Automatic Derivation Of Nouns From Adjectives In Pashto
Automatic Derivation of Nounsfrom Adjectives in PashtoTariq NaeemMuhammad Abid Khan
Outline IntroductionIn the WorldProblem StatementData Collection / AnalysisModeling of DataImplementationError AnalysisConclusion
Introduction Derivations, in computational linguistics, are also calledlexeme formation and word formation. Derivational affixes change completely the grammaticalcategory of the root words to which they are connected. FoolishFoolishness
In the World. There are quite a lot of Natural Language Processing (NLP)applications. Morphological analysis is a pre-imperative in differentareas of NLP, for instance: Lemmatizers Parsers Stemmers Spell checkers Part of Speech taggers Speech Recognition systems
Contd. We looked in to the work of several other researchers inthe field of derivation in other languages.1. Paumari verbs as derivational additions depictingparticular affixes.2. Lexemes in Malay was inspected to find the associationbetween morphologically stems which can adapt toderivational morphology in pairs
Problem Statement Does class changing derivational morphological analysisexist in Pashto language? Under what circumstances, words in Pashto languagechange their grammatical category from adjectives tonouns?
Contd.3. Arabic language showed that part of words are deducedfrom stems by insertion of affixes.4. Turkish grammar was examined for dealing withproductive derivational strategies.5. Hindi original root words were examined to find patternsarranged by understanding the properties of the affixes.
Data Collection Developing corpus of Pashto language was the first andforemost activity. Data was collected from different areas, like, news,memos, letters, research articles, books, fiction, sportsand magazines, making it a representative corpus.
Data Analysis Pashto corpus was created using the corpus improvementtool XML Aware Indexing and Retrieval Architecture(XAIRA). Tagging of Pashto text was done in Extensible MarkupLanguage (XML) XAIRA tagged files were used in SQL Server 2012 tables.
Contd.Table: Derivational stems and suffixes The table has entry against every grammatical class. The derivations were broke down into stems and affixes
Modeling of Data We model our corpus data into Finite State Transducers(FSTs). We used FSTs because they are mathematically derivedmodels which efficiently compute many useful NLPfunctions and weighted transitions on strings. They are efficient and more effective.
Contd.The eight (8) unique rules defined were modeled by FSTs.1. The adjective ( تور Black) to a noun ( توروالي Blackness) andadjective ( کلک hard) to a noun ( کلک والي hardness). It isshown as:
Contd.2. [ ُرونړ ronrr] or [ ُروڼ ronrr] ‘bright’, [ ّرنړا rarra] or ّرڼا [rarra] ‘brightness’.
Contd.3. [تور tor] ‘dark’ or ‘black’ [ ّتیاره tiyarah] or [ تیار tiyara]‘darkness’ or ‘blackness’.
Contd.4. ژوندي [zowande] ‘alive’ or ‘existing’, [ ژوندون zwande]ّ‘life’, ‘existence’; [ نښتي nkhate] ‘captive’, ‘prison’, نښتون [nakhtoon] ‘captivity’, ‘imprisonment’.
Contd.5. [ ښه khah] ‘good’, [ ښیګړه khaegarrah] ‘goodness’.
Contd.6. [ وږي wagey] ‘hungry’ or [ لوږ lawaga] ‘hunger’
Contd.7. [ بیل bael] ‘separate’, [ بیلتون baeltoon] ‘separation’; [ ځاي zaey] ‘a place’, [ ځایتون zaeytoon] ‘a dwellingplace’, ‘a home’, ‘a birthplace’;
Contd.8. [ مین maen] ‘affectionate’ [ مینتوب maentoob] ‘affection’,‘love’, [ لیوني lewane] ‘mad’, [ لیونتوب lewantoob]‘madness’;
Design and Flowchart of System
Implementation We used four (4) programming languages/tools.1. We used XEROX LEXC compiler to draw finite statediagrams from the lexicon of adjectives and nouns inPashto.2. After finite state diagrams, we used XEROX XFST to compilethe .txt files generated by LEXC.3. MS SQL Server was used for corpus and4. MS CSharp was used as a front end to fetch data from thedatabase.
Error Analysis A sample of 174 nouns and 169 adjectives were collectedfrom the written Pashto data and given to the system asinput. Out of this input 147 nouns and 142 adjective wordswere accurately examined. Consequently, the complete accuracy of the overallsystem is:((147 142) / (174 169)) * 100 84.25%
Conclusion The collation feature for Pashto language is not available inolder versions of MS SQL Server. It was made available in MS SQL Server 2012 to storePashto text Derivation is sensitive because a slight change of َ[zabbar], َ[zeir] and ُ [paish] can be disastrous.
Future Work Further work can be done on class maintainingderivatives in Pashto language. Also, the corpus size can be increase to achieve higheraccuracy for derivational rules.
Questions please.
Pashto corpus was created using the corpus improvement tool XML Aware Indexing and Retrieval Architecture (XAIRA). Tagging of Pashto text was done in Extensible Markup Language (XML)
Like collective nouns, unit nouns are also countable A slice/slices Unlike collective nouns, unit nouns refer to separate instances of a phenomenon Compare: A pack of wolves, but an item of news Unlike quantifying collectives, unit nouns are always followed by an of uncountable nouns
NUMBER NOUNS O Singular nouns name only one person, place, thing, or idea. (student, school, toy, lesson, dress) O These nouns stay singular. (corn, sugar) O Nouns that name more than one person, place, thing, or idea are called plural nouns.Most singular nouns can be changed to a plural noun by adding -s or -es.(students, schools, toys, lessons; dresses)
1. The Arabic Aphabet 4 2. Vowels 5 3. Other Accents 8 Part II: NOUNS 4. The Word 9 5. Prepositions 10. 6. The Nouns 12 7. Nouns: Singular Form 15 8. Nouns: Dual Form 16 9. Nouns: Plural Form: Unbroken 18 10. Nouns: Plural Form: Broken 20 11. Detached Pronouns I 22 12. Detached Pronouns II 2
Nouns that do not change between singular and plural Nouns with no singular form Nouns with no plural form Animal names with two plural forms Plurals of compound nouns Normal Nouns Normal nouns are made plural simply by adding an s. Because normal nouns are very numerous, this is a
The Definition of (Common) Nouns and Proper Nouns Abelard Saragossà DOI: 10.2436/15.8040.01.26 Abstract This article defines common and proper nouns (§5) on the basis of traditional grammars (Grevisse, 1936; Fabra, 1956; §3). The definitions offer two justifications: proper nouns are syntactically equivalent to common nouns
Unit 2: Nouns Lesson 8 Nouns: Proper and Common A noun is a word that names a person, place, thing, or idea. There are two basic kinds of nouns: common nouns and proper nouns. A common noun names any person, place, thing, or idea. A proper noun names a particular person, place, thing, or idea. It may consist of
Plural Nouns 1 Most nouns can be both singular and plural. Singular nouns name one thing, for example: cow, peach, mother-in-law. Plural nouns name more than one thing, for example: cows, peaches, mothers-in-law. Most plural nouns are forme d by adding 's' or 'es' to the singular noun, for
Dec 07, 2017 · Identifying nouns Nouns are defined as naming a person, place, thing, or an idea. Circle if the following are nouns: tree when beds glass said slowly factory ticket boxes almost More practice, underline the nouns
