An evaluation of a method to detect and correct erroneous characters in Japanese input through an OCR using Markov models

Abstract

The "Selective Error Correction Method" to judge these three types of errors, and correct them, using ra-th order Markov chain model for Japanese 'kanji-kana' characters, has been proposed and shown to be useful to detect and correct errors generated randomly (Araki et al., 1994).In this paper, this method is applied to detect and correct erroneous characters in Japanese text input through an OCR.. The method is confirmed to be also elfective to detect and correct the errors introduced by the OCR.

A Japanese OCR post-processing approach based on dictionary matching

“…After that, we will cut apart the Japanese written character strings with the maximal length word. When we do it, it will generate two results [3], the one is we really find the phrase from the character strings and so we can segment it and can apply it in the post-processing such as spelling check. But most of the time we find it can not get a matching from any character strings.…”

“…The key of this method is we must use the segmented result which gets from the Reverse Maximum Matching algorithm [5] to do the work. The increasing of recognition rate depends on the previous step.…”

Detection and correction of mutually interfered erroneous characters in Japanese texts

“…[2] The Accuracy Rate of Judging Types of Error Section Using Procedure 2 and Procedure 3 The average of accuracy rate for judging types of error section using Procedure 2 is about 90 %. The accumulative accuracy of correct candidates included in candidate lattice L obtained by Procedure 3, is shown in Fig.…”

“…Thie method diotinguisher the typea of erron and correcta them in accordance with the nature of each error charactera. It wae applied to Japantext input through an OCR [3]. The experiment8 showed that thlr method yield good resultr for error charactem located at dilhrent plaea each other (we call thir type of errore Y an iroloted emr) .…”

A revised method to detect erroneous characters wrongly substituted, deleted, and inserted at the end position in Japanese sentences and ‘bunsetsu’s

“…In order to solve this problem, by using the relation between the types of errors and the length of a chain in which the values of Markov joint probability remain small, a new method has been proposed to judge the three types of the errors, which are characters wrongly substituted, deleted, or inserted in Japanese sentences and ‘bunsetsu’s; to find the locations and the lengths of these erroneous characters; and to correct these errors in Japanese ‘kanji‐kana’ chains using m th‐order Markov chain model [11–18].…”