Workshop on Computer-Based Text Coding - Aug. 15-17, 2007

Text-based data has become increasingly fundamental for answering key theoretical concerns of Political Scientists and Sociologists in both Europe and the United States. In recent decades, social scientists have generated a variety of comprehensive text-based data sets, but for the most part they have done it "the old fashion way" -- that is, with painstaking human labor. At the same time, computer scientists have made great advances in the identification of appropriate text and the development and implementation of automated schemes for the classification of elements of that text. Different identification and classification systems may be most useful for different applications.

The Workshop on Automated Text Identification and Classification brings together a distinguished group of social science researchers who are actively generating text-based data for their own work, and a small group of innovators in the use of electronic means of identification and classification of text data. The computer science community has much to gain from greater access to large databases that have already been coded to high quality standards (thus allowing them to test the accuracy of their automated processes). Social scientists may benefit from vast increases in efficiency if parts of their work can be automated, allowing much larger research projects to be envisioned. This is particularly timely as vast text-based data archives are increasingly becoming available on-line, including mass media, government, and other print sources covering not only the current period, but long historical periods as well. The main goal of the workshop is to introduce these diverse communities to each other and to foster future collaborations across the disciplines.

For more information about the conference, please click the links on the right.