Abstract— Wordnet is a collection of words that interpret or
present a meaning, in its development Wordnet has an important
part, the Synonym Set or Synset. In making Synonym sets,
synonyms are needed and the commutative nature of words is
needed. To get word synonyms, the English language thesaurus
becomes the reference data for taking synonym data. Broadly
speaking, the difference between Wordnet and the dictionary is that
the meaning of the word is related to other words, to determine the
equation requires a commutative process. The process is made easy
by using commutative methods that will produce a candidate
synonym set. Candidates for the synonym set cannot be used for
word syntax, the grouping process of words which produces the
Synonym set as the final result must be carried out. The process of
grouping words can one of them use clustering techniques, in this
study will use Agglomerative Clustering techniques. In the process
of agglomerative clustering techniques there is a threshold value to
determine the number of repetitions or as a condition to stop the
iteration process. The clustering process in this study will use a
threshold value of 0.1 to 1 to test the best threshold value to
produce the best Synonym set and calculate its accuracy value.
Accuracy calculation and evaluation will use the F-measure
method to find the best results.