Notes:
An n-gram is a contiguous sequence of n items from a given sample of text or speech. N-grams are widely used in natural language processing and computational linguistics.
In the context of dialog systems, n-grams can be used to generate responses to user inputs. For example, a dialog system might use a model trained on a large dataset of text that includes n-grams up to 4-grams (four contiguous items). When the system receives a user input, it could use the n-grams in the input to generate a response by finding the most likely response based on the n-grams in the input. This approach can help the system generate more natural-sounding responses because it can take into account the context provided by the n-grams in the input.
N-grams can also be used to improve the performance of language models, which are used in many natural language processing tasks. For example, a language model might use n-grams to predict the next word in a sentence, or to classify a piece of text as spam or non-spam.
This references below mention that n-gram models are simple and dominant in speech recognition, but can only capture short-distance context dependency within a certain window of words. The text also discusses the importance of coping with out-of-vocabulary words and filled pauses in spoken dialogue systems and dictation systems, and presents a novel approach for improving the accuracy of spoken dialogue systems by integrating multi-layer concept information into the trigram language model. The text also mentions the use of language models in automatic speech recognition systems and the importance of language modeling in natural language continuous speech recognition. It discusses the challenges of recognizing name utterances accurately due to the large vocabulary size, and mentions the use of neural networks in speech recognition and the efficiency of statistical n-gram models in handling packet loss during transmission. The text also mentions the use of n-gram models in email classification tasks and the potential for n-gram models to be used in speech-driven question answering.
- Continuous speech recognition, also known as speaker-dependent speech recognition, is a type of speech recognition technology that is designed to recognize and transcribe speech spoken by a particular individual. One advantage of continuous speech recognition is that it can handle long stretches of uninterrupted speech, as the user does not need to pause between words or phrases.
- Name utterances refer to words or phrases that are used to refer to people or entities by name. In natural language processing, recognizing name utterances accurately can be a challenging task due to the large vocabulary size and the wide variety of ways that names can be expressed. Name utterances may play a crucial role in the meaning of a sentence or the context of a conversation.
- Statistical n-gram model is a type of statistical model that estimates the probability of a sequence of words occurring in a language. An n-gram model estimates the probability of any given word occurring based on the previous n-1 words in the sequence.
- Trigram language model is a statistical model that estimates the probability of a sequence of words occurring in a language. Specifically, a trigram model estimates the probability of any given word occurring based on the two previous words in the sequence (hence the name “trigram”). Trigram models are often more accurate than unigram models (which only consider the current word) or bigram models (which only consider the current word and the previous word) because they take into account more context. However, they may not be as accurate as higher-order models (such as 4-gram or 5-gram models) that consider even more context.
Wikipedia”
See also:
N-gram & Tag Clouds | N-gram Grammars | N-gram Timeline | N-gram Transducers (NGT)
Automatic language identification of telephone speech messages using phoneme recognition and n-gram modeling MA Zissman… – … , Speech, and Signal Processing, …, 1994 – ieeexplore.ieee.org Abstract The paper compares the performance of four approaches to automatic language identification (LID) of telephone speech messages: Gaussian mixture model classification (GMM), language-independent phoneme recognition followed by language-dependent … Cited by 98 – Related articles – BL Direct – All 4 versions
A unified context-free grammar and n-gram model for spoken language processing [PDF] from microsoft.comYY Wang, M Mahajan… – Acoustics, Speech, and …, 2000 – ieeexplore.ieee.org Abstract While context-free grammars (CFGs) remain as one of the most important formalisms for interpreting natural language, word n-gram models are surprisingly powerful for domain-independent applications. We propose to unify these two formalisms for both … Cited by 73 – Related articles – BL Direct – All 9 versions
Variable n-grams and extensions for conversational speech language modeling M Siu… – Speech and Audio Processing, IEEE …, 2000 – ieeexplore.ieee.org Abstract Recent progress in variable n-gram language modeling provides an efficient representation of n-gram models and makes training of higher order n grams possible. We apply the variable n-gram design algorithm to conversational speech, extending the … Cited by 75 – Related articles – BL Direct – All 6 versions
A maximum entropy language model integrating n-grams and topic dependencies for conversational speech recognition [PDF] from nthu.edu.twS Khudanpur… – … , Speech, and Signal Processing, 1999. …, 1999 – ieeexplore.ieee.org Abstract A compact language model which incorporates local dependencies in the form of N- grams and long distance dependencies through dynamic topic conditional constraints is presented. These constraints are integrated using the maximum entropy principle. Issues … Cited by 55 – Related articles – BL Direct – All 18 versions
[PDF] Interpolation of n-gram and mutual-information based trigger pair language models for Mandarin speech recognition [PDF] from suda.edu.cnZ GuoDong… – Computer Speech and Language, 1999 – nlp.suda.edu.cn Abstract While n-gram modeling is simple and dominant in speech recognition, it can only capture the short-distance context dependency within an n-word window where currently the largest practical n for natural language is three. However, many of the context … Cited by 38 – Related articles – BL Direct – All 6 versions
Dialog-context dependent language modeling combining n-grams and stochastic context-free grammars [PDF] from psu.eduK Hacioglu… – … , Speech, and Signal Processing, 2001. …, 2001 – ieeexplore.ieee.org Abstract We present our research on dialog dependent language modeling. In accordance with a speech (or sentence) production model in a discourse we split language modeling into two components; namely, dialog dependent concept modeling and syntactic modeling … Cited by 34 – Related articles – BL Direct – All 13 versions
Experiments in spoken document retrieval using phoneme n-grams [PDF] from mu.oz.auC Ng, R Wilkinson… – Speech Communication, 2000 – Elsevier In spoken document retrieval (SDR), speech recognition is applied to a collection to obtain either words or subword units, such as phonemes, that can be matched against queries. We have explored retrieval based on phoneme n-grams. The use of phonemes addresses the … Cited by 39 – Related articles – All 15 versions
Improving email speech acts analysis via n-gram selection [PDF] from upenn.eduVR Carvalho… – … Conversations in Text and Speech, 2006 – dl.acm.org Abstract In email conversational analysis, it is often useful to trace the the intents behind each message exchange. In this paper, we consider classification of email messages as to whether or not they contain certain intents or email-acts, such as” propose a meeting” or” … Cited by 32 – Related articles – All 22 versions
A discriminative HMM/N-gram-based retrieval approach for Mandarin spoken documents [PDF] from ntnu.edu.twB Chen, HM Wang… – ACM Transactions on Asian Language …, 2004 – dl.acm.org Abstract In recent years, statistical modeling approaches have steadily gained in popularity in the field of information retrieval. This article presents an HMM/N-gram-based retrieval approach for Mandarin spoken documents. The underlying characteristics and the various … Cited by 35 – Related articles – All 14 versions
Dealing with out-of-vocabulary words and speech disfluencies in an n-gram based speech understanding system [PDF] from nthu.edu.twA Kai, Y Hirose… – … Conference on Spoken Language …, 1998 – isca-speech.org In this study, we investigate the effectiveness of an unknown word processing (UWP) algorithm, which is incorporated into an N-gram language model based speech recognition system for dealing with filled pauses and out-of-vocabulary (OOV) words. We have … Cited by 27 – Related articles – All 3 versions
Backoff hierarchical class n-gram language models: effectiveness to model unseen events in speech recognition I Zitouni – Computer Speech & Language, 2007 – Elsevier In this paper, we introduce the backoff hierarchical class n-gram language models to better estimate the likelihood of unseen n-gram events. This multi-level class hierarchy language modeling approach generalizes the well-known backoff n-gram language modeling … Cited by 27 – Related articles – All 5 versions
Method and apparatus for a speech recognition system language model that integrates a finite state grammar probability and an N-gram probability JR Bellegarda – US Patent 6,154,722, 2000 – Google Patents A method and an apparatus for a speech recognition system that uses a language model based on an integrated finite state grammar probability and an n-gram probability are provided. According to one aspect of the invention, speech signals are received into a … Cited by 24 – Related articles – All 2 versions
Importance of high-order n-gram models in morph-based speech recognition [PDF] from tkk.fiT Hirsimaki, J Pylkkonen… – Audio, Speech, and …, 2009 – ieeexplore.ieee.org Abstract Speech recognition systems trained for morphologically rich languages face the problem of vocabulary growth caused by prefixes, suffixes, inflections, and compound words. Solutions proposed in the literature include increasing the size of the vocabulary … Cited by 22 – Related articles – All 6 versions
Multi-Class Composite N-gram language model for spoken language processing using multiple word clusters [PDF] from upenn.eduH Yamamoto, S Isogai… – … of the 39th Annual Meeting on …, 2001 – dl.acm.org Abstract In this paper, a new language model, the Multi-Class Composite N-gram, is proposed to avoid a data sparseness problem for spoken language in that it is difficult to collect training data. The Multi-Class Composite N-gram maintains an accurate word … Cited by 21 – Related articles – BL Direct – All 23 versions
An HMM/N-gram-based linguistic processing approach for Mandarin spoken document retrieval [PDF] from uni-hamburg.deB Chen, H Wang… – … European Conference on Speech …, 2001 – isca-speech.org In this paper an HMM/N-gram-based linguistic processing approach for Mandarin spoken document retrieval is presented. The underlying characteristics and different structures of this approach were extensively investigated. The retrieval capabilities were verified by … Cited by 15 – Related articles – All 18 versions
Hierarchical class n-gram language models: towards better estimation of unseen events in speech recognition [PDF] from cmu.eduI Zitouni, O Siohan… – … European Conference on Speech …, 2003 – isca-speech.org In this paper, we show how a multi-level class hierarchy can be used to better estimate the likelihood of an unseen event. In classical backoff n-gram models, the (n-1)-gram model is used to estimate the probability of an unseen n-gram. In the approach we propose, we use … Cited by 14 – Related articles – All 2 versions
Variable n-gram language modeling and extensions for conversational speech M Siu… – Fifth European Conference on Speech …, 1997 – isca-speech.org Recent progress in variable n-gram language modeling provides an efficient representation of n-gram models and makes training of higher order n-grams possible. In this paper, we apply the variable n-gram design algorithm to conversational speech, extending the … Cited by 11 – Related articles – All 3 versions
Error-responsive modifications to speech recognizers: negative n-grams L Chase, R Rosenfeld… – … Conference on Spoken …, 1994 – isca-speech.org We describe an error analysis technique that facilitates blame assignment among the various components of a speech recognizer and provides insight into their behavior. Tools are presented that help clarify how each of the component models and their interactions … Cited by 11 – Related articles
Language modeling by string pattern N-gram for Japanese speech recognition [PDF] from udel.eduA Ito… – Spoken Language, 1996. ICSLP 96. …, 1996 – ieeexplore.ieee.org Abstract This paper describes a new powerful statistical language model based on N-gram model for Japanese speech recognition. In English, a sentence is written word-by-word. On the other hand. A sentence in Japanese has no word boundary character. Therefore. A … Cited by 9 – Related articles – All 8 versions
A fast and memory-efficient N-gram language model lookup method for large vocabulary continuous speech recognition [PDF] from psu.eduX Li… – Computer Speech & Language, 2007 – Elsevier Recently, minimum perfect hashing (MPH)-based language model (LM) lookup methods have been proposed for fast access of N-gram LM scores in lexical-tree based LVCSR (large vocabulary continuous speech recognition) decoding. Methods of node-based LM … Cited by 11 – Related articles – All 7 versions
Dialog act classification using n-gram algorithms [PDF] from gsu.eduMM Louwerse… – … of the Florida Artificial Intelligence Research …, 2006 – aaai.org Abstract Speech act classification remains one of the challenges in natural language processing. This paper evaluates a classification system that assigns one of twelve dialog acts to an utterance from the Map Task Corpus. The dialog act classification system … Cited by 10 – Related articles – All 10 versions
[CITATION] A study of an N-Gram Language Model for Speech Recognition PL O’Boyle – 1993 – Queen’s University of Belfast Cited by 9 – Related articles – Library Search
Speech recognition using particle N-grams and content-word N-grams R Isotani… – Third European Conference on Speech …, 1993 – isca-speech.org This paper proposes a new stochastic language model for speech recognition based on particle N-grams and content-word N-grams. The conventional word N-gram model is considered as effective for speech recognition; however, it represents only local … Cited by 8 – Related articles – All 2 versions
Backoff hierarchical class n-gram language modelling for automatic speech recognition systems I Zitouni, O Siohan, HKJ Kuo… – … Conference on Spoken …, 2002 – isca-speech.org In this paper, we propose an extension of the backoff word n-gram language model that allows a better likelihood estimation of unseen events. Instead of using the (n-1)-gram to estimate the probability of an unseen n-gram, the proposed approach uses a class … Cited by 8 – Related articles – All 2 versions
N-gram language model adaptation using small corpus for spoken dialog recognition A Ito, H Saitoh, M Katoh… – … Conference on Speech …, 1997 – isca-speech.org This paper describes an N-gram language model adaptation technique. As an N-gram model requires a large size sample corpus for probability estimation, it is difficult to utilize N- gram model for a specific small task. In this paper, N-gram task adaptation is proposed … Cited by 7 – Related articles
[PDF] Combination of phone N-grams for a MPEG-7-based spoken document retrieval system [PDF] from psu.eduN Moreau, HG Kim… – in EUSIPCO, 2004 – Citeseer ABSTRACT In this paper, we present a phone-based approach of spoken document retrieval (SDR), developed in the framework of the emerging MPEG-7 standard. The audio part of MPEG-7 aims at standardizing the indexing of audio documents. It encloses a … Cited by 7 – Related articles – View as HTML – All 12 versions
Improving unsegmented dialogue turns annotation with n-gram transducers. CD Martínez-Hinarejos, V Tamarit… – 2009 – eprints.pascal-network.org Abstract The statistical models used for dialogue systems need annotated data (dialogues) to infer their statistical parameters. Dialogues are usually annotated in terms of Dialogue Acts (DA). The annotation problem can be attacked with statistical models, that avoid … Cited by 6 – Related articles – Cached – All 2 versions
Discriminative training of n-gram classifiers for speech and text routing [PDF] from microsoft.comC Chelba… – Eighth European Conference on Speech …, 2003 – isca-speech.org We present a method for conditional maximum likelihood estimation of N-gram models used for text or speech utterance classification. The method employs a well known technique relying on a generalization of the Baum-Eagon inequality from polynomials to rational … Cited by 6 – Related articles – All 4 versions
The use of word n-grams and parts of speech for hierarchical cluster language modeling [PDF] from sri.comW Wang… – … , Speech and Signal Processing, 2006. …, 2006 – ieeexplore.ieee.org Abstract We present extensions to the work of backoff hierarchical class n-gram language modeling of Zitouni et al.(2003) by studying the efficacy of exploring the use of parts of speech (POS) information in hierarchical word clustering. We propose two approaches. … Cited by 6 – Related articles – All 13 versions
[CITATION] TPOST: A Template-based, N-gram Part-of-Speech TAGGER for Tagalog V Rabo – MSCS Thesis, De la Salle University-Manila, …, 2004 Cited by 6 – Related articles
[PDF] Estimation method of user satisfaction using N-gram-based dialog history model for spoken dialog system [PDF] from lrec-conf.orgS Hara, N Kitaoka… – Proc. LREC, 2010 – lrec-conf.org Abstract In this paper, we propose an estimation method of user satisfaction for a spoken dialog system using an N-gram-based dialog history model. We have collected a large amount of spoken dialog data accompanied by usability evaluation scores by users in real … Cited by 5 – Related articles – View as HTML – All 2 versions
Rapid adaptation of n-gram language models using inter-word correlation for speech recognition K Sasaki, H Jiang… – … International Conference on Spoken …, 2000 – isca-speech.org In this paper, we study the fast adaptation problem of n-gram language model under the MAP estimation framework. We have proposed a heuristic method to explore inter-word correlation to accelerate MAP adaptation of n-gram model. According to their correlations, … Cited by 5 – Related articles – All 3 versions
A new word clustering method for building n-gram language models in continuous speech recognition systems [PDF] from uni-saarland.deM Bahrani, H Sameti, N Hafezi… – New Frontiers in Applied …, 2008 – Springer In this paper a new method for automatic word clustering is presented. We used this method for building n-gram language models for Persian continuous speech recognition (CSR) systems. In this method, each word is specified by a feature vector that represents the … Cited by 5 – Related articles – BL Direct – All 4 versions
Discriminative training of n-gram language models for speech recognition via linear programming [PDF] from yorku.caV Magdin… – … Speech Recognition & Understanding, …, 2009 – ieeexplore.ieee.org Abstract This paper presents a novel discriminative training algorithm for n-gram language models for use in large vocabulary continuous speech recognition. The algorithm uses Maximum Mutual Information Estimation (MMIE) to build an objective function that involves … Cited by 4 – Related articles – All 3 versions
Effectiveness of the backoff hierarchical class n-gram language models to model unseen events in speech recognition [PDF] from pitt.eduI Zitouni… – … Speech Recognition and Understanding, …, 2003 – ieeexplore.ieee.org Abstract Backoff hierarchical class n-gram language models use a class hierarchy to define an appropriate context. Each node in the hierarchy is a class containing all the words of the descendant nodes (classes). The closer a node is to the root, the more general the … Cited by 4 – Related articles – All 3 versions
Part of speech n-grams and Information Retrieval C Lioma, CJ van Rijsbergen – Revue française de linguistique …, 2008 – cairn.info Information Retrieval (IR) systems aim to locate and quantify information in data with respect to some user query. A common example of IR systems is World Wide Web (Web) search engines, in which a short keyword query is used to generate a ranked list from a pre- … Cited by 4 – Related articles – Library Search – All 3 versions
[CITATION] An HMM/N-gram-based Linguistic Approach for Mandarin Spoken Document Retrieval B Chen, H Wang… – … of the 7th European Conference on Speech …, 2001 Cited by 4 – Related articles
Integrating Layer Concept Inform ation into N-gram Modeling for Spoken Language Understanding [PDF] from nthu.edu.twNJC Wang, JL Shen… – … Conference on Spoken Language …, 2004 – isca-speech.org The paper presents a novel approach, integrating layer concept information into the trigram language model, to improve the understanding accuracy for spoken dialogue systems. With this approach, both the recognition accuracy and out-of-grammar problem can be largely … Cited by 3 – Related articles – All 7 versions
Parse tree n-grams for spoken language modelling A Wrigley – … , Applications and Alternatives, IEE Colloquium on, 1993 – ieeexplore.ieee.org Abstract A method is described for modelling natural language for speech recognition. Its aim is to incorporate the advantages of two previous types of approach; the statistical approach and the formal linguistic approach. The n-gram model (JK Baker, 1975), is a … Cited by 4 – Related articles – BL Direct
[PDF] Overcoming HMM time independence assumption using n-gram based modelling for continuous speech recognition [PDF] from eurasip.orgM Casar… – European Signal Processing Conference. …, 2008 – eurasip.org ABSTRACT The development of new acoustical models that overcome traditional HMM restrictions is an active field of research in automatic speech recognition. One possible approach to achieve this goal is to work with N-gram based augmented HMM. In this … Cited by 3 – Related articles – View as HTML – All 4 versions
[CITATION] Dialog-Context Dependent Language Modeling Using N-Grams and Stochastic Context-Free Grammars. 2001 K Hacioglu… – Proceedings of ICASSP Cited by 3 – Related articles
Spontaneous speech understanding in train timetable inquiry processing based on n-gram language models and finite state transducers L Jelinek… – ORLANDO, FL, USA, 2004 – kky.zcu.cz Abstract The presented paper concerns the spoken language understanding in an information retrieval dialogue system. There are described methods of use the finite state transducers for conceptual semantic parsing and meaning extraction from speaker’s … Cited by 2 – Related articles – Cached
[PDF] Reduction of the Temporal Complexity of N-gram Transducers for Dialogue Annotation [PDF] from psu.eduV Tamarit, CD Martinez-Hinarejos… – … of Spoken Dialog Systems …, 2009 – Citeseer Abstract. The annotation of dialogues with Dialogue Acts is important to develop dialogue systems. One way to do this annotation is using a method called N-gram Transducer (NGT), that has shown very good results in unsegmented turns compared with other models, but … Cited by 2 – Related articles – View as HTML – All 2 versions
[PDF] Integrating multiple layers of concept information into n-gram modeling for spoken language understanding [PDF] from korea.ac.krNJC Wang – Proc. ICASSP, 2005 – ispl.korea.ac.kr Abstract The paper presents a novel approach, integrating multi-layer concept information into the trigram language model, to improve the understanding accuracy for spoken dialogue systems. With this approach, both the recognition accuracy and out-of-grammar … Cited by 2 – Related articles – View as HTML – All 7 versions
Large margin estimation of n-gram language models for speech recognition via linear programming [PDF] from yorku.caV Magdin… – Acoustics Speech and Signal Processing ( …, 2010 – ieeexplore.ieee.org Abstract We present a novel discriminative training algorithm for n-gram language models for use in large vocabulary continuous speech recognition. The algorithm uses large margin estimation (LME) to build an objective function for maximizing the minimum margin … Cited by 4 – Related articles – All 6 versions
Jurilinguistic engineering in Cantonese Chinese: an N-gram-based speech to text transcription system [PDF] from uni-hamburg.deBK T’sou, KK Sin, SWK Chan, TBY Lai… – Proceedings of the 18th …, 2000 – dl.acm.org Abstract A Cantonese Chinese transcription system to automatically convert stenograph code to Chinese characters is reported. The major challenge in developing such a system is the critical homocode problem because of homonymy. The statistical N-gram model is … Cited by 2 – Related articles – All 24 versions
[PDF] Linearly interpolated hierarchical n-gram language models for speech recognition engines [PDF] from intechopen.comI Zitouni… – … Speech Recognition and Understanding, 2007 – intechopen.com Language modeling is a crucial component in natural language continuous speech recognition, due to the difficulty involved by continuous speech [1],[2]. Language modeling attempts to capture regularities in natural language for the purpose of improving the … Cited by 3 – Related articles – View as HTML – All 9 versions
[PDF] A n-gram approach to overcome time and parameter independence assumptions of HMM for speech recognition [PDF] from eurasip.orgM Casar… – Proceedings of the ISCA European Signal …, 2007 – eurasip.org ABSTRACT There is significant interest in developing new acoustic models for speech recognition that overcome traditional HMM restrictions. In this work, we propose to use a N- gram based augmented HMM. Two approaches are presented. The first one consists on … Cited by 2 – Related articles – View as HTML – All 6 versions
[CITATION] Combining N-grams and SCFGs in Speech Language Models A Stolcke – Proc IEEE Automatic Speech Recognition Workshop, …, 1995 Cited by 2 – Related articles
[CITATION] Characteristics of High Occurrence Frequency N-gram of Spoken Japanese Corpora N Ueda… – Proc. Spring Meet. Acoust. Soc. Jpn Cited by 2 – Related articles
[CITATION] Detection of errors n part-of-speech tagged corpora by bootstrapping generalized negative n-grams P Kveton… – Proceedings of 3rd International Conference on …, 2002 Cited by 2 – Related articles
Speech Recognition Using Function-Word< I> N</I>-Grams and Content-Word< I> N</I>-Grams R Isotani, S Matsunaga… – IEICE TRANSACTIONS on …, 1995 – search.ieice.org … Summary: This paper proposes a new stochastic language model for speech recognition based on function-word N-grams and content-word N-grams. The conventional word N-gram models are effective for speech recognition, but they represent only local constraints … Cited by 2 – Related articles – Cached – BL Direct – All 2 versions
[CITATION] Conversational speech recognition using sentence style related multi N-grams T Shimizu, S Kuroiwa… – Proc. ASRU, 1999 Cited by 1 – Related articles – All 3 versions
Using extra n-gram counts for statistical language model adaptation in speech-driven question answering T Akiba, K Itou, A Fujii… – IPSJ SIGNotes of Spoken …, 2002 – sciencelinks.jp Abstract; Aiming at speech-driven question answering, we propose two methods to produce statistical language models for recognizing spoken questions with a high accuracy. Both methods use a target collection (ie, a document set from which answers are derived) to … Cited by 1 – Related articles – Cached
Automatic detection of task-incompleted dialog for spoken dialog system based on dialog act N-gram [PDF] from nagoya-u.ac.jpS Hara, N Kitaoka… – … of the International Speech …, 2010 – isca-speech.org In this paper, we propose a method of detecting task-incompleted users for a spoken dialog system using an N-gram-based dialog history model. We collected a large amount of spoken dialog data accompanied by usability evaluation scores by users in real environments. … Cited by 1 – Related articles – All 3 versions
An Integrated Language Modeling with n-gram model and WA model for Speech Recognition [PDF] from mirlab.orgS Zhang… – Fifth European Conference on Speech …, 1997 – isca-speech.org As to traditional n-gram model, smaller n value is an inherent defect for estimating language probabilities in speech recognition, simply because that estimation could not be executed over farther word association but by means of short sequential word correlated information … Cited by 1 – Related articles – All 2 versions
Combination of a hidden tag model and a traditional n-gram model: A case study in Czech speech recognition [PDF] from cmu.eduP Krbec, P Podvesky… – … European Conference on Speech …, 2003 – isca-speech.org A speech recognition system targeting high inflective languages is described that combines the traditional trigram language model and an HMM tagger, obtaining results superior to the trigram language model itself. An experiment in speech recognition of Czech has been … Cited by 1 – Related articles – All 2 versions
Speech interface for name input based on combination of recognition methods using syllable-based N-gram and word dictionary [PDF] from nthu.edu.twH Oshikawa, N Kitaoka… – … Conference on Spoken …, 2004 – isca-speech.org We propose an interface for name input based on speech recognition using syllable-based N-gram and word dictionary. Name utterance is hard to recognize accurately because of the large vocabulary size, so the system uses continuous syllable recognition with syllable- … Cited by 1 – Related articles – All 11 versions
[PDF] Development of an N-Gram Based Language Model for Continuous Speech Recognition [PDF] from piconepress.comSP Given – … , The Language Modeling Group, Department of …, 1996 – isip.piconepress.com ABSTRACT An essential element of any speech recognition system is the language model. A language model attempts to identify and make use of the regularities in natural language to better define language syntax for easier recognition. One major obstacle in speech … Cited by 1 – Related articles – View as HTML – All 5 versions
[PDF] High speed spoken term detection by combination of n-gram array of a syllable lattice and LVCSR result for NTCIR-SpokenDoc [PDF] from nii.ac.jpK Iwami… – … of the Ninth NTCIR Workshop Meeting, 2011 – research.nii.ac.jp ABSTRACT For spoken document retrieval, it is very important to consider Out-of- Vocabulary (OOV) and mis-recognition of spoken words. Therefore, sub-word unit based recognition and retrieval methods have been proposed. This paper describes a Japanese … Cited by 1 – View as HTML
[PDF] NETWORK AND N-GRAM DECODING IN SPEECH RECOGNITION [PDF] from piconepress.comJ Zhao – 2000 – isip.piconepress.com I would like to thank Dr. Picone for his guidance through my work and study. I would also like to thank every one in ISIP (Institute for signal and information processing) for their help. A lot of work presented in this thesis is the result of team works. I am only one member of the … Cited by 1 – Related articles – View as HTML – All 4 versions
[CITATION] … Language Modeling Using N-Grams and Stochastic Context-Free Grammars”, to appear in IEEE International Conference on Acoustics, Speech, and … K Hacioglu… Cited by 1 – Related articles
[CITATION] Dialog-Context Dependent Modeling Combining N-Grams and Stochastic Context-Free Grammars K Hacioglu… – 26th International Conference on Acoustics, Speech … Cited by 1 – Related articles
Task extension with fusion of N-gram and grammar in public speech dialogue system T KITAMURA, T TODA, H KAWANAMI… – IEIC Technical Report …, 2005 – sciencelinks.jp Abstract; A suitable task domain for speech recognition and speech dialogue systems used in a real environment is important. It is indispensable to popularize such systems so that its technology can easily be applied or enhanced to a required task. When a new task is … Cached
Detection of task-incomplete dialogs based on utterance-and-behavior tag N-gram for spoken dialog systems K Takeda, N Kitaoka… – … Conference of the International Speech …, 2011 – ci.nii.ac.jp … Detection of task-incomplete dialogs based on utterance-and-behavior tag N-gram for spoken dialog systems. … ?????/??. ?????????(??). ?????. 12th Annual Conference of the International Speech Communication Association in Florence, Italy, on August 27 … Cached
Speech Interface for name input using syllable N-gram and word dictionary. H OSHIKAWA, N KITAOKA… – IEIC Technical Report ( …, 2003 – sciencelinks.jp Abstract; We propose an interface for a name input based on speech recognition using syllable-based N-gram and a word dictionary. User first utters a name and then chooses the correct word/syllables by pen touch from word/syllable candidates which were obtained … Cached
Speech Recognition Algorithm Strengthening N-gram Probability with Task Grammar. R TSURUMI, A RI, H SARUWATARI… – Joho Shori Gakkai …, 2003 – sciencelinks.jp Abstract; In speech dialogue systems, both word N-gram model and written network grammar are mainly used as language models. Since a word N-gram model is a statistical language model, it is practically feasible for various speech expressions and even for an … Cached
Detecting errors in Chinese spoken dialog system using ngram and dependency parsing W Zhou, B Yuan, Z Miao, W Zhu… – 2008 – link.aip.org In this paper, a hybrid method of detecting ASR error in spoken turns is developed. The erroneous text is locally analyzed first by neighbouring co-occurrence relations using ngram model. Then the text is globally analyzed by long distance dependency relations using a …
Spoken Language Processing. Dealing with Out-of-vocabulary Words and Filled Pauses in Word N-gram Based Speech Recognition System. A KAI, Y HIROSE… – Transactions, 1999 – sciencelinks.jp Abstract; For practical use of spoken dialog systems and dictation systems, it is important to cope with out-of-vocabulary words and filled pauses including the phenomena such as interjection, restart and hesitation. To address these problems, this study tries to use an … Cached
N-gram language models in JLASER neural network speech recognizer M Konopi´k, I Habernal… – Applied Electronics (AE), …, 2010 – ieeexplore.ieee.org Abstract In our recent research we have discovered that neural networks can be more efficient in speech recognition than the state of the art approach based on Gaussian mixtures. This statement is valid only for small corpora, however, many applications do not … Related articles
Method and apparatus for a speech recognition system language model that integrates a finite state grammar probability and an N-gram probability JR Bellegarda – The Journal of the Acoustical Society of America, 2001 – link.aip.org Method and apparatus for a speech recognition system language model that integrates a finite state grammar probability and an N-gram probability. [The Journal of the Acoustical Society of America 110, 24 (2001)]. Jerome R. Bellegarda. Keywords. Acoustics. … All 2 versions
[PDF] Evaluation of the incremental dialogue annotation using N-gram Transducers [PDF] from upm.esCD Martinez-Hinarejos, JMB Vicent Tamarit – lorien.die.upm.es Abstract The annotation of dialogues in terms of Dialogue Acts (DA) is an important task in the development of dialogue systems. Recently, the N-gram Transducers (NGT) technique showed a better performance than other techniques in the annotation of unsegmented … Related articles – View as HTML – All 2 versions
… on Importance SamplingA Measure of Dissimilarity for Evaluating Relational Structure of Learning TasksJapanese-English Bilingual Speech Recognition of Voice … K TAKAHASHI, M MIMURA, Y ISOBE… – search.ieice.org … pp.979-990 PAPER Japanese-English Bilingual Speech Recognition of Voice Command in Enroute Air Traffic Control Communication Atsunori … pp.1011-1018 PAPER Character String Recognition Based on Island-Driven Search Algorithm Using N-Gram Statistics Ryuji MINE … Cached – All 2 versions
Spoken Language Processing and Applications. Construction Method of Language Model Using Stochastic Switching n-gram. T NAGANO, M SUZUKI… – Transactions, 2002 – sciencelinks.jp Abstract; In traditional speech recognition systems, a single kind of n-gram is used for n- gram language model. If a task can divide into more small sub tasks, utilization of several kinds of n-gram gives better performance. In this paper, we propose a so-called SS ( … Cached – All 2 versions
[PDF] Louwerse, MM & Crossley, SA (2006). Dialog act classification using n-gram algorithms. In Proceedings of the 19th International Florida Artificial Intelligence … [PDF] from psu.eduM Louwerse – Citeseer Abstract Speech act classification remains one of the challenges in natural language processing. This paper evaluates a classification system that assigns one of twelve dialog acts to an utterance from the Map Task Corpus. The dialog act classification system … Related articles – View as HTML – All 3 versions
Spoken Language Understanding Using Layered N-gram Modeling [PDF] from pitt.eduNJC Wang – Ninth European Conference on Speech …, 2005 – isca-speech.org This paper presents an approach which integrates layer concept information into the trigram language model in order to improve the understanding accuracy for spoken dialogue systems and to improve the portability of the language modeling materials among different … Related articles – All 3 versions
On-line detection of task incompletion for spoken dialog systems using utterance and behavior tag N-gram vectors S Hara, N Kitaoka… – … and its Integration in Spoken Dialogue …, 2011 – Springer We propose a method of detecting the task incompletion in spoken dialog systems using N- gram-based dialog features. We used a database created during a field test in which inexperienced users used a client-server music retrieval system with a spoken dialog … Related articles – All 2 versions
[CITATION] A Method to Detect the Syllable Strings of Self-repair in Spontaneous Speech Using N-gram Model T Araki, S Ikehara… – MEMOIRS-FACULTY OF …, 1998 – FUKUI UNIVERSITY BL Direct
Generating compound words with high order n-gram information in large vocabulary speech recognition systems [PDF] from uiuc.eduJ Zhou, Q Shi… – … , Speech and Signal Processing ( …, 2011 – ieeexplore.ieee.org Abstract In this work we concentrate on generating compound words with high order n-gram information for speech recognition. In most existing compound words generation methods, only bi-gram information is considered. They are successful for improving the performance … Related articles – All 8 versions
Connectionist Models of Speech Segmentation and the Utterance Boundary Strategy: A Comparison of the SOM, SRN and N-GRAMS JA Hammerton – PROGRESS IN NEURAL PROCESSING, 2004 – books.google.com Some connectionist models of speech segmentation have exploited the utterance boundary strategy, where the fact that utterance endings are also word endings is used to infer where word boundaries are. In this paper, it is demonstrated that using a simple N-gram based … Related articles – BL Direct – All 2 versions
On the Use of N-Gram Transducers for Dialogue Annotation V Tamarit, CD Martínez-Hinarejos… – Spoken Dialogue Systems …, 2011 – Springer The implementation of dialogue systems is one of the most interesting applications of language technologies. Statistical models can be used in this implementation, allowing for a more flexible approach than when using rules defined by a human expert. However, … Related articles
Detection of task-incomplete dialogs based on utterance-and-behavior tag N-gram for spoken dialog systems [PDF] from nagoya-u.ac.jpS Hara, N Kitaoka… – … of the International Speech …, 2011 – ir.nul.nagoya-u.ac.jp We propose a method of detecting “task incomplete” dialogs in spoken dialog systems using N-gram-based dialog models. We used a database created during a field test in which inexperienced users used a client-server music retrieval system with a spoken dialog … Related articles – All 2 versions
Spoken Language Processing and Applications. A Metric Based on Likelihood Difference for n-gram Language Model Evaluation. A ITO… – Transactions, 2002 – sciencelinks.jp Abstract; Perplexity and cross entropy have been widely used as an evaluation metric of stochastic language model. Recently, several papers reported that correlation between these metrics and word error rate was poor when complicated language models were … Cached
An N-gram based Chinese syllable evaluation approach for speech recognition error detection X Wang… – … and Knowledge Engineering, 2009. NLP-KE …, 2009 – ieeexplore.ieee.org Abstract In order to find errors and correct words after Chinese speech recognition to improve its accuracy rate, an N-gram based phonetic syllable evaluation approach is proposed according to the conjunction rules in Chinese syllables. In this paper, Bigram … Related articles
[PDF] DEALING WITH OUT-OF-VOCABULARY WORDS AND SPEECH DISFLUENCIES IN AN N-GRAM BASED SPEECH UNDERSTANDING SYSTEM [PDF] from uiuc.eduYH AtsuhikoKAI… – mickey.ifp.uiuc.edu ABSTRACT In this study, we investigate the effectiveness of an unknown word processing (UWP) algorithm, which is incorporated into an N-gram language model based speech recognition system for dealing with filled pauses and outof-vocabulary (OOV) words. We … Related articles – All 4 versions
[PDF] TPOST: A Template-Based, n-gram Part-Of-Speech Tagger for Tagalog [PDF] from dlsu.edu.phCK Cheng… – dlsu.edu.ph We apologize for the inconvenience, but the page you are seeking for cannot be found in this location. It might have been removed, had its name changed, had been transferred to another directory, had it temporarily unavailable, or you might had it typed incorrectly. Related articles – View as HTML – All 2 versions
Part-of-Speech N-gram and Word N-gram Fused Language Model [PDF] from nthu.edu.twH Yamamoto… – … European Conference on Speech …, 1999 – isca-speech.org In th is paper, an accurate and com pact language model is proposed to cope robustly with data sparseness and task dependencies. This language model adopts new categories which are generated by continuously interpolating POS word-class categories and word … Related articles – All 3 versions
[CITATION] SPEECH-P6. 14: USE OF STATISTICAL N-GRAM MODELS IN NATURAL LANGUAGE GENERATION FOR MACHINE TRANSLATION FH Liu, L Gu, Y Gao… – … ON ACOUSTICS SPEECH AND …, 2003 – IEEE; 1999 BL Direct
[PDF] Modeling out-of-vocabulary words using multi class-based n-gram language model for automatic speech recognition [PDF] from tut.ac.jpW Naptali, M Tsuchiya… – slp.ics.tut.ac.jp Out-of-vocabulary (OOV) words cause a serious problem for automatic speech recognition (ASR) system. Not only it will be miss-recognized as an invocabulary word with similar phonetics, but the error will also affect nearby words to make errors. Language models ( … Related articles – View as HTML – All 2 versions
[CITATION] Part VI-Cross-Language Spoken Document Retrieval-N-grams for Translation and Retrieval in CL-SDR P McNamee, J Mayfield – Lecture Notes in …, 2004 – Berlin: Springer-Verlag, 1973-
[CITATION] SPEECH-P1. 10: SEMANTIC N-GRAM LANGUAGE MODELING WITH THE LATENT MAXIMUM ENTROPY PRINCIPLE S Wang, D Schuurmans, F Peng… – … ON ACOUSTICS SPEECH …, 2003 – IEEE; 1999 BL Direct
Part-of-Speech Class N-gram and Word N-gram Fused Language Model. H YAMAMOTO… – Reports of the Meeting. the …, 1999 – sciencelinks.jp TOP > J-EAST > List of Journal Titles (R) > Reports of the Meeting. the Acoustical Society of Japan(1999) > Part-of-Speech Class N-gram and Word N-gram Fused Language Model. … Part-of-Speech Class N-gram and Word N-gram Fused Language Model. Cached
Packet loss concealment based on statistical n-gram predictive models for use in voice-over-IP speech transmission M Lee, Q Zhou… – US Patent 7,701,886, 2010 – Google Patents A method for performing packet loss concealment of lost packets in Voice over IP (Internet Protocol) speech transmission. Statistical n-gram models are initially created with use of a training speech corpus, and then, packets lost during transmission are advantageously … Related articles – All 3 versions
… on Saliency and Gradient IntensityDetection of Task Incomplete Dialogs Based on Utterance Sequences N-gram for Spoken Dialog SystemDetecting Side Oncoming … K TANABE – search.ieice.org … Summary | Full Text(in Japanese):PDF (1.3MB). pp.497-500 LETTER Detection of Task Incomplete Dialogs Based on Utterance Sequences N-gram for Spoken Dialog System Sunao HARA Norihide KITAOKA Kazuya TAKEDA. Summary | Full Text(in Japanese):PDF (495KB). … Cached – All 2 versions
N-gram language models for Polish language. Basic concepts and applications in automatic speech recognition systems [PDF] from imcsit.orgB Rapp – … Science and Information Technology, 2008. IMCSIT …, 2008 – ieeexplore.ieee.org Abstract Usage of language models in automatic speech recognition systems usually give significant quality and certainty improvement of recognition outcomes. On the other hand, wrongly chosen or trained language models can result in serious degradation not only … Related articles – All 2 versions
… of a Phonetically Rich Speech CorpusBi-Spectral Acoustic Features for Robust Speech RecognitionLocal Peak Enhancement for In-Car Speech Recognition in Noisy … C LANG, D XU… – search.ieice.org … Model Compositions Using Multi-Pass Search with Multi-Label N-gram Models Takatoshi … PAPER-Acoustic Modeling Development of a Mandarin-English Bilingual Speech Recognition System … ASR System Architecture Selection of Optimum Vocabulary and Dialog Strategy for … Cached – All 2 versions
Packet loss concealment based on statistical n-gram predictive models for use in voice-over-IP speech transmission M Lee, Q Zhou… – US Patent App. 20,050/276,235, 2004 – freepatentsonline.com 1. A method for performing packet loss concealment in a packet-based speech communication system, the method comprising the steps of: receiving a sequence of one or more speech packets each comprising a set of speech parameters representative of … Cached
[CITATION] A comparison of N-grams in the spoken component of the” British National Corpus” and the” Longman Spoken American Corpus” C Ruoss – 2001 – Ref.: Gunnel Tottie. Library Search
A New Language Model by Using (n. GEQ. 4)-gram for Broadcast News Speech Transcription. N KATO, N URATANI, T EHARA… – IEICE Transactions on …, 2002 – sciencelinks.jp … 2- or 3-gram under carrying out adoptation to the native RM by modeling such character with n-gram (n is larger than or equal to 4). As when all of n-gram are simply … As a result of an evaluation test based on perplexity and speech recognition, a good result could be obtained. … Cached – All 2 versions
… Topic Transition ModelCooperative Spoken Dialogue Model Using Bayesian Network and Event HierarchyError Analysis of Field Trial Results of a Spoken Dialogue … Y WANG, JL XU, K INOUE… – search.ieice.org … pp.636-641 PAPER Error Analysis of Field Trial Results of a Spoken Dialogue System for Telecommunications Applications Shingo KUROIWA Kazuya … pp.692-697 PAPER Speech Recognition Using Function-Word N-Grams and Content-Word N-Grams Ryosuke ISOTANI … Cached – All 2 versions