Telugu To English Translation Using Direct Machine Translation Approach


International Journal ofScience and Engineering Investigationsvol. 2, issue 12, January 2013ISSN: 2251-8843Telugu to English Translation using Direct Machine TranslationApproachT. Venkateswara Prasad1, G. Mayil Muthukumaran212Dean of Computing Sciences, Visvodaya Technical Academy, Kavali, AP, IndiaTechnical Director, National Informatics Centre, Govt. of India, New Delhi, India(, The motivation behind working on a translationsystem from Telugu to English were based on the principlesthata) There are many translation systems for translating fromEnglish to Indian languages but very few for vice versa.Telugu is a language that exhibits very strong phrasal,word and sentence structures next to Sanskrit, whichmakes the work organized on one hand but complex inhandling on the other. This work demonstrates one suchmachine translation (MT) system for translating simpleand moderately complex sentences from Telugu toEnglish.b) Of the many MT approaches, the direct MT is used fortranslation between similar or nearly related languages.However, the direct MT has been used in this work forconversion from Telugu to English, which is quitecomplex compared to other Indian languages. Thepurpose of using direct MT for development of such atool was to have the flexibility in usage, keeping it simple,look for rapid development and primarily to have betteraccuracy than all the known system.c) There are very large numbers of elisions/ inflection rulesin Telugu requiring complex morphs, like those inSanskrit. A large number of rules for handling inflectionswere to be developed along with the grammar rules.The outcomes were compared with Google Translator, apublicly available translation web based system. The outcomeswere found to be much better, as much as 90 percent moreaccurate. This work shall bring forth deeper insights intoTelugu MT research.Keywords- Machine translation (MT), direct MT, Telugu toEnglish, natural language processing (NLP), elisions,inflections.I.INTRODUCTIONLanguages that are descent from Brahmi script are verygood in grammar. The sentences are constructed strictlyaccording to the norms laid out and there are very less chancesof any deviation or violation. All Indian languages havedescended from Brahmi script. Telugu is said to have splitfrom proto-Dravidian languages around 6th to 3rd centuryBCE [13].Telugu language is a highly structured, disciplined, suaveand rich in terms of expression, style and construction. Itexhibits clear and structured implementation of grammar inthe best possible manner while including present daycorruptions (or vulgarity) and foreign words. There is a clearand specific purpose and meaning of each letter. Slightmodification in the way a letter is written can change themeaning itself, e.g., kada and kaDa are two unique wordshaving different meaning. Similarly, rama, rāma and ramā arethree different usages. The language also provides largenumbers of exceptions in usage thus making it more complex,beautiful and expressive [1].The richness of Telugu language lies in the extremelylarge number of words representing different moods,expressions, contexts, etc. Ancient Telugu usage often knownas “Grāndhika” had well defined grammar, classes of words,morphology, etc. Telugu language currently encompasseswords of five categories, viz., a) of its own (purest form), b) ofSanskrit origin, c) of corrupt form of Sanskrit words, d) ofcolloquial usage and e) of other states/nations. Normally, thewords of colloquial usage are not considered to be part of theTelugu grammar since it is considered as vulgar, was onlyprevalent with working class people [11].Due to the modernization in the last century includingserious impact of the media and cinema, the colloquial usagehas taken centre stage of the grammar. When used in poeticsense, Telugu language exhibits very high level ofgrammatical usage. It is notable that each Telugu lettertogether with the consonants must be spoken very clearly withproper emphasis and intonation.Tools for machine translation (MT) from English to certainIndian languages and from one Indian language to another areavailable; however, such tools for MT from Indian language toEnglish are very few.Indian languages are many in number but have a similarsubject-object-verb (SOV) pattern of grammar, unlike theEnglish that has SVO pattern or the VSO pattern of Arabicand Japanese. It is worth notable that translation from English25

to any Indian language is a relatively easier process, whereasvice-versa is very complex.This research work brings forth the process of convertingTelugu sentences into its equivalent English sentences.Telugu grammar, vocabulary and style as documented by wellknown Telugu and British scholars during the British rule inIndia were studied in depth [1-2]. These books were selectedsince they were published during the mid 19th and early 20thcentury until when the Telugu language was relatively freefrom the heavy corruptions of the modern day literature.II.methods need a skilled linguist to carefully design thegrammar that they use.Following are the known approaches of MT:a)Rule-based: The rule-based MT paradigm includestransfer-based MT, interlingual MT and dictionary-basedMT paradigms. Transfer-based machine translation: To translatebetween closely related languages, a techniquereferred to as shallow-transfer machine translationmay be used. Interlingual: Interlingual MT is one instance of rulebased MT approaches. In this approach, the sourcelanguage, i.e. the text to be translated, is transformedinto an interlingual, i.e. source-/target-languageindependent representation. The target language isthen generated out of the interlingua. Dictionary-based: MT can use a method based ondictionary entries, which means that the words willbe translated as they are by a dictionary.MT SYSTEM APPROACHBernard Vauquois' pyramid is shown in Fig -1 depictingcomparative depths of intermediary representation,interlingual machine translation at the peak, followed bytransfer-based, then direct translation [3].Figure 1. Bernard Vauquois' pyramid showing generalized model of MTb) Statistical: Statistical MT tries to generate translationsusing statistical methods based on bilingual text corpora,such as the Canadian Hansard corpus, the English-Frenchrecord of the Canadian parliament and EUROPARL, therecord of the European Parliament. Where such corporaare available, good results can be achieved translatingsimilar texts, but such corpora are still rare for manylanguage pairs.Machine translation can use a method based on linguisticrules, which means that words will be translated in a linguisticway — the most suitable (orally speaking) words of the targetlanguage will replace the ones in the source language. It isoften argued that the success of machine translation requiresthe problem of natural language understanding to be solvedfirst.c)Rule-based methods parse a text, usually creating anintermediary, symbolic representation, from which the text inthe target language is generated. According to the nature of theintermediary representation, an approach is described asinterlingual MT or transfer-based MT. These methods requireextensive lexicons with morphological, syntactic, andsemantic information, and large sets of rules.d) Hybrid MT: Hybrid MT (HMT) leverages the strengthsof statistical and rule-based translation methodologies.Several MT organizations (such as Asia Online,LinguaSys, Systran, etc.) claim a hybrid approach thatuses both rules and statistics. The approaches differ in anumber of ways:Given enough data, MT programs often work well enoughfor a native speaker of one language to get the approximatemeaning of what is written by the other native speaker. Thedifficulty is getting sufficient data of right kind to support theparticular method. For example, the large multilingual corpusof data needed for statistical methods to work is not necessaryfor the grammar-based methods. But then, the grammarExample-based: Example-based MT (EBMT) approachwas proposed by Makoto Nagao in 1984. It is oftencharacterized by its use of a bilingual corpus as its mainknowledge base, at run-time. It is essentially a translationby analogy and can be viewed as an implementation ofcase-based reasoning approach of machine learning. Rules post-processed by statistics: Translations areperformed using a rules based engine. Statistics arethen used in an attempt to adjust/correct the outputfrom the rules engine. Statistics guided by rules: Rules are used to preprocess data in an attempt to better guide thestatistical engine. Rules are also used to post-processInternational Journal of Science and Engineering Investigations, Volume 2, Issue 12, January 2013ISSN: 2251-8843www.IJSEI.com26Paper ID: 21213-05

the statistical output to perform functions such asnormalization. This approach has a lot more power,flexibility and control when translating.There has been debate on the suitability of statistical basedMT on rule-based MT and vice versa for long; [19] concludesthat it is purely dependent on the kind of applications and thatthese days a hybrid approach is being used more widely so asto combine the goodness of both approaches. Rule based NLPfor demonstrating improvement in disease normalization inbiomedical texts was also used [17]. The rule-based approachfor MT of Arabic text was employed in [18]. Elaborateddetails on different approaches of MT and specific emphasiswere put on Knowledge based MT (KBMT) are given in [16].Latest views are also presented on the classification ofdifferent approaches in seminal work on English to TeluguMT [15].In addition to the above classification of approaches,researchers have used various other methods like neuralnetworks, fuzzy logic, genetic algorithms, hidden Markovmodels, etc. in different domains/languages for achievingbetter a) organization, b) rules and c) accuracy.III.DIRECT MACHINE TRANSLATIONThe direct MT system is considered to be the mostprimitive approaches of all carrying out replacement of thewords in the source language with words in the targetlanguage. This is carried out in the same sequence andwithout much linguistic analysis or processing. The onlyresource direct MT uses is a bilingual dictionary, and that iswhy it is also known as dictionary-driven MT.While certain researchers consider it to be quiteunsophisticated approach and obsolete for many years, whilesome believe that direct MT has been considered useful fortranslation between two similar or near related languages.Systems falling under such approach are used for translationbetween Sanskrit and Hindi, Punjabi and Hindi, and so on.Description of evaluation of direct MT approach betweenPunjabi and Hindi is given in [21]. Earlier, [20] used the directMT for English to Swedish translation.Rule-based translation is one of the forms of MT, the rulebased MT paradigm includes transfer-based MT, interlingualMT and dictionary-based MT paradigms. Some experts calldirect MT approach as part of the rule-based MT and considerit to be different from dictionary based MT approach. There isalso a scope of combining the features of two or moreapproaches together for bringing out better translation results.Of all these approaches, the direct MT approach waschosen for the proposed research on Telugu to English MT,keeping in view that the aspects of a) rapid softwareapplication development, b) higher accuracy, c) customizableMT, and d) provisioning of very simple and easilyunderstandable design.It is strongly believed that direct MT still has a place intoday’s automated translation tools. Such approaches are usedwhere both vocabulary and syntax are standardized, indomains like weather reports, financial profiles, and many ecommerce applications. For implementation of such approach,word-for-word or phrase-for-phrase substitution is all that isneeded.Records reveal that human translation projects provided anunacceptably high level of error rates. The direct MT hasproved to be very useful where initial tests had shown thatboth translation memories and rules-based machine translationsystems produced poor results with text that has little or norepetition on the sentence level; or even high repetition on theword/phrase level.Since direct MT does not require human post-editing inmost of the cases, using MT in this kind is highly welcomedby translators and buyers needing very quick, cheap andmoderately good quality of translation.Many of the words are formed by combining two or morerelated words. Sandhis are actually conjugations of two ormore words and elisions are reverse of sandhi, i.e. splitting ofa word into two or more components. The more is the usage ofelisions in Telugu, the structure of the sentence is consideredthe better [12].For Telugu, certain work has been done on MT to/fromTelugu related to handling of corpora and building of treebank [6-7]. Most of the work has been built around Hindilanguage and generalized to all Indian languages as theyfollow the same SVO structure [6] with slight variations inplacement of articles, pre/post-positions, etc. Morphologicalsynthesis of English – Telugu MT was done [8]. Very less isavailable for MT from Indian languages to English. Onerecent attempt has been documented for Malayalam to English[10]. A lucid account of various useful works done on MT onIndian languages is given in [5].Currently, there is only one known web based Telugu MTsystem available in the form of Google Translator [4]. A largenumber of experiments were conducted on the GoogleTranslator to obtain the translation of various simple andmoderately complex statements. Google Translator could notprovide good translation of many words since the elisionsection was not handled adequately.IV.EXPERIMENTAL WORKDue to the vastness of the subject, the scope was limited tions/initial boundaries made for the purpose are (a)translation for simple Telugu statements are to be undertaken,(b) more focus to be given on word morphology that forms themost complex part of the research.With these premises, a comprehensive software tool by thename “Telugu to English Translation Suite” was developed inAccess Basic on Windows platform. A limited dictionary ofTelugu to English database comprising of over 2000 wordsInternational Journal of Science and Engineering Investigations, Volume 2, Issue 12, January 2013ISSN: 2251-8843www.IJSEI.com27Paper ID: 21213-05

was developed. As the Telugu language comprises extremelylarge number of conjunctions/ elisions/ inflections or sandhiforms, over 650 of them were analyzed, grouped in 222paradigms and incorporated in the software suite, Table I.The test sentences/corpora were put into the MT systemdeveloped for MT from Telugu to English and were foundcomparatively to be very successful.TABLE I. TELUGU – ENGLISH DICTIONARYV.DescriptionQtyTelugu Verbs399Telugu Nouns908Telugu Pronouns2Telugu Adverbs247Telugu Adjectives125Telugu Prepositions299Telugu Ordinals40English Irregular verbs362Verb forms276Pronoun forms109Elision rules649Telugu being a free word-order structure language, MTfrom English to Telugu can be easy. However, the vice-versais very complex keeping in view the complexity of Englishlanguage structure.Handling of two elisions in Telugu text were successfullyimplemented with accuracy of translation as high as 90percent over the given test statements. Though the translationof idioms, style, feelings, handling synonyms of a word, etc.aspects have not been touched at this stage, the translationresults were over 60 percent better than the web based GoogleTranslator.Sample outcomes of the MT to English as well ascomparison with the outputs of Google Translator aretabulated in Table III. Some of the outcomes resultingtranslation specific to tenses have also been detailed in TableIII. Some examples of poor or bad translation are given inTable IV.Broadly, the system has been divided into five parts ormodules, Figure II, viz. Conversion to Roman Telugu form (by transliteration)Application of Telugu morphology on the wordsApplication of machine translation by replacing eachTelugu word by equivalent English wordMaintaining word orderApplication of English morphology (called here asreverse morphology) RESULTS AND DISCUSSIONSThere were 450 Telugu sentences categorized into fivegroups as listed in Table II, were taken from [1] and [14]. TheTETS system was tested basically for the first two categories.The developed software suite was rigourouslyexperimented with large number of different types/structuresof sentences. The outcomes of the software suite were alsocompared with the Google Translator (currently the onlyknown publicly available translation site). The results werevery encouraging as the accuracy of the developed softwarewas very much higher.TABLE II. CATEGORIZATION OF TELUGU TEST SENTENCESGroupDescription of test/example sentenceNumberIVery Simple Telugu Sentences346IISimple Telugu Sentences65IIIComplex Telugu Sentences29IVVery Complex Telugu Sentences15VFree Flowing Telugu ParagraphsManyThe TETS system was also tested using free flowingsentences from various websites of newspaper companies.The parsing of lexicon, splitting or stripping of suffices, andtheir translation to English was very much satisfactory. Onlythose words could not be translated accurately that form verycomplex elisions/ inflections, or those not available in thedictionary or those having many synonyms.It is most notable that the dictionary for Telugu to EnglishMT should be populated with words that are spoken/used asthey are. This means, there can be more words in thedictionary than predicted. For example, the Telugu equivalentfor December is represented commonly in day-to-day usageby the words DiseMbaru డిసెంబరు as well as Dishambarడిశెంబర్, however, if the dictionary is built only with thestandard version, it is sure that the accuracy of translation willdrastically reduce.VI.CONCLUSIONWith the present work, it was brought out that forsuccessful translation of Indian languages, special emphasishas to be done on handling inflections/ elisions. There arelarge numbers of words that have three or more elisions.For the first time, successful implementation of direct MTon two dissimilar languages was demonstrated through thiswork.Addition of more linguistic rules related to handling ofelisions/inflections and the word ordering system wouldenhance the accuracy of the proposed translation system.International Journal of Science and Engineering Investigations, Volume 2, Issue 12, January 2013ISSN: 2251-8843www.IJSEI.com28Paper ID: 21213-05

13][14]Arden A. H., “A Progressive Grammar of the Telugu Language withCopious Examples and Exercises”, 2ed, Society for Promoting ChristianKnowledge, Madras, 1905Brown C. P., The Grammar of the Telugu Language, W. H. Allen & Co.,London, 1857Machine Translation Approaches, Wikipedia, available atwww.wikipedia.comGoogle Translator, Google, available at, Seema Bawa, “Survey of Indian Machine TranslationSystems”, Int. J. of Computer Sc. & Tech., Vol. 3, Issue 1, Jan. - March2012Kolachina, P., Kolachina, S., Singh, A.K., Naidu, V., Husain, S., Sangal,R., Bharati, A., “Grammar Extraction from Treebanks for Hindi samar/data/grammarextraction-LREC10.pdfChaitanya Vempaty, Viswanatha Naidu, Samar Husain, Ravi Kiran,Lakshmi Bai, Dipti M Sharma, and Rajeev Sangal, “Issues in AnalyzingTelugu Sentences towards Building a Telugu Treebank”, Proc. of 12thInternational Conference on Intelligent Text Processing andComputational Linguistics (CICLing 2010), Lect. Notes in CS 6008, pp.50–59, 2010Suryakanthi T., Prasad S. V. A. V. and Sharma Kamlesh,“Morphological Synthesis in English to Telugu Machine TranslationSystem”, Proc. of International Conference on System Modeling &Advancement in Research Trends (SMART 2012), 2012Anusaaraka, Indian Institute of Information Technology, Hyderabad,available at anusaarakaLatha R Nair, David Peter & Renjith P Ravindran, “Design andDevelopment of a Malayalam to English Translator – A Transfer BasedApproach”, Int. J. of Computational Ling. (IJCL), Volume 3, Issue languageVenkateswara Prasad T. and Mayil Muthukumaran G., “HandlingElisions in Telugu to English Machine Translation”, Journal of NaturalLanguage Processing, 2012 (submitted)Wikipedia [Brahmic family of scripts], 2012, available family of scriptsLearn Telugu,, available at telugu.php[15] Sri Badri Narayanan R., English – Telugu Rule Based MachineTranslation System, Master’s Thesis, Amrita Vishwa project/mhrd/ ms/Final Thesis.pdf[16] Tripathi Sneha and Sarkhel J. K., Approaches to Machine Translation,Annals of Library and Info. Sc., Vol. 57, pp. 388-393, 2010, available ALIS%2057%284%29%20388-393.pdf[17] Ning Kang, Bharat Singh, Zubair Afzal, Erik M van Mulligen, and JanA Kors, Using rule-based natural language processing to improvedisease normalization in biomedical text, J Am Med Inform Assoc, 2012,doi:10.1136/amiajnl-2012-001173[18] Khaled Shaalan, Rule-based Approach in Arabic Natural LanguageProcessing, Int. J. of Info. & Comm. Tech., Vol. 2, No. 3, pp. 11-19,2010, available at aledl.pdf[19] [20] Ahrenberg Lars and Holmqvist Maria, Back to the Future? The Case I10.[21] Josan G. S. and Lehal G. S., Evaluation of Direct Machine pdf/directtrans.pdfMr. T. Venkateswara Prasad is a researcher working in the field ofnatural language processing for last few years. He is currentlypursuing his post graduate studies at CMJ University, Shillong,Meghalaya, India. His research areas include machine translation,human computer interaction, discourse analysis and translation.Mr. G. Mayil Muthukumaran is associated with NationalInformatics Centre, Ministry of Communications & IT, Govt. ofIndia in the capacity of Technical Director. He is currently directingmajor e-governance projects of the Central and State Governments.His research areas include data mining and data warehousing, designand analysis of algorithms, supercomputing, cloud computing, etc.International Journal of Science and Engineering Investigations, Volume 2, Issue 12, January 2013ISSN: 2251-8843www.IJSEI.com29Paper ID: 21213-05

Figure 2. Screen shot of the TETS systemTABLE III. TRANSLATION OF SIMPLE SENTENCES FROM TELUGU TO ENGLISHS.No.Telugu textRoman TeluguEnglish textTranslated byGoogleTranslated by ourSuite*1ఆమె చూచుచున్నదిaame chuuchuchunnadiShe is seeingShe cucucunnadiShe is seeing2వారు వెళ్ళుచున్ననరుvaaru veLLuchunnaaruthey are goingThey vellucunnaruThey are going3మీరు న్డుచుచున్ననరుmiiru naDuchuchunnaaruYou are walkingYou naducucunnaruyou are walkingaame vaarini tiTTuchunnadiShe is abusinghim/themShe tittucunnadithemShe is abusing themmiiru Telifoonu varakunaDuchiriyou walked tillthe telephoneYou can telephoneto naduciriYou till havewalked telephone45ఆమె �దిమీరు ట్ెలిఫో న్ు ��ు రేపు వాాయగలన్ుneenu reepu vraayagalanuI will (or can)write tomorrowI will writeI tomorrow canwrite7న్ేన్ు మాట్లాడుతనన్ుnenu matlaaDutaanuI will speakI am talkingI will speak8మీరు మాట్లాడుతనరుmeeru matlaaDutaaruyou will speakYou speakYou will speak9న్ేన్ు మాట్లాడన్ుnEnu maTlaDanuI don't speakI do not talkI do not speakvaaru ataniki salahaichchaaruthey advised himThey counseled himThey suggestiongave to him10వారు అతనికి సలహాఇచ్నారుInternational Journal of Science and Engineering Investigations, Volume 2, Issue 12, January 2013ISSN: 2251-8843www.IJSEI.com30Paper ID: 21213-05

S.No.Telugu textRoman TeluguEnglish textTranslated byGoogleTranslated by ourSuite*11న్ేన్ు దననిన తేగలన్ుneenu daanni teegalanuI can bring itI tegalanuI that can bringaame ataniniyennukuMTuMdishe chooses himShe made himyennukuntundiShe him will selectataDu daanni guriMchiyeDchaaDuhe cried about thatHe was about to cryHe that about weptmeeM daanni guriMchimaaTlaaDutaamwe talk about itWe're about itmatladutamWe that will speakabout121314ఆమె �్టెందిఅతడు దననిన � దననిన �నెంThis shell is washthis plate well washUmbrella, where itumbrella keep thereaa peTTe akkadapeTTavadduWash this platewellPut the umbrellathereDon’t put that boxthereThe box where thepettavadduthat box keep theredon'taa gurramu tvaragaaparugettutunnadiThat horse isgalloping quicklyThe horse quicklyparugettutunnadithat horse quicklyrunning is15ఈ చిపప బలగా కడుగుee chippa baagaa kadugu16గొడుగు అకకడ పట్టుgodugu akkada peTTu17ఆ పట్ెు అకకడ పట్ు వద్ుు18ఆ గురరము తవరగాపరుగెతు తతతన్నది19న్ే(న్ు) మాట్లాడనన్ుne(nu) matladaanuI spokeI talked toI spoke20న్ే(న్ు) వాాశాన్ుne(nu) vrasaanuI wroteI writeI wrote21న్ే తీసుకున్ననన్ుnee tiisukunnanee tiisukunnaanuI took theI have taken22అతన్ు తీసుకున్ననడుatanu tiisukunnaaDuI took (or)I have takenhe took (or)he has takenHe tookhe has taken23మెం మాట్లాడనెంmeeM maaTlaDaaMwe spokeWe matladamWe spoke24అతడు తీసుకుెంట్లడుataDu tiisukuntaaDuhe will takeHe was takingHe will take25మము మాట్లాడుతనముmemu maaTlaDutaamwe will speakWe matladutamuWe will speak26న్ేన్ు �neenu maaTlaaDutunnaanuI am speakingI was talkingI am speaking27న్ేన్ు వాాసుున్ననన్ుneenu vraastunnaanuI am writingI am writingI am writingatanu kaarunaDuputunnaaDuhe is driving carHe ran the carhe car is drivingmeemu premistunnaamuwe are lovingWe loveWe are lovingmeemu tiisukuMtunnaamuwe are takingWe takeWe are taking282930అతన్ు �ుమము ��International Journal of Science and Engineering Investigations, Volume 2, Issue 12, January 2013ISSN: 2251-8843www.IJSEI.com31Paper ID: 21213-05

TABLE IV. EXAMPLES OF BAD TRANSLATION TO ENGLISHSl.No.1Telugu QuestionsRoman Teluguవాడు ఎకకడ?vaaDu ekkaDa?atadu ekkaDa?అతడు ఎకకడ?EnglishQuestionsWhere is he?Translated by GoogleIf you know me?2మీకు న్ేన్ు తెలుసా?miiku neenu telusaa?3న్న పుసు కెం నీ(మీ) ద్గగ రnaa pustakaM nee(mii)daggara unnadaa?Do you knowme?Do you have mybook?ఉన్నదన?Where heTranslated byour Suite*exploit where?He where?Napustakam near you?4అది ఎెంత పద్ు ది?adi eMta peddadi?How big is it?It was so big?that how much5మీకు న్న సహాయెంmiiku naa sahaayaMkaavaalaa?Can I help you?If you want me to helpyou?to you my supportmiiru aaMglamulOmaTlaDagalaraa?Do you speakEnglish?Do you speak English?You in Englishspeakకావాలా?6మీరు ఆెంగా ది ఎెంత ద్ూరెం?idi eMta dooraM?How far is this?This is how far?this how much far8ట్ెైమెెంత?TaimeMta?Taimu eMta?What time is it?Taimenta?Time at how much?time how muchideMta?idi eMta?How much isthis?Identa?This is how much?this how muchmii(nee) pErEmi?nee pEru eemi?What is yourname?Miperemi?/ Niperemi?What is your name?you name whatట్ెైము ఎెంత?9ఇదెెంత?ఇది ఎెంత?10మీ(నీ) పరరేమి?నీ పరరు ఏమి?International Journal of Science and Engineering Investigations, Volume 2, Issue 12, January 2013ISSN: 2251-8843www.IJSEI.com32Paper ID: 21213-05

Telugu sentences into its equivalent English sentences. Telugu grammar, vocabulary and style as documented by well known Telugu and British scholars during the British rule in India were studied in depth [1-2]. These books were selected since they were published during the mid 19th and early 20th