Introduction to automated text analysis for research and evaluation

A practical overview of natural language processing techniques that can be used in research and evaluation projects

This short guide introduces three techniques that can be useful for automated analysis of text data in research and evaluation work: sentiment analysis, topic detection, and summarisation. The aim of the guide is to describe what is possible rather than technical details, and each technique is illustrated with a practical application. Data requirements and limitations of automated analysis are also discussed, so that researchers and evaluators can understand how to use these methods effectively and appropriately.

Download the guide (PDF)