Evaluation of latent vocabularies through Zipf’s law and heaps’law

Author

Sano, Yukie and Takayasu, Hideki and Takayasu, Misako

Abstract

We discuss about the number of latent distinct words through simulations by using Zipf’s law and Heaps’ law. From the standpoint of the number of latent distinct words which is estimated by our simulations, we can discuss about the difference among languages, author’s properties such as professional and amateur authors and so on. In addition, Zipf’s law and Heaps’ law can be observed various field, thereby our approach has benefit not only for linguistic word occurrences but also various fields such as ecology and society to estimate hidden system size.