Cleaning Corpus-generated Word Lists for an Indonesian Spelling-checker  

Oleh:

Hananto

Jenis: Article from Proceeding
Dalam koleksi: CONCORPS: The 3rd Atma Jaya Conference on Corpus Studies, Gaining Better Insights Into Language Through Corpora, Jakarta, August 21, 2015, page 9-14.
Topik: word list; corpus; Indonesian spelling checker
Fulltext: hal 9.pdf (11.76MB)

Isi artikel

This paper reports a work-in-progress project aiming at generating an Indonesian word list based on roughly more than 182.000.000 word-token-corpus. The resulting Indonesian word list later can be used as a custom dictionary in Microsoft Office package to identify the spelling mistakes of documents typed in Indonesian and to suggest possible corrections. This paper will discuss different approaches to making Indonesian spelling checkers, the background and the design of the present study, the problems encountered, and an attempt to overcome the problems to minimize the manual proofreading to clean of the word list.

Opini Anda

Klik untuk menuliskan opini Anda tentang koleksi ini!

Kembali