and appear
frequently in mythological and folklore stories. During bad weather conditions
or war times, caverns have provided shelter for humans and livestocks and have
been used for storing and mining resources (e.g. dairy products, water, minerals).
Caves narrate the history of earth, and due to their isolation and remoteness they
preserve important facts of the human history and host alien and fragile closed
ecosystems. As a result they are considered sensitive information islands for a
lot of scientific and academic disciplines, including but not limited to geology,
anthropology, archeology, paleontology, biology, hydrology, seismology and folk-
loristics, and as such they are protected by local and international legislations.
Currently, a lot of people are visiting horizontal caves for their beautiful dec-
orations or historical importance, and vertical ones for athletic and exploration
reasons. Unfortunately, public or restricted access to cave related information
is limited to either general information offered by local or global touristic por-
tals (e.g. tripadvisor) or hard to find scientific publications and books that use
3
http://archeologie.culture.fr/lascaux/en
2 Nikolaos Fanourakis and Panagiotis Papadakos

complex terminologies. In addition, a lot of precious, delicate and difficult to

extract information from previous expeditions is stored by local cave clubs in
adhoc digital or non-digital forms. Consequently, the preservation, retrieval, val-
idation, integration, and dissemination of information about caves, and the re-
lated human activities and their impact on them, is a rather cumbersome task.
Addressing the above issues will not only provide a useful entry point to such
information, but can also raise the awareness of the public opinion about the
importance, sensitivity and particularities of this unseen underworld.
In this demo we present the Speleothem system, an open-source4 web based
system for storing and retrieving information about caves and cave related activi-
ties. It is based on semantic web technologies and introduces the speleothem vo-
cabulary, especially designed for modeling cave related information. A SPARQL
endpoint and a REST API for easy access to the data are also provided. The
Speleothem system is designed in collaboration with the Speleological Club of
Crete5 where it is planned to become operational in the near future. Our aim
is to offer a reference caving information system, on top of a domain specific
knowledge base that stores information that can be used and linked by other
external resources, for the general public, speleologists, speleological clubs and
organizations.

2 Related Work

Software for caving has been limited to cave surveying, where by using specific
electronic devices (or not) the software can create 2D or 3D representation of a
cave (e.g. [1] and survex6 ), recording of locations, entrances and other cave re-
lated information over a map7 , or for showcasing the importance and fragility of
caves (iCavern app [3]). Currently important cave related information is stored
locally in cave clubs using adhoc digital or non-digital forms and vocabularies.
Unfortunately, well-known ontologies (e.g. DBpedia) model only general infor-
mation about caves8 and miss other important entities and information, while
others are too restrictive (e.g. the ontology described in [4] for mining equip-
ment). In this work we propose the speleothem vocabulary, especially designed
for modeling information about caves and cave related activities, that is at the
core of the demonstrated Speleothem system.

3 speleothem Vocabulary for Caves and Caving Activities

The current version of the proposed speleothem vocabulary consists of 50 classes,

25 object properties and 40 data properties. Using this vocabulary we can store
4
https://bitbucket.org/speleothem/speleothem
5
http://caves-crete.gr/
6
https://survex.com/
7
https://github.com/apgeo/silexgis, a relevant work based on a relational DB
8
http://dbpedia.org/ontology/Cave
 Lecture Notes in Computer Science: Authors’ Instructions 3

information about the physical properties of the cave, such as the type and the
size of the cavern, its location and entrances9 , the map of the cave (either in
textual representations that can be used by other software like therion, or as
sketches/images), collections of images of the cave itself and the outside envi-
ronment, information related to the contained rooms and the natural or human
created speleothems that it contains, along with their images. Each cave can
be associated with climate data, living organisms, folklore stories or historical
data. Regarding the caving activities, the vocabulary uses the foaf10 vocabulary
to store user profiles and caving clubs, and associates the profiles with caving
equipment, special abilities of cavers (e.g. rescue training) and access policies to
information. Furthermore, the vocabulary is able to hold important data about
cave expeditions. sameAs connections with classes of other ontologies like from
DBpedia11 are also provided. A screenshot of the vocabulary using the Web-
VOWL tool12 (a web application for the interactive visualization of ontologies)
is shown in Fig. 1. The current version of the speleothem vocabulary is online13
but is not considered stable yet. We expect to make further refinements based
on feedback from the SPOK club (where it is planned to become operational),
and publish it using the best practices and the five stars policy described in [2].

Fig. 1. The speleothem vocabulary as visualized by WebOWL

4 Architecture & Implementation

The architecture of the Speleothem system is given in Fig.2. The backend is
implemented using Java technologies while the front-end is based on common
web technologies and libraries (e.g. bootstrap, Google Maps API). The client
9
http://www.geonames.org
10
http://xmlns.com/foaf/spec/
11
http://wiki.dbpedia.org
12
http://www.visualdataweb.de/webvowl/
13
http://www.speleothem.org/vocabulary/alpha
4 Nikolaos Fanourakis and Panagiotis Papadakos

sends AJAX requests to a REST service14 that is implemented using spark-

Java15 , which exploits the SPARQL endpoint16 . We use RDF4J17 for storing
and querying the Speleothem knowledge base.

Fig. 2. The Speleothem architecture

5 Demo Scenarios
In the current demonstration we plan to showcase the following scenarios for
different user roles (i.e. unregistered, registered and administrator):
Tour of a Cave: Public data like physical properties and position, cave cli-
mate, biological and historical data, and images (Fig. 3 shows the search screen).
Moreover, registered users have access to detailed cave and rigging maps, past
expeditions, and any information about rooms and speleothems (check Fig. 4).
Expedition Overview: General information about a cave expedition, its aims,
the participating cavers and clubs, previous related expeditions, a calendar and
a report of the results, and possible revisions of cave data due to this mission.
Caving Club Overview: General club information, participated missions, mem-
bers, general/rigging equipment and photos/sketches (check Fig. 5).
Caver Personal Information & Log book: User profiles of cavers, includ-
ing caving clubs that they are members of, their personal equipment, important
skills like knowledge of cave rescue and first aid techniques, participation in cave
rescue exercises, and personal log book.
Cave Rescue: Organization and data logging during a cave rescue event.

References
1. Martin Budaj and Stacho Mudrák. Therion–digital cave maps therion–cartographie
souterraine digitale. In presented on the 4th European Speleological Congress. Banska
Bystrica, Slovakia, 2008.
2. Krzysztof Janowicz, Pascal Hitzler, Benjamin Adams, Dave Kolas, II Vardeman,
et al. Five stars of linked data vocabulary use. Semantic Web, 5(3):173–176, 2014.
3. Sandra Dianne Joop. icaverns: Interpretation, there’s an app for that! 2013.
4. Zhi Hai Liu, Qing Liang Zeng, Cheng Long Wang, and Yu Shan Li. Research
of equipment selection and matching expert system in fully mechanized caving face
based on ontology. In Key Engineering Materials, volume 419, pages 117–120. Trans
Tech Publ, 2010.
14
http://www.speleothem.org/rest
15
http://sparkjava.com/
16
http://www.speleothem.org/sparql
17
http://rdf4j.org/
 Lecture Notes in Computer Science: Authors’ Instructions 5

Fig. 3. Search Results Screen

‘’

Fig. 4. Cave Profile (Missions)

Fig. 5. Caving Club Profile