0% found this document useful (0 votes)
130 views6 pages

Al-Lamy An Automatic Conjugation System of Arabic Verbs: October 2015

The document describes an automatic conjugation system for Arabic verbs called Al-Lamy. It addresses deficiencies in other Arabic conjugation systems by using a dictionary of 8000 verbs and implementing rules for 70 verb classes. The system displays conjugations and verb characteristics in a user-friendly interface and tests showed satisfactory performance.

Uploaded by

Ayesha
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
130 views6 pages

Al-Lamy An Automatic Conjugation System of Arabic Verbs: October 2015

The document describes an automatic conjugation system for Arabic verbs called Al-Lamy. It addresses deficiencies in other Arabic conjugation systems by using a dictionary of 8000 verbs and implementing rules for 70 verb classes. The system displays conjugations and verb characteristics in a user-friendly interface and tests showed satisfactory performance.

Uploaded by

Ayesha
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd

See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/286459918

Al-Lamy An Automatic Conjugation System of Arabic Verbs

Conference Paper · October 2015

CITATIONS READS

0 3,372

2 authors, including:

Abdelaziz Mirad
University of Science and Technology Houari Boumediene
1 PUBLICATION   0 CITATIONS   

SEE PROFILE

All content following this page was uploaded by Abdelaziz Mirad on 11 December 2015.

The user has requested enhancement of the downloaded file.


Al-Lamy An Automatic Conjugation System of
Arabic Verbs
Mirad Abdelaziz and Bensaou Nacéra

University of Science and Technology Houari Boumedienne, Algiers, Algeria.

Abstract
An automatic conjugation system is an important tool for a language
learners. Though efficient conjugation systems dedicated to English, French,
Spanish and other languages exist, to the best of our knowledge,Arabic
conjucation systems present several deficiencies. In fact, [1, 2, 3] provide
false verbal declinations and conjugate inaccurate verbs and random char-
acter sequences.
In this paper, we describe "Al-Lamy" our new educational program for
conjugating Arabic verbs. It counters the problems observed in other
Arabic conjugation systems by using a dictionary of 8000 verbs and by
implementig the gramatical rules of 70 verbal classes that cover their de-
clinations. Al-Lamy displays conjugations and characteristics of Arabic
verbs in a user-friendly interfaces and tests show that it performs in a
satisfactory way.

Intoduction
An increasing interest toward Arabic natural language processing has been no-
ticed recently. Researchers developed different tools dedicated to Arabic. These
include morphological analyzers [6], named entity recognition systems [5, 8],
automatic conjugation systems [2, 3, 1], etc.
Arabic is a highly inflectional language. Almost all of its words connect
to roots of three, four or five letters which approach the meaning of all their
inflectional forms [9]. Interestingly, Arabic owns a regular conjugation system
[4] whose the automation would be beneficial to its learners.
Automatic conjugation systems represent an important tool for grammar
training. Efficient conjugation systems exist for English, French, Spanish and
other languages. For instance, in [7], the authors state that ONOMA1 could
provide reliable conjugations of Spanish verbs and also the approximate infini-
tive forms of non-infinitive verbs before conjugating them. But, to the best of
our knowledge, Arabic conjugation systems are dysfunctional [2, 3, 1]. They
conjugate random strings and provide inaccurate verb declinations.
In this paper, we present our conjugation system "Al-Lamy" which solves
the problems noticed in the other systems. Al-Lamy uses a dictionary which
comprises entries of verbs with informative structures that indicate their most
important grammatical characteristics.
1 http://www.onoma.es/

1
In the following, we present preliminary properties of the Arabic grammar in
section 2, outline the architecture of Al-Lamy in section 3, report an assessment
of Al-Lamy’s performance in section 4 and we conclude this paper in section 5.

1 Characteristics of the Arabic Grammar


The Arabic language is a Semitic language, its alphabet is composed of 28
letters, three of which are long vowels ’waw’ ð ’alif’ @ ’yaa’ ø and has three short

vowels. It is a highly inflectional language, except some particles, conjunctions,


prepositions, pronouns and words coming from other languages, Arabic words
relate to roots of three, four or five letters2 . A set of patterns related to the

verb ɪ ¯ (fa‘ala/ to do) can be attributed to Arabic words and Arabic verbal
patterns vary according to the personal pronoun to which they correspond, its
gender and its number in addition to the tense, the voice and the mode of the
verb.

Example: The pattern of the verb I . J» (kataba/ to write>) is ɪ ¯ (fa‘ala).
ª ® K (tuf‘alaAn/
àA J . JºK (tuktabaAn/ has been written) is associated to the pattern àC
has been done) and indicates the use of the present, the passive voice and the
third person dual feminine pronoun.

1.1 The Classification of Arabic Verbs


Arabic verbs can be classified according to different criteria as it is shown in
Table 1.

2 Al-Lamy: an Arabic Conjugation System of


Arabic Verbs
2.1 Data

Al-Lamy ú
×CË@ is an educational program which conjugates 70 classes of Arabic
3
verbs which cover the verbal declinations of 8000 verbs . Al-Lamy implements
those classes and uses a dictionary whose entries are Arabic roots to which we
associate information about their grammatical properties, including:
• a binary value that indicates whether the verb is strong or weak,

• a value specifying the transitivity of the verb,


• the pattern of the root and
• clues regarding the patterns of the infinitive verbal forms that could be
derived from the root and whether they are transitive or not.
2 Most Arabic roots are triliteral.
3 @ ­K Qå” €ñÓA¯ http://ia700608.us.archive.org/8/items/waq44108/44108.pdf
ɪ¯ 8000 ÈAª ¯B

2
Feature Definition Example

in the sentenceZAÜ Ï @ A ¯ (water
overowed), the verb A ¯ (overowed) is
an intransitive verb because it does
It is semantic characteristic not require an object. On the other
The transitivity of verbs. A verb is transitive
B@ 
hand, in the sentence P Q¢ÖÏ @ ÉÊK.
ø
YªJË@ if it admits an object andis
intransitive otherwise. (The rain wet the ground), the verb

ÉÊK. (wet) is a transitive verb because it

required the object B@
P (the
ground).
Y«ð (waada to promise) is a weak verb
Verbs that include the a long
because it contains the letter ð and
Strength and vowel ’a’ @ ’w’ ð or ’y’ ø are


I.J» (kataba to write) is a strong verb
weakness weak. Otherwise, they are
strong because it does not contain long
vowels.
- If a verb is exclusively The verb I 
. J» (to write) is a primitive

composed of the letters of its
root, it is primitive and verb whereas the verb I.KA¾K (to write
Primitiveness - If a verb is different from to) is an inflected verb whose the root
is I 
and inflection its root and includes . J» (to write) and its additional
different short vowels and/or letters are its first letter H (t) and its
additional letters, it is
inflected. third letter @ (a).

Table 1: Classification features of Arabic verbs.

2.2 Conjugation Process


Given an input verb v, Al-Lamy verifies if v is included in the dictionary Dict.
If v is not included in the dictionary, v is not conjugated. Otherwise, Al-
Lamy collects information about v’s category and displays on the screen the
conjugation of v.

3 Evaluation
In this section, we compare our application to the best known Arabic conjuga-
tion systems [1, 2, 3]4 . conjugate random strings such as I

. .K. ("bbb") and provide
inaccurate conjugations inasmuch as they do not use dictionaries. Moreover, [1]
does not indicate personal pronouns or information about the conjugated verbs
and displays verb declinations without short vowels. As well, [2] adapts incor-
¯ (qilna) instead of áÊ ¯ (qulna/
rect short vowels to verb declinations such as áÊ
we said). Qutrub [3] also conjugates nonexistent verbs such as ÈñË (lul) and its
4 There is also an application named Sarf available at http://sourceforge.net/projects/sarf/

to which a very elaborate documentation is associated but that does not function correctly.

3
conjugations are often incorrect (e.g. the conjugation of the verb ¨AK . (baa/ to
buy) s incorrect). In contrast to these applications, Al-Lamy5 verifies the input
verbs using a dictionary of 8000 different verbs that only the system adminis-
trator can update and implements reliable conjugation rules based on a manual
of Arabic grammar. Hence, to our knowledge, it is the first Arabic conjugation
system that does not conjugate random strings and that is able to conjugate
properly a considerable number of verbs6 . To strengthen the pedagogical aspect
of Al-Lamy, characteristics of the input verb such as its pattern and whether it
is transitive or not are indicated to the user. Nevertheless, our application has
to be improved in order to conjugate Arabic verbs preceded by specific parti-
cles7 and by integrating a stemming process in order to extract infinitive forms
of already conjugated input verbs as in [7].

4 Conclusion
We created "Al-Lamy" an educational tool dedicated to the automatic conju-
gation of Arabic verbs. To our knowledge, it is the first Arabic conjugation
system that uses an elaborate dictionary and that can provide reliable gram-
matical information about verbs and accurate conjugations. Therefore, it could
help children and Arabic learners in learning Arabic grammar.
Al-Lamy would be improved by the integration of other conjugation modes.
Additionally, it could be extended by incorporating a module dedicated to the
verification and the correction of input verbs, in particular by the implementa-
tion of a stemming process that would extract infinitive forms of non-infinitive
input verbs to help novice Arabic learners.

References
[1] Azkary El Hussein, https://app.box.com/shared/bny6rt35dx

[2] Baykal Erol, http://acon.baykal.be/


[3] Arabeyes.org, http://qutrub.arabeyes.org/
[4] Amar, S., Dichy, J.: Collection Bescherelles, Al Shamel Fy TaSryf Al
Afaal Al Arabiya. Hatier (1999)

[5] Benajiba, Y.: Arabic Named Entity Recognition. Procesamiento del


Lenguaje Natural 44, (2010)
[6] Habash, N., al: MADA+TOKAN: A Toolkit for Arabic Tokenization,
Diacritization, Morphological Disambiguation, POS Tagging, Stemming
and Lemmatization. In: Proceedings of the Second International Confer-
ence on Arabic Language Resources and Tools. (2009)
5 http://sourceforge.net/projects/allamy2/
6 all verbs included in hyperlinks of the wiktionary page of Arabic verbs :

http://ar.wiktionary.org/wiki/%D8%AA%D8%B5%D9%86%D9%8A%D9%81%D8%A3%D9%81%D8%B9%D8%A7%D9%84
7 .Ð Qm Ì '@ð I’JË@
. H@ @ specific particles of nasb and jazm.
 ðX
.

4
[7] Rello, L., Basterrechea, E.: Grid Automatic Conjugation and Iden-
tification of Regular and Irregular Verb Neologisms in Spanish. In: Pro-
ceedings of the NAACL HLT 2010 Second Workshop on Computational
Approaches to Linguistic Creativity. (2010)
[8] Shaalan, K., Raza, H.: NERA: Named Entity Recognition for Arabic.
JASIST 60(8), 16521663 (2009)
[9] Watson, J. : The Phonology and Morphology of Arabic. OUP Oxford
(2007)

View publication stats

You might also like