scispace - formally typeset
Search or ask a question
Proceedings ArticleDOI

A thematic approach to emerging narrative structure

19 Jun 2008-pp 41-45
TL;DR: The proposed thematic underpinning could provide the narrative direction which can often be a problem with stories from existing narrative generation methods, and a thematic model of narrative built of narrative atoms and their features, motifs and themes is presented.
Abstract: In this paper we look at the possibility of using a thematic model of narrative to find emergent structure in tagged collections. We propose that a thematic underpinning could provide the narrative direction which can often be a problem with stories from existing narrative generation methods, and present a thematic model of narrative built of narrative atoms and their features, motifs and themes. We explore the feasibility of our approach by examining how collaborative tags in online collections match these properties, and find that while tags match across the model the majority are higher level (matching broader themes and motifs rather than more specific features) which may require further investigation into their utility.

Summary (3 min read)

Introduction

  • Categories and Subject Descriptors H.1 [Models and Principles]: General.
  • Keywords Narrative, Narrative Generation, Thematics, Emergent Structures.

1. I%TRODUCTIO%

  • User generated content on the web (such as blog entries, photos, videos, etc) are often accompanied by explicit virtual structures in the form of tags (overlapping named collections) and other domain-specific collections such as albums (for photos), channels (for videos) and reading lists (for books).
  • These explicit structures give rise to other emergent structures sometimes refereed to as folksonomies, such as emergent vocabularies (e.g. tag clouds) and taxonomies (e.g. Wikipedia categories).
  • Narrative generation is a field that seeks to explore alternative representations of narrative, and investigate the possibility of automatically generating custom stories from information collections.
  • It is their belief that the thematic approach would generate richer stories that benefit from a thematic subtext.

2.1.1 Structuralism

  • Narratology is the study of narrative within literature.
  • Structuralism is an approach to narrative analysis that aims to deconstruct narrative and to learn about the components from which a story is built and how they are connected and contrasted against each other within a narrative work.
  • The discourse however represents what parts of the story are told and how it is told; if the collection is the story then the result of narrative generation (telling the story) is the discourse.
  • A motif is the smallest atomic thematic element and refers to an individual element with the narrative which connotes in some way the theme.

2.1.2 Semiotics

  • Semiotics or semiology is the study of signs and how the authors extract meaning from them.
  • Saussure wrote that all signs are built of two parts [9], a signifier (the physical signal from the sign such as the appearance of an apple) and a signified (the denotation of that sign such as the concept of ‘apple-ness’ or ‘fruit’).
  • Barthes made a distinction between denotative signs (signifiers that lead directly to their signified, such as a word having a literal meaning) and connotative signs (signifiers that lead indirectly to some contextual or culturally important signified, such as the red light implying Stop to a driver) [5].
  • Barthes goes on to point out that should a sign connote something then the signifier of such a sign would itself be built out of a denotative sign (a picture of a red light denotes a red light, red light connotes Stop).
  • In such a way the authors can draw contextual cultural concepts from static basic objects that in a particular context have a greater meaning.

2.2 %arrative Systems

  • Narrative generation has been the focus of a wide range of systems, varying in both their methods and application.
  • Some systems use narrative generation techniques as a way of adding more meaning to information, for example Topia [3] where search results are presented as a discourse.
  • While existing techniques often succeed in generating narratives they have several drawbacks.
  • Narratives generated from story grammars are heavily bound to the rules of a given genre and become very formulaic, and emergent narratives can seem like a bland account of a set of actions as the generation is based on a simple report of what happened in sequence, and as such lacks emphasis and flavor.
  • If direction, emphasis, or the authorial voice could be incorporated into generated narratives then it would lead to less bland or formulaic stories.

3.1 The Model

  • This subtext gives a narrative direction beyond merely communicating a chronology leading to deeper narratives and giving an authorial voice to stories.
  • The authors assume a situation where a story is compiled with many small segments of narrative that are structured together, in this case the selection of these small atomic segments and their content are key to communicating a theme.
  • Themes are connoted by other themes and motifs Features denote Motifs because motifs are directly associated with the feature (normally as a generalized version of it).
  • Using an appropriately populated thematic model the authors could examine the features of those natoms in order to identify motifs and thus potential themes.

3.2 An Example

  • There is a hierarchy of themes (white boxes), motifs (grey boxes) and features (dark grey boxes) under the overall themes of Celebration and Spring.
  • Furthermore motifs may be denoted by any number of features, the example shows how a party motif could be denoted from either a Champagne feature or a Balloon feature.

4. %ARRATIVES AS EMERGE%T STRUCTURES

  • To evaluate the feasibility of applying this model and generating narratives from virtual collections the authors decided to survey some existing collections to see how natoms were tagged, with features, motifs, or themes.
  • Using the example above the authors searched for images on Flickr1 with each feature, motif, and 1 http://www.flickr.com theme, searched for tags of items only, and comprised a table of the average number of results for each main theme (Celebration and Spring) from the example.
  • The results are in Table 1, Averages rounded to nearest value.
  • Any motif can be denoted by a wide range of features, so for any given motif there may be more feature tags in total then tags for the motif itself.

6. REFERE%CES

  • Automatic Ontology-based Knowledge Extraction and Tailored Biography Generation from the Web. IEEE Inteligent Systems, 18, 14-21. [2].
  • Folksonomies versus Automatic Keyword Extraction: An Empirical Study.
  • Procceedings of the fourteenth ACM conference on Hypertext and Hypermedia, 81-84. [8].
  • Story Creation by Inteligent Agents, also known as The Virtual Storyteller.
  • Four Essays, also known as Russian Formalist Criticism.

Did you find this useful? Give us your feedback

Content maybe subject to copyright    Report

A Thematic Approach to Emerging Narrative Structure
Charlie Hargood
Learning Societies Lab
School of Electronics and
Computer Science
University of Southampton
+44 (0)23 8059 7208
cah07r@ecs.soton.ac.uk
David E. Millard
Learning Societies Lab
School of Electronics and
Computer Science
University of Southampton
+44 (0)23 8059 5567
dem@ecs.soton.ac.uk
Mark J. Weal
Learning Societies Lab
School of Electronics and
Computer Science
University of Southampton
+44 (0)23 8059 9400
mjw@ecs.soton.ac.uk
ABSTRACT
In this paper we look at the possibility of using a thematic model
of narrative to find emergent structure in tagged collections. We
propose that a thematic underpinning could provide the narrative
direction which can often be a problem with stories from existing
narrative generation methods, and present a thematic model of
narrative built of narrative atoms and their features, motifs and
themes. We explore the feasibility of our approach by examining
how collaborative tags in online collections match these
properties, and find that while tags match across the model the
majority are higher level (matching broader themes and motifs
rather than more specific features) which may require further
investigation into their utility.
Categories and Subject Descriptors
H.1 [Models and Principles]: General.
General Terms
Standardization, Human Factors, Experimentation.
Keywords
Narrative, Narrative Generation, Thematics, Emergent Structures
1. I%TRODUCTIO%
User generated content on the web (such as blog entries, photos,
videos, etc) are often accompanied by explicit virtual structures in
the form of tags (overlapping named collections) and other
domain-specific collections such as albums (for photos), channels
(for videos) and reading lists (for books). These explicit structures
give rise to other emergent structures sometimes refereed to as
folksonomies, such as emergent vocabularies (e.g. tag clouds) and
taxonomies (e.g. Wikipedia categories).
Folksonomies have several advantages over generated metadata as
they permit richer semantic annotation of collections as opposed
to auto-generated metadata [2].
Narratives (or stories) are a commonly prevalent form of
information representation that are well established as an
engaging way of representing an experience. Narrative generation
is a field that seeks to explore alternative representations of
narrative, and investigate the possibility of automatically
generating custom stories from information collections. There are
a wide variety of different techniques for narrative generation
ranging from structured narrative grammars to emergent
narratives. However the narratives generated can seem flat,
lacking engagement and direction.
In our work we are exploring a thematic approach to solving some
of the problems with narrative generation. The thematic approach
focuses on themes within a story to give narratives a sense of
direction and purpose. For example, rather than simply recounting
photographs taken during a holiday in chronological order, it
might emphasize photos with themes such as relaxation or
celebration to create alternative narratives based on the same
resources and the same events. It is our belief that the thematic
approach would generate richer stories that benefit from a
thematic subtext.
In this paper we present a model representing the thematic
element of narratives. We also explore whether these elements
map to the tags found on shared online resources in order to
explore the feasibility of our approach as a method of creating
emergent narrative structure from collaboratively tagged
materials.
2. BACKGROU%D
2.1 %arratology
2.1.1 Structuralism
Narratology is the study of narrative within literature. It is
primarily focused on narrative analysis and on deconstructing
existing narratives but it also provides a rich theoretical basis for
narrative generation techniques.
Structuralism is an approach to narrative analysis that aims to
deconstruct narrative and to learn about the components from
which a story is built and how they are connected and contrasted
against each other within a narrative work. For narrative
generation this approach is particularly attractive as it defines
tangible objects within a narrative that can be modeled and used
to represent parts of a generated narrative.
Structuralism asserts that a narrative may be deconstructed into a
story and a discourse [4] where the story represents a chronology
of all the information to be communicated and the discourse
represents what parts of the story are told and how those parts are
presented (shown in Figure 1).
Permission to make digital or hard copies of all or part of this work for
personal or classroom use is granted without fee provided that copies are
not made or distributed for profit or commercial advantage and that
copies bear this notice and the full citation on the first page. To copy
otherwise, or republish, to post on servers or to redistribute to lists,
requires prior specific permission and/or a fee.
WebScience’08, June XXX, 2008, Pittsburg, Pennsylvania, USA.
Copyright 2008 ACM 978-1-59593-XXX-X/08/06…$5.00.

The story element of this narrative is constructed by the
observations and experiences that make up the subject of the
narrative. In a virtual collection of resources the story represents
the collection itself, containing all observations and experiences.
The discourse however represents what parts of the story are told
and how it is told; if the collection is the story then the result of
narrative generation (telling the story) is the discourse.
Figure 1. A %arrative can be deconstructed into Story and
Discourse
The discourse is the result of a multitude of different mechanics
including how the story is presented, what medium is used, the
style, the genre, and the themes of the narrative. Thematics is the
discipline of approaching themes within narrative in a structuralist
way, deconstructing and analyzing the relations between the
components that communicate a theme within a narrative.
Tomashevsky deconstructed thematic elements into themes (broad
ideas such as ‘politics’ or ‘drama’) and motif’s (more atomic
elements directly related to the narrative such as ‘the helpful
beast’ or ‘the thespian’) [11]. He describes how themes are
constructed from other themes (sub themes) and motifs. A motif is
the smallest atomic thematic element and refers to an individual
element with the narrative which connotes in some way the
theme. Themes may always be deconstructed into other themes or
motif’s whereas a motif may not be deconstructed.
2.1.2 Semiotics
Semiotics or semiology is the study of signs and how we extract
meaning from them. Saussure wrote that all signs are built of two
parts [9], a signifier (the physical signal from the sign such as the
appearance of an apple) and a signified (the denotation of that
sign such as the concept of ‘apple-ness’ or ‘fruit’).
Barthes made a distinction between denotative signs (signifiers
that lead directly to their signified, such as a word having a literal
meaning) and connotative signs (signifiers that lead indirectly to
some contextual or culturally important signified, such as the red
light implying Stop to a driver) [5]. Barthes goes on to point out
that should a sign connote something then the signifier of such a
sign would itself be built out of a denotative sign (a picture of a
red light denotes a red light, red light connotes Stop). In such a
way we can draw contextual cultural concepts from static basic
objects that in a particular context have a greater meaning.
2.2 %arrative Systems
Narrative generation has been the focus of a wide range of
systems, varying in both their methods and application. Some
systems use narrative generation techniques as a way of adding
more meaning to information, for example Topia [3] where search
results are presented as a discourse. Using narrative as a way of
representing information in this way is similar to various
hypertext projects such as AHA! [7] where the omission,
emphasis, and spatial presentation of information creates a
discourse and gives the information presented additional meaning.
In other systems, such as the virtual storyteller [10], the aim is to
completely generate an entertaining story rather the represent
existing content.
There are many different methods used to generate narrative but
they largely fall into two types; grammar narratives, and emergent
narratives. Grammar narratives work by extensively modeling the
rules of a given genre and using structuralism to create a grammar
of narrative elements. A story is then generated by fitting
prewritten narrative segments together by identifying which
narrative element they match and following the rules of the
grammar. An example of such a system is Artequakt [1] and to an
extent Card Shark [6]. Emergent narratives generate a story by
simulating the story setting and simply presenting what occurs.
Often this is achieved by using agents to play the parts of
characters within a story that follow the rules of the environment
and using a director agent to influence the actor agents into a
creative narrative. Examples of such emergent narratives are
Façade [8] and the Virtual Storyteller [10].
While existing techniques often succeed in generating narratives
they have several drawbacks. Narratives generated from story
grammars are heavily bound to the rules of a given genre and
become very formulaic, and emergent narratives can seem like a
bland account of a set of actions as the generation is based on a
simple report of what happened in sequence, and as such lacks
emphasis and flavor. Both techniques generate narratives that can
tend to lack any authorial voice, leading to narratives without any
emphasis, creating stories without an objective that can seem
directionless. A human author imbeds meaning, subtle themes,
and his/her own goals into a piece these are lacking in any
computer generated narratives. If direction, emphasis, or the
authorial voice could be incorporated into generated narratives
then it would lead to less bland or formulaic stories.
3. A THEMATIC APPROACH
3.1 The Model
Authors use themes to communicate a subtext within a narrative.
This subtext may be an agenda or simply an emphasis of a
particular part of the narrative or even simply an emphasis of the
authors own style. This subtext gives a narrative direction beyond
merely communicating a chronology leading to deeper narratives
and giving an authorial voice to stories. We propose a thematic
under pinning to narrative generation techniques so that richer
narratives with direction may be generated.
To do this we go back to Tomashevsky’s structurist work on
thematics. Features within the narrative denote Motifs and from
these Themes can be identified.
We assume a situation where a story is compiled with many small
segments of narrative that are structured together, in this case the
selection of these small atomic segments and their content are key
to communicating a theme. We use the term Narrative-Atoms or
atoms to describe these segments; small atomic pieces of

narrative that cannot be further broken down, for example a
single photo or paragraph. The content of these natoms is rich
with information, however only some of it visible to a machine
(such as generated metadata and authored tags), we call these
visible computable elements Features. Natoms contain any
number of features which may or may not work towards
connoting a theme in a story. Features can each denote a motif, a
basic thematic object that has connotations within the story, for
example the feature cake denotes the motif of food. These motifs
in turn connote broader themes in the context in which they are
presented, for example food in the context of a gathering may
connote feasting. These themes, when combined with other
themes or motifs could in turn be used to further connote other
themes, for example feasting might connote celebration.
This forms the foundation of our thematic model of a narrative:
Natoms contain tagged features
Features denote motif’s
Themes are connoted by other themes and motifs
Figure 2. The Thematic Model
The model is shown in Figure 2, which also shows how the parts
of the model map to Barthes’ idea of denotative signs as the
signifiers for connotative signs. Features denote Motifs because
motifs are directly associated with the feature (normally as a
generalized version of it). Themes are broader concepts
communicated over the entirety of the narrative, typically by
numerous motifs. By their nature they cannot be denoted as they
rely on some cultural context which cannot be contained within
a natom, as such a theme is a connotation of the motifs, and by
extension the features, within the narrative.
This model is but one part of a narrative generation system, it
contains no rules for the presentation of elements or the
narrative structure. However it can be used to select natoms to
be used within a discourse. As such we could use themes
constructed from this model to influence the story selection in
grammar or emergent narratives to give them a thematic subtext.
When a narrative is formed a part of the story is selected and
then presented as a discourse [4]. We can consider virtual
collections of resources as our story, and should we want to
create a discourse to tell a story of Tuesday it would select all
the natoms (photos, blog entries, etc.) of that day. Using an
appropriately populated thematic model we could examine the
features of those natoms in order to identify motifs and thus
potential themes. Natoms that connote these popular themes
could then be selected or emphasized to create a final discourse
that felt more purposeful. If the virtual collections were very
large we could set out to look for natoms that supported
particular themes, for example, by using public photo collections
to create a discourse (a photo montage) with the themes of
family, winter and Christmas.
Because features could be tagged in any way for such a system
to work every motif object would need a list of features that
could denote the motif. In turn theme object will also require
some way of knowing what motif’s are suitable for them,
however in this case it is less simple as themes are contextual
things not simply denoted. It seems likely that a theme should be
described as having core thematic elements that are required for
a theme to be communicated, such as a wedding theme requiring
a bride motif, as well as optional thematic elements that
exaggerate or promote the theme but are not essential (such as a
religious theme). Themes would need to keep a set of required
and optional thematic elements (both motif’s and sub themes).
The power of the thematic approach will be proportional to the
quantity and richness of these feature-motif and motif-theme
connections.
3.2 An Example
Figure 3 shows an example of an instantiated model. There is a
hierarchy of themes (white boxes), motifs (grey boxes) and
features (dark grey boxes) under the overall themes of
Celebration and Spring. As you can see each of these themes is
made up of 2 sub-themes and a motif, the sub-themes in turn are
made up of a few motifs and each motif is implied by a feature
that could easily be tagged in a natom.
Figure 3 also shows that motifs can be part of a theme but are in
no way bound to it. In the example the inclusion of a party motif
could be used to connote either the Birthday or Easter themes.
Furthermore motifs may be denoted by any number of features,
the example shows how a party motif could be denoted from
either a Champagne feature or a Balloon feature.

4. %ARRATIVES AS EMERGE%T
STRUCTURES
The thematic model described in Section 3 could be used to
influence narrative generation given a set of natoms with
appropriate metadata, but how well might current tagging
behavior support this?
Any user generated virtual collection is an account of some
human experience and as such should contain a potential
narrative; in a sense every blog, photo album, and video has a
story to tell. But the generation of our own custom narratives
from these collections depends on the quality, quantity and
nature of the metadata available. For example if natoms were to
be tagged mainly with themes then a narrative generation system
could find itself starved of features to connote other themes.
While the tagging is still relevant to the thematic approach a
theme has very few connections, whereas a natom tagged with a
feature could be used to denote many different motifs and as
such connote many themes. Referring to our example in Figure
3 if a photo of a bottle of champagne was tagged as
“Celebration!” it would be accurate, however in this case the
photo could only be used to connote a theme of celebration
whereas if the same photo were tagged as “Champagne” it could
also denote the motif of “Party” and subsequently connote the
themes of both Celebration” and “Spring”. As such it is
important to measure where tags fall within the model so that
their utility at constructing themes may be assessed.
To evaluate the feasibility of applying this model and generating
narratives from virtual collections we decided to survey some
existing collections to see how natoms were tagged, with
features, motifs, or themes. Using the example above we
searched for images on Flickr
1
with each feature, motif, and
1
http://www.flickr.com
theme, searched for tags of items only, and comprised a table of
the average number of results for each main theme (Celebration
and Spring) from the example. We also modeled a few other
super themes (Winter, Hedonism, and Childhood) and surveyed
the results for them also. The results are in Table 1, Averages
rounded to nearest value.
Main Theme Themes Motifs Features
Celebration 1,915,532 1,929,864 44,557
Spring 503, 078 1,830,234 214,397
Winter 1,601,127 1,365,610 39,866
Hedonism 8,940 1,800,366 73,384
Childhood 615,775 346,701 204,390
The results shown are varied, all thematic elements appear as
tags, however themes and motifs seem more popular than
features (with the exception of Hedonism). It is possible that this
is because users tag collections more frequently with broader
concepts (themes and motifs) rather than identifying specific
elements (features). If true this could make using our thematic
model difficult as the relatively few number of features tagged
might prove insufficient to build suitable themes.
However this could also be due to our search being driven from
the top of the model (by themes). Any motif can be denoted by a
wide range of features, so for any given motif there may be
more feature tags in total then tags for the motif itself. If we
were to search for a full range of features that could denote these
motifs then we would almost certainly find a more even spread
of feature and motif/theme tags.
Table 1. Average Tag Types
Figure 3 Worked Example

5. CO%CLUSIO% A%D FUTURE WORK
In this paper we have presented a thematic model of narrative
based on the work of structuralism in narratology. Our model
consists of narrative atoms (natoms), features, motifs and
themes. We believe that the model could be used to create
emergent narrative structures from virtual collections that may
have more focus and direction than narratives created just
through existing approaches.
The thematic model can be used to select natoms that together
promote given themes, but this will only result in collections or
montages without some additional structure. We believe that the
thematic underpinnings when coupled with discourse generation
will create greater coherence and causality within generated
narratives; other natom meta-data (such as format and creator)
could also be considered in order to create a stronger sense of
authorial voice and style.
Our initial survey of online collections has indicated that a
thematic approach should be possible, although it is not yet clear
whether there would be enough tags relating to features (which
are more flexible than tags that relate to motifs/themes) to create
narratives with an arbitrary theme. More work needs to be done
to measure whether it would be viable to do this from existing
collaborative tagging collections or if such an approach would
be limited to collections that had been formally tagged in a
specific fashion.
Our intention is to build a thematic narrative generation system
based on online collections in order to investigate the effect
empirically, and to evaluate the effectiveness of thematic
selection in creating narratives with a perceived focus and
direction
6. REFERE%CES
[1] Alani, H. Kim, S. Millard, D. Weal, M. Hall, W. Lewis, P.
and Shadbolt, N. 2003. Automatic Ontology-based
Knowledge Extraction and Tailored Biography Generation
from the Web. IEEE Inteligent Systems, 18, 14-21.
[2] Al-Khalifa, H and Davis, H. 2006. Folksonomies versus
Automatic Keyword Extraction: An Empirical Study.
IADIS International Journal on Computer Science and
Information Systems (IJCSIS), Col 1, 132-143
[3] Alberink, M. Rutledge, L. and Veenstra, M. 2003.
Sequence and Emphasis in Automated Domain-
Independent Discourse Generation. Information Systems,
1-10.
[4] Barthes, R. 1966. Introduction to the Structural Analysis of
Narrative. A Roland Barthes Reader. Heath, S, Trans.
Sontag, S, Ed. 1994. London: Vintage.
[5] Barthes, R. 1996. Mythologies. Lavers, A, Trans. London:
Vintage.
[6] Bernstein, M. 2001. Card shark and thespis: exotic tool for
hypertext narrative. Proceedings of the twelth ACM
conference on Hypertext and Hypermedia, 41-50.
[7] DeBra, P. Aerts, A. Berden, B. de Lange, B. Rousseau, B.
Santic, T. Smits, D. and Stash, N. 2003. AHA! The
Adaptive Hypermedia Architecture. Procceedings of the
fourteenth ACM conference on Hypertext and Hypermedia,
81-84.
[8] Mateas, M. and Stern, A. 2003. Façade: An Experiment in
Building a Fully-Realized Interactive Drama. Game
Developers Conference 2003.
[9] Saussure, F. 1974. Course in General Linguistics. Baskin,
W, Trans. Glasgow: Fontana.
[10] Theune, M. Fass, S. Nijholt, A. and Heylen, D. 2003. The
Virtual Storyteller: Story Creation by Inteligent Agents.
TIDSE 2003: Technologies for Interactive Digital
Storytelling and Entertianment.
[11] Tomashevsky, B. 1965. Russian Formalist Criticism: Four
Essays. Thematics. Lemon, L, T. and Rees, R, J, Ed.
University of Nebraska Press, 66-68
Citations
More filters
01 Jun 2009
TL;DR: In this article, the authors propose that the inclusion of themes will enrich the resulting narrative and explore how a prototype thematic system might be integrated with existing methods of narrative generation, revealing that integration at the generation of story elements is important to avoid constraints on desired themes, that the detailed characters fundamental to character centric narrative generation make integration difficult and that integration must be done carefully to avoid damaging the resulting narratives.
Abstract: Adaptive hypermedia aims to create a tailored experience for users and with narrative generation this dynamic content can be in the form of engaging stories. Narrative generation systems range from the automatic generation of bespoke stories to the representation of existing information as narrative. However narrative generation systems often produce bland and unvaried stories. In this paper we propose that the inclusion of themes will enrich the resulting narrative and explore how a prototype thematic system might be integrated with existing methods of narrative generation. This investigation reveals that integration at the generation of story elements is important to avoid constraints on desired themes, that the detailed characters fundamental to character centric narrative generation make integration difficult, and that integration must be done carefully to avoid damaging the resulting narrative.

37 citations

Proceedings ArticleDOI
13 Jun 2010
TL;DR: This paper evaluates the performance of a system utilising a thematic model in order to generate simple narratives in the form of photo montages compared to a keyword based system that does not and shows that the relevance of the resulting narratives to the titles used to generate them is higher in the thematic system than those generated by the other system.
Abstract: A wide variety of systems could be considered 'narrative systems', either directly working towards generating rich narratives or, more frequently, because they present or handle information in a narrative context. These narratives, generated or otherwise handled, may contain themes; an essential part of the subtext of narrative communicating important concepts outside the capabilities of the literal meaning of the content and forming the thematic cohesion that aids the flow of the presented narrative. However despite this very little work has been undertaken to understand of take advantage of these themes, particularly in narrative generation where the presence of well defined themes may improve the richness of those generated narratives. In this paper we evaluate the performance of a system utilising a thematic model in order to generate simple narratives in the form of photo montages compared to a keyword based system that does not. The experiment demonstrates that the system utilising the thematic model is capable of successfully connoting themes within these narratives. It also shows that the relevance of the resulting narratives to the titles used to generate them is higher in the thematic system than those generated by the other system.

15 citations


Cites methods from "A thematic approach to emerging nar..."

  • ...We propose the use of the thematic model presented in [12] to improve the thematic cohesion in systems that utilise narrative which will then enrich the resulting narratives, which can make them more relevant thematically, and giving them more subtextual depth....

    [...]

  • ...This model was first presented in [12], and then further explored how it could be utilised by narrative systems in [13]....

    [...]

06 Jun 2011
TL;DR: This paper presented a five variable approach to measure narrative cohesion and found that beyond linguistic connections narrative elements are coherently bound together through other concepts and structures such as themes, genre, narrator, and style.
Abstract: In this paper we present a five variable approach to measuring narrative cohesion. Increasingly narratives are dynamically adapted for presentation to enhance personalisation or fit different presentational objectives. Narrative generation systems seek to either automatically generate stories from scratch or, create stories from predefined conditions. With the rise of machines as co-authors it is important to consider what the affect is upon the cohesion of the narratives represented or created in this way. Typically, in existing work, this is limited to an analysis of the use of textual language within the narrative to communicate a coherent message. However we find that beyond linguistic connections narrative elements are coherently bound together through other concepts and structures such as themes, genre, narrator, and style. We present these variables, and features that may be used to identify their presence, as an alternative approach to measuring narrative cohesion and demonstrate their application on two generated narratives.

12 citations


Cites background from "A thematic approach to emerging nar..."

  • ...Thematics however, is an aspect largely un-modelled by most narrative generation systems and as such something we would expect to score lowly, we are pursuing this area in our own work elsewhere [8][9][10]....

    [...]

  • ...In previous work we have explored the concept of modelling themes in narrative [8][9][10] and how we might model individual motifs and their connection to different themes to enable embedding themes in narrative generation or thematic analysis....

    [...]

Proceedings ArticleDOI
29 Jun 2009
TL;DR: This paper describes how a system which generated story selections in the form of photo montages was developed using a thematic model of narrative, and shows that the positive impact of thematic selection increases when the images are presented together.
Abstract: Narrative systems attempt to present users with media collections that include some element of structure or story, however these collections can lack an authorial voice and seem bland as a result. In this paper we explore how themes could be used to enrich automatically generated narratives, and describe how a system which generated story selections in the form of photo montages was developed using a thematic model of narrative. This was achieved by selecting narrative atoms, in this case photographs, from a selection of images on a specific subject with relevance to a desired theme. Our pilot study shows that our thematic system selects images with greater relevance to desired titles, and that the positive impact of thematic selection increases when the images are presented together. We hope that our thematic work will inform others working on narrative systems, and will lead to richer automated narratives.

11 citations


Cites background from "A thematic approach to emerging nar..."

  • ...However, while .ltered Permission to make digital or hard copies of all or part of this work for personal or classroom use is granted without fee provided that copies are not made or distributed for pro.t or commercial advantage and that copies bear this notice and the full citation on the .rst…...

    [...]

Journal Article
TL;DR: This study presents a method for building an automatic knowledge extraction system for filling in biography forms from Turkish texts and shows it to be a promising process to be further developed especially for creating forms in the Turkish language.
Abstract: This study presents a method for building an automatic knowledge extraction system for filling in biography forms from Turkish texts. Several biographies are analyzed in order to choose the set of biography categories to be studied. The fields of the biography form to be created are also defined based on this analysis. Information extraction techniques are used for implementation. A separate testing platform is designed to evaluate the accuracy of the extracted data. Results of the testing platform have shown this study to be a promising process to be further developed especially for creating forms in the Turkish language.

6 citations


Cites methods from "A thematic approach to emerging nar..."

  • ...A separate testing platform is designed to evaluate the accuracy of the extracted data....

    [...]

References
More filters
Book
01 Jan 1916
TL;DR: A brief survey of the history of Linguistics can be found in this paper, with a focus on synchronic and diachronic Linguistic information. But it is not a comprehensive survey of all of the major aspects of linguistics.
Abstract: Introduction to the Bloomsbury Revelations Edition Preface to the First Edition Preface to the Second Edition Preface to the Third Edition Editor's Introduction, Roy Harris Introduction 1. A Brief Survey of the History of Linguistics 2. Data and Aims of Linguistics: Connexions with Related Sciences 3. The Object of Study 4. Linguistics of Language Structure and Linguistics of Speech 5. Internal and External Elements of a Language 6. Representation of a Language by Writing 7. Physiological Phonetics Appendix: Principles of Physiological Phonetics 1. Sound Types 2. Sounds in Spoken Sequences Part One: General Principles 1. Nature of the Linguistic Sign 2. Invariability and Variability of the Sign 3. Static Linguistics and Evolutionary Linguistics Part Two: Synchronic Linguistics 1. General Observations 2. Concrete Entities of a Language 3. Identities, Realities, Values 4. Linguistic Value 5. Syntagmatic Relations and Associative Relations 6. The Language Mechanism 7. Grammar and Its Subdivisions 8. Abstract Entities in Grammar Part Three: Diachronic Linguistics 1. General Observations 2. Sound Changes 3. Grammatical Consequences of Phonetic Evolution 4. Analogy 5. Analogy and Evolution 6. Popular Etymology 7. Agglutination 8. Diachronic Units,Identities and Realities Appendices Part Four: Geographical Linguistics 1. On the Diversity of Languages 2. Geographical Diversity: Its Complexity 3. Causes of Geographical Diversity 4. Propagation of Linguistic Waves Part Five: Questions of Retrospective Linguistics Conclusion 1. The Two Perspectives of Diachronic Linguistics 2. Earliest Languages and Prototypes 3. Reconstructions 4. Linguistic Evidence in Anthropology and Prehistory 5. Language Families and Linguistic Types Index

4,601 citations

Journal ArticleDOI

2,450 citations


"A thematic approach to emerging nar..." refers background in this paper

  • ...Saussure wrote that all signs are built of two parts [9], a signifier (the physical signal from the sign such as the appearance of an apple) and a signified (the denotation of that sign such as the concept of ‘apple-ness’ or ‘fruit’)....

    [...]

Journal ArticleDOI
TL;DR: In the literature, there is a variety of genres, each of which branches out into a wide variety of media, as if all substances could be relied upon to accommodate man's stories as discussed by the authors.
Abstract: of all, there is a prodigious variety of genres, each of which branches out into a variety of media, as if all substances could be relied upon to accommodate man's stories. Among the vehicles of narrative are articulated language, whether oral or written, pictures, still or moving, gestures, and an ordered mixture of all those substances; narrative is present in myth, legend, fables, tales, short stories, epics, history, tragedy, drame [suspense drama], comedy, pantomime, paintings (in Santa Ursula by Carpaccio, for instance), stained-glass windows, movies, local news, conversation. Moreover, in this infinite variety of forms, it is present at all times, in all places, in all societies; indeed narrative starts with the very history of mankind; there is not, there has never been anywhere, any people without narrative; all classes, all human groups, have their stories, and very often those stories are enjoyed by men of different and even opposite cultural backgrounds: narrative remains largely unconcerned with good or bad literature. Like life itself, it is there, international, transhistorical, transcultural. Are we to infer from such universality that narrative is insignificant? Is it so common that we can say nothing about it, except for a modest description of a few highly particularized species, as literary history sometimes does? Indeed how are we to control such variety, how are we to justify our right to distinguish or recognize them? How can we tell the novel from the short story, the tale from the myth, suspense drama from tragedy (it has been done a thousand times) without reference to a common model? Any critical attempt to describe even the most specific, the most historically oriented narrative form implies such a model. It is, therefore, understandable that thinkers as early as Aristotle should have concerned themselves with the study of narrative forms, and not have abandoned all ambition to talk about them, giving

1,260 citations


"A thematic approach to emerging nar..." refers background in this paper

  • ...Structuralism asserts that a narrative may be deconstructed into a story and a discourse [4] where the story represents a chronology of all the information to be communicated and the discourse represents what parts of the story are told and how those parts are presented (shown in Figure 1)....

    [...]

  • ...When a narrative is formed a part of the story is selected and then presented as a discourse [4]....

    [...]

01 Jan 2003
TL;DR: This research presents a meta-game architecture that automates the very labor-intensive and therefore time-heavy and therefore expensive and expensive process of building and interacting with player avatars in real-time.
Abstract: Contemporary games are making significant strides towards offering complex, immersive experiences for players. We can now explore sprawling 3D virtual environments populated by beautifully rendered characters and objects with autonomous behavior, engage in highly visceral action-oriented experiences offering a variety of missions with multiple solutions, and interact in ever-expanding online worlds teeming with physically customizable player avatars.

525 citations


"A thematic approach to emerging nar..." refers background in this paper

  • ...Examples of such emergent narratives are Façade [8] and the Virtual Storyteller [10]....

    [...]

  • ...Examples of such emergent narratives are Façade [8] and the Virtual Storyteller [10]....

    [...]

  • ...Façade: An Experiment in Building a Fully­Realized Interactive Drama....

    [...]

Journal ArticleDOI
TL;DR: The Artequakt project is considered, which links a knowledge extraction tool with an ontology to achieve continuous knowledge support and guide information extraction and is further enhanced using a lexicon-based term expansion mechanism that provides extended ontology terminology.
Abstract: To bring the Semantic Web to life and provide advanced knowledge services, we need efficient ways to access and extract knowledge from Web documents. Although Web page annotations could facilitate such knowledge gathering, annotations are rare and will probably never be rich or detailed enough to cover all the knowledge these documents contain. Manual annotation is impractical and unscalable, and automatic annotation tools remain largely undeveloped. Specialized knowledge services therefore require tools that can search and extract specific knowledge directly from unstructured text on the Web, guided by an ontology that details what type of knowledge to harvest. An ontology uses concepts and relations to classify domain knowledge. Other researchers have used ontologies to support knowledge extraction, but few have explored their full potential in this domain. The paper considers the Artequakt project which links a knowledge extraction tool with an ontology to achieve continuous knowledge support and guide information extraction. The extraction tool searches online documents and extracts knowledge that matches the given classification structure. It provides this knowledge in a machine-readable format that will be automatically maintained in a knowledge base (KB). Knowledge extraction is further enhanced using a lexicon-based term expansion mechanism that provides extended ontology terminology.

490 citations

Frequently Asked Questions (7)
Q1. What would be the way to generate narratives?

If direction, emphasis, or the authorial voice could be incorporated into generated narratives then it would lead to less bland or formulaic stories. 

Structuralism is an approach to narrative analysis that aims to deconstruct narrative and to learn about the components from which a story is built and how they are connected and contrasted against each other within a narrative work. 

Somesystems use narrative generation techniques as a way of adding more meaning to information, for example Topia [3] where search results are presented as a discourse. 

In other systems, such as the virtual storyteller [10], the aim is to completely generate an entertaining story rather the represent existing content. 

Their initial survey of online collections has indicated that a thematic approach should be possible, although it is not yet clear whether there would be enough tags relating to features (which are more flexible than tags that relate to motifs/themes) to create narratives with an arbitrary theme. 

There are many different methods used to generate narrative but they largely fall into two types; grammar narratives, and emergent narratives. 

These explicit structures give rise to other emergent structures sometimes refereed to as folksonomies, such as emergent vocabularies (e.g. tag clouds) and taxonomies (e.g. Wikipedia categories). 

Trending Questions (1)
Thematic Structure in a short story rrl?

The paper discusses the use of a thematic model of narrative to find emergent structure in tagged collections, but it does not specifically mention the analysis of thematic structure in a short story.