Arabic language is still facing some difficulties in automatic processing relating to the richness, morphology,
phonetic and lexicon. This paper presents a new strategy for building a morphological field dictionary for Arabic
language. Our strategy is divides into two parts. The first, extracts an efficient Field Association FA words for each
domain specific. Second, generates the Part Of Speech (POS) tagging for these (FA) word and collect them in one
frame. After that, the FA words with its frame collected in alphabetic order. The method of building the automatic
morphological field dictionary using a main algorithm is discussed and studied. The advantage of our approach is to
build an extended and updated automatic Arabic morphological field dictionary. The average of the accuracy measures
(F measures) of the experimental results is up to 76 %.