Case Study: London School of Economics achieves faster, less biased data labeling with Appen

London School of Economics Takes Agile Approach to Data Labeling with Appen

The London School of Economics (LSE) needed a faster, more scalable, and less biased way to label political text data for research in political science. Traditional labeling by expert researchers was slow, expensive, difficult to scale, and limited by the availability of multilingual experts. LSE partnered with Appen, using its data annotation platform and global crowd to support its research projects.

Appen provided a user-friendly platform with validation metrics, reporting features, and access to a diverse global workforce of contributors. This enabled LSE to complete labeling tasks in hours instead of weeks, annotate 20,000 sentences across six political parties, and replicate studies in multiple languages. The results supported published research in top journals and helped LSE build a machine learning model to predict the readability and sophistication of political texts.

London School Of Economics

Kenneth Benoit

Director of the Data Science Institute