=Paper=
{{Paper
|id=Vol-3322/short1
|storemode=property
|title=Building a French Revolution Narrative from Wikidata
|pdfUrl=https://ceur-ws.org/Vol-3322/short1.pdf
|volume=Vol-3322
|authors=Inès Blin
|dblpUrl=https://dblp.org/rec/conf/ijcai/Blin22
}}
==Building a French Revolution Narrative from Wikidata==
https://ceur-ws.org/Vol-3322/short1.pdf
Building a French Revolution Narrative from Wikidata
Inès Blin1,2
1
SONY Computer Science Laboratories, Paris, France
2
Department of Computer Science, Vrije Universiteit Amsterdam, The Netherlands
Abstract
Historical reconstructions can help us better understand important events. This is essential for bettering our
knowledge on specific topics and this enables us to build coherent sequences of information. Knowledge graphs
like Wikidata contain information about generic knowledge that includes historical events, but the information
is often entity-centric rather than event-centric. This makes the analysis and understanding of events related
to a same topic harder, since it is less easily accessible. In this demonstration paper, we developed a system
to build a timeline about the French Revolution in the form of a graph of a sequence of events, also called a
narrative in this work. The system comes with an interface, where a user can first select the types of events
to be retrieved, and is then walked through the steps of the system. The output of the system is a timeline of
the French Revolution.
1. Introduction 2. Related Work
We constantly create narratives to provide explana- A narrative is one way to understand or explain a
tions or justifications for why and how somethings series of related events. In this work, a narrative is
happens. This capability of creating such narratives technically defined as a sequence of events in the
is even suggested to be part of what makes us human form of a graph. A narrative in this context is thus
[1, 2]. AI systems able to create narratives could a type of knowledge graph. Therefore, our work is
therefore help in the realisation of human-centric related to narratives and to how events are struc-
systems. In this demonstration paper, we present a turally represented, especially in event knowledge
system that can build a timeline of the French Rev- graphs structures. The output of our system is a
olution by exploring Wikidata. The system takes timeline generated from a knowledge graph, hence
the user’s inputs to select nodes in Wikidata, walks our work is also related to timeline generation from
the users through the different steps and outputs a knowledge bases. We describe the aforementioned
timeline. research domains in this section.
The system presented in this demonstration pa- Narratives are a means for better understanding
per builds a narrative in the form of a graph. A how a situation unfolds over time. The events are
narrative is defined here as a sequence of events. the atomic concepts underlying these narratives.
The main novelty of the paper is a system that walks Towards formally defining such structures, Bartalesi
the user through the different steps of the process and Meghini [5] for example propose to define formal
and lets the user choose the events to include for components of narratives using event calculus theory
the narrative. Earlier systems could let the user [6].
navigate through events and entities without a fi- From a resource perspective, the closest to our
nal timeline output [3], or could build the timeline work would be event knowledge graphs that focus
without interacting with the user, for example by on describing events and links between them. Guan
retrieving most significant events for a biography et al. [7] survey event knowledge graphs from four
description [4]. different perspectives: history, ontology, instance
and application. More recent research has focused
on providing better access to semantic structured
data about events, and has emphasised the need to
have such event-centric information for applications
such as timeline generation or history reconstruc-
IJCAI 2022: Workshop on semantic techniques for
tion. Available resources now include (i) EventKG
narrative-based understanding, July 24, 2022, Vienna, Aus-
tria [8] built from generic knowledge graphs, (ii) a knowl-
$ ines.blin@sony.com (I. Blin) edge graph built from textual news data [9] (iii) a
0000-0003-0956-9466 (I. Blin) knowledge graph built from a Sherlock Holmes novel
© 2022 Copyright for this paper by its authors. Use permitted under
Creative Commons License Attribution 4.0 International (CC BY [10] (iv) semantic data for cultural heritage [11] and
4.0).
CEUR
CEUR Workshop Proceedings (CEUR-WS.org)
Workshop
Proceedings
http://ceur-ws.org
ISSN 1613-0073
(v) GDELT [12] and ICEWS [13].
1
From an application perspective, the closest to with a description of each event as well as event
our work would be systems that use knowledge bases type and participants. The demonstrator currently
for timeline or hypothesis generation. The aim of outputs a timeline for the French Revolution only,
such systems is to help a user understand better one but could be extended to other types of revolutions
topic. For quality, Althoff et al. [14] define three for instance. The code is publicly available1 .
criteria for a timeline quality: relevance, tempo- The backend is written in Python 3.9.4, the inter-
rally diverse and content diverse. For applications, face was developed with Streamlit2 . Wikidata was
Metilli et al. [4] use Wikidata to build biographies, accessed through the Wikidata SPARQL endpoint.
e.g. Dante’s one. Kroll et al. [15] furthermore build Figure 1 shows the home page of the demonstrator
narratives as hypotheses on top of different knowl- that explains the purpose of the application and the
edge bases, with an application in the biomedical main steps to extract the timeline.
domain.
3.1. Event Collection from Wikidata
3. Demonstrator Features The first step to build the timeline is to collect
events from Wikidata. In the demonstrator, the
user can select a collection of paths to extract events
from, depicted in Figure 2. The user can then press
the button “Collect events” to query the Wikidata
knowledge graph.
Events and links to Wikidata and Wikipedia are
stored in the backend, and the user is provided with
some additional information, such as the number of
events retrieved.
3.2. Extracting narrative features
Once the events have been extracted from Wikidata
depending on the user’s input paths, the next step
is to extract narrative features for the events. A
narrative feature for an event contains information
about either a participant, a timestamp, a location,
a type of event, or a temporal or causal link to
another event. In Section 2. and 3. displayed in
Figure 1, the user can press a button to extract
these features.
In the demonstrator, the user can see the features
that were extracted, as well as some additional
statistics on the distribution of the types of features.
We found that features extracted from Wikidata
contained mostly information about time and lo-
cation, whereas features extracted from Wikipedia
contained more information about participants and
causal links.
Figure 1: Home page of the application. Features in Wikipedia also contain hyper links
to other Wikipedia pages, which can therefore be
linked back to Wikidata. We use and retrieve these
In the demonstration system, the user is walked corresponding Wikidata pages to have consistent
through the different steps to build a narrative URIs for the final graph.
graph: 1) Collect events from Wikidata and se-
lect paths to extract nodes from 2) Extract features
from Wikidata 3) Extract additional information
from Wikipedia info boxes 4) Convert the data
1
into RDF triples compatible with the Simple Event https://github.com/SonyCSLParis/
building-fr-narrative-from-wikidata
Model 5) Display the narrative graph as a timeline, 2
https://streamlit.io
2
Figure 3: Event description example for 10 August
were therefore discarded from the timeline.
Figure 3 shows one event description for the 10
August. The user can read a summary of what hap-
Figure 2: Collecting events from Wikidata pened during this event, as well as some additional
information retrieved from Wikidata and Wikipedia.
The user therefore understands that 10 August was
3.3. Building the network a riot implying different actors like Louis XVI of
France as a commanding officer, or the French First
The next step is to build a narrative graph of Republic as a combatant. Figure 4 lastly displays
the French Revolution events and their descrip- the interactive timeline overview. The user can
tion, in the form of RDF triples. The rules manually slide events over time and click on each of
to convert the extracted features for each event them to better understand what happened. On the
to triples were designed manually. The ontol- bottom of the figure, one can see that it is also pos-
ogy used is the Simple Event Model [16]. The sible to see which events happened during a coarser
four main classes of this model are: sem:Event period of time, like for example the Kingdom of
(what), sem:Actor (who), sem:Place (where), France.
and sem:Time (when). Further constraints classes
sem:Role, sem:Temporary and sem:View can
add information on the role of an actor, a temporal 4. Conclusion
constraint or on a specific viewpoint respectively.
In this demonstration paper, we presented a sys-
tem that is able to retrieve events and features of
3.4. Timeline output the French Revolution to construct a timeline of
The final output in this demonstrator is an interac- ordered events and descriptions. During this pro-
tive timeline. This timeline contains ordered events cess, the user can choose how events can be selected
extracted from Wikidata, as well as three addi- and is walked through the different steps of the
tional components: (i) a brief description taken from process, that makes the system transparent. The
Wikipedia (ii) event type information (iii) partici- user lastly has access to an interactive timeline to
pant information. The events with no timestamps better understand the series of events during the
3
Acknowledgments
The work reported in this paper was
funded by the European MUHAI project
from the Horizon 2020 research and in-
novation programme under grant number
951846 and the Sony Computer Science Laborato-
ries Paris.
We thank Ilaria Tiddi, Annette ten Teije, Frank
van Harmelen and Remi van Trijp for their time
and support.
Figure 4: Output timeline overview
References
French Revolution. [1] B. Boyd, On the origin of stories: Evolu-
The system described in this paper furthermore tion, cognition, and fiction, Harvard University
helped to identify some first challenges for building Press, 2010.
narratives from knowledge graphs. These challenges [2] J. Gottschall, The storytelling animal: How
are linked to the main steps described in Section 3: stories make us human, Houghton Mifflin Har-
1) collect data 2) extract features 3) build the net- court, 2012.
work. The main challenge for collecting data is to [3] V. d. Boer, L. Melgar, O. Inel, C. M. Ortiz,
assess how to best choose the set of events that L. Aroyo, J. Oomen, Enriching media collec-
optimally describe the coarser event. Should such tions for event-based exploration, in: Research
a set contain all events that happened, or only the Conference on Metadata and Semantics Re-
most relevant, or only the ones that are interesting search, Springer, 2017, pp. 189–201.
for one user? When extracting features, there is the [4] D. Metilli, V. Bartalesi, C. Meghini, A
challenge of completing missing information, like wikidata-based tool for building and visualising
participants or locations. What additional resources narratives, International Journal on Digital
or algorithms are best to complete the narrative Libraries 20 (2019) 417–432.
graph? Lastly, the challenge when building the [5] V. Bartalesi, C. Meghini, Formal components
graph network is to find or define the ontology that of narratives, in: Italian Research Conference
is best suited for the application, and to populate on Digital Libraries, Springer, 2016, pp. 112–
this ontology. 121.
Takeaways from this paper are also about possible [6] R. Kowalski, M. Sergot, A logic-based calculus
future work directions. One first direction could be of events, in: Foundations of knowledge base
to study the temporality of the narrative, to better management, Springer, 1989, pp. 23–55.
understand how events unfold. In the demonstrator, [7] S. Guan, X. Cheng, L. Bai, F. Zhang, Z. Li,
the timeline output provides a sequence of events or- Y. Zeng, X. Jin, J. Guo, What is event
dered in time, but it does not provide explanations knowledge graph: A survey, arXiv preprint
for how one event might have triggered changes arXiv:2112.15280 (2021).
that in turn caused another event. One second [8] S. Gottschalk, E. Demidova, Eventkg: A
challenge and direction is about the evaluation of multilingual event-centric temporal knowledge
the narrative. For instance, one assumption here is graph, in: European Semantic Web Confer-
that Wikidata and Wikipedia contain objective non- ence, Springer, 2018, pp. 272–287.
contradictory facts, but one interesting track would [9] M. Rospocher, M. van Erp, P. Vossen,
be to compare different input resources, or to add A. Fokkens, I. Aldabe, G. Rigau, A. Soroa,
confidence scores to each triple. Another direction T. Ploeger, T. Bogaard, Building event-centric
would be to see how well the system would scale to knowledge graphs from news, Journal of Web
other events such as other revolutions. Lastly, one Semantics 37 (2016) 132–151.
direction could be linked to pattern identification [10] T. Kawamura, S. Egami, K. Tamura, Y. Hoka-
and event forecasting, to generalise from instanti- zono, T. Ugai, Y. Koyanagi, F. Nishino, S. Oka-
ated examples. jima, K. Murakami, K. Takamatsu, et al., Re-
port on the first knowledge graph reasoning
challenge 2018, in: Joint International Seman-
4
tic Technology Conference, Springer, 2019, pp.
18–34.
[11] I. C. Dorobăt, , V. Posea, Enriching the cultural
heritage metadata using historical events: a
graph-based representation, in: International
Conference on Theory and Practice of Digital
Libraries, Springer, 2019, pp. 344–347.
[12] K. Leetaru, P. A. Schrodt, Gdelt: Global data
on events, location, and tone, 1979–2012, in:
ISA annual convention, volume 2, Citeseer,
2013, pp. 1–49.
[13] E. Boschee, J. Lautenschlager, S. O’Brien,
S. Shellman, J. Starz, M. Ward, ICEWS Coded
Event Data, 2015. URL: https://doi.org/10.
7910/DVN/28075. doi:10.7910/DVN/28075.
[14] T. Althoff, X. L. Dong, K. Murphy, S. Alai,
V. Dang, W. Zhang, Timemachine: Time-
line generation for knowledge-base entities, in:
Proceedings of the 21th ACM SIGKDD Inter-
national Conference on Knowledge Discovery
and Data Mining, 2015, pp. 19–28.
[15] H. Kroll, D. Nagel, W.-T. Balke, Modeling
narrative structures in logical overlays on top
of knowledge repositories, in: International
Conference on Conceptual Modeling, Springer,
2020, pp. 250–260.
[16] W. R. Van Hage, V. Malaisé, R. Segers,
L. Hollink, G. Schreiber, Design and use of
the simple event model (sem), Journal of Web
Semantics 9 (2011) 128–136.
5