Creating a phraseme matrix based on a Tertium Comparationis

Diachronic exploration of linguistic resources like collections and dictionaries from different time periods allows researchers to get first impressions on language change and define specific research questions to investigate further, for example by integrating empirical data. However, manual inspection of large collections is exhausting and error prone. Automatic extraction and comparison of the keywords of dictionary entries from several dictionaries can be used to create a combined index, allowing to easily access respective dictionary entries to extract related information. As a case in point we consider information on German phrasemes in dictionaries and collections from the 18th to the 21st century. We use a concept-driven semi-automatic approach to create a matrix based on a Tertium Comparationis to allow users to easily look up related phrasemes.