10 Million International Dyadic Events

This page has been superceeded. Please see the data in the more convenient form on my dataverse:

Can You Spot the Android? These graphs show estimated conflict-cooperation scores (plotted horizontally) by errors (plotted vertically). The machine's performance appears in one of these graphs and the performance of trained human coders appear in the other three. Can you tell which is which? For the answer, see Figure 3 in King and Lowe [ PDF ].

When the Palestinians launch a mortar attack into Israel, the Israeli army does not wait until the end of the calendar year to react. Yet, most modern data collections are aggregated to the month or year. The data available here include almost 10 million individual events, each coded to the exact day they occur or become known. Each event is summarized in the data as "Actor A does something to Actor B", with Actors A and B recording about 450 countries and other (within-country) actors and "does something to" coded in an ontology of about 200 types of actions. The data are coded by computer from millions of Reuters news reports. The software system (produced by VRA) that performs this task has been independently evaluated by King and Lowe (2003). This article found that for the numbers of events it was possible to convince humans (trained Harvard undergraduates) to code by hand, the machine did as well as the humans. For much larger numbers of events for which no expert coder could keep up, the machine dominates.

• Improved and extended data for 1990-2004 for the entire world - almost 10 million events in total. [ Data: ZIP 121MB | Documentation: PDF ]
• Aggregated forms of the above data in MS Access files. [1990-1995 Events: ZIP126MB | 1995-1999 Events: ZIP 185 MB | 2000-2004 Events: ZIP 161 MB | Documentation: DOC ]
• These data are made available as a companion to the article that conducted the independent evaluation: Gary King and Will Lowe. 2003. "An Automated Information Extraction Tool For International Conflict Data with Performance as Good as Human Coders: A Rare Events Evaluation Design," International Organization, 57, 3 (July, 2003): Pp. 617-642. [ PDF | Abstract ]
• The first version of the data, including 3.7 million international dyadic events are still available. [ Documentation & data in MS Access: ZIP 150MB | ASCII text data: ZIP 43 MB ] from the last decade. (Dale Thomas generously contributed this Windows program [ ZIP ] to extract data from these large files) and additional documentation on the event codes [ XLS ].
• Methods for the analysis of rare events: Gary King and Langche Zeng. "Explaining Rare Events in International Relations," International Organization, 55, 3 (Spring, 2001): 693-715; Gary King and Langche Zeng. "Logistic Regression in Rare Events Data," Political Analysis, Vol. 9, No. 2, (Spring, 2001): Pp. 137--63; Gary King and Langche Zeng. 2002. "Estimating Risk and Rate Levels, Ratios, and Differences in Case-Control Studies," Statistics in Medicine, Vol. 21, Pp. 1409-1427; and Gary King and Langche Zeng. 2004. "Inference in Case-Control Studies," in Shein-Chung Chow, ed., Encyclopedia of Biopharmaceutical Statistics, 2nd edition. New York: Marcel Dekker.
• ReLogit Software for analyzing rare events: for Stata and for Zelig and R.
• Concluding comment in a symposium on the analysis of dyadic international conflict data: Gary King. "Proper Nouns and Methodological Propriety: Pooling Dyads in International Relations Data," International Organization, 55, 2 (Fall, 2001): 497-507.