• Forecasting Turmoil in Indonesia: An Application of Hidden Markov Models (2004)
• Integrated Data for Events Analysis (IDEA): An Event Form Typology for Automated Events Data Development (2003)
• The Integrated Data for Events Analysis (IDEA) Website (2002)
• Casualty Assessment and the IDEA Framework (2002)
• Operationalizing and Assessing “TRACE:” A Tool for the Rapid Assessment of Complex Emergencies
• An Automated Information Extraction Tool For International Conflict Data with Performance as Good as Human Coders: A Rare Events Evaluation Design (2002)
• Integrated Data for Events Analysis (IDEA): An Event Form Typology for Automated Events Data Development (2001)
• Conflict-Carrying Capacity, Political Crisis, and Reconstruction: A Framework for the Early Warning of Political System Vulnerability (2001)
• Conflict-Cooperation for Interstate and Intrastate Interactions: An Expansion of the Goldstein Scale (2001)
• Charting the Korean Peninsula: A 1990-2000 Assessment of North Korean Event Interactions (2000)
• Mapping Mass Political Conflict and Civil Society: Issues and Prospects for the Automated Development of Event Data (1997)
• Profiles of International "Hotspots" (1995)
• Recovering Events from Events Data (1995)

Forecasting Turmoil in Indonesia: An Application of Hidden Markov Models (2004)

Joe Bond, Vladimir Petroff, Sean O'Brien, Doug Bond

Abstract

In this paper we illustrate an application of hidden Markov models to the problem of short term forecasting of conflict escalation/de-escalation using Indonesia as an example. Utilizing Integrated Data for Events Analysis (IDEA), the Baum-Welch estimation procedure and the Kullback-Leibler Distance measure trained, estimated, and evaluated the relevancy of the models. Employing a MATHEMATICA©-based simulator, we generate the most likely monthly scenarios as well as the symbol distributions for Indonesia for the year 1999. We conclude with a discussion of some unexpected results derived from the fifty two generated forecasts.

Link to file of the paper

Integrated Data for Events Analysis (IDEA): An Event Form Typology for Automated Events Data Development (2003)

Doug Bond, Joe Bond, Churl Oh, J. Craig Jenkins, Charles Lewis Taylor

Abstract

This article outlines the basic parameters and current status of the Integrated Data for Event Analysis (IDEA) project. IDEA provides a comprehensive events framework for the analysis of international interactions by supplementing the event forms from all earlier projects with new event forms needed to monitor contemporary trends in civil and interstate politics. It uses a more flexible multi-leveled event and actor/target hierarchy that can be expanded to incorporate new event forms and actors/targets, and adds dimensions that can be employed to construct indicators for early warning and assessing conflict escalation. IDEA is currently being used in the automated coding of news reports (Reuters Business Briefs) and, in collaboration with other projects, in the analysis of field reports. We summarize the conceptual framework being used in this data development effort, its major variables, and its geographic and temporal coverage.

Link to file of the paper

The Integrated Data for Events Analysis (IDEA) Website (2002)

Abstract

The IDEA protocol and the VRA Knowledge Manager software system operate together to automatically generate social, economic, enviornmental and political events data and to display them in summary from in terms of event counts and various scales. The IDEA conceptual framework, though optimized for machine coding, offers a useful extension of a human events coding tradtion that extends back nearly forty years. We have sought throughout our development process to preserve backwards compatibility as well as extensibility. Descriptions, examples and usage notes for each of the roughly 200 political, social, economic and environmental events are provided.

Link to the IDEA site

Casualty Assessment and the IDEA Framework (2002) 

Joe Bond (Harvard University), Doug Bond (Harvard University) and Marianne Abbott (The Ohio State Universtiy)

Abstract

This memo reviews the recent and ongoing research activities of the IDEA project, with a focus on the challenge of assessing the intensity of acute conflict in general and the casualties of armed conflict in particular.

Link to file of the paper

Operationalizing and Assessing TRACE: Tool for the Rapid Assessment of Complex Emergencies

Doug Bond, Joe Bond and Churl Oh

Abstract

“TRACE” (Tool for the Rapid Assessment of Complex Emergencies) is a rapid assessment tool for use by analysts seeking real-time data necessary to identify, track and evaluate crisis situations that may evolve into or have already become CHEs. Virtual Research Associates, Inc. (VRA) first outlined the TRACE approach in the final report of a study that was completed in December 1999. TRACE is designed for use in rapidly assessing a range of complex emergency situations, both natural and human-made in their origin.

Link to file of the paper

An Automated Information Extraction Tool For International Conflict Data with Performance as Good as Human Coders: A Rare Events Evaluation Design (2002)

Gary King (Harvard University) and Will Lowe (Harvard University)

Abstract

Despite widespread recognition that aggregated summary statistics on international conflict and cooperation miss most of the complex interactions among nations, the vast majority of scholars continue to employ annual, quarterly, or occasionally monthly observations. Daily events data, coded from some of the huge volume of news stories produced by journalists, have not been used much for the last two decades. We offer some reason to change this practice, which we feel should lead to considerably increased use of these data. We address advances in event categorization schemes and software programs that automatically produce data by "reading" news stories without human coders. We design a method that makes it feasible for the first time to evaluate these programs when they are applied in areas with the particular characteristics of international conflict and cooperation data, namely even categories with highly unequal prevalences, and where rare events (such as highly conflictual actions) are of special interest. We use this rare events design to evaluate one existing program, and find it to be as good as trained human coders, but obviously far less expensive to use. For large scale data collections, the program dominates human coding. Our new evaluative method should be of use in international relations, as well as more generally in the field of computational linguistics, for evaluating other automated information extraction tools. We believe that the data created by programs similar to the one we evaluated should see dramatically increased use in international relations research. To facilitate this process, we are releasing this article data on 4.3 million international events, covering the entire world for the last decade.

Link to Adobe .pdf file of the paper

Integrated Data for Events Analysis (IDEA): An Event Form Typology for Automated Events Data Development (2001)

Doug Bond, Joe Bond, Churl Oh (Harvard University and Virtual Research Associates, Inc.), J. Craig Jenkins (The Ohio State University) and Charles Lewis Taylor (Virginia Polytechnic Institute and State University)

This paper outlines the heritage, basic parameters and current status of an automated approach to events data development that evolved from the PANDA project at Harvard University. The IDEA project has roots in the World Handbook for Political and Social Indicators project as well as the Mass Protest in East Europe and KEDS projects. The ten-year old PANDA project now uses a protocol dubbed IDEA, for the Integrated Data for Events Analysis, which is used to guide and inform the automated coding of news, field and other reports. The paper includes a description of the variables used in the data development effort, along with a description of the current geographic coverage, availability and temporal scope of the data sets being produced. It then reviews the IDEA protocol, focusing on the unit and level of analysis, its reliability and its main assumptions and its implications for use. Also included is a general discussion of the strengths and limitations of automated coding for applied and basic research.

Link to Adobe .pdf of the paper (superceded by Integrated Data for Events Analysis (IDEA): An Event Form Typology for Automated Events Data Development (2003)

Conflict-Carrying Capacity, Political Crisis, and Reconstruction: A Framework for the Early Warning of Political System Vulnerability (2001)

J. Craig Jenkins (The Ohio State University) and Doug Bond (Harvard University)

Abstract

This study provides a conceptual framework for the analysis of conflict-carrying capacity (or CCC) defined as the ability of political systems to regulate intense internal conflicts. CCC is indexed by the multiplicative interaction between the proportions of civil contention, state repression, and violence. Data from the PANDA project is used to illustrate the usefulness of this CCC index in capturing system instability in an institutionalized democracy (the United States), a bureacratic-authoritarian regime (Mexico), an institutionalized Communist regime (China), and a peaceful democratic transition (Poland). It provides early warning signals of civil war (Algeria, Sri Lanka) and moves toward political stability (Peru). Civil contention and state repression are not destablizing per se. Rather it is the simultaneous combination of these with violent contention that leads to internal political crises and, alternatively, to political stabilization.

Link to Adobe .pdf of the paper (available soon)

Conflict-Cooperation for Interstate and Intrastate Interactions: An Expansion of the Goldstein Scale (2001)

Charles Lewis Taylor (Virginia Polytechnic Institute and State University), Joe Bond (Harvard University), Doug Bond (Harvard University), J. Craig Jenkins (The Ohio State University) and Zeynep Benderlioglu Kuzucu (The Ohio State University)

This paper presents the current state of our project to expand and to replicate Goldstein's Conflict-Cooperation Scale for WEIS Event Data using a technique and focusing on intrastate as well as interstate interactions. We report here the results of a pilot study conducted with the assistance of a small expert panel. Our intention is to take advantage of a larger and more diverse set of judges representing the governmental, national security, commercial and academic communities while controlling for major demographic characteristics. In our Integrated Data for Event Analysis (IDEA), classic WEIS categories are supplemented with verb cues drawn from the World Handbook and other interaction event taxonomies. Panelists are asked to rank these event categories, with regard to Contentiousness-accommodation, coercion-altruism, and physical violence and to overall conflict-cooperation. No assumption is made with regard to the three hypothesized dimensions within conflict and cooperation. Interrelationships will be sought empirically. This expansion of the Goldstein scale will be a useful tool in the analysis of conflict and cooperation in international and intra-national political interactions.

Link to Adobe .pdf of the paper

Charting the Korean Peninsula: A 1990-2000 Assessment of North Korean Event Interactions (2000)

Joe Bond, Doug Bond and Churl Oh (Harvard University) This paper introduces the VRAâ Knowledge Manager, a real-time data development and analysis software system designed to process large volumes of text-based information. The system is designed to illuminate countries' vulnerabilities to natural and social hazards by monitoring global news wire service reports, assessing them and presenting up-to-the-minute intelligence graphs, maps and tables. Following a short description of the system, a decade of North Korean events data is analyzed using indicators that tap dimensions of conflict carrying capacity, forceful action and conflict and cooperation.

Link to Adobe .pdf of the paper

Mapping Mass Political Conflict and Civil Society: Issues and Prospects for the Automated Development of Event Data (1997)

Doug Bond (Harvard University), J. Craig Jenkins (The Ohio State University), Charles Lewis Taylor (Virginia Polytechnic Instititute and State University) and Kurt Schock (Rutgers University)

Abstract

Mass political conflict is typically examined in terms of violence and in isolation from routine civil interactions. We argue that mass conflict is multi-dimensional and that violence should be treated as an outcome of conflict, as well as a form of action. We define three dimensions of conflict--contentiousness, coerciveness, and change goals--and indices of the civil society that are central to mapping global trends in mass conflict. We then outline a strategy for mapping mass conflict and civil interactions using the PANDA protocol to generate highly reliable event data and then use these indices to trace two democratic transitions (in Poland and South Korea), a conflict crisis that was repressed (China) and a conflict escalation that flared into a civil war (the former Yugoslavia). Automation has major advantages over human coding in terms of transparency, integration with existing event data series, real time availability and long-term maintenance costs. It also opens up new ways of thinking about event data and the assessment of reliability.

URL to Adobe .pdf of the paper (available soon)

Profiles of International "Hotspots" (1995)

Doug Bond and William B. Vogele (Harvard Univeristy)

Abstract

Three profiles are presented and discussed - conflict intensity, violence signals and conflict carrying capacity. The examination of the level of sanctions being wielded in the conflict and consideration of the context of the news reports in terms of violence results in the conflict intensity profile. In the violence signals profile we consider reports of threats, reported demonstrations and reductions in relationships. The threats and demonstrations signal the possibility of violence and/or mobilization and decreased bonds and relationships weaken the social prohibitions against violence. The third profile reveals the conflict carrying capacity of a system. This profile is designed to illuminate the nature (violent or not) and locus (within or outside the political system) of conflict and is presented in two parts to highlight the distinct roles of the two components. As a backdrop to these three profiles, we graphically present the conflict context or world news attention, focusing on the volume of Reuters reporting on the conflict situation over time. This conflict context is offered to facilitate interpretation of the three profiles, especially to alert us when the volume of reports over any year in a particular country is adequate. We identify three countries or "hotspots" deemed to be most at risk for eruption or escalation into violence: Afghanistan, Algeria and Turkey. Conflict event data profiles were created of these armed conflicts, tracing their evolution from January 1984 through April 1995. These event data were coded from more than a half million worldwide news reports from Reuters. The criteria employed to select these three countries most at risk include recent peak intensities in their level of reported sanctions, report context and proportion of reported threat; recent three year continuous rises in their reported demonstrations or reductions in relationships; and recent conflict carrying capacity scores continuously approaching or falling below, their proportion of foreful action.

Link to Adobe .pdf of the paper (available soon).

Recovering Events from Events Data (1995)

Karen Rothkin (MIT) and Doug Bond (Harvard Univeristy)

Abstract

In this paper we draw upon a large (N=530,718) global event data set to examine source coverage bias to model, in two equations, contextual (i.e. country) differences while trying to predict violent crises. Given various patterns of reporting patterns across countries, we draw upon three kinds of variables--exogenous data on countries; the endogenous event report stream, including theoretically derived indicators of impending crises; and historically threaded events, such as elections, that are not usefully approximated as discrete events. It should be emphasized that this paper reports on an ongoing, theoretically informed yet largely inductive modeling exercise. In other words, we do not test an antecedently derived hypothesis; as the model is refined, we plan to conduct such tests, as well as encourage independent evaluation. Toward this end, and in the spirit of the September 1995 PS Symposium on "Verification/Replication," all of our event data is publicly available.

Link to Adobe .pdf of the paper.