Post-Editing Error Correction Algorithm For Speech Recognition using Bing Spelling Suggestion
Journal Title: International Journal of Advanced Computer Science & Applications - Year 2012, Vol 3, Issue 2
Abstract
ASR short for Automatic Speech Recognition is the process of converting a spoken speech into text that can be manipulated by a computer. Although ASR has several applications, it is still erroneous and imprecise especially if used in a harsh surrounding wherein the input speech is of low quality. This paper proposes a post-editing ASR error correction method and algorithm based on Bing’s online spelling suggestion. In this approach, the ASR recognized output text is spell-checked using Bing’s spelling suggestion technology to detect and correct misrecognized words. More specifically, the proposed algorithm breaks down the ASR output text into several word-tokens that are submitted as search queries to Bing search engine. A returned spelling suggestion implies that a query is misspelled; and thus it is replaced by the suggested correction; otherwise, no correction is performed and the algorithm continues with the next token until all tokens get validated. Experiments carried out on various speeches in different languages indicated a successful decrease in the number of ASR errors and an improvement in the overall error correction rate. Future research can improve upon the proposed algorithm so much so that it can be parallelized to take advantage of multiprocessor computers.
Authors and Affiliations
Youssef Bassil , Mohammad Alwani
Keywords
Related Articles
- EP ID EP124811
- DOI -
- Views 141
- Downloads 0
An Analysis of Cloud Computing Adoption Framework for Iraqi e-Government
This paper presents an analysis of the factors which could have possible affect over the adoption of cloud computing via the Iraqi e-government. A conceptual framework model for cloud computing within Iraqi e-government...
Theoretical and numerical characterization of continuously graded thin layer by the reflection acoustic microscope
This article presents a theoretical and numerical study by the reflection acoustic microscope of the surface acoustic waves propagation at the interface formed by a thin layer and the coupling liquid (water). The thin la...
Case Study of Named Entity Recognition in Odia Using Crf++ Tool
NER have been regarded as an efficient strategy to extract relevant entities for various purposes. The aim of this paper is to exploit conventional method for NER in Odia by parameterizing CRF++ tool in different ways. A...
Proposal for Scrambled Method based on NTRU
Scrambling is widely used to protect the security of data files such as text, image, video or audio files; however, it is not the most efficient method to protect the security of the data files. This article uses NTRU pu...
Towards the Algorithmic Detection of Artistic Style
The artistic style of a painting can be sensed by the average observer, but algorithmically detecting a painting’s style is a difficult problem. We propose a novel method for detecting the artistic style of a painting th...