Nepali Lexicon Development

Similar documents
1. Introduction. 2. The OMBI database editor

Comprehension Recognize plot features of fairy tales, folk tales, fables, and myths.

Mercer County Schools

ELA/ELD Standards Correlation Matrix for ELD Materials Grade 1 Reading

BULATS A2 WORDLIST 2

Modeling full form lexica for Arabic

Opportunities for Writing Title Key Stage 1 Key Stage 2 Narrative

What the National Curriculum requires in reading at Y5 and Y6

Derivational and Inflectional Morphemes in Pak-Pak Language

Prentice Hall Literature: Timeless Voices, Timeless Themes Gold 2000 Correlated to Nebraska Reading/Writing Standards, (Grade 9)

Procedia - Social and Behavioral Sciences 154 ( 2014 )

Senior Stenographer / Senior Typist Series (including equivalent Secretary titles)

Prentice Hall Literature: Timeless Voices, Timeless Themes, Platinum 2000 Correlated to Nebraska Reading/Writing Standards (Grade 10)

Primary English Curriculum Framework

Literature and the Language Arts Experiencing Literature

The Internet as a Normative Corpus: Grammar Checking with a Search Engine

CLASSIFICATION OF PROGRAM Critical Elements Analysis 1. High Priority Items Phonemic Awareness Instruction

Parsing of part-of-speech tagged Assamese Texts

PAGE(S) WHERE TAUGHT If sub mission ins not a book, cite appropriate location(s))

FOREWORD.. 5 THE PROPER RUSSIAN PRONUNCIATION. 8. УРОК (Unit) УРОК (Unit) УРОК (Unit) УРОК (Unit) 4 80.

Reading Grammar Section and Lesson Writing Chapter and Lesson Identify a purpose for reading W1-LO; W2- LO; W3- LO; W4- LO; W5-

National Literacy and Numeracy Framework for years 3/4

Books Effective Literacy Y5-8 Learning Through Talk Y4-8 Switch onto Spelling Spelling Under Scrutiny

Grade 7. Prentice Hall. Literature, The Penguin Edition, Grade Oregon English/Language Arts Grade-Level Standards. Grade 7

English Language and Applied Linguistics. Module Descriptions 2017/18

Correspondence between the DRDP (2015) and the California Preschool Learning Foundations. Foundations (PLF) in Language and Literacy

SINGLE DOCUMENT AUTOMATIC TEXT SUMMARIZATION USING TERM FREQUENCY-INVERSE DOCUMENT FREQUENCY (TF-IDF)

Developing a TT-MCTAG for German with an RCG-based Parser

ScienceDirect. Malayalam question answering system

Stefan Engelberg (IDS Mannheim), Workshop Corpora in Lexical Research, Bucharest, Nov [Folie 1] 6.1 Type-token ratio

Specification and Evaluation of Machine Translation Toy Systems - Criteria for laboratory assignments

Intra-talker Variation: Audience Design Factors Affecting Lexical Selections

Subject: Opening the American West. What are you teaching? Explorations of Lewis and Clark

LING 329 : MORPHOLOGY

Dickinson ISD ELAR Year at a Glance 3rd Grade- 1st Nine Weeks

Controlled vocabulary

Content Language Objectives (CLOs) August 2012, H. Butts & G. De Anda

BANGLA TO ENGLISH TEXT CONVERSION USING OPENNLP TOOLS

The development of a new learner s dictionary for Modern Standard Arabic: the linguistic corpus approach

Enhancing Unlexicalized Parsing Performance using a Wide Coverage Lexicon, Fuzzy Tag-set Mapping, and EM-HMM-based Lexical Probabilities

The College Board Redesigned SAT Grade 12

Improved Effects of Word-Retrieval Treatments Subsequent to Addition of the Orthographic Form

Language Acquisition Chart

Program Matrix - Reading English 6-12 (DOE Code 398) University of Florida. Reading

A Simple Surface Realization Engine for Telugu

CX 105/205/305 Greek Language 2017/18

Development of the First LRs for Macedonian: Current Projects

Coast Academies Writing Framework Step 4. 1 of 7

AQUA: An Ontology-Driven Question Answering System

Houghton Mifflin Reading Correlation to the Common Core Standards for English Language Arts (Grade1)

Taught Throughout the Year Foundational Skills Reading Writing Language RF.1.2 Demonstrate understanding of spoken words,

Learning Disability Functional Capacity Evaluation. Dear Doctor,

Introduction to HPSG. Introduction. Historical Overview. The HPSG architecture. Signature. Linguistic Objects. Descriptions.

MULTILINGUAL INFORMATION ACCESS IN DIGITAL LIBRARY

Emmaus Lutheran School English Language Arts Curriculum

MISSISSIPPI OCCUPATIONAL DIPLOMA EMPLOYMENT ENGLISH I: NINTH, TENTH, ELEVENTH AND TWELFTH GRADES

correlated to the Nebraska Reading/Writing Standards Grades 9-12

Loughton School s curriculum evening. 28 th February 2017

Advanced Grammar in Use

GERM 3040 GERMAN GRAMMAR AND COMPOSITION SPRING 2017

Teaching Vocabulary Summary. Erin Cathey. Middle Tennessee State University

LANGUAGE IN INDIA Strength for Today and Bright Hope for Tomorrow Volume 11 : 12 December 2011 ISSN

Florida Reading Endorsement Alignment Matrix Competency 1

A Minimalist Approach to Code-Switching. In the field of linguistics, the topic of bilingualism is a broad one. There are many

Underlying and Surface Grammatical Relations in Greek consider

Constructing Parallel Corpus from Movie Subtitles

Target Language Preposition Selection an Experiment with Transformation-Based Learning and Aligned Bilingual Data

1 st Quarter (September, October, November) August/September Strand Topic Standard Notes Reading for Literature

Words come in categories

CS 598 Natural Language Processing

Cross Language Information Retrieval

CORPUS ANALYSIS CORPUS ANALYSIS QUANTITATIVE ANALYSIS

ENGBG1 ENGBL1 Campus Linguistics. Meeting 2. Chapter 7 (Morphology) and chapter 9 (Syntax) Pia Sundqvist

Universiteit Leiden ICT in Business

Language Acquisition Fall 2010/Winter Lexical Categories. Afra Alishahi, Heiner Drenhaus

4 th Grade Reading Language Arts Pacing Guide

Ch VI- SENTENCE PATTERNS.

Linguistic Variation across Sports Category of Press Reportage from British Newspapers: a Diachronic Multidimensional Analysis

Test Blueprint. Grade 3 Reading English Standards of Learning

TABE 9&10. Revised 8/2013- with reference to College and Career Readiness Standards

Writing a composition

Project in the framework of the AIM-WEST project Annotation of MWEs for translation

Myths, Legends, Fairytales and Novels (Writing a Letter)

Formulaic Language and Fluency: ESL Teaching Applications

Vocabulary Usage and Intelligibility in Learner Language

Basic concepts: words and morphemes. LING 481 Winter 2011

Constraining X-Bar: Theta Theory

Multilingual Sentiment and Subjectivity Analysis

5. UPPER INTERMEDIATE

Florida Reading for College Success

Proof Theory for Syntacticians

Exploiting Phrasal Lexica and Additional Morpho-syntactic Language Resources for Statistical Machine Translation with Scarce Training Data

The role of the first language in foreign language learning. Paul Nation. The role of the first language in foreign language learning

Semantic Modeling in Morpheme-based Lexica for Greek

Heritage Korean Stage 6 Syllabus Preliminary and HSC Courses

Course Outline for Honors Spanish II Mrs. Sharon Koller

The Effect of Extensive Reading on Developing the Grammatical. Accuracy of the EFL Freshmen at Al Al-Bayt University

South Carolina English Language Arts

HinMA: Distributed Morphology based Hindi Morphological Analyzer

Computer Software Evaluation Form

Transcription:

Nepali Lexicon Development 1 Sanat Kumar Bista, 1 Birendra Keshari 2 Laxmi Prasad Khatiwada, 2 Pawan Chitrakar, 2 Srihtee Gurung 1 Information and Language Processing Research Lab Kathmandu University, Dhulikhel, Kavre, Nepal. 2 Madan Puraskar Pustakalaya Lalitpur, Nepal sana@ku.edu.np, birendra@ku.edu.np, laxmi@mpp.org.np Abstract This report on Nepali Lexicon Development highlights the process followed in the development of Nepali Lexicon for the Nepal component of the PAN localization project. The scope of this development initiative lies within its use in spell checking system for Nepali. The target is to have a minimum measure of twenty five thousand words with at least four hundred noun headwords, one hundred and fifty verbs, one hundred adverbs and adjective and all the pronouns of Nepali grammar in the entries of lexicon database. This report also describes briefly the tool used in development, and some problems and issues faced during the lexicon development process. 1. Introduction Lexicon is the heart of any natural language processing systems. All NLP Systems like Spelling Checker, Syntax Checker, Machine Translation and others require a lexicon to work. The format of the lexicons for these systems may differ according to their specific need but the basic concept of Lexicon remains the same for any NLP systems. A lexicon may be defined as a list of words with additional word-specific information, i.e. a dictionary. The word has a Greek origin, which means vocabulary. The study of linguistics involves several issues like what words are, the structure of the vocabulary in the language, the methods of using and storing words, the way words are learnt, the origin of words, relationships between words, the methods of words creation etc. [2]. words belonging to a certain category, the vocabulary of the language is built. Then for generating certain new words (both simple and complex), lexical rules are used. For example, the suffix -able which clearly applies to transitive verbs like measure to get a new word, measurable does not necessarily apply to some other verbs like sing and dance [2]. Two important terminologies worth introducing here are Lexicology and Lexicography. Lexicology is the science of dictionary compilation and Lexicography the art of lexicon building [1]. Some linguistic aspects of concern to a Lexicographer are: a. Linguistic information such as phonemic, morphological, and semantic analysis. b. Language, context relative as well as grammatical information of that Language. c. World knowledge and extra linguistics information such as cultural conventions, date, and currency format etc. Lexicons are distributed according to several features. It can be language specific, subject specific, style specific etc. The Nepali Lexicon intended for development falls under the category of Language Lexicon designed to suit those readers who intend to use Nepali Language in writing texts. The following figure categorizes the lexicon distribution and also shows the position of Nepali Lexicon in it. Linguistics, however, has a more specialized definition of the lexicon as it deals with lexemes to actualize the words. Lexemes are formed with the help of morpho-syntactic rules. Being constrained on a set of certain principles, say for instance grouping all

Nepali Lexicon 3.1.1. Planning. Language Subject Style This phase consist of the overall planning for the lexicon development process. Some of the specific activities are choice of using tools for development, frequency count utility, file format choice etc. Monolingual Multilingual Select a word Consult Dictionaries General Class find common usage of the selected word in all the consulted dictionaries Nepali Lexicon Figure 1: Nepali Lexicon in lexicon distribution Accumulation of words Y commo n matchi N Reject Choose another word 2. Objective and Scope Select a word The main objective of the development of Nepali Lexicon is for its use in Nepali Spell Checking System. The lexicon is supposed to facilitate word level spell checking and is monolingual in nature. The lexicon targets to accommodate at least twenty five thousand words and is expected to contain all the words from dictionaries and some recently emerging words from technical fields like information and communication technology that are not enlisted by the existing Nepali dictionaries. 3. Development Process The process chart given in Figure 2 is an overview of the overall process of Nepali Lexicon Development. The following specific steps have been followed in the development of Nepali Lexicon: 3.1. Step I The first step consists of planning, identifying the types, finalizing the source area and the volume to be managed by the author. This first step can further be divided into the following phases: 312 Assignment of grammatical Category for the word Fragmentation of word into lexical and functional category Deciding appropriate annotations Planning the layout and the order of the word Entry of the word according to the predefined attributes Figure 2: Process Chart for Nepali Lexicon Development 3.1.2. Material Collection and Indexing This is an important phase of the lexicon development process. Material collection largely depends upon the category of the lexicon. As the intended lexicon is of general type, the initial inventory of words is derived from sources like

Working Papers 2004-2007 dictionaries and survey of the vocabularies commonly found in grammatical text books. The dictionaries referred for the word selection are: Brihad Nepali Sabdakosh [3], Practical Nepali Sabdakosh [4], Sajha Sabdakosh [5] and Technical Nepali Sabdakosh [6]. Other sources such as fiction texts, newspapers, periodicals etc were also referred in this phase. Earlier, in absence of a reliable tool for frequency counting of a word occurrence in texts, word selection was done on the basis of linguistic experience of a linguist. Later, a frequency count tool was available and hence was used in including words in lexicon according to the frequency measure of the word. 3.1.3. Entry Selection This is the lexical phase of lexicon development. After material collection proper citation work is needed. The selection of entry words for citation is directed by the scope and the objective of this lexicon development. Relations of direct, inflection, derivation have been calculated properly to choose the proper head word. Some words that have been considered are: a. Technical words b. Newly coined word c. Obsolete words d. Eco-empty words e. Relational words like prepositions and post position. f. Prefix suffix and infix g. Compound Words h. Idioms 3.2. Step II In this part entry layout and order are addressed. Headword, lexical unit, grammar, meaning and other lexicon type sensitive features are described properly. The following lexical attributes have been established for Nepali Lexicon: a. Root word b. Head word c. Pronunciation d. Syllable break e. Meaning f. Parts of Speech g. Synonym h. Idiom 3.3. Step III This phase primarily focuses on lexicon formatting and enables the lexicon with proper shape and sequence. This is the final step of lexicon editing. Before the publication or bounding level (in electronic lexicon) every element should be checked. In this part the following aspects have been addressed properly. a. Entry sequence according to the text variety b. Collation sequence c. Subject division d. Index e. Bibliography 4. LexTool: A Tool developed and Used for Nepali Lexicon development In order to ease the building and maintenance of Lexicon for Nepali language, LexTool has been developed and used throughout. Though the main aim of this tool is to assist the development of Nepali Lexicon, it can be used to develop the lexicon of any other languages because this system is Unicode aware which means this system supports multi-language. The system stores the lexicon database in a XML file format (Unicode). The main reason for this is that XML is easily portable. The extensibility and the structured nature of XML allows it to be used for communication between different systems, which otherwise would be unable to communicate. The Lexicon in XML format can be exported to different formats according the specific need of the NLP systems. The system requires manual entries of the stem words. After that it is feed with some morphological rules. The rules are stored as a separate database. The rules can be loaded and then the system can generate some other entries by manipulating the rules. Thus, it provides facility of semi-automatic expansion of the existing lexicon. 313

Nepali The following figure provides an overview of the architecture of LexTool. e. Corpus: This is the corpus that is to be feed to the Rule Interpreter Module for semi-automatic expansion of the existing Lexicon Database. The corpuses are generally newspaper, journals and magazines in digital form. f. Rule Interpreter Module: This module interprets the rules that is loaded by Rule Maintainer Module and reads the Corpus. Depending upon the morphological rules in the Rule Database, this module generates some entries and stores them in the Lexicon Database. Figure 3: LexTool Architecture LexTool consists of three major modules; Lexicon Maintainer Module, Rule Maintainer Module and Rule Interpreter Module. These modules interact with Lexicon Database, Rules Database and Corpus. The detailed description of each of these modules and the entity they interact with are given below: a. Lexicon Database: This is the lexicon database in XML format that contains several lexical entries. Each entry contains the language head word and all related attributes. There can be separate lexicon databases for different languages. Similarly, same language can have different lexicon database may be each for a specific category of words that share common attributes. b. Lexicon Maintainer Module: This module interacts with the Lexicon Database. It loads, modifies and saves the Lexicon Database. c. Rules Database: This database contains the morphological rules that can help in the generation of other entries in the Lexicon Database. It is also in XML format. d. Rule Maintainer Module: This module interacts with Rules Database. It can load the Rule Database, modify it and save it. 314 4.1. Nepali Lexicon Development Using LexTool: Nepali is a morphologically rich language. Both inflectional and derivational morphology are found in the words of this language. This information can be very nicely used in the semi-automatic generation of Nepali Lexicon using LexTool. The morphological information found in this language can be put as rules database in LexTool. For example in Nepali language plural forms of nouns are generated by adding haruu at the end. Such inflectional morphology can be put in the Rules Database. While reading the corpus, if the Rule Interpreter Module finds a token (word) with ending, haruu it can detect that the root of the word is a noun and thus it can store the root as noun in Lexicon Database. Similarly, other rules can govern the correct determination of the attributes of stem words by morphological analysis. The exceptions found in the language can cause the errors in the lexical entries and they can be manual removed. 5. Results The output that is currently available is a Nepali Lexicon of approximately five hundred root words. The result is categorized into different grammatical parts such as noun, pronoun, verb, adverb, preposition etc. The lexicon entries have been annotated according to the contemporary linguistics norms.

Working Papers 2004-2007 6. Conclusion The initial phase of Nepali Lexicon development process suffered a lot due to the unavailability of technical tools to analyze several things like word frequency. Also unavailability of appropriate data of Nepali corpus still remains as a major in the lexicon development process. However, with some tools that have been developed on our own effort we are able to initiate the development process of Nepali Lexicon. The targeted lexicon is of 25,000 head words. 7. References [1] N. Krishnaswamy, S.K Verma, and M.Nagarajan, Modern Applied Linguistics, Mac Millan, India, 2000 [2] ] http://www.wikipedia.org [3] Ghimire M.et al., Brihad Nepali Sabdakosh, Rayal Nepal academy, Nepal, 5 th Reprint, 2004 [4] Adhikari H., Prayogatmak Nepali Sabdakosh, Bidthyarthi Prakashan, Nepal, 2005 [5] Dixit N.A.,Sajha Sabdakosh, SajhaPrakashan, Nepal, 5 th Reprint 2001 [6] Bigyan Nepali Sabdakosh, Royal Nepal Academy, Nepal, 2002. 315