Adding Dialectal Lexicalisations to Linked Open Data Resources: the Example of Alsatian

• Main Author: Bernhard, Delphine
• Other Authors: Linguistique, Langues et Parole (LILPA), Université de Strasbourg (UNISTRA), COPAL
• Format: Conference Object
• Language: English
• Published: HAL CCSD 2014
• Subjects: Alsatian; spelling variants; lexicon alignment; [INFO.INFO-TT]Computer Science [cs]/Document and Text Processing; Reykjavík; Iceland
• Online Access: https://hal.archives-ouvertes.fr/hal-00966820; https://hal.archives-ouvertes.fr/hal-00966820/document; https://hal.archives-ouvertes.fr/hal-00966820/file/7_Camera_ready_paper.pdf

International audience This article presents a method to align bilingual lexicons in a resource-poor dialect, namely Alsatian. One issue with Alsatian is that there is no standard and widely-acknowledged spelling convention and a lexeme may therefore have several different written variants. Our proposed method makes use of the double metaphone algorithm adapted to Alsatian in order to bridge the gap between different spellings. Once variant citation forms of the same lexeme have been aligned, they are mapped to BabelNet, a multilingual semantic network (Navigli and Ponzetto, 2012). The mapping relies on the French translations and on cognates for Alsatian words in the English and German languages.