Exploring Topics in Historical Sources

Latent Dirichlet Allocation (LDA) topic modeling is a method for analyzing the distribution of semantic word clusters, so-called “topics” in collection of texts. In history it can be used for exploring the content of a corpus of digital sources.

Topic modeling is based on a statistical analysis of co-occurrence for words. It does not use dictionaries or external training data. Topic modeling is a flexible method for computational text analysis.

The Topics Explorer of DARIAH-DE is designed as a very user friendly tool. It allows researchers to experiment with topic modeling on their own computers.

The work flow from plain text to visualized results is done in a graphical user interface. The software allows users to load a collection of plain-text or xml-files and to analyze it by means of the LDA algorithm.

It runs on Windows, MacOS and Linux without installation. Try it here: https://dariah-de.github.io/TopicsExplorer/

Guido Koller

Senior Historian, Swiss Federal Archives, CH-3003 Berne, Switzerland

More Posts - Website

Follow Me: