No full-text available
To read the full-text of this research,
you can request a copy directly from the authors.
Toward Generative Narrative Models of the Course and Resolution of Conflict
Abstract
Over the last 15 years, interest in narrative as a concept for supporting national defense has grown considerably. But efforts to win the battle of the narrative have yielded limited results because of a reliance on a definition of narrative rooted in the spoken or written word. In this chapter, we adopt a contemporary view of narrative as a cognitive process of comprehension and argue that narratives can be modeled as complex systems, providing insight into how actors understand consequences of events in the real world. To accomplish this, we propose a graph‐based generative modeling framework and illustrate its application to well‐known fairy tale. We demonstrate that standard network measures can discern structural details about the narrative that would normally require reading the text. We also show that various graph modeling methods can potentially identify information that may be missing in the graph, a necessary feature for applying the technique in situations of less‐than‐complete information. We conclude by identifying the challenges the technique faces in terms of analysis, scale, and sensitivity.
... A formal narrative modeling, simulation, and analysis environment can enable users to (1) identify salient narratives in the operating context, (2) anticipate which narratives are likely to be connected to actions, (3) recognize favorable and unfavorable narratives to objectives, and (4) create opportunities for more favorable narratives. Discovering narratives used by the population is difficult, and including them for analysis in modeling and simulation is challenging but rewarding (Bar-Tal et al., 2014;Corman et al., 2019;Paul, 2019). The strategy presented in this article is predicated on the belief that by embracing the complexity of human understanding via cognitive computational models of narrative comprehension, one can successfully develop more in-depth insight into the formation and evolution of complex intergroup conflicts and explore strategies to mitigate or counter them in a methodical manner. ...
... According to Paul (2019), computational models can facilitate the identification of narratives salient in a given situation and analyzing their dynamics concerning competing and adversarial narratives in an evolving context. In Corman et al. (2019), authors present a generative modeling framework that views a narrative as a conceptual map in terms of a directed graph that links events, people, places, actions, goals, and so on. Similarly, narrative-based knowledge is increasingly finding its place in cognitive architectures and systems (León, 2016). ...
Humans make sense of the world through narratives. Therefore, adversaries often use conflict-sustaining narratives to maintain dominance and delegitimize the actions of the rivals. To better understand narratives' role and influence in such intractable conflicts, a computational framework and methodology are introduced. The computational cognitive model and its underlying inference mechanism allow analysts to simulate and analyze narratives in relation to opposing narratives. The ability to simulate the interaction of adversarial stories with a set of micronarratives shared by members of a group opens new avenues to counter conflict-sustaining narratives. The methodology and its application to a concrete conflict scenario demonstrate how to conduct simulation-driven exploratory analysis over a complex adaptive narrative space to discern how narratives are matched to unfolding events and how they can be used to facilitate favorable change.
This chapter summarizes our view of priority challenges for social‐behavioral modeling. Some challenges are inherent: individuals behave in complicated ways and social systems are complex adaptive systems (CAS). Such systems pose wicked problems for policy‐makers. Other challenges relate to specific scientific problems. We discuss a selection of these: rebalancing the portfolio of models and methods, confronting uncertainty, synthesis, the dynamic of connecting theory and data, and redefining model validity. We summarize briefly elements of a way ahead.
This chapter describes narrative‐related research for the US Marine Corps and then highlights corresponding challenges for social‐behavioral modeling.
Proceedings of International Virtual Conference on Recent Trends and Techniques in Mathematical and Computer Science.
In this paper we present a broad overview of the last 40 years of research on cognitive architectures. Although the number of existing architectures is nearing several hundred, most of the existing surveys do not reflect this growth and focus on a handful of well-established architectures. While their contributions are undeniable, they represent only a part of the research in the field. Thus, in this survey we wanted to shift the focus towards a more inclusive and high-level overview of the research in cognitive architectures. Our final set of 86 architectures includes 55 that are still actively developed, and borrow from a diverse set of disciplines, spanning areas from psychoanalysis to neuroscience. To keep the length of this paper within reasonable limits we discuss only the core cognitive abilities, such as perception, attention mechanisms, learning and memory structure. To assess the breadth of practical applications of cognitive architectures we gathered information on over 700 practical projects implemented using the cognitive architectures in our list. We use various visualization techniques to highlight overall trends in the development of the field. Our analysis of practical applications shows that most architectures are very narrowly focused on a particular application domain. Furthermore, there is an apparent gap between general research in robotics and computer vision and research in these areas within the cognitive architectures field. It is very clear that biologically inspired models do not have the same range and efficiency compared to the systems based on engineering principles and heuristics. Another observation is related to a general lack of collaboration. Several factors hinder communication, such as the closed nature of the individual projects (only one-third of the reviewed here architectures are open-source) and terminological differences.
Several variants of the Long Short-Term Memory (LSTM) architecture for
recurrent neural networks have been proposed since its inception in 1995. In
recent years, these networks have become the state-of-the-art models for a
variety of machine learning problems. This has led to a renewed interest in
understanding the role and utility of various computational components of
typical LSTM variants. In this paper, we present the first large-scale analysis
of eight LSTM variants on three representative tasks: speech recognition,
handwriting recognition, and polyphonic music modeling. The hyperparameters of
all LSTM variants for each task were optimized separately using random search
and their importance was assessed using the powerful fANOVA framework. In
total, we summarize the results of 5400 experimental runs (about 15 years of
CPU time), which makes our study the largest of its kind on LSTM networks. Our
results show that none of the variants can improve upon the standard LSTM
architecture significantly, and demonstrate the forget gate and the output
activation function to be its most critical components. We further observe that
the studied hyperparameters are virtually independent and derive guidelines for
their efficient adjustment.
The article presents a model-oriented approach to how third-person narratives are read. Building on Minsky's (1979 [1975]) theory of frames, Jackendoff's (1983; 1987) concept of preference rules, Perry's (1979) theory of literary dynamics, and Sternberg's (1982b) Proteus Principle, its main aim is to conceptualize third-person narrative situations (Stanzel 1984) in terms of cognitive models, and to explore the mechanics of top-down/bottom-up hermeneutic processes. Avoiding classical "low-structuralist" narratology with its "normal case" approach, the essay also proposes new ways of analyzing protean phenomena like description, free indirect discourse, and parenthetical discourse. It presents an integrative account of primacy/recency conflicts and sketches the possible direction of a genuinely reading-oriented narratology.
Narrative structure is an ubiquitous and intriguing phenomenon. By virtue of structure we recognize the presence of Villainy or Revenge in a story, even if that word is not actually present in the text. Narrative structure is an anvil for forging new artificial intelligence and machine learning techniques, and is a window into abstraction and conceptual learning as well as into culture and its in influence on cognition. I advance our understanding of narrative structure by describing Analogical Story Merging (ASM), a new machine learning algorithm that can extract culturally-relevant plot patterns from sets of folktales. I demonstrate that ASM can learn a substantive portion of Vladimir Propp's in influential theory of the structure of folktale plots. The challenge was to take descriptions at one semantic level, namely, an event timeline as described in folktales, and abstract to the next higher level: structures such as Villainy, Stuggle- Victory, and Reward. ASM is based on Bayesian Model Merging, a technique for learning regular grammars. I demonstrate that, despite ASM's large search space, a carefully-tuned prior allows the algorithm to converge, and furthermore it reproduces Propp's categories with a chance-adjusted Rand index of 0.511 to 0.714. Three important categories are identied with F-measures above 0.8. The data are 15 Russian folktales, comprising 18,862 words, a subset of Propp's original tales. This subset was annotated for 18 aspects of meaning by 12 annotators using the Story Workbench, a general text-annotation tool I developed for this work. Each aspect was doubly-annotated and adjudicated at inter-annotator F-measures that cluster around 0.7 to 0.8. It is the largest, most deeply-annotated narrative corpus assembled to date. The work has significance far beyond folktales. First, it points the way toward important applications in many domains, including information retrieval, persuasion and negotiation, natural language understanding and generation, and computational creativity. Second, abstraction from natural language semantics is a skill that underlies many cognitive tasks, and so this work provides insight into those processes. Finally, the work opens the door to a computational understanding of cultural in influences on cognition and understanding cultural differences as captured in stories.
A semi-automated approach to evidential reasoning uses template-based structured argumentation. Graphical depictions convey
lines of reasoning, from evidence through to conclusions.
Many intellectual disciplines contribute to the field of cognitive science: psychology, linguistics, anthropology, and artificial intelligence, to name just a few. Cognitive science itself originated in the desire to integrate expertise in these traditionally separate disciplines in order to advance our insight into cognitive phenomena, phenom ena like problem solving, decision making, language, memory, and learning. Each discipline has a history of asking certain types of questions and accepting certain types of answers. And that, according to Allen Newell, a founder of the field of artificial intelligence, is both an advantage and a problem.
This paper presents a visual editor which supports authors to define the narrative macrostructure of non-linear interactive dramas. This authoring tool was used to represent Propp's narrative macrostructure of Russian fairy tales in non-linear plot graphs. Moreover, Propp's thorough characterization of basic nar- rative constituents by explanations, discussions and examples of their different re- alizations in his corpus is utilized to construct an automatic classification model. A semi-automatic classification supports (i) authors to associate new scenes with basic narrative constituents and (ii) players to control the narrative flow in the story engine. For the latter task, the selection of an appropriate plot element and behavioral pattern within the dialog model in response to player interactions is controlled by similarities between stimuli and known realizations of basic narra- tive constituents or behavioral patterns. This approach tackles the main challenge of interactive drama — to balance interactivity and storyness.
Impressive image captioning results (i.e., an objective description for an image) are achieved with plenty of training pairs. In this paper, we take one step further to investigate the creation of narrative paragraph for a photo stream. This task is even more challenging due to the difficulty in modeling an ordered photo sequence and in generating a relevant paragraph with expressive language style for storytelling. The difficulty can even be exacerbated by the limited training data, so that existing approaches almost focus on search-based solutions. To deal with these challenges, we propose a sequence-to-sequence modeling approach with reinforcement learning and adversarial training. First, to model the ordered photo stream, we propose a hierarchical recurrent neural network as story generator, which is optimized by reinforcement learning with rewards. Second, to generate relevant and story-style paragraphs, we design the rewards with two critic networks, including a multi-modal and a language-style discriminator. Third, we further consider the story generator and reward critics as adversaries. The generator aims to create indistinguishable paragraphs to human-level stories, whereas the critics aim at distinguishing them and further improving the generator by policy gradient. Experiments on three widely-used datasets show the effectiveness, against state-of-the-art methods with relative increase of 20.2% by METEOR. We also show the subjective preference for the proposed approach over the baselines through a user study with 30 human subjects.
This volume introduces the concept of Islamist extremist ‘master narratives’ and offers a method for identifying and analyzing them. Drawing on rhetorical and narrative theories, the chapters examine thirteen master narratives and explain how extremists use them to solidify their base, recruit new members, and motivate actions. © Jeffry R. Halverson, H. L. Goodall Jr., and Steven R. Corman, 2011.
Objective
This essay explores the narratives that are used to justify terrorist attacks by perpetrators and their supporters.
Method
The author examines the ways in which narratives rationalizing violence have been constructed and repeated by Islamist extremists and, in the case of Oklahoma City bombing, by McVeigh and other members of the Patriot movement.
Results
The analysis of these two case studies demonstrates the power of narrative rationality to motivate terrorist acts and suggests narrative counterstrategies are key to undermining that power.
Islamic extremism is the dominant security concern of many contemporary governments, spanning the industrialized West to the developing world. Narrative Landmines explores how rumors fit into and extend narrative systems and ideologies, particularly in the context of terrorism, counter-terrorism, and extremist insurgencies. Its concern is to foster a more sophisticated understanding of how oral and digital cultures work alongside economic, diplomatic, and cultural factors that influence the struggles between states and non-state actors in the proverbial battle of hearts and minds. Beyond face-to-face communication, the authors also address the role of new and social media in the creation and spread of rumors. As narrative forms, rumors are suitable to a wide range of political expression, from citizens, insurgents, and governments alike, and in places as distinct as Singapore, Iraq, and Indonesia-the case studies presented for analysis. The authors make a compelling argument for understanding rumors in these contexts as "narrative IEDs," low-cost, low-tech weapons that can successfully counter such elaborate and expansive government initiatives as outreach campaigns or strategic communication efforts. While not exactly the same as the advanced technological systems or Improvised Explosive Devices to which they are metaphorically related, narrative IEDs nevertheless operate as weapons that can aid the extremist cause. Copyright © 2012 by Daniel Leonard Bernardi, Pauline Hope Cheong, Chris Lundry, and Scott W. Ruston. All rights reserved.
The Handbook of Rational Choice Social Research offers the first comprehensive overview of how the rational choice paradigm can inform empirical research within the social sciences. This landmark collection highlights successful empirical applications across a broad array of disciplines, including sociology, political science, economics, history, and psychology. Taking on issues ranging from financial markets and terrorism to immigration, race relations, and emotions, and a huge variety of other phenomena, rational choice proves a useful tool for theory- driven social research. Each chapter uses a rational choice framework to elaborate on testable hypotheses and then apply this to empirical research, including experimental research, survey studies, ethnographies, and historical investigations. Useful to students and scholars across the social sciences, this handbook will reinvigorate discussions about the utility and versatility of the rational choice approach, its key assumptions, and tools.
Real-world physical and abstract data objects are interconnected, forming gigantic, interconnected networks. By structuring these data objects and interactions between these objects into multiple types, such networks become semi-structured heterogeneous information networks. Most real-world applications that handle big data, including interconnected social media and social networks, scientific, engineering, or medical information systems, online e-commerce systems, and most database systems, can be structured into heterogeneous information networks. Therefore, effective analysis of large-scale heterogeneous information networks poses an interesting but critical challenge. In this book, we investigate the principles and methodologies of mining heterogeneous information networks. Departing from many existing network models that view interconnected data as homogeneous graphs or networks, our semi-structured heterogeneous information network model leverages the rich semantics of typed nodes and links in a network and uncovers surprisingly rich knowledge from the network. This semi-structured heterogeneous network modeling leads to a series of new principles and powerful methodologies for mining interconnected data, including: (1) rank-based clustering and classification; (2) meta-path-based similarity search and mining; (3) relation strength-aware mining, and many other potential developments. This book introduces this new research frontier and points out some promising research directions. Table of Contents: Introduction / Ranking-Based Clustering / Classification of Heterogeneous Information Networks / Meta-Path-Based Similarity Search / Meta-Path-Based Relationship Prediction / Relation Strength-Aware Clustering with Incomplete Attributes / User-Guided Clustering via Meta-Path Selection / Research Frontiers
Measurement error, an inherent quality of any empirical data collection technique, is discussed in the context of sociometric data. These data have long been assumed to possess face validity and to be the data of choice in any study of the sentiment structure of small scale social systems. However, it is argued that while methods of sociometric analysis have become increasingly more sophisticated they have failed to yield unequivocal results because they do not distinguish structural complexity from measurement error. Through a discussion of increasingly more complex examples the distortion laden character of most sociometric data is illustrated. This distortion is introduced by the formalities of the sociometric test and it will not be removed by developing increasingly more sophisticated structural models or throwing out some of the data. Instead, when issues concerning the nature of specific relational networks are raised data of much higher quality than those which are commonly available are required. A technique for generating high quality sociometric data is briefly discussed. On the other hand, it is suggested that the extant body of sociometric data ought to be adequate when sizeable aggregations are examined for evidence of statistical tendencies in structure.
This essay proposes a theory of human communication based on a conception of persons as homo narrans. It compares and contrasts this view with the traditional rational perspective on symbolic interaction. The viability of the narrative paradigm and its attendant notions of reason and rationality are demonstrated through an extended analysis of key aspects of the current nuclear war controversy and a brief application to The Epic of Gilgamesh. The narrative paradigm synthesizes two strands in rhetorical theory: the argumentative, persuasive theme and the literary, aesthetic theme.
Given Q nodes in a social network (say, authorship net- work), how can we find the node/author that is the center- piece, and has direct or indirect connections to all, or most of them? For example, this node could be the common ad- visor, or someone who started the research area that the Q nodes belong to. Isomorphic scenarios appear in law en- forcement (find the master-mind criminal, connected to all current suspects), gene regulatory networks (find the pro- tein that participates in pathways with all or most of the given Q proteins), viral marketing and many more. Connection subgraphs is an important first step, handling the case of Q=2 query nodes. Then, the connection sub- graph algorithm finds the b intermediate nodes, that provide a good connection between the two original query nodes. Here we generalize the challenge in multiple dimensions: First, we allow more than two query nodes. Second, we allow a whole family of queries, ranging from 'OR' to 'AND', with 'softAND' in-between. Finally, we design and compare a fast approximation, and study the quality/speed trade-off. We also present experiments on the DBLP dataset. The experiments confirm that our proposed method naturally deals with multi-source queries and that the resulting sub- graphs agree with our intuition. Wall-clock timing results on the DBLP dataset show that our proposed approximation achieve good accuracy for about 6 : 1 speedup.
Making you cry is the folk wisdom test for strong narrative involvement, and has long been proposed as a reachable but distant
goal for digital games. Immersion is a related phenomenon which has been difficult to achieve in Virtual Reality and Mixed
Reality research projects, but reliably achieved in theme park rides by integrating interactivity with strong narrative elements.
For Virtual and Mixed Reality environments to achieve narrative power and strong immersion we can adopt the strategies of
theme park rides, with special attention to the creation of threshold objects that take the interactor into the virtual space.
Spatial presentation of the world, interactions with characters, and differentiating real from virtual events are design challenges
without clear genre conventions. Design choices that locate the interactor in a coherent point of view, limit natural language
expectations, and conventionalize the boundaries of the virtual world will reinforce immersion and dramatic agency.
Node centrality measures are important in a large number of graph applications, from search and ranking to social and biological network analysis. In this paper we study node centrality for very large graphs, up to billions of nodes and edges. Various definitions for centrality have been proposed, ranging from very simple (e.g., node degree) to more elaborate. However, measuring centrality in billion-scale graphs poses several challenges. Many of the "traditional" definitions such as closeness and betweenness were not designed with scalability in mind. Therefore, it is very difficult, if not impossible, to compute them both accurately and efficiently. In this paper, we propose centrality measures suitable for very large graphs, as well as scalable methods to effectively compute them. More specifically, we propose effective closeness and LINERANK which are designed for billion-scale graphs. We also develop algorithms to compute the proposed centrality measures in MAPREDUCE, a modern paradigm for large-scale, distributed data processing. We present extensive experimental results on both synthetic and real datasets, which demonstrate the scalability of our approach to very large graphs, as well as interesting findings and anomalies.
We review the literature on semi-supervised learning, which is an area in machine learning and more generally, artificial intelligence. There has been a whole
spectrum of interesting ideas on how to learn from both labeled and unlabeled data, i.e. semi-supervised learning. This document is a chapter excerpt from the author’s
doctoral thesis (Zhu, 2005). However the author plans to update the online version frequently to incorporate the latest development in the field. Please obtain the latest
version at http://www.cs.wisc.edu/~jerryzhu/pub/ssl_survey.pdf
How closely related are two nodes in a graph? How to compute this score quickly, on huge, disk-resident, real graphs? Random walk with restart (RWR) provides a good relevance score between two nodes in a weighted graph, and it has been successfully used in numerous settings, like automatic captioning of images, generalizations to the "connection subgraphs", personalized PageRank, and many more. However, the straightforward implementations of RWR do not scale for large graphs, requiring either quadratic space and cubic pre-computation time, or slow response time on queries. We propose fast solutions to this problem. The heart of our approach is to exploit two important properties shared by many real graphs: (a) linear correlations and (b) block- wise, community-like structure. We exploit the linearity b y using low-rank matrix approximation, and the community structure by graph partitioning, followed by the Sherman- Morrison lemma for matrix inversion. Experimental results on the Corel image and the DBLP dabasets demonstrate that our proposed methods achieve significant savings over the straightforward implementations: they can saveseveral orders of magnitude in pre-computation and storage cost, and they achieve up to 150x speed up with 90%+ quality preservation. second one is to compute the matrix inversion on the fly, say, through power iteration ("OnTheFly" method). The first method is fast on query time, but prohibitive on space (quadratic on the number of nodes on the graph), while the second is slow on query time. Here we propose a novel solution to this challenge. Our approach, B LIN, takes the advantage of two prop- erties shared by many real graphs: (a) the block-wise, community-like structure, and (b) the linear correlations across rows and columns of the adjacency matrix. The pro- posed method carefully balances the off-line pre-processing cost (both the CPU cost and the storage cost), with the re- sponse quality (with respect to both the accuracy and the response time). Compared to PreCompute, it only requires pre-computing and storing the low-rank approximation of a large but sparse matrix, and the inversion of some small size matrices. Compared with OnTheFly, it only need a few matrix-vector multiplication operations in on-line respo nse process.
We consider variations of two well-known centrality measures, betweenness and closeness, witha different model of information spread. Rather than along shortest paths only, it is assumed that information spreads efficiently like an electrical current. We prove that the current-flow variant of closeness centrality is identical with another known measure, information centrality, and give improved algorithms for computing bothmeasures exactly. Since running times and space requirements are prohibitive for large networks, we also present a randomized approximation scheme for current-flow betweenness.
Agent-based social simulation modeling, a social phenomena on the basis of models of autonomous agents has grown tremendously in recent decades. Researchers use this approach to study a wide range of social and economic issues, including social beliefs and norms, resource allocation, traffic patterns, social cooperation, stock market dynamics, group interaction and dynamics, and organizational decision making. Agent-based social simulation with multiagent systems in social computing can benefit from incorporating cognitive architectures. A cognitive architecture is a domain-generic computational cognitive model that captures essential structures and processes of the individual mind for the purpose of a broad (multiple-domain) analysis of cognition and behaviour.
This paper provides a brief description of the project goals and design requirements, discusses the problem of autonomy in the context of story-based believable agents, and finally describes an architecture that uses the dramatic beat as a structural principle to integrate plot and character
Handbook of Ethnography
- M. Cortazzi
Reframing the narrative of the global war on terrorism
- Sloggett D.
Proceedings of the SIAM international conference on data mining
- L. Akoglu
- D. Chau
- F. Faloutsos
Rational Choice Theory: Potential and Limits
- L. Eriksson
Proceedings of the 3rd International Conference on Virtual Storytelling
- K. Hartmann
- S. Hartmann
- M. Feustel
Narrative led operations
- Nissen T.
Understanding and Communicating Through Narratives
- G. Paruchabutr
Hypertextes et Hypermédias
- J. Sánchez
- M. Lumbreras
Proceedings of the SIAM International Conference on Data Mining
- U. Kang
- S. Papadimitriou
- J. Sun
- H. Tong
Proceedings of the 6th International Conference on Data Mining
- H. Tong
- C. Faloutsos
- J. Pan
- Eriksson
- Wittek