Only Connect: Text mining for historians

Nick Poyntz looks at the opportunities offered to historians by text mining, the use of computer programmes to examine concordances and divergences within and between documents and texts.

Text mining is the process of deriving previously unknown information, patterns or trends from large amounts of text. At its simplest, it can involve counting how many times a particular word or phrase occurs, or listing the differences between two texts. The most advanced types of text mining can involve complicated linguistic and statistical techniques. What all types of text mining have in common is that they are not just about searching for information that is already known to exist. At its most powerful, text mining is an example of new research techniques yielding findings that were unknown or unthinkable to any previous generation of historians.

To continue reading this article you will need to purchase access to the online archive.

Buy Online Access  Buy Print & Archive Subscription

If you have already purchased access, or are a print & archive subscriber, please ensure you are logged in.

Please email digital@historytoday.com if you have any problems.