Clustering of Nouns Using the Adjectives that Describe Them
Name
Birgit Sõrmus
Abstract
The usage of adjectives gives additional information about the nouns they describe. That allows the grouping of similar nouns. The aim of this thesis is to use three different methods to cluster nouns using adjectives that describe them in texts. For clustering, the models used are Spectral Clustering on data of Jaccard similarities, Non-Negative Matrix Factorization and Latent Dirichlet Allocation. The resulting clusters are analysed to determine the relatedness of nouns in clusters and which adjectives are significant for which nouns in language use.
Graduation Thesis language
Estonian
Graduation Thesis type
Bachelor - Computer Science
Supervisor(s)
Sven Laur
Defence year
2021