Text input aids such as automatic correction systems play an increasingly important role in facilitating fast text entry and efficient communication between text message users. Although these tools are beneficial when they work correctly, they can cause significant communication problems when they fail. To improve its autocorrection performance, it is important for the system to have the capabi...

This paper explores an approach for extracting scene text from a sequence of images with relative motion between the camera and the scene. It is assumed that the scene text lies on planar surfaces, whereas the other features are likely to be at random depths or undergoing independent motion. The motion model parameters of these planar surfaces are estimated using gradient based methods, and mul...

OCR (Optical Character Recognition) scanners do not always produce 100% accuracy in recognizing text documents, leading to spelling errors that make the texts hard to process further. This paper presents an investigation for the task of spell checking for OCR-scanned text documents. First, we conduct a detailed analysis on characteristics of spelling errors given by an OCR scanner. Then, we pro...

Phrase-based statistical machine translation (SMT) systems have previously been used for the task of grammatical error correction (GEC) to achieve state-of-the-art accuracy. The superiority of SMT systems comes from their ability to learn text transformations from erroneous to corrected text, without explicitly modeling error types. However, phrase-based SMT systems suffer from limitations of d...

Because of different designs, different text input devices have different error patterns. If we consider these aspects when designing an error correction mechanism, we can obtain significantly lower error rates. In this paper, we propose and evaluate a spelling algorithm specifically designed for a five-key chording keyboard. It is based on the maximum a posteriori probability (MAP) criterion, ...

In this paper we describe the technical implementation of our system that participated in the Helping Our Own 2012 Shared Task (HOO-2012). The system employs a number of preprocessing steps and machine learning classifiers for correction of determiner and preposition errors in non-native English texts. We use maximum entropy classifiers trained on the provided HOO-2012 development data and a la...

This paper includes the information about the technique used to detect Skew which are introduced during the scanning of the documents. It also discusses about the tool which have been used to implement the technique. The algorithm has been implemented on various scripts. The method provides a very efficient way to calculate the Skew. Correction in the skewed scanned document image is very impor...

The errors usually made by authors during text preparation are classified. The notion of semantic errors is elaborated, and malapropisms are pointed among them as “similar” to the intended word but essentially distorting the meaning of the text. For whatever method of malapropism correction, we propose to beforehand compile dictionaries of paronyms, i.e. of words similar to each other in letter...

We describe the ’TILB’ team entry for the CONLL-2013 Shared Task. Our system consists of five memory-based classifiers that generate correction suggestions for center positions in small text windows of two words to the left and to the right. Trained on the Google Web 1T corpus, the first two classifiers determine the presence of a determiner or a preposition between all words in a text. The sec...

We describe the Valkuil.net team entry for the HOO 2012 Shared Task. Our systems consists of four memory-based classifiers that generate correction suggestions for middle positions in small text windows of two words to the left and to the right. Trained on the Google 1TB 5gram corpus, the first two classifiers determine the presence of a determiner or a preposition between all words in a text i...