Identifying Temporal Trends Based on Perplexity and Clustering: Are We Looking at Language Change?

Publikation: Bidrag til bog/antologi/rapportKonferencebidrag i proceedingsForskningfagfællebedømt


  • W19-4711

    Forlagets udgivne version, 302 KB, PDF-dokument

In this work we propose a data-driven methodology for identifying temporal trends in a corpus of medieval charters. We have used perplexities derived from RNNs as a distance measure between documents and then, performed clustering on those distances. We argue that perplexities calculated by such language models are representative of temporal trends. The clusters produced using the K-Means algorithm give an insight of the differences in language in different time periods at least partly due to language change. We suggest that the temporal distribution of the individual clusters might provide a more nuanced picture of temporal trends compared to discrete bins, thus providing better results when used in a classification task.
TitelProceedings of the 1st International Workshop on Computational Approaches to Historical Language Change
ForlagAssociation for Computational Linguistics
StatusUdgivet - 2019
BegivenhedComputational Approaches to Historical Language Change 2019: Workshop co-located with ACL 2019 - Florence, Italien
Varighed: 2 aug. 20192 aug. 2019


WorkshopComputational Approaches to Historical Language Change 2019

Antal downloads er baseret på statistik fra Google Scholar og

Ingen data tilgængelig

ID: 227472498