DOCSLIB.ORG

Inferring Parts of Speech for Lexical Mappings Via the Cyc KB

Total Page：16

File Type：pdf, Size：1020Kb

Download full-text PDF Read full-text
Inferring Parts of Speech for Lexical Mappings Via the Cyc KB to appear in Proceeding 20th International Conference on Computational Linguistics (COLING-04), August 23-27, 2004 Geneva Inferring parts of speech for lexical mappings via the Cyc KB Tom O’Hara‡, Stefano Bertolo, Michael Witbrock, Bjørn Aldag, Jon Curtis,withKathy Panton, Dave Schneider,andNancy Salay ‡Computer Science Department Cycorp, Inc. New Mexico State University Austin, TX 78731 Las Cruces, NM 88001 {bertolo,witbrock,aldag}@cyc.com [email protected] {jonc,panton,daves,nancy}@cyc.com Abstract in the phrase “painting for sale.” In cases like this, semantic or pragmatic criteria, as opposed We present an automatic approach to learn- to syntactic criteria only, are necessary for deter- ing criteria for classifying the parts-of-speech mining the proper part of speech. The headword used in lexical mappings. This will fur- part of speech is important for correctly identifying ther automate our knowledge acquisition sys- phrasal variations. For instance, the first term can tem for non-technical users. The criteria for also occur in the same sense in “paint for money.” the speech parts are based on the types of However, this does not hold for the second case, the denoted terms along with morphological since “paint for sale” has an entirely different sense and corpus-based clues. Associations among (i.e., a substance rather than an artifact). these and the parts-of-speech are learned us- When lexical mappings are produced by naive ing the lexical mappings contained in the Cyc users, such as in DARPA’s Rapid Knowledge For- knowledge base as training data. With over mation (RKF) project, it is desirable that tech- 30 speech parts to choose from, the classifier nical details such as the headword part of speech achieves good results (77.8% correct). Ac- be inferred for the user. Otherwise, often complex curate results (93.0%) are achieved in the and time-consuming clarification dialogs might be special case of the mass-count distinction for necessary in order to rule out various possibilities. nouns. Comparable results are also obtained For example, Cycorp’s Dictionary Assistant was using OpenCyc (73.1% general and 88.4% developed for RKF in order to allow non-technical mass-count). users to specify lexical mappings from terms into the Cyc knowledge base (KB). Currently, when a 1 Introduction new type of activity is described, the user is asked In semantic lexicons, the term lexical mapping de- a series of questions about the ways of referring to scribes the relation between a concept and a phrase the activity. If the user enters the phrase “painting used to refer to it (Onyshkevych and Nirenburg, for money,” the system asks whether the phrases 1995; Burns and Davis, 1999). Lexical mappings “paint for money” and “painted for money” are include associated syntactic information, in partic- suitable variations in order to determine whether ular, part of speech information for phrase head- ‘painting’ should be treated as a verb. Users find words. The term lexicalize will refer to the process such clarification dialogs distracting, since they are of producing these mappings, which are referred more interested in entering domain rather than lin- to as lexicalizations. Selecting the part of speech guistic knowledge. Regardless, it is often very diffi- for the lexical mapping is required so that proper cult to produce prompts that make the distinction inflectional variations can be recognized and gen- intelligible to a linguistically naive user. erated for the term. Although producing lexi- A special case of the lexicalization speech part calizations is often a straightforward task, there classification is the handling of the mass-count dis- are many cases that can pose problems, especially tinction. Having the ability to determine if a con- when fine-grained speech part categories are used. cept takes a mass or count noun is useful not only For example, the headword ‘painting’ is a verb for parsing, but also for generation of grammat- in the phrase “painting for money” but a noun ical English. For example, automatically gener- PhysicalDevice Collection: “Dubya” refers to President George W.Bush). The Microtheory: ArtifactGVocabularyMt KB also includes a broad-coverage English lexicon isa: ExistingObjectType mapping words and phrases to terms throughout genls: Artifact ComplexPhysicalObject the KB. A subset of the Cyc KB including parts of SolidTangibleProduct the English lexicon has been made freely available Microtheory: ProductGMt as part of OpenCyc (www.opencyc.org). isa: ProductType 2.1 Ontology Figure 1: Type definition for PhysicalDevice. Central to the Cyc ontology is the concept collec- tion, which corresponds to the familiar notion of ated web pages (e.g., based on search terms) oc- a set, but with membership intensionally defined casionally produce ungrammatical term variations (so distinct collections can have identical members, because this distinction is not addressed properly. which is impossible for sets). Every object in the Although learner dictionaries provide informa- Cyc ontology is a member (or instance,inCyc tion on the mass-count distinction, they are not parlance) of one or more collections. Collection suitable for this task because different senses of a membership is expressed using the predicate (i.e., word are often conflated in the definitions for the relation-type) isa, whereas collection subsumption sake of simplicity. In cases like this, the word or is expressed using the transitive predicate genls sense might be annotated as being both count and (i.e., generalization). These predicates correspond mass, perhaps with examples illustrating the dif- to the set-theoretic notions element of and subset ferent usages. This is the case for ‘chicken’ from of respectively and thus are used to form a partially the Cambridge International Dictionary of English ordered hierarchy of concepts. For the purposes of (Procter, 1995), defined as follows: this discussion, the isa and genls assertions on a a type of bird kept on a farm for its eggs or Cyctermconstituteitstype definition. its meat, or the meat of this bird which is Figure 1 shows the type definition for Physi- cooked and eaten calDevice, a prototypical denotatum term for count nouns. The type definition of PhysicalDevice indi- This work describes an approach for automati- cates that it is a collection that is a specialization cally inferring the parts of speech for lexical map- of Artifact, etc. As is typical for terms referred to pings, using the existing lexical assertions in the by count nouns, it is an instance of the collection CycKB.Wearespecificallyconcerned with select- ExistingObjectType. ing parts of speech for entries in a semantic lexi- Figure 2 shows the type definition for Water,a con, not about determining parts of speech in con- prototypical denotation for mass nouns. Although text. After an overview of the Cyc KB in the next the asserted type information for Water does not section, Section 3 discusses the approach taken to convey any properties that would suggest a mass inferring the part of speech for lexicalizations. Sec- noun lexicalization, the genls hierarchy of collec- tion 4 then covers the classification results. This is tions does. In particular, the collection Chemical- followed by a comparison to related work in Section CompoundTypeByChemicalSpecies is known to be 5. a specialization of the collection ExistingStuffType, via the transitive properties of genls. Thus, by 2 Cyc knowledge base virtue of being an instance of ChemicalCompound- In development since 1984, the Cyc knowledge base TypeByChemicalSpecies, Water is known to be an (Lenat, 1995) is the world’s largest formalized rep- instance of ExistingStuffType. This illustrates that resentation of commonsense knowledge, contain- the decision procedure for the lexical mapping ing over 120,000 concepts and more than a mil- speech parts needs to consider not only asserted, lion axioms.1 Cyc’s upper ontology describes but also inherited collection membership. the most general and fundamental of distinctions (e.g., tangibility versus intangibility). The lower 2.2 English lexicon ontology contains facts useful for particular appli- Natural language lexicons are integrated directly cations, such as web searching, but not necessar- into the Cyc KB (Burns and Davis, 1999). Though ily required for commonsense reasoning (e.g., that several lexicons are included in the KB, the English 1These figures and the results discussed later are lexicon is the only one with general coverage. The basedonCycKBversion576andOpenCycKBversion mapping from nouns to concepts is done using one 567. of two general strategies, depending on whether the 2 Collection: Water Usage Microtheory: UniversalVocabularyMt Predicate OpenCyc Cyc isa: ChemicalCompoundTypeByChemicalSpecies multiWordString 1123 24606 Microtheory: UniversalVocabularyMt denotation 2080 16725 genls: Individual compoundString 318 2226 Microtheory: NaivePhysicsVocabularyMt headMedialString 200 942 genls: Oxide total 3721 44499 Figure 2: Type definition for Water. Table 1: Denotational predicate usage in Cyc English lexicon. This excludes slang and jargon. mapping is from a name or a common noun phrase. Several different binary predicates indicate name- Usage to-term mappings, with the name represented as a SpeechPart OpenCyc Cyc string. For example, CountNoun 2041 21820 MassNoun 566 9993 (nameString HEBCompany “HEB”) Adjective 262 6460 Verb 659 2860 A denotational assertion maps a phrase into a AgentiveNoun 81 1389 concept, usually a collection. The phrase is spec- ProperCountNoun 16 906 ified via a lexical word unit (i.e., lexeme concept) Adverb 50 310 with optional string modifiers. The part of speech ProperMassNoun 1 286 is specified via one of Cyc’s SpeechPart constants. GerundiveNoun 7 275 Syntactic information, such as the wordform vari- other 39 185 ants and their speech parts, is stored with the Cyc total 3721 44499 constant for the word unit. For example, Device- TheWord, the Cyc constant for the word ‘device,’ Table 2: Most common speech parts in deno- has a single syntactic mapping since the plural tational assertions. The other entry covers 20 form is inferable: infrequently used cases.
Load more

Recommended publications
  • Problem of Creating a Professional Dictionary of Uncodified Vocabulary
    Utopía y Praxis Latinoamericana ISSN: 1315-5216 ISSN: 2477-9555 [email protected] Universidad del Zulia Venezuela Problem of Creating a Professional Dictionary of Uncodified Vocabulary MOROZOVA, O.A; YAKHINA, A.M; PESTOVA, M.S Problem of Creating a Professional Dictionary of Uncodified Vocabulary Utopía y Praxis Latinoamericana, vol. 25, no. Esp.7, 2020 Universidad del Zulia, Venezuela Available in: https://www.redalyc.org/articulo.oa?id=27964362018 DOI: https://doi.org/10.5281/zenodo.4009661 This work is licensed under Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International. PDF generated from XML JATS4R by Redalyc Project academic non-profit, developed under the open access initiative O.A MOROZOVA, et al. Problem of Creating a Professional Dictionary of Uncodified Vocabulary Artículos Problem of Creating a Professional Dictionary of Uncodified Vocabulary Problema de crear un diccionario profesional de vocabulario no codificado O.A MOROZOVA DOI: https://doi.org/10.5281/zenodo.4009661 Kazan Federal University, Rusia Redalyc: https://www.redalyc.org/articulo.oa? [email protected] id=27964362018 http://orcid.org/0000-0002-4573-5858 A.M YAKHINA Kazan Federal University, Rusia [email protected] http://orcid.org/0000-0002-8914-0995 M.S PESTOVA Ural State Law University, Rusia [email protected] http://orcid.org/0000-0002-7996-9636 Received: 03 August 2020 Accepted: 10 September 2020 Abstract: e article considers the problem of lexicographical fixation of uncodified vocabulary in a professional dictionary. e oil sublanguage, being one of the variants of common language realization used by a limited group of its medium in conditions of official and also non-official communication, provides interaction of people employed in the oil industry.
    [Show full text]
  • Towards a Conceptual Representation of Lexical Meaning in Wordnet
    Towards a Conceptual Representation of Lexical Meaning in WordNet Jen-nan Chen Sue J. Ker Department of Information Management, Department of Computer Science, Ming Chuan University Soochow University Taipei, Taiwan Taipei, Taiwan [email protected] [email protected] Abstract Knowledge acquisition is an essential and intractable task for almost every natural language processing study. To date, corpus approach is a primary means for acquiring lexical-level semantic knowledge, but it has the problem of knowledge insufficiency when training on various kinds of corpora. This paper is concerned with the issue of how to acquire and represent conceptual knowledge explicitly from lexical definitions and their semantic network within a machine-readable dictionary. Some information retrieval techniques are applied to link between lexical senses in WordNet and conceptual ones in Roget's categories. Our experimental results report an overall accuracy of 85.25% (87, 78, 89, and 87% for nouns, verbs, adjectives and adverbs, respectively) when evaluating on polysemous words discussed in previous literature. 1 Introduction Knowledge representation has traditionally been thought of as the heart of various applications of natural language processing. Anyone who has built a natural language processing system has had to tackle the problem of representing its knowledge of the world. One of the applications for knowledge representation is word sense disambiguation (WSD) for unrestricted text, which is one of the major problems in analyzing sentences. In the past, the substantial literature on WSD has concentrated on statistical approaches to analyzing and extracting information from corpora (Gale et al. 1992; Yarowsky 1992, 1995; Dagan and Itai 1994; Luk 1995; Ng and Lee 1996).
    [Show full text]
  • Symbols Used in the Dictionary
    Introduction 1. This dictionary will help you to be able to use Einglish - English Dictionary well & use it in your studying. 2. The exercies will provide you with what information your dictionary contains & how to find as well use them quickly & correctly. So you MUST: 1. make sure you understand the instructuion of the exercise. 2. check on yourself by working out examples, using your dictionay. 3. Do it carefully but quickly as you can even if you think you know the answer. 4. always scan the whole entry first to get the general idea, then find the part that you’re interested in. 5. Read slowly & carefully. 6. Help to make the learning process interesting. 7. If you make a mistake, find out why. Remember, the more you use the dictionary, the more uses you will find for it. 1 Chapter one Dictionary Entry • Terms used in the Dictionary. •Symbols used in the Dictionary. •Different typfaces used in the Dictionary. •Abbreviation used in the Dictionary. 2 1- Terms used in the Dictionay: -What is Headword? -What is an Entry? -What is a Definition? -What is a Compound? -What is a Derivative? -What information does an entry contain? 3 1- What is Headword? It’s the word you look up in a dictionary. Also, it’s the name given to each of the words listed alphabetically. 4 2- What is an Entry? It consists of a headword & all the information about the headword. 5 3- What is a Definition? It’s a part of an entry which describes a particular meaning & usage of the headword.
    [Show full text]
  • DICTIONARY News
    Number 17 y July 2009 Kernerman kdictionaries.com/kdn DICTIONARY News KD’s BLDS: a brief introduction In 2005 K Dictionaries (KD) entered a project of developing We started by establishing an extensive infrastructure both dictionaries for learners of different languages. KD had already contentwise and technologically. The lexicographic compilation created several non-English titles a few years earlier, but those was divided into 25 projects: 1 for the French core, 8 for the were basic word-to-word dictionaries. The current task marked dictionary cores of the eight languages, another 8 for the a major policy shift for our company since for the first time we translation from French to eight languages, and 8 more for the were becoming heavily involved in learner dictionaries with translation of each language to French. target languages other than English. That was the beginning of An editorial team was set up for developing each of the nine our Bilingual Learners Dictionaries Series (BLDS), which so (French + 8) dictionary cores. The lexicographers worked from far covers 20 different language cores and keeps growing. a distance, usually at home, all over the world. The chief editor The BLDS was launched with a program for eight French for each language was responsible for preparing the editorial bilingual dictionaries together with Assimil, a leading publisher styleguide and the list of headwords. Since no corpora were for foreign language learning in France which made a strategic publicly available for any of these languages, each editor used decision to expand to dictionaries in cooperation with KD. different means to retrieve information in order to compile the The main target users were identified as speakers of French headword list.
    [Show full text]
  • Japanese-English Cross-Language Headword Search of Wikipedia
    Japanese-English Cross-Language Headword Search of Wikipedia Satoshi Sato and Masaya Okada Graduate School of Engineering Nagoya University Chikusa-ku, Nagoya, Japan [email protected] masaya [email protected] Abstract glish article. To realize CLHS, term translation is required. This paper describes a Japanese-English cross-language headword search system of Term translation is not a mainstream of machine Wikipedia, which enables to find an appro- translation research, because a simple dictionary- priate English article from a given Japanese based method is widely used and enough for query term. The key component of the sys- general terms. For technical terms and proper tem is a term translator, which selects an ap- nouns, automatic extraction of translation pairs propriate English headword among the set from bilingual corpora and the Web has been in- of headwords in English Wikipedia, based on the framework of non-productive ma- tensively studied (e.g., (Daille et al., 1994) and chine translation. An experimental result (Lin et al., 2008)) in order to fill the lack of en- shows that the translation performance of tries in a bilingual dictionary. However, the qual- our system is equal or slightly better than ity of automatic term translation is not well exam- commercial machine translation systems, ined with a few exception (Baldwin and Tanaka, Google translate and Mac-Transer. 2004). Query translation in cross-language infor- mation retrieval is related to term translation, but it concentrates on translation of content words in 1 Introduction a query (Fujii and Ishikawa, 2001). Many people use Wikipedia, an encyclopedia on In the CLHS situation, the ultimate goal of the Web, to find meanings of unknown terms.
    [Show full text]
  • Download in the Conll-Format and Comprise Over ±175,000 Tokens
    Integrated Sequence Tagging for Medieval Latin Using Deep Representation Learning Mike Kestemont1, Jeroen De Gussem2 1 University of Antwerp, Belgium 2 Ghent University, Belgium *Corresponding author: [email protected] Abstract In this paper we consider two sequence tagging tasks for medieval Latin: part-of-speech tagging and lemmatization. These are both basic, yet foundational preprocessing steps in applications such as text re-use detection. Nevertheless, they are generally complicated by the considerable orthographic variation which is typical of medieval Latin. In Digital Classics, these tasks are traditionally solved in a (i) cascaded and (ii) lexicon-dependent fashion. For example, a lexicon is used to generate all the potential lemma-tag pairs for a token, and next, a context-aware PoS-tagger is used to select the most appropriate tag-lemma pair. Apart from the problems with out-of-lexicon items, error percolation is a major downside of such approaches. In this paper we explore the possibility to elegantly solve these tasks using a single, integrated approach. For this, we make use of a layered neural network architecture from the field of deep representation learning. keywords tagging, lemmatization, medieval Latin, LSTM, deep learning INTRODUCTION Latin —and its historic variants in particular— have long been a topic of major interest in Natural Language Processing [Piotrowksi 2012]. Especially in the community of Digital Humanities, the automated processing of Latin texts has always been a popular research topic. In a variety of computational applications, such as text re-use detection [Franzini et al, 2015], it is desirable to annotate and augment Latin texts with useful morpho-syntactical or lexical information, such as lemmas.
    [Show full text]
  • Phonemic Similarity Metrics to Compare Pronunciation Methods
    Phonemic Similarity Metrics to Compare Pronunciation Methods Ben Hixon1, Eric Schneider1, Susan L. Epstein1,2 1 Department of Computer Science, Hunter College of The City University of New York 2 Department of Computer Science, The Graduate Center of The City University of New York [email protected], [email protected], [email protected] error rate. The minimum number of insertions, deletions and Abstract substitutions required for transformation of one sequence into As grapheme­to­phoneme methods proliferate, their careful another is the Levenshtein distance [3]. Phoneme error rate evaluation becomes increasingly important. This paper ex­ (PER) is the Levenshtein distance between a predicted pro­ plores a variety of metrics to compare the automatic pronunci­ nunciation and the reference pronunciation, divided by the ation methods of three freely­available grapheme­to­phoneme number of phonemes in the reference pronunciation. Word er­ packages on a large dictionary. Two metrics, presented here ror rate (WER) is the proportion of predicted pronunciations for the first time, rely upon a novel weighted phonemic substi­ with at least one phoneme error to the total number of pronun­ tution matrix constructed from substitution frequencies in a ciations. Neither WER nor PER, however, is a sufficiently collection of trusted alternate pronunciations. These new met­ sensitive measurement of the distance between pronunciations. rics are sensitive to the degree of mutability among phonemes. Consider, for example, two pronunciation pairs that use the An alignment tool uses this matrix to compare phoneme sub­ ARPAbet phoneme set [4]: stitutions between pairs of pronunciations. S OW D AH S OW D AH Index Terms: grapheme­to­phoneme, edit distance, substitu­ S OW D AA T AY B L tion matrix, phonetic distance measures On the left are two reasonable pronunciations for the English word “soda,” while the pair on the right compares a pronuncia­ 1.
    [Show full text]
  • Towards Structuring an Arabic-English Machine-Readable Dictionary Using Parsing Expression Grammars
    Towards Structuring an Arabic-English Machine-Readable Dictionary Using Parsing Expression Grammars Diaa Mohamed Fayed1 Aly Aly Fahmy2 Mohsen Abdelrazek Rashwan3 Wafaa Kamel Fayed4 1 Department of Computer Science, Faculty of Computers and Information, 2 Department of Computer Science, Faculty of Computers and Information, 3 Department of EECE, Faculty of Engineering, 4 Department of Arabic Language and Literatures, Faculty of Arts, Cairo University, [email protected], [email protected], [email protected], [email protected] ABSTRACT: Dictionaries are rich sources of lexical information about words that is required for many applications of human language technology. However, publishers prepare printed dictionaries for human usage not for machine processing. This paper presents a method to structure partly a machine-readable version of the Arabic-English Al-Mawrid dictionary. The method converted the entries of Al-Mawrid from a stream of words and punctuation marks into hierarchical structures. The hierarchical structure or tree structure expresses the components of each dictionary entry in explicit format. The components of an entry are subentries and each subentry consists of defining phrases, domain labels, cross-references, and translation equivalences. We designed the proposed method as cascaded steps where parsing is the main step. We implemented the parser through the parsing expression grammar formalism. In conclusion, although Arabic dictionaries do not have microstructure standardization, this study demonstrated that it is possible to structure them automatically or semi- automatically with plausible accuracy after inducing their microstructure. Keywords: Al-Mawrid, Arabic-English, Machine-Readable Dictionary, Parsing Expression Grammars Received: 16 November 2013, Revised 19 December 2013, Accepted 24 December 2013 © 2014 DLINE.
    [Show full text]
  • Effects of Mora Phonemes on Japanese Word Accent
    PACLIC 18, December 8th-10th, 2004, Waseda University, Tokyo Effects of Mora Phonemes on Japanese Word Accent Yasuyo TOKUHIRO Shizuo HIKI Graduate School of Japanese Applied Waseda University Linguistics, Waseda University 2-579-15 Mikajima 1-7-14, Nishi-Waseda, Shinjuku-ku Tokorozawa, 359-1192 Japan Tokyo, 169-8050 Japan [email protected] [email protected] Abstract Effects of mora phonemes on Japanese word accent was analyzed statistically, utilizing a set of about 124,000 frequently used common nouns derived from the Japanese Word Dictionary edited by EDR (the Japan Electronic Dictionary Research Institute, Ltd., Japan). In this analysis, Japanese syllable was defined as preceding consonant (+semi-vowel) +following vowel, accompanied / not accompanied by mora phoneme. Mora phonemes adopted here were elongated portion of vowel: /H/, chocked sound: /Q/, syllabic nasal: /N/, and /i/ portion of diphthongs: /I/. The average ratio of attachment of accent to vowel type syllables was 13% and to semi-vowel +vowel type syllables 16%. In both types, the difference according to the kind of following vowels was not significant. As for the effect of mora phonemes, the ratios of accent attached to the syllables accompanied by mora phonemes were about several percents higher than their average for any kind of following semi-vowels and vowels. The average ratio of accent attached to vowel and semi-vowel +vowel type syllables was 14%, but in syllables of any kind of preceding consonants accompanied by mora phoneme, the ratios were more than several percents higher than the average. The words which involve the mora phonemes were categorized by each kind of mora phoneme, and compiled into the list in the order of Japanese syllabary, and the lists of different number of syllables composing each word in the order of Japanese syllabary.
    [Show full text]
  • From Machine Readable Dictionaries to Lexicons for NLP: the Cobuild Dictionaries - a Different Approach
    Carol Peters Istituto di Elaborazione dell'Informazione, CNR, Pisa Stefano Federici Istituto di Linguistica Computazionale, CNR, Pisa Simonetta Montemagni Istituto di Linguistica Computazionale, CNR, Pisa Nicoletta Calzolari Istituto di Linguistica Computazionale, CNR, Pisa From Machine Readable Dictionaries to Lexicons for NLP: the Cobuild Dictionaries - a Different Approach Abstract We describe the results of a syntactic-semantic parser for Cobuild dictionary definitions. Unlike previous work on the automatic analysis of machine readable dictionaries, the particular structure of the Cobuild definition allows us to derive information that classifies the lexical item mainly in terms of the selectional restrictions or preferences encoded on its arguments. The resulting formalized lexical entries contain data that has generally been lacking in other lexical representations but which is expected to be very useful in a wide range of NLP purposes. We show how this information can be used in dictionary sense disambiguation by creating links throughout the lexicon both on the paradigmatic and the syntagmatic axes. 1. Introduction It is widely recognized that machine readable dictionaries are important sources of lexical data which, suitably analysed, extracted and formalized, could be used effectively in the construction of lexical components for NLP applications. Recent studies in this direction have concentrated on classifying the headword in terms of the particular semantic features that can be derived from the genus items and their differentiae, and representing the lexicon as a hierarchical inheritance network (see, for example, the ACQUILEX project as described in Calzolari et al., forthcoming). In the paper we describe a different strategy that has been adopted to acquire lexical information from the Cobuild Student's dictionary.
    [Show full text]
  • Data for Lexicography the Central Role of the Corpus
    Wacana, Vol. 12 No. 2 (October 2010): 219—242 Data for lexicography The central role of the corpus ALLAN F. LAUDER Abstract This paper looks at the nature of data for lexicography and in particular on the central role that electronic corpora can play in providing it. Data has traditionally come from existing dictionaries, citations, and from the lexicographer’s own knowledge of words, through introspection. Each of these is examined and evaluated. Then the electronic corpus is considered. Different kinds of corpora are described and key design criteria are explained, in particular the size of corpus needed for lexicography as well as the issue of representativeness and sampling. The advantages and disadvantages of corpora are weighed and compared against the other types of data. :hile each of these has Eenefits, it is argued that corpora are a requirement, not an option, as data for dictionary making. Keywords Corpus linguistics, lexicography, data, linguistic intuition, citations, reading program, corpora, lexical database, type, token, corpus size, corpus representativeness, corpus balance, headword, lemma, lexeme, word form, word, general corpus, specialized corpus, monitor corpus, reference corpus, synchronic corpus Introduction For a dictionary to be reliable, it should provide generalizations about word behaviour that closely approximate the ways people generally use language for actual communication (Atkins and Rundell 2008: 45-46). These generalizations should be derived from evidence. Lexicographers have a number of sources of ALLAN F. LAUDER took his PhD at the University of Atma Jaya in Jakarta with a corpus linguistic study of keywords and collocates in an Indonesian paper. He has been a guest lecturer in Postgraduate Linguistics at the Faculty of Humanities at Universitas Indonesia teaching Psycholinguistics but currently holds a full time position in the central administration of the university as Consultant to the Rector.
    [Show full text]
  • Improving Dictionaries by Measuring Atypical Relative Word-Form Frequencies
    Proceedings of eLex 2019 Improving Dictionaries by Measuring Atypical Relative Word-form Frequencies Kristian Blensenius, Monica von Martens University of Gothenburg, Dept. of Swedish, Lundgrensgatan 1B, Gothenburg (Sweden) E-mail: [email protected], [email protected] Abstract In this article, we discuss and give examples of how word-form frequency information derived from existing corpora statistics can be used to improve dictionary content. The frequency information is used in combination with rule-based morphological data based on derivational and inflectional information from the Swedish Morphological Database compiled at the University of Gothenburg, and the lexical database owned by the Swedish Academy. The method currently used in the ongoing project for updating the monolingual Contemporary Dictionary of the Swedish Academy is described, and some examples of dictionary entries identified as candidates for update based on frequency measures are given. Different aspects of morphological dictionary content are discussed and highlighted by comparison between the above-mentioned definition dictionary and a learner’s dictionary. The role of headword or lemma as well as cross-referencing methods in a digital dictionary as compared to a printed dictionary is also discussed. Finally, a few examples of suggested modifications and enhancements are given. Keywords: morphology; frequency; word forms 1. Introduction In this article, we discuss how dictionary content can be improved by (re)using morphological information and enhancing it with corpus frequency information. Two contemporary Swedish monolingual dictionaries are used to illustrate how this method can be used to enhance dictionary content. Morphological frequency matters have been much discussed from various perspectives, for example from the point of view of learning and producing word forms (e.g.
    [Show full text]