This thesis investigates N-best hypotheses reranking techniques for improving speech recognition accuracy. We have focused on improving the accuracy of a speech recognizer used in a dialog system. Our post-processing approach uses a linear regression model to predict the error rate of each hypothesis from hypothesis features, and then outputs the one that has the lowest (recomputed) error rate....

This paper proposes a new automatic speech summarization method having two stages: important sentence extraction and sentence compaction. Relatively important sentences are extracted from the results of large-vocabulary continuous speech recognition (LVCSR) based on the amount of information and the confidence measures of constituent words. The set of extracted sentences is compressed by our se...

Manually annotated data is expensive, so manually covering a large terminological resource like the UMLS Metathesaurus is infeasible. In this paper, we evaluate two approaches used to improve the quality of an automatically extracted corpus to train statistical learners to performWSD. The first one contributes to more specific terms while the second filters out false positives. Using both appro...

In this paper, we explore a novel bilingual word alignment approach based on DNN (Deep Neural Network), which has been proven to be very effective in various machine learning tasks (Collobert et al., 2011). We describe in detail how we adapt and extend the CD-DNNHMM (Dahl et al., 2012) method introduced in speech recognition to the HMMbased word alignment model, in which bilingual word embeddin...

Twitter is a type of social media that contains diverse user-generated texts. Traditional models are not applicable to tweet data because the text style is not as grammaticalized as that of newswire. In this paper, we construct word embeddings via canonical correlation analysis (CCA) on a considerable amount of tweet data and show the efficacy of word representation. Besides word embedding, we ...

Query term misrecognition caused by the speech recognizer is one of the important issues in the spoken query information retrieval. The misrecognized term in the transcribed query leads to the retrieval of irrelevant documents. To raise the correct ranking of the retrieved documents, we use a speech recognition confidence score based on word posterior probability to weight the term in the infer...

Recent work has shown success in using continuous word embeddings learned from unlabeled data as features to improve supervised NLP systems, which is regarded as a simple semi-supervised learning mechanism. However, fundamental problems on effectively incorporating the word embedding features within the framework of linear models remain. In this study, we investigate and analyze three different...

Previous functional magnetic resonance imaging (fMRI) studies have investigated the role of phonological processing by utilizing nonword rhyming decision tasks (e.g., Pugh et al., 1996). Although such tasks clearly engage phonological components of visual word recognition, it is clear that decision tasks are more cognitively involved than the simple overt naming tasks, which more closely map on...

We evaluate some recent developments in recurrent neural network (RNN) based speech enhancement in the light of noise-robust automatic speech recognition (ASR). The proposed framework is based on Long Short-Term Memory (LSTM) RNNs which are discriminatively trained according to an optimal speech reconstruction objective. We demonstrate that LSTM speech enhancement, even when used ‘näıvely’ as f...

In this paper we present a system named EasyCmd that provides voice navigation on the desktop of Microsoft Window 9x system. Speech recognition engine for EasyCmd is much similar to that for dictation machine. Statistical Knowledge Based Frame Synchronous Search algorithm (SKBFSS) and Word Search Tree (WST) technologies are applied for acoustic decoding. Recognition Score Gap (RSG) is used for ...