=Paper=
{{Paper
|id=Vol-3322/short3
|storemode=property
|title=On Design Principles for Narrative Information Systems
|pdfUrl=https://ceur-ws.org/Vol-3322/short3.pdf
|volume=Vol-3322
|authors=Hermann Kroll,Wolf-Tilo Balke
|dblpUrl=https://dblp.org/rec/conf/ijcai/KrollB22
}}
==On Design Principles for Narrative Information Systems==
https://ceur-ws.org/Vol-3322/short3.pdf
On Design Principles for Narrative Information Systems
Hermann Kroll1 , Wolf-Tilo Balke1
1
Institute for Information Systems, TU Braunschweig, Mühlenpfordtstr. 23, Braunschweig, 38106, Germany
Abstract
Information systems enable users to represent, manipulate, and access data effectively. Currently, the development and
availability of extensive knowledge bases receive a lot of attention for supporting a variety of intelligent applications like
smart assistants, question answering, and knowledge discovery. But building such knowledge bases generally means breaking
down information into manageable pieces (e.g., in the form of triples in RDF) that tend to lose connections and extraction
contexts. In contrast, humans typically share and exchange knowledge by interweaving different pieces into narratives that
are plausible and easy to grasp. In this position paper, we summarize the main research directions of narratives in computer
sciences. Moreover, we propose basic design principles that narrative information systems should consider in practice. In
particular, we take a closer look at narrative representations, possible bindings between narratives and real-world data, the
context-compatibility of information, and finally a narrative’s plausibility.
Keywords
Narratives, Information Systems, Plausibility, Context, Knowledge Representation
1. Introduction to Narratives sentation of (mostly entity-centric) knowledge [2] that
is adapted in most of today’s knowledge bases. The cen-
Supporting narratives for knowledge exchange is a new tral idea is to store knowledge in the form of triples, i.e.,
challenge for information systems research, as humans subject-predicate-object tuples like (Albert Einstein, was
tend to process information easier whenever they are born in, Germany). These triples are often called facts
presented as (a part of) coherent narratives [1]. The term or statements. A collection of such facts is then called a
narrative is, however, hard to grasp: Some understand knowledge base. In recent years, extensive knowledge
the term narrative synonymously to the concept of a bases have been collected mostly from Web data: The
story, in essence referring to a temporally or causally Wikidata [3] project, for example, includes knowledge
ordered sequence of events. Some specifically distinguish from many different domains. Wikidata users can either
between fictional and non-fictional narratives. Others access the knowledge in a browsing fashion or retrieve
talk about the intention of a narrative, e.g., to convince it using the structured query language SPARQL.
others of a particular stance. Besides storing knowledge about entities, there are
Our research is focused on the exchange of knowl- also knowledge bases that capture knowledge about other
edge through meaningful patterns, i.e., we understand a types of things. In particular, there is an increasing inter-
narrative as a possible form to share knowledge. In this est in knowledge bases focusing on events, e.g., Knowly-
paper, we focus on the exchange and transfer of knowl- wood captures information about activities mined from
edge through plausible narratives. Therefore, we first Hollywood narratives [4], and the EventKG captures all
examine how established systems represent knowledge kinds of information about a variety of events [5]. To
and discuss their limitations. We continue with a brief efficiently build such collections, methods have been pro-
introduction to narratives in computer science and move posed to extract events from textual sources, e.g., events
forward to our position: Narratives need to be handled from news [6], temporal facts and events [7], or events
as first-class citizens in information systems. from Wikipedia [8]. For a good overview on event ex-
traction, see [9].
Knowledge Bases. So how do current systems rep- In brief, the resulting knowledge bases store factual
resent knowledge? In contrast to classical relational knowledge about entities or events, e.g., an entity’s name,
databases or warehouses for well-structured data, the age, and type or some event’s date, location, and partici-
development of the Resource Description Framework pants. Moreover, it also reflects typed relations between
(RDF) within the standardization efforts of the Semantic pairs of entities and events. So is this knowledge repre-
Web has allowed a structured and simple-to-use repre- sentation sufficient for information systems? From our
perspective, knowledge bases indeed effectively repre-
IJCAI-ECAI 2022 workshop: Semantic Techniques for Narrative-based sent and allow access to factual knowledge, in the sense
Understanding, July 24, 2022, Vienna, Austria of universally valid statements. In strict correspondence
$ kroll@ifis.cs.tu-bs.de (H. Kroll); balke@ifis.cs.tu-bs.de (W. Balke)
0000-0001-9887-9276 (H. Kroll); 0000-0002-5443-1215 (W. Balke)
to the real world, such statements can consistently co-
© 2022 Copyright for this paper by its authors. Use permitted under Creative
Commons License Attribution 4.0 International (CC BY 4.0).
exist within a knowledge base. But what happens to
CEUR
Workshop
Proceedings
http://ceur-ws.org
ISSN 1613-0073
CEUR Workshop Proceedings (CEUR-WS.org) knowledge that is not universally valid? For instance,
11
Narrative
CACM prerequisite Implementation of a
Substory: “Invention” for relational DB
published in
developed participant
Edgar Codd Relational Model
turns into Successful
Oracle
tech-company
Bindings Context-Compatibility
|timeframe=1980s
|timeframe=1980s
Figure 1: Design Principles for Narrative Information Systems by an example: The narrative structures the story that Oracle has
turned into a successful tech-company after implementing a relational database, for which the development and publishing of
the relational model by Edgar Codd was an essential prerequisite. The different relationships can be bound against knowledge
of DBLP, Wikipedia, and stock values. The bindings share a common context in the sense of time, i.e., timeframe 1980s (the
remaining bindings are valid in general and thus are compatible too). These bindings are thus context-compatible and cover
the whole narrative – making it plausible.
some statements may either reflect subjective opinions and were collected in a respective schema database [14].
or they may be only valid in certain contexts or under Previous works generally assume an order and relation
certain conditions or even only with a certain probability. of events in texts that needs to be mined. Tang et al. pro-
The resulting knowledge containing also such facts base posed neural methods to learn the respective relations
might easily become inconsistent or contradictory. Thus, between events from text directly [15]. Another approach
while knowledge bases are effective for factual knowl- proposed the mining of so-called plot graphs [16]. In con-
edge, they suffer in representing the connection between trast to describing plots as conjunctions of events, such
statements. plot graphs also allowed disjunctions between events,
A good overview of the limitation of today’s knowl- optional, and mutually exclusive events. In brief, a plot
edge bases is given in [10]. Suchanek argues for moving graph is a template to craft a certain narrative, i.e., the
beyond triple-based structures to support negations, dy- plot graph determines which sequences of events are
namic behavior, and beliefs. For example, a certain state- allowed.
ment might be accepted by a group of people and rejected
by some other group: how to deal with both pieces of Narrative Applications. Various applications have
information within the same repository? As another ex- been proposed to utilize narratives in information sys-
ample take two individual statements being prerequisite tems. Narrative schemas and plot graphs are used for
and reason for a third statement: currently this cannot narrative planning, i.e., to write specific stories by com-
be represented in a knowledge base, at least not with puters [17, 18]. Another application is the alignment of
a triple-like structure. Suchanek argues for represent- narratives and movie scripts, i.e., linking narrative com-
ing such situations as frames, i.e., scopes that connect ponents (e.g., fall in love) to performed actions (e.g., A
knowledge pieces in different ways. In contrast, we ar- kisses B) in movies [19].
gue to reconsider on how humans exchange knowledge: The digital library project Europeana utilizes nar-
through plausible narratives. ratives to compose events involving artists into time-
lines [20]. The narratives are then shown to users as
Narratives. Narratives have been studied in computer overviews of certain artists, i.e., to tell a story about them.
science for a while. Sequences of events or actions to The benefits of such a representation have already been
model human behavior and situation awareness go back studied [21]. Moreover, knowledge bases like Wikidata
to Schank and Abelson [11] in the form of scripts. [12] can be utilized to populate narratives with additional
proposed the representation of narratives as a sequence factual information [22].
of events: Narrative schemas have been introduced to Finally, Shahaf et al. propose to connect the dots be-
represent a set of events and their semantic roles [13] tween news articles in so-called metro maps [23]. Metro
12
maps consist of articles as stations and topics as metro different kinds of relationships: factual and narrative
lines. Users may then follow a line (topic) across different relationships. Like knowledge bases, factual relation-
articles. In addition to news items, Shahaf et al. have ships describe properties involving entities and literals,
demonstrated how scientific articles can be arranged as e.g., that Edgar Codd developed the relational model. These
metro maps of sciences, too [24]. In summary, such Metro relationships can be put between entities and literals. We
maps allow users to understand better the spatial, causal, compose the actual progress of a narrative by a set of
or temporal structure of available resources. We under- narrative relationships, e.g., the implementation of a rela-
stand such maps as a possible application of narratives tional database system by Oracle that has since turned into
aligning similar textual sources in a meaningful way. a successful tech-company. In brief, narrative relations
feature special, non-factual labels. They can be placed
between events or between events and entities, but not
2. Narratives and Plausibility between entities. In this way, narrative can represent
static knowledge and interweave such facts into dynamic
Our main objective is to integrate narratives in infor-
progress (e.g., some causal structure, some temporal flow,
mation systems to allow a meaningful representation
etc.).
of knowledge. Therefore, we discuss four central de-
The last thing a narrative representation must con-
sign principles that narrative information systems must
sider is nesting. With nesting, also called recursion or
consider: narrative representation, bindings, context-
induction, we refer to using narratives inside narratives,
compatibility, and plausibility.
e.g., a single narrative leads to another narrative. Nesting
When is a narrative meaningful? While we agree
supports recursive elements as components, known from
that meaningfulness is strongly connected to the actual
spoken languages [26]. In our example, the development
structure of a narrative, our research is focused on the
of the relational model forms a narrative on its own that
plausibility of narratives. We argue that plausible nar-
is, in turn, the prerequisite for the actual implementation.
ratives have a meaning. Consider our example in Fig. 1:
While we agree that nesting makes the representation
The development of the relational model by Edgar Codd
more challenging, especially concerning precise seman-
was a prerequisite for the implementation of a relational
tics, we argue that nesting allows a higher expressiveness
database system by Oracle. Oracle has then turned into
of narratives, e.g., the occurrence of narrative a’s events
a successful tech-company. The narrative is considered
together finally lead to the whole narrative b.
plausible because of two reasons: First, all of its parts
have evidence in knowledge repositories. Second, the
timeframe of the implementation matches the timeframe Narrative Bindings. For now, narratives are rather
of Oracle’s stock value increase. With that in mind, we artificial structures. However, the question remains, how
first argue on narrative representations and then con- can we estimate whether the narrative is plausible? We
tinue with key concepts to assess the plausibility of a argue that evidence for a narrative is one way to go: Evi-
narrative. dence allows us to estimate if a narrative is grounded by
some real-world data. If a structure is based on real-world
data, the narrative itself is supported and has thus evi-
Narrative Representation. We understand a narra-
dence. Such a grounding leads to a better understanding
tive as a structure that describes some progress, e.g., of
of whether the narrative is plausible.
temporal nature (one event might follow another event),
The next step for a plausibility assessment is thus to
of conditional nature (one event might depend on the
connect narratives with knowledge repositories. In our
other event being observed before), or of causal nature
understanding, a knowledge repository can be any col-
(one event leads to another event). In previous work, we
lection of knowledge, be it a collection of articles, a
designed narratives as logical overlays on top of knowl-
structured database, a data set repository, etc. We call
edge repositories [25]. Here, we summarize what a nar-
the connection between a narrative’s relationship and a
rative representation should consider.
knowledge repository a binding [25]. In other words, the
Components of a narrative are entities (relevant
binding binds a certain relationship against real-world
things of interest, concepts, etc.), events (something
data. It gives, in this way, evidence for it. In our example,
which happens, some observed state, a change of some
the development of the relational model by Edgar Codd
state, etc.) and literals (values). We made the distinction
can be bound against DBLP (a bibliographic database of
between entities and events to distinguish between rather
computer science). The implementation of a relational
static things from the real world and dynamic happenings
database by Oracle could be bound against a Wikipedia
that have a temporal dimension or represent some state
article. In brief, different knowledge repositories may
or state change.
require different computations of bindings [27]. To give
A narrative then puts these components in relation
a few examples:
to each other. Therefore, we distinguish between two
13
Structured Repositories like databases and knowl- ships of narratives against knowledge that is valid under
edge bases usually support structured query lan- the same context conditions.
guages such as SPARQL, Gremlin, or SQL: Bind- The question remains how contexts can be represented:
ings for those knowledge sources may be com- On the one hand, explicit context models allow one to
puted by translating narrative components to re- model every single condition explicitly. For example, Mc-
spective queries. Carthy proposed a model based on first-order predicate
logic [30]. Temporal-restricted or spatial-restricted state-
Relational Data Sets Tabular data formats comprise ments might be a good example of such explicit models.
a large amount of knowledge. However, the Suppose each statement is attached with a temporal in-
computation of bindings against those data re- terval defining the statement’s validity. In that case, a
quires a certain amount of knowledge about meta- reasoning process could then only consider statements
data [28]. A binding must first determine which that are valid in the same time interval, e.g., a compatible
columns of a table represent a component in a timeframe as introduced in our running example.
narrative and afterward find a statistical measure- On the other hand, explicitly modeling every condition
ment to determine if a narrative’s relationship is can be cost-intensive and, in some domains, may be close
supported. In our example, an application of this to impossible. We, therefore, proposed an implicit context
procedure is the measurement of the success of model [31]: If statements are used in textual sources, e.g.,
Oracle against stock values. scientific publications or articles, we assume that these
Textual Sources Computing bindings for textual articles should implicitly include the relevant context
sources can roughly be done in two ways: 1. conditions. We then proposed textual or metadata-based
preprocessing of the text into a structured (e.g., authors) metrics to estimate whether two statements
representation (e.g., graphs) [29] or 2. text-based are context-compatible, e.g., if the articles, they belong
retrieval methods like traditional keyword- to, are similar enough.
based retrieval or textual entailment methods. While explicit models allow a controlled fusion, in
Especially domains that can be described by the sense of correctness, they might be cost-intensive to
controlled vocabularies (e.g., the biomedical develop and maintain. Implicit models are cheaper and
domain), benefit from graph representations [29]. easier to use. However, they do not guarantee correctness
and may lag behind in quality and explainability. For an
extensive discussion on contexts, we refer the reader to
Context-Compatibility. Briefly speaking, bindings
our previous article on context models [32].
allow us to find evidence for a narrative. But can we
just compute bindings for each relationship separately to
ground the whole narrative? While some pieces of infor- Plausibility. With the basic requirements of narrative
mation are universally applicable and easy to connect to structures, bindings, and context-compatibility discussed,
other pieces, such as the birth date of some person, there we can now define when we call a narrative plausible.
may be pieces that are only true within specific semantic
Definition 1. We call a narrative 𝑛 plausible, iff the fol-
settings. For instance, the capital of some country may
lowing conditions hold:
change over time and thus should only be connected to
pieces valid in the same time frame. Thus, when extract- 1. There exists a set of bindings NB that bind each
ing pieces of information from knowledge repositories, relationship of 𝑛 against real-world data.
it is necessary to consider potential contexts. Contexts 2. This set of bindings NB must be context-compatible,
are given by constraints on environment variables that i.e., there is at least a single context to which all
describe under which condition a certain piece of infor- bindings agree (are valid in).
mation is valid.1
Narrative information systems must consider contexts We call the task to find such an answer narrative query
to support a valid information fusion of pieces, i.e., to processing [32]. The central idea is that if we can bind
fuse only pieces that are valid under the same context the whole narrative structure, and all bindings are also
conditions. In our example, the turning into a successful context-compatible, then there is decisive evidence that
tech-company should be valid in the same timeframe as the narrative is plausible – again in the sense of having
the implementation of the actual relational database. We evidence.
call bindings context-compatible if they bind relation- However, although we defined narrative plausibility,
there still remain challenges for a practical application.
1
Please note that for the scope of this paper, we are abstracting We proposed four major dimensions that influence the
from the actual truth or correctness of each piece of information, overall plausibility of a narrative [33]:
which may rather reflect on the respective repository’s trustworthi-
ness than provide a context in the sense of validity.
14
Narrative Structure: A slight change of a narrative, Narrative Information Access. Narrative informa-
e.g., adding, editing, or leaving out some compo- tion access allows user to formulate their information
nent, can affect the result of a plausibility assess- need as a narrative. Then query processing takes place
ment, i.e., different narratives might end up in to make the narrative plausible, i.e., finding context-
different assessments (plausible/not plausible). compatible bindings that cover the whole narrative. We
already demonstrated the benefits of narrative informa-
Validation Approach: Which knowledge repositories, tion access to digital libraries in [32, 29].
and especially, which methods are used to com-
pute the bindings? What do the methods guaran-
Linking Data Sets to Narratives. Today, research
tee? And thus, what could a user then expect?
data management has become more apparent than ever.
Types of Evidence: Are we just looking for direct bind- But connecting published data sets and claims from sci-
ings? What about counter-examples, i.e., state- entific publications remains challenging. If we could link
ments that contradict a narrative’s relationship? both, we, on the one hand, could explain what a data
set tells, and on the other hand, verify claims of new
Confidence of Bindings: Each binding should be con- publications by existing data sets [28].
nected to confidence, i.e., a score that describes
how certain we are that the relationship can be Narrative Event Aspects. For now, the previous ar-
bound against some knowledge repository. There- gumentation focused on the exchange of knowledge
fore, confidence is related to two properties: the through narratives. How a narrative, or its events, are
trustworthiness of knowledge repositories and perceived by an audience also plays a central role today
the quality of the binding process (e.g., the confi- (think about fake news and social media perception here).
dence of a retrieval or NLP method). Attributions and roles could further extend the narrative
representation for richer semantics to better describe
3. Narrative Information Systems these events [34, 35].
We do not believe there will be a unified narrative model
that rules all purposes. Nevertheless, we have seen deci-
4. Conclusions
sive properties that should be considered. From our per- The design and implementation of narrative informa-
spective, a narrative information system must thus tion systems promise to support a wide range of novel
consider: and exciting applications to support human-centered
Narrative Representation that makes a distinction workflows, e.g., by accessing and explaining knowledge
between factual knowledge (e.g., entities and their through plausible narratives in the sense of storytelling.
properties) and dynamic progress (e.g., causation Knowledge shared in this way is bound to allow for a
and temporality). better and easier understanding of how a piece of knowl-
edge is placed in the respective domain. And of course,
Bindings that connect artificial narrative structures such models could even be extended further, for instance,
with real-world data and events to give evidence by adding attributions also to support subjective aspects
for the narrative. of knowledge in the sense of opinions, reflect special
user intents in how knowledge can or should be used, or
Context-Compatibility that embeds information express emotional stances towards entities, events, and
pieces into contexts and defines rules when temporal or causal developments.
contexts are compatible to support a valid In summary, the central argument of this paper is that
information fusion. in order to build effective narrative information systems,
Plausibility of narratives by binding the whole narra- we need to treat narratives as first-class citizens within
tive with context-compatible bindings. these systems. Arguably there might not be a single
model that applies perfectly to all use cases. Nevertheless,
So, why should we implement a narrative information there are some key properties and essential problems that
system? In contrast to pure knowledge bases, narra- every narrative information system needs to consider,
tive information systems allow us to interweave factual such as extraction methods, representational issues, or
knowledge into plausible patterns. This way, the knowl- the possibility and validity of information fusion. This
edge is represented in a pattern like humans would ex- paper opens up the relevant design dimensions and points
change their thoughts. Reaching this target then allows to some early-stage solutions.
a set of useful applications.
15
5. Acknowledgments Narrative Extraction From Texts co-located with
42nd European Conference on Information Re-
Supported by the Deutsche Forschungsgemeinschaft trieval, Text2Story@ECIR 2020, Lisbon, Portugal,
(DFG, German Research Foundation): PubPharm – the April 14th, 2020 [online only], volume 2593 of CEUR
Specialized Information Service for Pharmacy (Gepris Workshop Proceedings, CEUR-WS.org, 2020, pp. 95–
267140244). 104. URL: http://ceur-ws.org/Vol-2593/paper12.pdf.
[11] R. Schank, R. Abelson, Scripts, Plans, Goals, and
Understanding, Routledge, 1977.
References [12] N. Chambers, D. Jurafsky, Unsupervised Learn-
[1] J. László, The Science of Stories: An Introduction ing of Narrative Event Chains, in: ACL 2008, Pro-
to Narrative Psychology, Routledge, 2008. doi:10. ceedings of the 46th Annual Meeting of the As-
4324/9780203894934. sociation for Computational Linguistics, June 15-
[2] F. Manola, E. Miller, B. McBride, et al., Rdf primer, 20, 2008, Columbus, Ohio, USA, The Association
W3C recommendation 10 (2004) 6. for Computer Linguistics, 2008, pp. 789–797. URL:
[3] D. Vrandecic, M. Krötzsch, Wikidata: a free collab- https://aclanthology.org/P08-1090/.
orative knowledgebase, Commun. ACM 57 (2014) [13] N. Chambers, D. Jurafsky, Unsupervised Learning
78–85. doi:10.1145/2629489. of Narrative Schemas and their Participants, in:
[4] N. Tandon, G. de Melo, A. De, G. Weikum, Know- ACL 2009, Proceedings of the 47th Annual Meet-
lywood: Mining Activity Knowledge From Holly- ing of the Association for Computational Linguis-
wood Narratives, in: Proceedings of the 24th ACM tics and the 4th International Joint Conference on
International on Conference on Information and Natural Language Processing of the AFNLP, 2-7
Knowledge Management, CIKM ’15, Association for August 2009, Singapore, The Association for Com-
Computing Machinery, New York, NY, USA, 2015, puter Linguistics, 2009, pp. 602–610. URL: https:
p. 223–232. doi:10.1145/2806416.2806583. //aclanthology.org/P09-1068/.
[5] S. Gottschalk, E. Demidova, EventKG - the hub [14] N. Chambers, D. Jurafsky, A Database of Nar-
of event knowledge on the web - and biographical rative Schemas, in: Proceedings of the Inter-
timeline generation, Semantic Web 10 (2019) 1039– national Conference on Language Resources and
1070. doi:10.3233/SW-190355. Evaluation, LREC 2010, 17-23 May 2010, Valletta,
[6] E. Kuzey, J. Vreeken, G. Weikum, A Fresh Look on Malta, European Language Resources Association,
Knowledge Bases: Distilling Named Events from 2010. URL: http://www.lrec-conf.org/proceedings/
News, in: Proceedings of the 23rd ACM Interna- lrec2010/summaries/58.html.
tional Conference on Conference on Information [15] J. Tang, H. Lin, M. Liao, Y. Lu, X. Han, L. Sun, W. Xie,
and Knowledge Management, CIKM ’14, Associ- J. Xu, From Discourse to Narrative: Knowledge Pro-
ation for Computing Machinery, New York, NY, jection for Event Relation Extraction, in: Proceed-
USA, 2014, p. 1689–1698. doi:10.1145/2661829. ings of the 59th Annual Meeting of the Association
2661984. for Computational Linguistics and the 11th Interna-
[7] E. Kuzey, G. Weikum, Extraction of Temporal tional Joint Conference on Natural Language Pro-
Facts and Events from Wikipedia, in: Proceed- cessing (Volume 1: Long Papers), Association for
ings of the 2nd Temporal Web Analytics Work- Computational Linguistics, Online, 2021, pp. 732–
shop, TempWeb ’12, Association for Computing 742. doi:10.18653/v1/2021.acl-long.60.
Machinery, New York, NY, USA, 2012, p. 25–32. [16] B. Li, S. Lee-Urban, G. Johnston, M. O. Riedl,
doi:10.1145/2169095.2169101. Story Generation with Crowdsourced Plot Graphs,
[8] A. Spitz, M. Gertz, Terms over LOAD: Leveraging in: Proceedings of the Twenty-Seventh AAAI
Named Entities for Cross-Document Extraction and Conference on Artificial Intelligence, July 14-
Summarization of Events, in: Proceedings of the 18, 2013, Bellevue, Washington, USA, AAAI
39th International ACM SIGIR conference on Re- Press, 2013. URL: http://www.aaai.org/ocs/index.
search and Development in Information Retrieval, php/AAAI/AAAI13/paper/view/6399.
SIGIR 2016, Pisa, Italy, July 17-21, 2016, ACM, 2016, [17] J. R. Meehan, The Metanovel: Writing Stories by
pp. 503–512. doi:10.1145/2911451.2911529. Computers, Yale University, 1976.
[9] W. Xiang, B. Wang, A Survey of Event Extraction [18] Mark O. Riedl and Robert Michael Young, Narrative
From Text, IEEE Access 7 (2019) 173111–173137. Planning: Balancing Plot and Character, J. Artif.
doi:10.1109/ACCESS.2019.2956831. Intell. Res. 39 (2010) 217–268. doi:10.1613/jair.
[10] F. M. Suchanek, The Need to Move beyond Triples, 2989.
in: Proceedings of Text2Story - Third Workshop on [19] P. Mirza, M. Abouhamra, G. Weikum, AligNarr:
Aligning Narratives on Movies, in: Proceedings
16
of the 59th Annual Meeting of the Association for due to Covid-19 outbreak), volume 2860 of CEUR
Computational Linguistics and the 11th Interna- Workshop Proceedings, CEUR-WS.org, 2021, pp. 57–
tional Joint Conference on Natural Language Pro- 63. URL: http://ceur-ws.org/Vol-2860/paper7.pdf.
cessing, ACL/IJCNLP 2021, (Volume 2: Short Pa- [28] D. Nagel, T. Affeldt, W. Balke, Data Narrations -
pers), Virtual Event, August 1-6, 2021, Association Using flexible Data Bindings to support the Repro-
for Computational Linguistics, 2021, pp. 427–433. ducibility of Claims in Digital Library Objects, in:
doi:10.18653/v1/2021.acl-short.54. Proceedings of the Workshop on Digital Infrastruc-
[20] C. Meghini, V. Bartalesi, D. Metilli, F. Benedetti, In- tures for Scholarly Content Objects (DISCO 2021)
troducing Narratives in Europeana: Preliminary co-located with ACM/IEEE Joint Conference on
Steps, in: New Trends in Databases and Infor- Digital Libraries 2021(JCDL 2021), Online (Due to
mation Systems - ADBIS 2017 Short Papers and the Global Pandemic), September 30, 2021, volume
Workshops, AMSD, BigNovelTI, DAS, SW4CH, DC, 2976 of CEUR Workshop Proceedings, CEUR-WS.org,
Nicosia, Cyprus, September 24-27, 2017, Proceed- 2021, pp. 19–23. URL: http://ceur-ws.org/Vol-2976/
ings, volume 767 of Communications in Computer short-2.pdf.
and Information Science, Springer, 2017, pp. 333–342. [29] H. Kroll, J. Pirklbauer, J. Kalo, M. Kunz, J. Ruth-
doi:10.1007/978-3-319-67162-8\_33. mann, W. Balke, Narrative Query Graphs for Entity-
[21] C. Meghini, V. Bartalesi, D. Metilli, F. Benedetti, In- Interaction-Aware Document Retrieval, in: To-
troducing narratives in Europeana: A case study, wards Open and Trustworthy Digital Societies -
Int. J. Appl. Math. Comput. Sci. 29 (2019) 7–16. 23rd International Conference on Asia-Pacific Dig-
doi:10.2478/amcs-2019-0001. ital Libraries, ICADL 2021, Virtual Event, Decem-
[22] D. Metilli, V. Bartalesi, C. Meghini, A Wikidata- ber 1-3, 2021, Proceedings, volume 13133 of Lecture
based tool for building and visualising narratives, Notes in Computer Science, Springer, 2021, pp. 80–95.
Int. J. Digit. Libr. 20 (2019) 417–432. doi:10.1007/ doi:10.1007/978-3-030-91669-5\_7.
s00799-019-00266-3. [30] J. McCarthy, Notes on Formalizing Context, in: Pro-
[23] D. Shahaf, C. Guestrin, Connecting the dots be- ceedings of the 13th International Joint Conference
tween news articles, in: Proceedings of the 16th on Artificial Intelligence. Chambéry, France, August
ACM SIGKDD International Conference on Knowl- 28 - September 3, 1993, Morgan Kaufmann, 1993,
edge Discovery and Data Mining, Washington, DC, pp. 555–562. URL: http://www-formal.stanford.edu/
USA, July 25-28, 2010, ACM, 2010, pp. 623–632. jmc/context3/context3.html.
doi:10.1145/1835804.1835884. [31] H. Kroll, J. Kalo, D. Nagel, S. Mennicke, W. Balke,
[24] D. Shahaf, C. Guestrin, E. Horvitz, Metro maps Context-Compatible Information Fusion for Scien-
of science, in: The 18th ACM SIGKDD Interna- tific Knowledge Graphs, in: Digital Libraries for
tional Conference on Knowledge Discovery and Open Knowledge - 24th International Conference
Data Mining, KDD ’12, Beijing, China, August 12- on Theory and Practice of Digital Libraries, TPDL
16, 2012, ACM, 2012, pp. 1122–1130. doi:10.1145/ 2020, Lyon, France, August 25-27, 2020, Proceed-
2339530.2339706. ings, volume 12246 of Lecture Notes in Computer
[25] H. Kroll, D. Nagel, W. Balke, Modeling Nar- Science, Springer, 2020, pp. 33–47. doi:10.1007/
rative Structures in Logical Overlays on Top of 978-3-030-54956-5\_3.
Knowledge Repositories, in: Conceptual Model- [32] H. Kroll, F. Plötzky, J. Pirklbauer, W.-T. Balke, What
ing - 39th International Conference, ER 2020, Vi- a Publication Tells You—Benefits of Narrative Infor-
enna, Austria, November 3-6, 2020, Proceedings, mation Access in Digital Libraries, in: Proceedings
volume 12400 of Lecture Notes in Computer Sci- of the 22nd ACM/IEEE Joint Conference on Digital
ence, Springer, 2020, pp. 250–260. doi:10.1007/ Libraries, JCDL ’22, ACM, New York, NY, USA, 2022.
978-3-030-62522-1\_18. doi:10.1145/3529372.3530928.
[26] M. D. Hauser, N. Chomsky, W. T. Fitch, The Faculty [33] H. Kroll, N. Mainzer, W. Balke, On Dimensions
of Language: What Is It, Who Has It, and How of Plausibility for Narrative Information Access
Did It Evolve?, Science 298 (2002). doi:10.1126/ to Digital Libraries, in: Linking Theory and
science.298.5598.1569. Practice of Digital Libraries - 26th International
[27] H. Kroll, D. Nagel, M. Kunz, W. Balke, Demon- Conference on Theory and Practice of Digital Li-
strating Narrative Bindings: Linking Discourses braries, TPDL 2022, Padua, Italy, September 20-23,
to Knowledge Repositories, in: Proceedings of 2022, Proceedings, volume 13541 of Lecture Notes
Text2Story - Fourth Workshop on Narrative Ex- in Computer Science, Springer, 2022, pp. 433–441.
traction From Texts held in conjunction with the doi:10.1007/978-3-031-16802-4\_43.
43rd European Conference on Information Retrieval [34] F. Plötzky, W. Balke, What it Boils Down to...
(ECIR 2021), Lucca, Italy, April 1, 2021 (online event The Case for Event Analogies to Combat Com-
17
plexity in Digital Libraries, in: Proceedings of
the Workshop on Digital Infrastructures for Schol-
arly Content Objects (DISCO 2021) co-located with
ACM/IEEE Joint Conference on Digital Libraries
2021(JCDL 2021), Online (Due to the Global Pan-
demic), September 30, 2021, volume 2976 of CEUR
Workshop Proceedings, CEUR-WS.org, 2021, pp. 24–
28. URL: http://ceur-ws.org/Vol-2976/short-3.pdf.
[35] F. Plötzky, W. Balke, It’s the Same Old Story! En-
riching Event-Centric Knowledge Graphs by Nar-
rative Aspects, in: WebSci ’22: 14th ACM Web
Science Conference 2022, Barcelona, Spain, June
26 - 29, 2022, ACM, 2022, pp. 34–43. doi:10.1145/
3501247.3531565.
18