Introducing Sanskrit Wordnet
Malhar Kulkarni Chaitali Dangarikar Irawati Kulkarni Department of Humanities Center for Indian Lan- Center for Indian Lan- and Social Sciences, guage Technology, guage Technology, Indian Institute of Tech- Indian Institute of Tech- Indian Institute of Tech- nology Bombay nology Bombay nology Bombay [email protected] chaita- irawatikulkar- li.dangarikar@gmai [email protected] l.com
Abhishek Nanda Pushpak Bhattacharyya Center for Indian Language Technology, Center for Indian Language Technology, Indian Institute of Technology Bombay Indian Institute of Technology Bombay abhi[email protected] [email protected]
guages range from 10 million (Konkani) to 500 Abstract million (Hindi/Urdu). 2. Being a heritage language, there is need to How does one build the wordnet of a lan- digitize and preserve ancient texts in Sanskrit. guage that has a rich lexical tradition span- This activity is greatly helped by word lists. An ning over millennia? The sheer volume of Optical Character Recognition Device (OCR) for words and their nuances, the rich, deep and Sanskrit, for example, would need spell correc- diverse grammatical tradition, the pressure of tion after scan, and this would need an exhaus- modern developments on the language- all tive lexicon. these factors and more combine to pose unique challenges in creating lexical re- 3. Simlarly, there exists real need for trans- sources for such languages. This present pa- lating ancient texts to preserve traditional culture per describes the construction of Sanskrit and knowledge. An online wordnet would no wordnet, being built using the expansion ap- doubt be a great help to a translator. proach . It presents the processes and chal- 4. Machine aided translation (MAT) is ma- lenges involved in this task that purports to turing fast, and automatic translation of Sanskrit uncover the intimate linkage that underlies text is a challenging problem needing wordnet. Indian languages most of which have speaker 5. There is an enormous amount of Sanskrit population numbering 20 to 500 million. text which should be available in keyword based searchable form. Text search is greatly helped by 1 Introduction wordnets. Sanskrit is historically an Indo-Aryan language 6. The tradition of developing lexical resource is very old in Sanskrit. There are diverse koshas (Deshpande 1992 ) and one of the 22 official , (traditional and rich monolingual dictionaries) in languages of India. It has a vast literature and the Sanskrit (see section 1.2 below). Sanskrit word- interest in analyzing and translating these texts is net will serve as the single reference point always on the rise, worldwide. representing and pointing to all these resources. Specifically, our motivation for building Sanskrit wordnet arises from the following facts: 1.1 Sanskrit language 1. For all languages in the Indo European Indian subcontinent is inhabited by a very family in India, the roots can be traced to San- large population who speak languages belong- skrit. A large part of the vocabulary of these lan- guages is derived from Sanskrit which can, there- ing to 4 major families, Indo-Aryan (a sub- fore, provide the pivot resource for many Indian family of Indo-European), Dravidian, Tibeto- Burman and Austro-Asiatic. Sanskrit is the languages. The speaker population for these lan- oldest member of the Indo-Aryan language
family, a sub branch of Indo-Iranian, which in 1.2 Rich lexical tradition of Sanskrit turn is a branch of Indo European language Sanskrit has a rich tradition of creating léxica family. 4 There is a traditional fourfold division of lex- (Kulkarni, 2008). Nighantu (700BC) on which ical units of Indian languages into: Yaska is believed to have written a commentary 1 Nirukta 1. tatsama - words having their origin called is the oldest known treatise that तसम arranged lexical material from the point of view in Sanskrit and accepted in the modern Indo- of synonymy as well as homonymy , and this tradi- Aryan languages without any change in their tion continued to Pali5 tradition as well. The first phonology. 2 and the foremost popular name of lexicon work 2. तव tadbhava - words which have their in classical Sanskrit is Amarasimha’s Amarako- origin in Sanskrit but their phonological forms sha (6th century AD) (Oka, 1913). The Cata- are changed as per the rules of the modern Indo- logous Catalogorum lists at least 40 commenta- Aryan languages. ries on Amarkosha alone, which shows how im- 3. देशी desh• - words which are the native portant and popular this synonyms dictionary in words of the particular language and ancient India was. There were many other léxica created more 4. Bवदेशी videsh• - words borrowed from for- or less in the style of Amarakosha which are giv- eign languages. en in Appendix A (11 of them). The links to तसम tatsama and तव tadbhava The first modern-day dictionary of Sanskrit words, in particular, will be a great pan-Indian was the Sanskrit-English Dictionary compiled by linguistic resource for computational purposes. Professor H.H. Wilson and published in 1819 Table 1 below lists some examples of Sanskrit 3 (Wilson, 1819)Two Indian dictionaries came out words in Hindi wordnet . soon after, namely, the Shabdakalpadruma 6 (Deb , 1988 ) of Pt. Sir Raja Radhakanta Dev HWN Synset Tatsam HWN English 7 word synset meaning and Vacasptyam (Bhattacharya, 2003 ) com- basil {तुलसी , पावनी , बहुमंजर/ , वृंदा , तुलसी तुलसी piled by Pt Taranatha Tarkavacaspati. वृदा , वैंणवी , भारवी , मंजर/क, वृदा वृदा So far the electronic lexical resources availa- 8 Bव#पावन , Bव# -पूDजता , पुंपसारा , वैंणवी वैंणवी ble for Sanskrit are mainly online dictionaries. Bऽदशमंजर/ , Bऽदशम जर/ , तीोा , पावनी पावनी The linguistic resources like Shabdakalpadruma पऽपुंपा , ौीमंजर/ ,ौीम जर/ , पऽपुंपा पऽपुंपा } अमृता 4 Nighantu is Sanskrit term for the collection of words, {भWह ,भW ,ॅू,भृकुट/ , ॅू ॅू eyebrow, brow, superci- grouped thematic categories with brief annotations तेवर ,कोदंड ,कोडंड ,अब } भृकुट/ भृकुट/ lium 5 Pali is a Middle Indo-Aryan language (or Prakrit) of India. {पेशी ,माँस -पेशी ,मांस - पेशी muscle, mus- It is best known as the language of the earliest extant Budd- culus पेशी ,माँसपेशी ,मांसपेशी ,माँस मांसपेशी hist scriptures . 6 पेशी ,मांस पेशी ,नस } Shabdakalpadruma is a first Sanskrit uni-lingual dictio- बSगन ,बैगन ,भंटा ,भाँटा , शाकBबव बSगन eggplant, nary arranged in the modern alphabetical principles. It gives aubergine, full quotations and definitions from the original Koshas शाकBबव ,शाकBबवक , mad_apple शाकौे%ा बSगन which were unavailable in print at that time. Sets of syn- वृंताक ,वृताक ,नीलवृषा , onymous words from the traditional Koshas are arranged
शाकौे%ा ,वृंताक3 , वागुण ,वरा , िचऽफला बSगन under the headword, followed by the brief gloss. Each entry िचऽफला ,रकंठ , रकठ ,िनिालु, वृताक बSगन in the lexicon includes headword, its category, meaning, नीलफला ,नटपBऽका usages in the Sanskrit texts . 7 िनिालु बSगन Vacasptyam is a modern mono-lingual Sanskrit lexicon. It नीलफल बSगन arranges words in the Sanskrit alphabetical order and gives grammatical information with word derivations as per the Table 1 : Tatsama words in the HWN traditional Sanskrit grammar. It contains about 46970 unique words. Each entry in the lexicon includes headword, These representative examples show that the its category, meaning, set of synonymous words, usages and some other information. synsets in Hindi wordnet contain 60-70% tatsa- 8 ma (directly borrowed from Sanskrit) words. The online dictionaries available for Sanskrit are-(1) Monier Williams dictionary < http://webapps.uni- koeln.de/tamil/>, (2) Apte’s Sanskrit-English Dictionary < http://www.aa.tufs.ac.jp/~tjun/sktdic/>, (3) Apte’s English- 1 Tatsama Shabda Kosha (Tatsama words dictionary) is Sanskrit Dictionary < http://www.sanskrit-lexicon.uni- published by Kendriya Hindi Nideshalaya, Shiksha Vibha- koeln.de/aequery/index.html> and (4) Spoken Sanskrit Dic- ga, Manava Samsadhana Vikasa Mantralaya, Bharata Sara- tionary: an online hypertext dictionary for Sanskrit - English kara in 1988. and English - Sanskrit.< http://spokensanskrit.de/>. Apart 2 from that various scanned versions of the printed dictiona- See Hindi ki Tadbhava Shabdavali (Sarma, 1968 ). ries prepared by European scholars are available at < 3 www.cfilt.iitb.ac.in/wordnet/webhwn. http://www.sanskrit-lexicon.uni-koeln.de/>.
and Vaacaspatyam are vast . For example, a 1.4 Expansion approach for Indian lan- comparison of the entries for the word war in guage wordnets these electronic dictionaries with the synsets of Wordnet construction activities in India started in the same word in the Sanskrit Wordnet is a good 9 indicator of the richness of this lexical tradition 2000 and the Hindi wordnet (Narayan et al., in Sanskrit. 2002) is the first one which got released on the Web in 2006. It was built ab initio using words from available lexical resources of Hindi. The 1. Spoken Sanskrit Dictionary: (7 words) यु , युध,् design of the Hindi wordnet follows the famous संमाम , समर , आयोधन, आहव , रय . English WordNet 10 . 2. Apate’s Sanskrit-English Dictionary: (7 While following the expand method, the words) Bवमहः , संमहारः , वैरारंभः , वैरं, संमामः , युं, रणं Sanskrit wordnet follows the hierarchy preserva- 3. Monier Williams Dictionary: (56 words) tion principle (HPP) (Tufis et al., 2008). In the
अनीक ,अयामदG ,अबर/ष ,अरर ,आDज ,आनतG, आयोधन , hierarchy of the Hindi wordnet, if synset H 2 is a hyponym of synset H , and the translation equi- आहव , आहाव , कठाल , कदल , खज , न, नदनु, िनमहण , 1 valents in the Sanskrit wordnet for H and H are पुंकर ,ूBवदारण ,ूसर ,बलज ,भडन ,भर ,भीमर ,युकार , 1 2 S1 and S 2 respectively, then in the hierarchy of यु ,योध ,योधन ,रय ,राAट , ,वराक ,Bवदथ ,Bवदार , Sanskrit wordnet S 2 should be a hyponym of syn-
Bवदारण ,BवमदG ,BवमदGन ,शबर ,िशलीमुख ,संयत,् संयुग , set S 1. Thus, in the expansion approach lexico- संःफोट ,संःफेट ,संबद ,संबदन ,संय ,संगथ ,समनीक , graphers are spared the task of establishing afresh semantic relations for the synsets of San- समाघात ,समुदय ,समुदाय ,समुषG ,समोह ,समर ,समृित , skrit wordnet. Appendix 2 describes and shows सपराय ,हाऽ ,हार and the screenshots of lexicographers’ interface for 4. Sanskrit Wordnet: ( 97 words) युम ्, संमामः , creating the Sanskrit wordnet. समरः , रणः , समरम,् आयोधनम,् आहवम,् रयम,् अनीकः , 1.5 Synset creation in Sanskrit wordnet अनीकम,् अिभसपातः , अयामदGः , अररः , आबदः , योधनम,् Domains: Initially the Sanskrit wordnet started जयम,् ूधनम,् ूBवदारणम,् मृधम,् आःकदनम,् संयम,् creating synsets with random synsets from the समीकम,् सायराियकम,् कलहः , Bवमहः , संूहारः , किलः , Hindi Wordnet. Later on, lists of important San- संःफोटः , संयुगः , समाघातः , अयागमः , आहवः , समुदायः , skrit words were acquired from different sources. संयत,् सिमितः , आDजः , सिमत,् युत,् संरावः , आनाहः , University of Hyderabad provided a list of most सपरायकः , Bवदारः , दारणम,् संBवत,् सपरायः , बलजम,् frequent words in their Sanskrit corpus. It con- sisted of 8338 words. Another word list available , , , , , , , आनGः अिभमरः समुदयः Bववाक् Bवखादः नदनुः भरः on the indology forum 11 contains a list of 127796 आबदः , पृतनायम,् अभीकम,् समीकम,् ममसयम,् नेमिधता , unique words from two major epics of Sanskrit 12 13 सकाः , समनम,् मीऴ_् हे, पृतनाः , ःपृत,् ःपृ , मृत,् मृ , पृत,् literature: Ramayana and Mahabharata. The पृ , समत,् समयGः , समरणम,् समोहः , सिमथः , सखः , सगः , third list is prepared based on the lexicon called Bharatiya Vyavahara Kosha (Naravane, 1961). संयुगम,् सगथः , सगमः , वृऽतूयGम,् पृ>ः , आDणः , शीरसाितः , Table 2 shows the part of speech distribution of वाजसाितः , समनीकम,् खलः , खजः , पWःयम,् महाधनः , वाजः , Naravane’s lexicon. It contains 2766 words अजम,् स , संयत,् संय , संवतः which are used for 1969 concepts related to the day to day life. Table 3 shows a comparison be- 1.3 The process of building the Sanskrit tween the lists of Sanskrit words gleaned from wordnet various sources mentioned above. There are two methods to develop a Wordnet: (1) Expand method and (2) Merge method (Vos- sen, 2002). In the first method, a wordnet is con- 9 www.cfilt.iitb.ac.in/wordnet/webhwn structed based on an existing wordnet. In the 10 Wordnet.princetoon.edu second method, sub-Wordnets for specific do- 11
The above mentioned words are organized kUla-vyApAraH where the members of the com- 14 into 52 domains . Omitting function words, a pounds are अय (anya ) ,ःथान (sthAna ) ,संयोग (sa- 17 core set of concepts was prepared and then by Myoga ), अनुकूल (anukUla ) , यापार (vyApAra ) . and Sept. 2009 synsets for all these core concepts 15 they are indicated by inserting hyphen. For ex- were created. ample- the gloss of a verb in Sanskrit is generally
created using technical terms like यापार vy ApAra Nouns Verbs Adjectives Adverbs ‘action’, जय janya ‘produced,’ अनुकूल anukUla 1512 225 180 52 ‘helpful,’ etc. 18 Table 2: POS distribution of the synsets created (core concepts) 2 Problems faced in the expansion ap- proach Hindi List Sanskrit List 1 Sanskrit List 2 Sanskrit List 3 1 In this section we enumerate the challenges faced Sanskrit Word Number of Hindi Univ. of Hyderabad list Sanskrit wordnet in creating the synsets of Sanskrit wordnet in most frequent (Based on Ra- Words in Nara- Total num- consonance with those of Hindi. words in Sanskrit mayana vane's ber of (Amba Kulkarni) and Mahabhara- Bhasha Vyava- unique ta) har Kosh words 8338 127796 2766 105157 17 This way of giving definitions is typical of Sanskritic Table 3 : Sanskrit word list tradition which used to strongly emphasise precision. The long compound simply defines the act of going. 18 So using these expressions, Hindi Wordnet gloss is While creating synsets the following considera- adapted in following ways- (1){ रोना , दन करना ,आँसू tions are kept in mind: बहाना ,बंदन करना ronA, rudana karanA, AMsu bhAnA, kran-
dana karanA } HWN AMkha se AMsu Inserting concepts or glosses in the Sanskrit आँख से आँसू िगराना girAnA SWN wordnet: A combination of the glosses given in सुख-दःखयोःु भावनावेगात ् नेऽायाम ् अौुपतन- dictionaries like Shabdakalpadruma and the पः यापारः । sukha-duHkhayoH bhAvanAvegAt netrAb- translation of the gloss of the Hindi wordnet syn- hyAm aZrupatan-rUpaH vyApAraH , (2){ मारना ,पीटना ,ूहार set is used to create the Sanskrit synset glosses. करना , ठUकना ,ठोकना ,Bपटाई करना ,धुनना ,धुनाई करना , While writing the gloss, complicated सDध s ताड़ना ,ूताड़ना ,रसीद करना mAranA, pITanA, prahAra karanA, 16 ThokanA, piTAI karanA, dhunanA, dhunAI karanA, tADa- sandhis and समास s samAsas (compounds) are nA, pratADanA, rasIda karanA } HWN Aकसी पर Aकसी वःतु avoided. Whenever lengthy compounds (having kisi par kisI vastu Adi se AghAta kara- 5-6 members) became necessary, the members of आAद से आघात करना nA SWN - the compounds were invariably joined with the कDःमन ् अBप केन अBप वःतुना आहनन पूवGकः यापारः। kasmin api kena api vastunA Ahanana-pUrvakaH hyphen symbol (-) as in: ‘‘ अय-ःथान-संयोगानुकूल- vyApAraH (3) {खर/दना ,बय करना ,मोल लेना ,लेना kharIdanA, यापार meaning the activity that is helpful in kraya karanA, mola lenA, lenA } HWN पैसे आAद देकर Aकसी reaching a place ’’ anya-sthAna-saMyogAnu- दकानु , यB आAद से कुछ सौदा मोल लेना paise Adi dekar kisI dukAna, vyakti Adi se kuch saudA mol lenA SWN आपणे 14 These domains are: 1) Grains and Cereals, 2) Limbs of वःतु तथा च तमूयम ् एतयोः आदान -ूदानामकः यापारः । Humans, 3) Medical treatment, 4) Tools & implements, 5) ApaNe vastu tathA cha tanmUlyam etayoH AdAna- Worms & Insects, 6) Minerals, 7) Food and Drinks, pradAnAtmakaH vyApAraH, (4) { ठना , $ होना , अनखना , 8)Games & sports, 9) Ornaments & Trinkets, 10) House- , , , , , rUThanA, hold articles, 11) Limbs of animals, 12) Post office, 13) सना @रसाना फूलना अनसाना अनखाना अनैसना Vegetables, 14) Directions, 15) Country, 16) Religion, 17) ruSTa honA, anakhanA, rUsanA, risAnA, phUlanA, anasA- Court, 18) Birds, 19) Trees & plants, 20) Dress, 21) Nature, nA, anakhAnA } HWN अूसन होकर उदासीन ,चुप या अलग हो 22) Animals, 23) Fruits, 24) Flowers, 25) Young-ones of जाना aprasanna hokara udAsIna, cupa yA alaga ho jAnA animals, 26) Amusement, 27) Spices, 28) Weights & meas- ures, 29) Colours, 30) Relatives, 31) Diseases, 32) Reptiles, SWN अूसनताहेतुजयः Bवयोग पः औदासीयफलजनकः वा 33) Conveyances, 34) Occupations, 35) Education, 36) यापारः । aprasannatAhetujanyaH viyogarUpaH audAsInya- Time, 37) Government, 38) Verbs, 39) Adverbs, 40) Ab- phalajanakaH vA vyApAraH (5) {आना:1, पहुँचना , पहुंचना , stract nouns, 41) Adjectives, 42) Prepositions, 43) Numer- als, 44) Conjunctions, 45) Collective words, 46) Pronouns, पधारना , अवना , आगमना AnaA, pahuMcanA, pahucanA, pad- 47) Ordinals, 48) Feminines, 49) Interjections, 50) War, 51) hAranA, avanA, AgamanA} HWN एक ःथान से आकर दसरेू House, and 52) Miscellaneous. ःथान पर उपDःथत होना eka stAna se Akara dUsare stAna 15 From this time Sanskrit Wordnet became a part of Indo- WordNet activity which provided a common platform for para upasthita honA SWN अय-ःथान-Bवयोग -पूवGकः अय- the lexicographers working on various Indian language ःथान-संयोगानुकूल - यापारः । anya-sthAna-viyoga-pUrvakaH Wordnets. anya-sthAna saMyogAnukUla-vyApAraH. 16 Phonological conjoining
Difficulty of finding equivalent words: (A plant- dry leaves of which Sometimes it is difficult to find a Sanskrit equiv- are boiled in the hot water and alent for a Hindi word. For example; the word a drink is prepared) and this gloss was modified in SWN as: {चाय } cAya (tea ) is very widely used. The con- cept of tea is explained as follows in the Hindi (4) चायः चहा एवंBवधैः शदैः भारतीय - wordnet: भाषासु ूिसः >ुपः - यःय शुंक -पणाGनां (1) चाय के पौधे क3 पBयU को पानी मQ डालकर चूणY उंणजले अिभपय तDःमन ् िवे शकGरा - चीनी , दधू आAद िमलाकर बनाया हआ पेय पदाथG ु दधाद/न ् संिमJय उंणपेयं िनमDयते। cAya ke paudhe kI pattiyon ko ु cAyaH cahA evaMvidhaH shabdaiH pAnI mein DAlkar cinI dUdha Adi milAkar banAyA huA peya padAr- bhAratIya-bhASAsu prasiddhaH kSupaH yasya shuSk-parNAnAm tha (A drink prepared by mixing the cUrNaM uSNajals Abhipacya tas- leaves of the Tea-plant with min drave sharkarA-dugdhAdIn sugar, milk and water) saMmIshrya uSNapeyaM nirmIyate (A plant, which is famous by
But Sanskrit does not have a word of its own for the names like चहा , चाय , etc. in this concept. Monier Williams in his Sanskrit- the Indian languages- dry English dictionary (MW hereafter) suggests that leaves of which are boiled in ‘‘ ’’ cahA (which is actually a Marathi word) the hot water and a drink is चहा prepared) should be used as a borrowed word. In the dic- tionary of spoken Sanskrit we find two different Difficulties with examples: regional words ‘‘ चाय ’’ cAya and ‘‘ चाया ’’ cAyA Generally, examples associated with Hindi syn- belonging to the North and South regions of In- sets are translated only if they read sensible dia. The gloss field in the synset of { कषायपेयम,् when translated into Sanskrit. In some cases, qu- चायः , चाया , चहा } {kaSAyapeyaM, cAyaH, cAyA, otations from the Sanskrit texts are included in cahA } in the Sanskrit wordnet is modified as fol- the example field. A special field has been lows: created to record the source of the quotations. This citation field is incorporated in the lexico- (2) चायः चहा एवंBवधैः शदैः भारतीय -भाषासु grapher's interface: ूिसःय >ुपःय शुंकपणाGनां चूणGम ् उंणजले अिभपय तDःमन ् िवे शकGरादधाद/नु ् संिमJय िनिमGतम ् उंणपेयम।् cAyaH cahA evaMvidhaiH shabdaiH bhAratIya-bhASAsu prasiddhasya kSupasya shuSka-parNAnAM cUrNam uSNajale abhipacya tasmin drave sharkarA-dugdhAdIn saMmishrya nirmitam uSNapeyam
(A hot drink which is prepared Figure 1. Lexicographer's interface to record ci- by first mixing the leaves of the a plant, which is famous tations by the names like cahA चहा , चाय cAya , etc. in the Indian lan- guages, into hot water and then mixing it with sugar and milk) The example with the citation is inserted in this This change is needed to translate the simple format: Hindi wordnet gloss. Similarly, for the tree plant, the Hindi wordnet gloss is: (5) "शिश -AदवाकरयोमGहपीडनम। ् " (3) एक पौधा Dजसक3 पBयाँ उबलते हए पानी मQ ु 19 [भतृG.2.91] डालकर एक पेय बनाते हS। eka paudhA jisakI pattiyAn uba- late hue pAnI mein DAlakar eka 19 peya banAte hein Here, [भतृG.2.91 ] indicates the place of the quotation in the original Sanskrit text authored by Bhartrhari.
shashi-divAkarayor grahapIDanaM AnAhaH, saMparAyakaH, vidAraH, dAraNacd, saM- [bhartR 2.91] vit, saMparAyaH, balajam, AnartaH, abhimAraH, (the eclipse of Sun and Moon). samudayaH, vivAkd, vikhAdaH, nadanuH, bharaH, Sometimes, apart from the translation of Hindi AkrandaH, pRtanAjyam, abhIkam, samIkam, mama- example sentence, an alternative example from satyam, nemadhitA, saGkAH, samanam, mIL_he, the Sanskrit text is provided. Multiple examples pRtanAH, spRt, spRd, mRt, mRd, pRt, pRd, samat, samaryaH, samaraNam, samohaH, samithaH, saGk- are separated with the ‘‘/’’ symbol. The sources haH, saGgaH, saMyugacd, saGgathaH, saGgamaH, of the examples are indicated in square brackets. vRtratUryam. pRkSaH, ANiH, ZIrsAtiH, vAjasAtiH, In some cases the translation of the Hindi exam- samanIkam, khalaH, khajaH, pauMsyam, mahAdha- ple sentence becomes problematic due to the un- naH, vAjaH, ajaM, sadma, saMyat, saMyad, saMva- naturalness of the sentence in Sanskrit. taH }
Coverage of words in Sanskrit wordnet: Tak- The problem of meaning attestation: Sanskrit ing into consideration the linguistic change and has a rich tradition of lexical resources. But the time, it is possible to classify Sanskrit language downside of this fact is that the lexicographer has into three periods- (1) Vedic period-beginning of to verify the consistency of word definitions at Vedic Sanskrit can be traced as early as around every step from multiple sources. For example, 1500 BCE and Vedas are written using literals of following words are mentioned in Shabdakalpa- that time, (2) Classical Sanskrit- A significant druma , but other dictionaries prepared by mod- form of post-Vedic Sanskrit is found in the San- ern scholars like Monier Williams (MW) make skrit beginning with the Hindu Epics-----the Ra- the following remarks in the gloss of these mayana and the Mahabharata. (3) Modern San- words. All of them are used in the Vedic litera- skrit. The usage of the words changed during ture for the concept of "war". these periods. The general policy adopted for synset making is to start with the most frequent सखे MW- सख -not found in MW and शदकपिमु words of modern Sanskrit and to close the synset MW- सग is not found in sense of यु in MW and सगे with the least frequent word of Vedic Sanskrit. शदकपिमु
The example of the synset of यु (yuddha) war- MW- संयुग n. conflict, battle, war MBh. Ka1v. &c (cf. Naigh. ii संयुगे shown below- is an illustrative case in point. The , 17) words in the synset of war are arranged from the सगथे MW- सगथ m. conflict, war Naigh. most common modern Sanskrit words to least सगमे MW- सग म does not have the sense of यु used Vedic Sanskrit words. वृऽतूयB MW- वृऽतूयG n. conquest of enemies or वृऽ , battle , victory RV. { , , , , , , , युम ् संमामः समरः रणः समरम ् आयोधनम ् आहवम ् पृ> MW- पृ> m. = संमाम Naigh. ii , 57. रयम,् अनीकः , अनीकम,् अिभसपातः , अयामदGः , अररः , MW- आDण m. (cf. अDण ( the pin of the axle of a cart RV. i , 35 आणौ आबदः , योधनम,् जयम,् ूधनम,् ूBवदारणम,् मृधम,् , 6 ; 63 , 3 ([" battle " Naigh. ii , 17]) and v , 43 , 8 आःकदनम,् संयम,् समीकम,् सायराियकम,् कलहः , Bवमहः , शीरसातौ शीरसाित is not found in MW and शदकपिमु , , , , , , , MW- समनीक n. battle, war RV. ( Naigh. ii , 17) Ba1lar. Vii , संूहारः किलः संःफोटः संयुगः समाघातः अयागमः आहवः समनीके 60÷61. समुदायः , संयत,् सिमितः , आDजः , सिमत,् युत,् संरावः , आनाहः , खले MW- खल m. contest, battle Naigh. Nir. सपरायकः , Bवदारः , दारणम,् संBवत,् सपरायः , बलजम,् खजे MW- खज m. contest, war (cf. -क्/ऋत ् &c ) Naigh. ii , 1 आनGः , अिभमरः , समुदयः , Bववाक्, Bवखादः , नदनुः , भरः , पWःये MW- पWःय is not mentioned in the sense of यु , , , , , , आबदः पृतनायम ् अभीकम ् समीकम ् ममसयम ् नेमिधता महाधने MW- महाधन m. a great contest, great battle ib. Naigh. सकाः , समनम,् मीऴ_् हे, पृतनाः , ःपृत,् ःपृ , मृत,् मृ , पृत,् MW- वाज m. the prize of a race or of battle, booty , gain , पृ , समत,् समयGः , समरणम,् समोहः , सिमथः , सखः , सगः , वाजे reward , any precious or valuable possession , wealth , t reasure RV. VS. AV. Pan5cavBr. संयुगम,् सगथः , सगमः , वृऽतूयGम,् पृ>ः , आDणः , शीरसाितः , MW- is not found in the sense of वाजसाितः , समनीकम,् खलः , खजः , पWःयम,् महाधनः , वाजः , अजम ् अजम ् यु स MW- स n. war , battle (= सं-माम ( ib.ii , 17 अजम,् स , संयत,् संय , संवतः } {yuddham, saMgramaH, raNaH, samaraH, samaram, Ayodhanam, Ahavam, संयत ् MW- संयत ् संय f. contest, strife, battle, war (generally found raNyam, anIkaH, anIkam, abhisampAtaH, abhyA- संय in loc. or comp.) MBh. Ka1v. &c mardaH, araraH, AkrandaH, yodhanam, jamyam, संवतः MW- not found in the sense of यु pradhanam, pravidAraNam, mRdham, Askandanam, Table 4. Verification of meaning of words stand- saMkhyam, samIkam, sAmyarAyikam, kalahaH, vi- ing for war . grahaH, saMprahAraH, kaliH, saMsphoTaH, saMyu- gaH, samAghAtaH, abhyAgamaH, AhavaH, samu- dAyaH, saMyat, samitiH, AjiH, samit, yut, saMrAvaH,
3 Special features of Sanskrit wordnet lexical data and the amount of lexicalization needs to be investigated. Verbal concepts: In Hindi wordnet, verbs are The future work is proposed to be carried out not inserted in their root forms. Instead, their in the following directions: dictionary forms like होना honA (to be) , करना karanA (to do) , खाना khAnA (to eat) , पीना pInA Use of ontology of न यन य----याययाय (Navya-NyAya) (to drink) etc. are included in the synset. The last The traditional Sanskrit Texts on Philosophy as ना nA is dropped through suffix stripping in verb well as Medicine contain various discussions on ontological categories and hierarchies. These morphology and the verb forms are generated texts are closely related to the grammar of the using only the initial parts like हो ho , कर kara , Sanskrit Language. The comparison of these on- खा khA , पी pI . Sanskrit lexicographers have tological structures and hierarchies to the exist- not conformed to this practice and have inserted ing one coming from the Hindi wordnet may shed light on new Indowordnet specific issues. the root forms of verbs like भू bhU (to be) , कृ kR (to do) , खा khAd (to eat) , पा pA (to drink), in धातु (dhAtu) based WN verbal synsets. There are theories in Sanksrit texts which adhere Gender: Sanskrit has grammatical gender. The to the view that all nouns are derived from verbal following practice is followed for tackling the roots. It is the actions denoted by the verbal roots issue of gender in Sanskrit wordnet: (1) In case that can be considered as the base of various ob- of nouns all gender variations are included in the jects denoted by nouns. There is a need to test synset. (2) Adjectives in Sanskrit have no gender this theory and build a lexical structure where all of their own. They take the gender of the nouns the verbal roots will be at the nodal level with which they qualify. Hence in the synset of adjec- connected nouns at the leaf level. A brief intro- tives only root forms are included. (3) Adverbs- duction of this is available in (Kulkarni and Technically adverbs in Sanskrit do not get con- Bhattacharyya, 2009). jugated as nouns and adjectives. But, we find that References some adverbs have BवभB (case ending) suffixes attached to them indicating the closed form of Abhishek G. Nanda. 2009. Tools and interfaces for wordnet construction, linking and maintenance. B. the word in that particular .(case ending). BवभB tech project report, Indian Institute of Technology In such cases, they are included as they are, i.e., Bombay, Mumbai. in the closed BवभB form. For example- Dan Tufis, Radu Ion, Luigi Bozianu, Alexandru Ceu- (6) सDनधौ sannidhau ‘near’ su, and Dan Stefaescu. 2008. Romanian wordnet: (which is actually a locative Cirrent state, new applications and proposals. In Attila Tanáces, Dóra Csendes, Vernoika Vincze, form of सDनिध sannidhi), Christaine Fellbaum, and Piek Vossen, editors, िनकटे nikaTe ‘near’ (which is ac- Proceedings of the Foruth Global WordNet Confe- tually a locative form of िनकट rence :441---445. nikaTa ), and H. H. Wilson, editor. 1819. A Dictionary in Sanskrit अदरेू adUre ‘near’ (which is ac- and English . Calcutta. tually a locative form of अदरू Krsnaji Govinda Oka, editor. 1913. Amarakosha of adUra ). Amarasinha . Law Printing Press. Malhar Kulkarni and Pushpak Bhattacharyya. 2009. 4 Conclusions and future work Verbal roots in the Sanskrit wordnet. In G. Huet, One of main challenges in creating the Sanskrit Amba Kulkarni, and Peter Scharf, editors, Sanskrit Computational Linguistics , Lecture Notes in Com- wordnet is dealing with the sheer volume of lexi- puter Science:328---338, Berlin/Heidelberg. Sprin- cal knowledge accumulated over at least 2000 ger-Verlag. years. The synsets tend to become long to ac- commodate coverage of words for a concept. The Malhar Kulkarni. 2008. Lexicographic traditions in other challenge is the extremely rich morphology India and Sanskrit. Journal of Language Technolo- gy, (1):160---165. of Sanskrit which produces new words from simple elements. The question of trade-off be- P. Vossen. 2002. Euro WordNet: General Document. tween a complex morphological interface to the University of Amsterdam.
Raja Radhakanta Deb, editor. 1988. Shabdakalpa- The linker tool druma , volume 1-5. Nag Publishers, 2003 edition. (Figure A. 2) is in- Delhi. tegrated in the inter- Saranamasimha Sarma. 1968. Hindi ki Tadbhava face for cross- Shabdavali . College Book Depo. linkage between the literals of source and Taranatha Tarkavacaspati Bhattacharya, editor. 2003. Vacaspatyam , volume 1-6 of Chaukhamba Sanskrit target synsets. It Book Series . Chaukhamba, Banares. allows a lexico- grapher to link a Vishwanath Dinkar Naravane. 1961. Bharatiya Vya- literal of the source vahara Kosha: Solah Bhasao ka kosha . Triveni language to one or Samgama. [In HIndi.]. more literals in the Appendix A: Early works on Sanskrit lex- corresponding target Figure A. 2 Linker ical knowledge bases (besides Amarako- language synset. sha) Transitivity- Value for the entire synset 1. Naamamaalikaa of Bhoja (11 C) Values for individual words 2.SiddhashabdaarNava of Sahajakirti- (17th C) in the synset 3. Shaaradiiyaakhyaanaamamaalaa of Harsakirti- (17th C) 4. Paryaayashabdaratna of Dhananjaya-Bhatta. Noun forms derived from the same root 5. Koshakalpataru 6. Naanaartharatnamaalaa of Irugapa Dandadhinatha उपसगG (prefix ) added to the root to (14th C) modify its sense 7. Naanaarthamañjarii of Raghava 8. DharaNiikosha of Dharanidas a (12th C) 9. Shivakosa of Sivadatta-Misra 10. Ekaarthanaamamaalaa-vyaksharanamamaalaa of Figure A.3 Morphological elements in the SWN Saubhari 11. Paramaanandiiyanaamamaalaa of Makrandadasa Appendix 2: Lexicographer's Interface for Sanskrit wordnet building
Figure A. 1 Lexicographer’s interface.
To create a lexical resource like wordnet, one needs a user friendly tool. Sanskrit wordnet team uses the MultiDict tool developed at the Center for Indian Language Technology, Computer Science Department, IIT Bombay (Figure A. 1). The tool provides an interface for linking the synsets that express the same meaning in differ- ent language (Nanda, 2009).