Maryam Zafar et al / International Journal on Computer Science and Engineering Vol.

1(3), 2009, 275-282

Interactive English to Urdu Machine Translation

using Example-Based Approach
Maryam Zafar Asif Masood
Department of Computer Science Department of Computer Science
Military College of Signals, NUST Military College of Signals, NUST
Rawalpindi, Pakistan Rawalpindi, Pakistan
[email protected] [email protected]

Abstract—This work is first attempt towards English to Urdu Third approach for MT is Example Based Machine
Machine Translation (MT) using example based approach. We Translation (EBMT). It was introduced by Nagao [6] in 1984.
have developed an interactive MT system to facilitate the user to Using EBMT, translation is carried out by decomposing
customize the translation to his needs, thereby improving the English sentence into fragments, then finding corresponding
performance of the translation. Our MT system supports idioms, translation for those fragments in Urdu and then recombining
homographs, and some other features in addition to the ability of translated fragment into Urdu sentence.
the bilingual corpus to evolve. In the end, we compare the
features of the MT system developed by the center for research in English and Urdu are structurally different languages [1].
Urdu Language Processing (CRULP) with those of our MT For structurally different languages EBMT is a better choice
system. than others [7]. This suggests us to use Example based
approach for our MT system.
Keywords-Machine Translation; Bilingual Corpus
A quality translation cannot be developed now unless the
user gives a feedback to the MT system as discussed in [8], and
I. INTRODUCTION [9]. An interactive system that takes the feedback of the user to
Knowledge is the key to progress and English is one improve the translation quality is developed in our MT system.
language that preserves a tremendous amount of knowledge.
There is a huge literature of Sciences and Engineering available The scheme of this paper is as follows. We present related
in English. Rightly so, English is believed to be an international work in section 2, our proposed method in section 3, discuss
language. It becomes more important to be able to understand bilingual corpus in section 4, compare the work that we have
English. done with that done by CRULP in section 5, after which
conclusion and future work is provided in section 6.
Urdu is a language that is spoken in all over Pakistan and
many parts of India and some other South Asian countries [1]. II. RELATED WORK
This makes Urdu a very important language.
EBMT has been attempted for many languages, including
Only 5% to 10% people of Pakistan are familiar with English, French, Spanish, German, Japanese, Chinese, Turkish,
English [2]. To be able to move with the world, people of Arabic, Indian, and even Sign language, as for instance in [10]
Pakistan must get the latest knowledge but the current standing – [16].
in terms of literacy and the understanding of English finds a
gulf that needs to be abridged. This suggests that there is a need An MT system from English to Urdu is recently developed
of the interface between the two languages, i.e. English and by CRULP that is available online [17]. This MT system uses
Urdu. rule based technique, i.e. this MT system uses syntactic parsing
trees, the parts of speech tagger, and the grammatical rules for
Solution comes in the form of Machine Translation (MT). the translation purpose.
MT refers to the use of computing resources to facilitate the
translation of the content available in one language to its Work has been done on a web-based interactive MT system
equivalent content in another language [3]. [18]. The system is a chat-style MT. The idea is to provide a
broad coverage machine translation using the user’s response to
There are three main techniques for MT known as Rule improve the results inspired by [8], and [9].
based Machine translation (RBMT), Statistical Machine
Translation (SMT), and Example Based Machine Translation To the best of our knowledge, no work is available in
(EBMT). RBMT depends upon linguistic rules to carry out literature for English to Urdu MT using example based
translation from one language to another. It performs approach. Also, interactive features are not available for an MT
morphological, syntactic and semantic analysis on the language system from English to Urdu. This has motivated us to build an
and then transfers it into target language [4]. SMT is a corpus interactive EBMT system for the translation from English to
based approach that uses statistical models to carry out Urdu. Also, we aim to overcome the discrepancies that are
translation [5]. present in the CRULP MT system. The proposed method is
discussed in the next section.

275 ISSN : 0975-3397

 Maryam Zafar et al / International Journal on Computer Science and Engineering Vol.1(3), 2009, 275-282
III. PROPOSED METHOD For loop ends
The proposed method utilizes the phrase based set of EXAMPLE1:
examples accompanied by interactivity with the user. The
methodology is divided into the following main phases. I cannot back out of my promise.
• Sentence Fragmentation This sentence contains an idiom ‘back out of’; therefore this
is first tagged as an idiom, and the word ‘of’ in the idiom is not
• Searching in Corpus considered as a cutter point. Thus, the sentence is fragmented
• N-ary Product based Retrieval into the following phrases.

• Ordering of Translated Text • I

• Cannot back out of
A. Sentence Fragmentation
Fragmentation of sentences into phrases is important to • My promise
improve the scope of the sentences that the translator can
handle. This result can be obtained alternatively by keeping EXAMPLE2:
sentences in corpus and by gaining a broad coverage by
I love to play football because it thrills.
fragmentation and combination to get new sentences using the
genetic algorithm at run time [19]. The problem of fragmenting This sentence is first decomposed into two sentences
a sentence into simpler sentences and phrases is handled using namely,
the concept of idioms, the connecting words and the cutter
points that are explained as follows. • I love to play football
• Because it thrills
1) Idioms
An idiom is a phrase whose meaning is not determined The first sub-sentence is further divided into phrases using
word for word, but is instead related to the common use of the cutter points as follows.
native speakers, for details consult any standard text in English • I love
grammar, for instance [20]. A list of idioms is developed to
handle idiomatic phrases. The idioms are picked from [20]. • To play football
Its fragmented phrases are:
2) Connecting Words
Connecting words are the words that separate two or more • I love
sentences that are present in a large sentence e.g. and, because
etc available at [21]. • To play football
• Because it thrills
3) Cutter Points
Cutter points are the words that separate different B. Source Text
meaningful phrases in a sentence. These words are auxiliary The source text in English is input and is fragmented into
verbs, possessive pronouns, and prepositions. phrases using the fragmentation algorithm.

FRAGMENTATION ALGORITHM C. Searching in Corpus

Input: A sentence in English Searching comprises of the idea of finding whether or not
Output: A set of English phrases an input phrase is available in the bilingual corpus, discussed in
section IV. If it doesn’t find the exact match, it attempts to find
Algorithm body:
the close match. The measure of closeness is done using the
For each idiom, in the list of Idioms threshold at two levels; one for the exact match, the other for
Find idiom in the input sentence the close match.
If idiom is found
Tag it as an idiom This is done in two ways using Levenshtein Algorithm, and
If ends semantic distance algorithm. The user is given the option to
For loop ends choose from the both approaches.
Find connecting words
For each connecting word LEVENSHTEIN ALGORITHM
Separate the fragments of sentences into This is a character based algorithm that compares an input
new sentences string with the target string [22]. We have used the concept of
For loop ends Levenshtein algorithm for the word based comparison of two
Find cutter points phrases or two sentences as follows.
For each cutter point
Separate the fragments of sentences into TABLE I. DESCRIPTION OF USING LEVENSHTEIN ALGORITHM
phrases

276 ISSN : 0975-3397

 Maryam Zafar et al / International Journal on Computer Science and Engineering Vol.1(3), 2009, 275-282
Is Eating found in the dictionary accompanied by the word whose
0 1 2
meanings are not found.

Was 1 1 2
SEARCHING ALGORITHM
Eating 2 2 1
Input: English Phrase
Output: OK or SENTENCE IS NOT SUPPORTED
1 in the bottom right corner indicates that one operation is Algorithm body:
needed to convert the source text ‘is eating’ into target text For each phrase
‘was eating’. Compare the phrase with the set of phrases
in the corpus
The 0 Levenshtein distance means the exact match. A
If exact match found
positive finite Levenshtein distance means that it requires some
Return OK
finite operations to get the target text from the input text. A
threshold 0<T<n/2 means a closer match between two strings. Else
We provide interactivity for the string having closer match. Show options similar to the phrase
taken from Corpus
Semantic distance measures how close the two strings are If user selects an option
semantically. Semantic distance algorithm is an open source Return OK
project [23]. This algorithm returns score, a coefficient measure Else
of semantic closeness between two strings. A score of 1 means Return NOT FOUND
exact match, a score 0.9<s<1 means a synonymous word, If ends
where closeness is towards 1. We provide interactivity for If ends
0.8<s<0.98. These thresholds have been set after testing on
For loop ends
some examples. They can be adjusted in some cases as
suitable. If total number of OK responses divided by total
phrases >= 0.75
For each word of the phrase
TABLE II. TRADE-OFF BETWEEN USING LEVENSHTEIN ALGORITHM AND Find meaning of the word from the
SEMANTIC DISTANCE ALGORITHM
dictionary
Processing Coverage If all found
Levenshtein Fast Low
Add the phrase to the
corpus
Semantic Slow High
Return OK
Else
Return the found words’
meanings or words
If ends
For loop ends
Else
Return SENTENCE IS NOT SUPPORTED
If ends

The tool returns a string “SENTENCE IS NOT

SUPPORTED” if the ratio of successful matches < 0.75.

D. N-ary Product Based Retrieval

This phase comprises of the steps used to retrieve the
translation of the source text. There is a possibility of many
translations available for an input sentence. We gather the
Figure 1. Screenshot for interactive searching of phrases
possibilities and we use the idea of n-ary product to list all the
possible sentences. The idea is discussed as follows.
We calculate the number of phrases for which a match is
found divided by the total number of phrases in the input
sentence, if the ratio is greater than or equal to 0.75, the RETRIEVAL ALGORITHM
program attempts to find the meanings of all the words in the Input: Set of all English phrases of the input sentence
remaining phrases from the dictionary. If the meanings of all Output: Set of all possible Urdu translations
the words in a phrase are found, the phrase is added to the Algorithm body:
corpus with its Urdu equivalent, otherwise, the phrase is Set i=1 // i is a counter of the number of phrases
returned to the user with whatever portion of the phrase is For each English phrase ei
Find the phrase ei in the corpus

277 ISSN : 0975-3397

 Maryam Zafar et al / International Journal on Computer Science and Engineering Vol.1(3), 2009, 275-282
Make a set Ui
Find all the translations of ei and add them
to Ui
Add 1 to i
For loop ends
Find the n-ary product of all Ui’s

The problem is that the number of phrases for a particular

input sentence is not known before program’s execution and it
becomes available only at runtime. Also, the number of
translation equivalents in Urdu for a particular phrase gets
available only at execution time. This poses a great problem in
implementation.
We resolve it by considering a set E of English phrases
e1 , e 2 ,…, e n and corresponding to each phrase e i of English,
Figure 2. Screenshot for interactive retrieval phase

a set of its Urdu equivalents is considered as follows. We have two triplets (3-tuples) in this case, i.e. we n1*n2*n3
E = Set of English phrases = {e1 , e2 ,..., en }
= 1*2*1 = 2 triplets (3-tuples).
For details of n-ary product, consult any standard text in
{
U1 = Set of Urdu translations of e1 = u 11 , u 12 ,…, u 1n1 } Discrete Mathematics; see for instance [24].

U2 = Set of Urdu translations of e2 = {u 21 , u 22 ,…, u 2n 2 } E. Ordering of Translated Phrases

This one aspect is usually ignored in the research articles.
However, the problem is important in the translation from
M English to Urdu and other distant language pairs.
Un = Set of Urdu translations of en = {u n1 , u n2 ,…, U nn n } . The ordering algorithm is as under:

U = n-ary product of all the sets of Urdu translations of all 1) Ordering Rule 1
If there are e1, e2,,…, en are the phrases of English whose
English phrases = U1 × U 2 × …× U n = Urdu equivalents are u1, u2, …, un, then the translation of the
{(u 11 , u 21 ,…u n1 ),…, (u 1n1 , u 2n 2 ,…, u nn n )} sentence [e1 e2 … en] is given by [u1 un un-1 … u3 u2].

The total number of Urdu translations is given by EXAMPLE 4:

n1 * n2 * ... * nn , where ni is the cardinality of Ui. She will work hard to pass the exam.
• She
EXAMPLE 3:
A student, ‘aik talib ilm’ • Will work hard

Has been studying, ‘parrh raha hai’, ‘parrh rahi hai’, • To pass the exam

For two hours, ‘do ghante se’. ‘Woh imtihan mein kamiabi ke liye mehnat kare gi’

E= {e1=A student, e2=has been playing, e3=for two hours} • Woh

U1 = {u11 = aik talib e ilm} • Mehnat kare gi
U2 = {u21 = parrh raha hai, u22 = parrh rahi hai} • Imtihan mein kamiabi ke liye
U3 = {u31 = do ghante se} By applying the rule above we get
U = U1*U2*U3 = {(u11, u21, u31), (u11, u22, u31)} = {(aik talib Woh imtihan me kamiabi ke liye mehnat kare gi
e ilm, parrh raha hai, do ghante se), (aik talib e ilm, parrh rahi
2) Ordering Rule 2
hai, do ghante se)}
For complex sentences, it is observed that the pattern
repeats when some English sentences are combined to form a
larger sentence using connecting words.

EXAMPLE 5:
My aunt is a doctor and my uncle works in a factory that
makes television sets.

278 ISSN : 0975-3397

 Maryam Zafar et al / International Journal on Computer Science and Engineering Vol.1(3), 2009, 275-282
In this sentence, ‘and’, and ‘that’ are connecting words that A list of idioms is separately developed to supplement the
split the sentence into three sub-sentences as under. bilingual corpus, as discussed in III A 1). Also, the
homographs, are handled inspired by [27] as in example 6,
• My aunt is a doctor these are the words that have same spelling but different
• And my uncle works in a factory meanings. Gender differences are also incorporated for the
phrases of neutral gender as in example 7, and the singularity
• That makes television sets differences are handled for the phrases of neutral singularity as
These sentences are decomposed into the following way in example 8.

• My aunt, is a doctor EXAMPLE 6:

• And my uncle works, in a factory • He gets an apple, ‘use aik saib mila’,
• That makes television sets • He gets an idea, ‘use aik tarkeeb soojhi’,
Using the rule stated above the translation is given as • He gets to home, ‘woh ghar ko pahuncha’.
Meri aunti doctor hain, aur mere uncle kaam karte hain In these sentences, ‘gets’ is the homograph that gives
factory mein, jo television set banati hai. different meanings in different contexts.
For more complex sentences in the sense of ordering of
their translations, we provide an interactive tree structure that a EXAMPLE 7:
user can utilize to adjust the phrases to get a correct order of A player is playing, ‘aik khilarri khel raha hai’, ‘aik khilarri
Urdu translations. khel rahi hai’. In this example, ‘player’ is the neutral gender
word.

EXAMPLE 8:
You are eating; ‘aap kha rahe hain’, ‘tum kha rahe ho’ etc.
In this example, ‘you’ is the neutral singular word.
Since the homographs are included in the corpus, and also
the gender differences are recognized, therefore, the corpus size
is not homogeneous, i.e. the column size is variable because
one English phrase may have many Urdu equivalents as in
example 9.

EXAMPLE 9:
Enjoys cricket; ‘cricket pasand karta hai’, ‘cricket pasand
karti hai’.
In this example, enjoys cricket is stored in one row, and the
corresponding equivalents are placed in the same row in
different columns.
There is another interesting feature of the tool associated
with the corpus construction. There is a separate module that
ensures that no phrase is duplicated in the corpus, although one
phrase can have more than one translation equivalents as has
Figure 3. Screenshot for interactive ordering of translated text been discussed above. However, in case a phrase occurs more
than once in the construction phase, its frequency is added
IV. BILINGUAL CORPUS thereby improving its probability in the sense of Bayesian
Bilingual corpus is one of the most important parts of this probability theory [28]. This idea is useful when there are
tool, since the tool relies mainly on the corpus and a set of rules scarce computing resources at our disposal as in the case of
to combine the phrases in the corpus. embedded devices.

For the construction of the corpus, we take the textbooks of A. Dictionary

the secondary school [20], [25], and [26], pick random A dictionary can be used in an EBMT approach to support
sentences, and then we fragment them based on the idea of translation process as in [29] The idea is to let the program
connecting words, and the cutter points, details of which are learn the new words from the dictionary whenever the program
discussed in section III A. Corpus is built using a phrase-based fails to translate a sentence by a small fraction of the
set of examples, rather than sentence based examples to unrecognized phrases to the total number of phrases in the
incorporate more sentences and scenarios. sentence. We build a specific dictionary whose words are taken

279 ISSN : 0975-3397

 Maryam Zafar et al / International Journal on Computer Science and Engineering Vol.1(3), 2009, 275-282
from [30]. This helps in evolving the bilingual corpus at ‘use mila aik khayal’
runtime.
EXAMPLE 13:
V. FEATURES ANALYSIS OF OUR MT SYSTEM
Input:
COMPARED WITH CRULP MT SYSTEM
CRULP system is a good attempt towards English to Urdu • He works in a bank
MT and it facilitates the correct translations in many situations.
However, there are some discrepancies that we have attempted • He is waiting near the bank of a river
to overcome in our system. CRULP MT system response:
Examples illustrating idioms: ‘woh bank mein kaam karta hai’
‘daria ke bank ke qareeb he is waiting’
EXAMPLE 10:
Input: Our MT system response (as options):

The conspiracy was brought to light by policeman Translation options for first sentence:

CRULP MT system response: ‘Woh kaam karta hai bank mein’

‘saazish police ke afsar ke paas roshni ki taraf layi gayi’ ‘Woh kaam karta hai kinare mein’

Our MT system response: Translation options for second sentence:

‘saazish police afsar se manzar e aam par aayi’ ‘Woh dariya ke kinare ke qareeb intizaar kar raha hai’
‘Woh daria ke bank ke qareeb intizaar kar raha hai’
EXAMPLE 11: In example 12, the word ‘get’ is taken in two senses, i.e. to
Input: come into possession of, and to perceive. In example 13, the
word ‘bank’ has two meanings i.e. bank as a financial
He has come of age today
institution, and bank as the slope beside a body of water.
CRULP MT system response: CRULP MT system doesn’t support the multiple uses of the
word bank and get in these examples, but our MT system
‘woh aaj umer ka aaya hai’ provides such support.
Our MT system response: Examples illustrating the gender and the words taken in
‘woh aaj baaligh hua hai’ singular and plural sense together:
In these examples, the idioms ‘brought to light’, and ‘come
of age’ have been considered for which CRULP MT system EXAMPLE 14:
attempts to translate word for word, whereas, our MT system Input:
attempts the idiomatic phrase correctly.
They are playing in the garden
Examples illustrating homographs:
CRULP MT system response:
EXAMPLE 12: ‘woh baagh mein khel rahe hain’
Input: Our MT system response:
• He gets an apple ‘woh baagh mein khel rahe hain’
• He gets an idea ‘who baagh mein khel rahi hain’
CRULP MT system response:
EXAMPLE 15:
‘use saib milta hai’ Input:
‘use khayal milta hai’ It is his work
Our MT system response (as options): CRULP MT system response:
Translation options for first sentence: ‘yeh uss ka kaam hai’
‘use mila ailk saib’ Our MT system response:
‘use soojha aik saib’ ‘yeh uss ka kaam hai’
Translation options for second sentence: ‘yeh unn ka kaam hai’
‘use soojha aik khayal’

280 ISSN : 0975-3397

 Maryam Zafar et al / International Journal on Computer Science and Engineering Vol.1(3), 2009, 275-282
In example 14, the word ‘they’ is taken in the masculine VII. REFERENCES
sense as well as feminine sense. In example 15, the word ‘his’ [1] Tafseer Ahmed Sadaf Alvi, English to Urdu translation system,
is taken in two senses in Urdu, i.e. when we want to give regard University of Karachi, Pakistan, 2002.
some third person, and when no such regard is needed. In these [2] Sarmad HUSSAIN “Urdu Localization Project: Lexicon, MT and TTS
(ULP)” Center for Research in Urdu Language Processing,. National
examples, the CRULP MT system does not support the words University of Computer and Emerging Sciences.
having the two gender senses, and it also does not discriminate
[3] W. John Hutchins and Harold L. Somers, “An Introduction to Machine
between the word senses for the regard, which is usual in Urdu. Translation”. A CADEMIC PRESS. (London),1992.
[4] Cyfieithu Peirianyddol a’r Gymraeg: y Ffordd Ymlaen, “Machine
TABLE 3: COMPARISON OF CRULP MT SYTEM WITH OUR MT SYSTEM translation and welsh: the way forward”, A Report for The Welsh
Language Board, July 2004.
CRULP MT system Proposed system [5] http://en.wikipedia.org/wiki/Statistical_machine_translation
Idioms Not supported Supported [6] M. Nagao, “A framework of a mechanical translation between Japanese
and English by analogy principle,” in Proceedings of the international
Homographs Not supported Supported NATO symposium on Artificial and human intelligence, pp. 173–
Gender Not supported Give options 180,1984.
[7] Chris Brockett, Takako Aikawa, Anthony Aue, Arul Menezes, Chris
Processing time High Low
Quirk, & Hisami Suzuki, “English-Japanese example-based machine
Up gradation Not supported Upgrades its corpus translation using abstract linguistic representation”, Coling-2002
workshop "Machine translation in Asia" , Taipei,Taiwan, 1 September
2002.
[8] Mark Seligman. 1997. Six issues in speech trans- lation. In Steven
Kra.uwer et al., editors, Spo- ken Language Translation Workshop,
Our MT system fails to provide a correct ordering of the pages 83--89, Madrid, July.
translated text when there are connecting words that also lie in [9] Robert Frederking , Alexander Rudnicky , Er Rudnicky , Christopher
the category of cutter points, e.g. ‘and’, ‘or’ etc, or when a Hogan, “Interactive speech translation in the DIPLOMAT project”, In
conjunction is simply a separator in the list of some nouns or Proceedings of the Spoken Language Translation Kluwer Academic
verbs. This also doesn’t handle the ordering of the translated Publishers Hingham, MA, USA, pp 27-42, 2000.
text correctly when there are phrases like ‘Government of [10] Yang, Muyun; Jiang, Hongfei; Tiejun, Zhao; Li, Sheng; Liu, Daxin,
Pakistan’, in which the word ‘of’ is not to be considered as a “Domain Sensitive Chinese-English Example Based Machine
Translation”, Fifth International Conference on Fuzzy Systems and
cutter point. Such ordering problems are ignored in the Knowledge Discovery, 2008. FSKD '08, Volume 2, Page(s):453 - 456
automated translations, as for instance discussed in [23]. 18- 20 Oct 2008.
However, we provide support to handle such issues as [11] Morrissey S. and A. Way. “An Example-Based Approach to Translating
discussed in [ordering of translated text]. Sign Language”, In Proceedings of the Second Workshop on Example-
Based Machine Translation, Phuket, Thailand, pp.109—116, 2005.
VI. CONCLUSION AND FUTURE WORK [12] Alp, N. D.; Turhan, C, “English to Turkish Example-Based Machine
We have developed an interactive MT system, thereby Translation with Synchronous SSTC”, Fifth International Conference on
Information Technology: New Generations, Page(s):674 – 679, 7-9 April
providing support for idioms, homographs, gender, the words 2008.
having plural and singular senses together, ability of the corpus [13] Ralf D. Brown, "Example-Based Machine Translation in the Pangloss
to evolve for a broader coverage, and also the support for the System", In Proceedings of the 16th International Conference on
problem of ordering of the translated text. Ordering is a Computational Linguistics (COLING-96), p. 169-174. Copenhagen,
problem that is hard for computer but easier for humans. The Denmark, August 5-9, 1996.
interactive system brings ease to the users thereby complying [14] Nicolas Stroppa, Declan Groves, Andy Way and Kepa Sarasola.
with the basic goal of the research in this direction, i.e. “Example-Based Machine Translation of the Basque Language”.In
Proceedings of the 7th biennial conference of the Association for
facilitating the user in the translation process and improving the Machine Translation in the Americas, pages 232-241, Cambridge,
efficiency of the task. Massachusetts, August 2006.
Our MT system works particularly well for the situations [15] Vamshi Ambati & Rohini U: “A hybrid approach to example based
machine translation for Indian languages”. ICON-2007: 5th International
that it has been trained to handle. This suggests us that an MT Conference on Natural Language Processing, IIIT Hyderabad, India, pp
system for domain specific needs can be built using the ideas 4-6, January 2007.
discussed here. Also, the size of our corpus suggests that it can [16] Kfir Bar, Y.Choueka, & N.Dershowitz: “ An Arabic to English example-
be used in embedded devices with low memory resources. The based translation system”. ICTIS 2007: Information and Communication
phrase based set of examples are used in this MT system, this Technologies International Symposium. Workshop on Arabic natural
gives it a broader coverage. language processing, pp. 355-359, 3-5 April 2007.
[17] Center of Research for Urdu Language Processing (CRULP),
Work is in progress to extend the bilingual corpus, and to http://www.crulp.org (Last accesss 10/07/2009).
improve the algorithm for the ordering of the translated text. [18] Christopher Hogan, and Robert Frederking, “WebDIPLOMAT: A Web-
The extended corpus and the improved ordering algorithm Based Interactive Machine Translation System”, Proceedings of the 18th
improve the scope of our MT system, enhance the quality of conference on Computational linguistics - Volume 2. PP 1041 – 1045,
2000.
the translation, and optimize the efficiency of the translation
[19] Hiroshi Echizen-ya , Kenji Araki , Yoshio Momouchi , Koji Tochinai, “
process. Machine translation method using inductive learning with genetic
algorithms” Proceedings of the 16th conference on Computational
linguistics, August 05-09, 1996, Copenhagen, Denmark.

281 ISSN : 0975-3397

 Maryam Zafar et al / International Journal on Computer Science and Engineering Vol.1(3), 2009, 275-282
[20] Afzal Anwar mufti, “English Grammer and Composition”, Qaumi Kutub [26] John Jackman, and Wendy Wren, “Book 2”,Sunrise/ Nelson publishers.
Khana, Lahore, Pakistan. [27] Michael McCarthy, Felicity O’Dell, English Vocabulary in Use
[21] http://daphne.palomar.edu/mlloyd- Advanced, Cambridge University Press.
jones/handouts/CONNECTING%20WORDS.pdf [28] Stuart Russell , and Peter Norvig, “Artificial intelligence: A modern
[22] http://en.wikipedia.org/wiki/Levenshtein_distance approach” Second Edition. Prentice Hall series in Artificial Intelligence.
[23] Troy Simpson, Thanh Dao; WordNet-based semantic similarity [29] Sadao Kurohashi, Toshiaki Nakazawa, Kauffmann Alexis, Daisuke
measurement, Kawahara: “Example-based Machine Translation Pursuing Fully
http://www.codeproject.com/KB/string/semanticsimilaritywordnet.aspx Structural NLP”, In Proceedings of International Workshop on Spoken
[24] James L Hein, “Discrete Mathematics”, Jones and Bartlett books in Language Translation (IWSLT'05), pp.207-212, Pittsburgh, Oct 2005.
mathematics. [30] Apni Urdu, http://www.apniurdu.com (Last accesss25/06/2009).
[25] John Jackman , and Wendy Wren, “Book 5”, Sunrise/ Nelson publishers.

282 ISSN : 0975-3397