Lab 5: Morphological Analysis
Dr. Amena Helmy
Objective
• Develop a morphological analysis tool that can perform a deep
functional analysis for the following Arabic words
"أسقيناكموه"" ,وسنكتبهم"" ,أسنحاربهم"" ,أعطيناه"" ,قتلوهم"" ,رأينا"" ,وسنستقبلهم",
"يرون"" ,كتب"" ,يكتب"" ,كتبوا"" ,أفسنقتلهم"" ,أفستقتلونهم"" ,أفسيعطيكموه",
"تعاون"" ,أوبالقلم"" ,لإلنسان"" ,ولي"" ,فهم"" ,وهم" ",سنلعب"" ,المدرسات",
"كتاباتهم"" ,مديرتنا"" ,طالبيهم"" ,المسلمين"" ,طالباهم",
"مدينتهم"" ,تطورات"" ,وللطالبان"" ,مدرسوهم"" ,اثنين"" ,أفيهم"" ,وعليه"" ,المحبون",
"بهم"" ,أوبهم"" ,ومنهما"" ,لهم"" ,ولك"" ,كاتبهم"
1. Perform a manual morphological analysis to
figure out the morphological patterns that the
Arabic words follow
2. Prepare the dictionaries:
• Proclitic (4 for each type)
• Verb prefix
• Perfective verb suffixes
Steps • Imperfective verb suffixes
• Noun suffixes
• Enclitic
3. Prepare the lists:
• List of stems
• 3 Lists of possible patterns as follows: verb
patterns, noun patterns, preposition patterns
Step 1: Manual Morphological Analysis
وسنكتبهم
هم+ كتب+ ن+ س + و
enclitic + stem+ prefix + part proc + Conj proc
لإلنسان
إنسان+ ل+ ل
stem + Det + part proc
وعليه
ه+ علي+ و
enclitic + stem + Conj proc
Step 2: Prepare the Dictionaries of clitics and
Affixes
Prepare dictionaries of Divide them to ensure Make sure to add the
clitics and affixes 2 compatibility 3 different forms of the
same affix:
• Proclitic (4 for each • for example, the
type) gender suffix of “”ة
• Verb prefix has also the from
• Perfective verb “”ت
suffixes • the number suffix of
• Imperfective verb “ ”ونhas also the
suffixes from “”و
• Noun suffixes
• Enclitic
Dictionary Example
Prepare the lists
A. List of stems
• Make sure that you write the stems as they are after clitics and affixation
removal
• Example:
< مدينت-- مدينتهم
B. Prepare the list of Patterns
Figure out the morphological patterns of each syntactic category on its own, i.e., work on
verbs first, then nouns, then preps. And make a list for each category
Make sure to make one pattern that matches the whole word.
Write these patterns in the form of a regular expression
Always remember that at the end of the analysis, we will need to check if the stem is in any
of the stem list, thus we need to keep for each pattern the group number of the stem.
Patterns are to be stored in a list of tuples, where the first item provides the pattern and the
next item is an integer that represents the stem’s group number
B. Prepare the list of Patterns
• verbPattern= [(pat, number),(pat, number),(pat, number)]
Stem A number indicating the position
of the stem in the pattern
Example: وسنكتبهم
enc+ stem+ prefix+ part_Proc+ conj_proc
Dear students
Prepare the dictionaries & Patterns
on python for next lab