Semi-automatic building of large-scale digital dictionaries

Kód výsledku v IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F00216224%3A14330%2F21%3A00136004" target="_blank" >RIV/00216224:14330/21:00136004 - isvavai.cz</a>

Výsledek na webu

—

DOI - Digital Object Identifier

—

Jazyk výsledku

angličtina

Název v původním jazyce

Semi-automatic building of large-scale digital dictionaries

Popis výsledku v původním jazyce

This paper presents a novel way of creating dictionaries by using a particular post-editing workflow, all of which is carried out in the context of building a set of three bilingual dictionaries - Tagalog, Urdu and Lao dictionaries with translations into English and Korean. The dictionaries were created completely from scratch without reusing any existing content and in a completely automatic manner, amounting to 50, 000 headwords, out of which 15, 000 headwords were subject to subsequent manual post-editing. In the paper we discuss the post-editing methodology that we used and its impact on the overall lexicographic workflow. We describe the web corpora that were built specifically for the purpose of building these three dictionaries as well as their annotations (such as PoS tagging and lemmatisation) and tools that were used for the corpus annotation and for automating individual entry parts and the post-editing thereof. Most of the automatic drafting and post-editing relied on a backbone consisting of the Sketch Engine corpus management system and Lexonomy dictionary editor We also detail the overall amount of work involved in each post-editing step, the technical and managerial difficulties faced alongside in the project, and the major technological issues that still need improvement in the post-editing scenario.

Název v anglickém jazyce

Semi-automatic building of large-scale digital dictionaries

Popis výsledku anglicky

This paper presents a novel way of creating dictionaries by using a particular post-editing workflow, all of which is carried out in the context of building a set of three bilingual dictionaries - Tagalog, Urdu and Lao dictionaries with translations into English and Korean. The dictionaries were created completely from scratch without reusing any existing content and in a completely automatic manner, amounting to 50, 000 headwords, out of which 15, 000 headwords were subject to subsequent manual post-editing. In the paper we discuss the post-editing methodology that we used and its impact on the overall lexicographic workflow. We describe the web corpora that were built specifically for the purpose of building these three dictionaries as well as their annotations (such as PoS tagging and lemmatisation) and tools that were used for the corpus annotation and for automating individual entry parts and the post-editing thereof. Most of the automatic drafting and post-editing relied on a backbone consisting of the Sketch Engine corpus management system and Lexonomy dictionary editor We also detail the overall amount of work involved in each post-editing step, the technical and managerial difficulties faced alongside in the project, and the major technological issues that still need improvement in the post-editing scenario.

Druh

D - Stať ve sborníku

CEP obor

—

OECD FORD obor

10201 - Computer sciences, information science, bioinformathics (hardware development to be 2.2, social aspect to be 5.8)

Projekt

<a href="/cs/project/LM2018101" target="_blank" >LM2018101: Digitální výzkumná infrastruktura pro jazykové technologie, umění a humanitní vědy</a><br>

Návaznosti

P - Projekt vyzkumu a vyvoje financovany z verejnych zdroju (s odkazem do CEP)

Rok uplatnění

2021

Kód důvěrnosti údajů

S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů

Název statě ve sborníku

Electronic lexicography in the 21st century. Proceedings of the eLex 2021 conference

ISBN

—

ISSN

2533-5626

e-ISSN

—

Počet stran výsledku

12

Strana od-do

396-407

Název nakladatele

Lexical Computing CZ s.r.o.

Místo vydání

Brno

Místo konání akce

virtual

Datum konání akce

5. 7. 2021

Typ akce podle státní příslušnosti

WRD - Celosvětová akce

Kód UT WoS článku

—