Title

Supervised And Unsupervised Automatic Spelling Correction Algorithms

Abstract

We present two algorithms for automatically improving the quality of texts which contain a large number of spelling errors 1. A supervised algorithm, which automatically corrects unknown words that are generated primarily from typing errors, is presented first. The second algorithm is an unsupervised approach to automatically correcting typing errors, individual words that have been split, multiple words which have been concatenated, and a combination of these errors. The algorithms have been developed and tested on a large source of real-world, human- and machine-generated spelling errors. © 2004 IEEE.

Publication Date

12-1-2004

Publication Title

Proceedings of the 2004 IEEE International Conference on Information Reuse and Integration, IRI-2004

Number of Pages

530-535

Document Type

Article; Proceedings Paper

Personal Identifier

scopus

Socpus ID

16244383879 (Scopus)

Source API URL

https://api.elsevier.com/content/abstract/scopus_id/16244383879

This document is currently not available here.

Share

COinS