Filters








20,901 Hits in 7.7 sec

N-Gram Based Test Sequence Generation from Finite State Models [chapter]

Paolo Tonella, Roberto Tiella, Cu D. Nguyen
2014 Lecture Notes in Computer Science  
In this paper, we propose a novel test case derivation strategy, based on the computation of the N -gram statistics.  ...  Models can be inferred from observations of real executions and test cases can be derived from models, according to various strategies (e.g., graph or random visits).  ...  Acknowledgments This work has been funded by the European Union FP7 project FITTEST (grant agreement n. 257574).  ... 
doi:10.1007/978-3-319-14121-3_4 fatcat:s2ktfojfwveh3h26bva7o4jnoi

N-Gram Based Test Sequence Generation from Finite State Models [chapter]

Paolo Tonella, Roberto Tiella, Cu D. Nguyen
2014 Lecture Notes in Computer Science  
In this paper, we propose a novel test case derivation strategy, based on the computation of the N -gram statistics.  ...  Models can be inferred from observations of real executions and test cases can be derived from models, according to various strategies (e.g., graph or random visits).  ...  Acknowledgments This work has been funded by the European Union FP7 project FITTEST (grant agreement n. 257574).  ... 
doi:10.1007/978-3-319-07785-7_4 fatcat:dmfar5lvm5fzxlcqwzocyxx6ze

Language model combination and adaptation usingweighted finite state transducers

X. Liu, M.J.F. Gales, J.L. Hieronymus, P.C. Woodland
2010 2010 IEEE International Conference on Acoustics, Speech and Signal Processing  
In this paper an alternative and more general approach based on weighted finite state transducers (WFSTs) is investigated for LM combination and adaptation.  ...  In speech recognition systems language model (LMs) are often constructed by training and combining multiple n-gram models.  ...  Many types of modelling information used in speech recognition systems, such as HMM topology, lexicon and n-gram LMs, involve a stochastic finite-state mappings between symbol sequences.  ... 
doi:10.1109/icassp.2010.5494941 dblp:conf/icassp/LiuGHW10 fatcat:ahnbnyfyara3vbtttv5ax6zgl4

Statistical modeling for unit selection in speech synthesis

Cyril Allauzen, Mehryar Mohri, Michael Riley
2004 Proceedings of the 42nd Annual Meeting on Association for Computational Linguistics - ACL '04  
We show that the concatenation cost can be accurately estimated from this corpus using a statistical n-gram language model over units.  ...  We present a new unit selection system based on statistical modeling.  ...  Mark also generated the training corpora and set up the listening test used in our experiments.  ... 
doi:10.3115/1218955.1218963 dblp:conf/acl/MohriAR04 fatcat:xvgyh6okyregfcmwoip67q4vrm

Evaluation of Technical Measures for Workflow Similarity Based on a Pilot Study [chapter]

Andreas Wombacher
2006 Lecture Notes in Computer Science  
Service discovery of state dependent services has to take workflow aspects into account.  ...  In this paper different similarity measures are presented and evaluated based on a pilot of an empirical study. In particular the different measures are compared with the study results.  ...  To represent the potentially infinite sequence by a finite n-gram list repeated occurrences of n-grams (states) are removed.  ... 
doi:10.1007/11914853_16 fatcat:fxvnzianxfdapmk77aizrtsu5u

Spontaneous handwriting recognition and classification

A.H. Toselli, A. Juan, E. Vidal
2004 Proceedings of the 17th International Conference on Pattern Recognition, 2004. ICPR 2004.  
Finite-state models are used to implement a handwritten text recognition and classification system for a real application entailing casual, spontaneous writing with large vocabulary.  ...  HMMs and n-grams are used for text recognition and n-grams are also used for text classification. Experimental results are reported which, given the extreme difficulty of the task, are encouraging. *  ...  Although the recognition n-gram models were trained using TS-1, the test set recognition WER was determined by filtering the punctuation signs from all recognized word sequences and their respective test  ... 
doi:10.1109/icpr.2004.1334151 dblp:conf/icpr/ToselliJV04 fatcat:oowqg62bvvcarajlzjov5jhxfe

Machine Translation with Inferred Stochastic Finite-State Transducers

Francisco Casacuberta, Enrique Vidal
2004 Computational Linguistics  
e.g., an n-gram) is inferred.  ...  One of these areas is machine translation, in which the approaches that are based on building models automatically from training examples are becoming more and more attractive.  ...  Figure 6 4 64 A finite-state transducer built from the n-gram ofFigure Figure 7 7 A finite-state transducer built from the n-gram of Figure 5. the transitions and the final states of the finite-state  ... 
doi:10.1162/089120104323093294 fatcat:iqli5nbt4zfzpak5qoltb7ucb4

Discriminatively estimated joint acoustic, duration, and language model for speech recognition

Maider Lehr, Izhak Shafran
2010 2010 IEEE International Conference on Acoustics, Speech and Signal Processing  
In the framework of finite state machines, a general model for speech recognition G is a finite state transduction from acoustic state sequences to word sequences (e.g., search graph in many speech recognizers  ...  The resulting model can be factored as corrections for the input and the output sides of the general model G. This formulation allows us to incorporate duration cues seamlessly.  ...  Acknowledgments We thank the Advanced Large Vocabulary Speech Recognition Group at IBM for making their tools and models available for this work and specifically Brian Kingsbury for facilitating the collaboration  ... 
doi:10.1109/icassp.2010.5495227 dblp:conf/icassp/LehrS10 fatcat:ybpu76pbcbff5oftjslkwhjmaa

Simple Variable Length N-grams for Probabilistic Automata Learning

Fábio Natanael Kepler, Sérgio L. S. Mergen, Cléo Zanella Billa
2012 Journal of machine learning research  
This paper proposes the usage of n-gram models with variable length.  ...  The main goal of the competition was to obtain insights about which techniques and approaches work best for sequence learning based on different kinds of automata generating machines.  ...  Our approach to solve these problems is based on an n-gram model with variable length.  ... 
dblp:journals/jmlr/KeplerMB12 fatcat:lozj5avvvvgvtokrauugorgauy

A Hybrid Seq-2-Seq ASR Design for On-Device and Server Applications

Cyril Allauzen, Ehsan Variani, Michael Riley, David Rybach, Hao Zhang
2021 Conference of the International Speech Communication Association  
Our experiments show decent improvements in WER over common speech phrases and significant gains on uncommon ones compared to the state-of-the-art approaches.  ...  The different strategies are designed with special attention to the choice of modeling units and to the integration of different types of external language models during first-pass beam-search or second-pass  ...  22M n-gram 7.2 6.3 6.5 6.1 6.2 5.8 16M n-gram 7.3 6.3 6.6 6.1 6.3 5.8 11M n-gram 7.5 6.4 6.7 6.1 6.4 5.8 8M n-gram 7.6 6.4 6.8 6.1 6.5 5.8 5.6M n-gram 8.1 6.6 7.0 6.3 6.7 5.8  ... 
doi:10.21437/interspeech.2021-658 dblp:conf/interspeech/AllauzenV0RZ21 fatcat:tmpqhp7jdbf77hkhzde4hro5dq

Diagnosis of Stochastic Discrete Event Systems Based on N-Gram Models with Wildcard Characters

Kunihiko HIRAISHI, Koichi KOBAYASHI
2016 IEICE Transactions on Fundamentals of Electronics Communications and Computer Sciences  
We are developing a method, called sequence profiling, based on N -gram models. The information necessary for sequence profiling is only event logs from the target system.  ...  From event logs in the normal situation, N-gram models are constructed through a simple statistical analysis.  ...  a system model from G i , i = 0, 1, · · · , m that most likely generates w test .  ... 
doi:10.1587/transfun.e99.a.462 fatcat:f2p46tq45vfhbois6hk5cz3y3m

Interpolated n-grams for model based testing

Paolo Tonella, Roberto Tiella, Cu Duy Nguyen
2014 Proceedings of the 36th International Conference on Software Engineering - ICSE 2014  
In this paper, we propose a method, based on the computation of the N-gram statistics, to increase the likelihood of deriving feasible test cases from a model.  ...  Models -in particular finite state machine models -provide an invaluable source of information for the derivation of effective test cases.  ...  Acknowledgements This work has been funded by the European Union FP7 project FITTEST (grant agreement n. 257574).  ... 
doi:10.1145/2568225.2568242 dblp:conf/icse/TonellaTN14 fatcat:64neev33b5bgndng5pz4la3n6i

Probabilistic Deterministic Infinite Automata

David Pfau, Nicholas Bartlett, Frank D. Wood
2010 Neural Information Processing Systems  
Posterior predictive inference in this model, given a finite training sequence, can be interpreted as averaging over multiple PDFAs of varying structure, where each PDFA is biased towards having few states  ...  We test PDIA inference both on PDFA structure learning and on both natural language and DNA data prediction tasks.  ...  n-gram models.  ... 
dblp:conf/nips/PfauBW10 fatcat:mouocuevmbgrtd3wvzhfozhtja

Syllable language models for Mandarin speech recognition: Exploiting character language models

Xunying Liu, James L. Hieronymus, Mark J. F. Gales, Philip C. Woodland
2013 Journal of the Acoustical Society of America  
To test this idea, word and character level n-gram LMs were trained on 2.8 billion words (equivalent to 4.3 billion characters) of texts from a wide collection of text sources.  ...  Both hypothesis and model based combination techniques were investigated to combine word and character level LMs.  ...  Many types of modeling information used in speech recognition systems, such as HMM topology, lexicon, and n-gram LMs, involve a stochastic finite-state mapping between symbol sequences.  ... 
doi:10.1121/1.4768800 pmid:23297923 fatcat:eist5x5hhrcntn53pvp62awhw4

Statistical dialog management applied to WFST-based dialog systems

Chiori Hori, Kiyonori Ohtake, Teruhisa Misu, Hideki Kashioka, Satoshi Nakamura
2009 2009 IEEE International Conference on Acoustics, Speech and Signal Processing  
A scenario WFST for dialog management was automatically created from an N-gram model of a tag sequence that was annotated in the corpus with Interchange Format (IF).  ...  We have proposed an expandable dialog scenario description and platform to manage dialog systems using a weighted finite-state transducer (WFST) in which user concept and system action tags are input and  ...  set (3 dialgos) #tag/turn 1.84 (46/25) 2.50 (70/28) Table 5 : 5 Test-set Perplexity of N-gram models Bigram Trigram 4-gram J-J 28.7 23.6 23.7 E-J 43.1 39.0 37.3 Table 6 : 6 Size  ... 
doi:10.1109/icassp.2009.4960703 dblp:conf/icassp/HoriOMKN09 fatcat:7rcayjn6ujakzkw6j6teiqr3ru
« Previous Showing results 1 — 15 out of 20,901 results