Japanese and Korean voice search
The DOI matches its Crossref record; 7 of the 8 references it cites were found in a registry or an index.
Abstract
This paper describes challenges and solutions for building a successful voice search system as applied to Japanese and Korean at Google. We describe the techniques used to deal with an infinite vocabulary, how modeling completely in the written domain for language model and dictionary can avoid some system complexity, and how we built dictionaries, language and acoustic models in this framework. We show how to deal with the difficulty of scoring results for multiple script languages because of ambiguities. The development of voice search for these languages led to a significant simplification of the original process to build a system for any new language which in in parts became our default process for internationalization of voice search.
Overview
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
In the citing paper
“LogRobust [8], which encodes log templates into semantic vectors using the FastText pre-trained model [31], cannot work well on 2 out of 4 datasets.”
From Log-based Anomaly Detection Without Log Parsing (2021) · cites this paper as [8]
In the citing paper
“Their empirical study on a Microsoft online service system shows that up to 30.3% logs are changed in the latest version.”
From Log-based Anomaly Detection Without Log Parsing (2021) · cites this paper as [8]
In the citing paper
“Zhang et al. [8] indicated that log data is unstable, meaning that new log events often appear due to software evolution at its lifetime.”
From Log-based Anomaly Detection Without Log Parsing (2021) · cites this paper as [8]
In the citing paper
“Another option is to use domain knowledge to reduce the input space of a model — such strategies include patch embeddings [Dosovitskiy et al. 2020] as well as fixed or learned tokenization [Schuster and Nakajima 2012; Sennrich et al. 2016; Kudo and Richardson 2018; van den Oord et al. 2017; Ramesh et al. 2021]. These methods may alleviate some challenges facing generative approaches, but have the”
From On the Opportunities and Risks of Foundation Models (2021) · cites this paper as Schuster and Nakajima 2012
Citations
Cites
Semi-tied covariance matrices for hidden Markov models
Gales, M. J. F. (1999) · IEEE Transactions on Speech and Audio Processing · doi:10.1109/89.759034 · Cited by 537
Boosted MMI for model and feature-space discriminative training
Povey, D., Kanevsky, D., Kingsbury, B., Ramabhadran, B., Saon, G., & Visweswariah, K. (2008) · 2008 IEEE International Conference on Acoustics, Speech and Signal Processing · doi:10.1109/icassp.2008.4518545 · Cited by 341
“Your Word is my Command”: Google Search by Voice: A Case Study
Schalkwyk, J., Beeferman, D., Beaufays, F., Byrne, B., Chelba, C., Cohen, M., Kamvar, M., & Strope, B. (2010) · Advances in Speech Recognition · doi:10.1007/978-1-4419-5951-5_4 · Cited by 232
Building transcribed speech corpora quickly and cheaply for many languages
Hughes, T., Nakajima, K., Ha, L., Vasu, A., Moreno, P. J., & LeBeau, M. (2010) · Interspeech (USB) · doi:10.21437/interspeech.2010-551 · Cited by 75
Self-citationGoogle Search by Voice: A Case Study
Schalkwyk, J., Beeferman, D., Beaufays, F., Byrne, B., Chelba, C. I., Cohen, M., Garrett, M., & Strope, B. P. (2010) · Cited by 63
Query language modeling for voice search
Chelba, C. I., Schalkwyk, J., Brants, T., Ha, V., Harb, B., Neveitt, W., Parada, C., & Xu, P. (2010) · IEEE Spoken Language Technology Workshop (SLT) · doi:10.1109/slt.2010.5700834 · Cited by 28
Search by voice in Mandarin Chinese
Shan, J., Wu, G., Hu, Z., Tang, X., Jansche, M., & Moreno, P. J. (2010) · Interspeech (USB) · doi:10.21437/interspeech.2010-129 · Cited by 18
Research works printed but not linked: 1
- sung, 0, Deploying Search by Voice in Cantonese, Proc Interspeech 2011
Cited by
Items 901–933 of 933
Predicting the Next Words From Shannon's Entropy to Large Language ModelsPreprint
Osterrieder, J. R. (2026) · SSRN Electronic Journal · doi:10.2139/ssrn.6616078
Surface-Form Neural Sparse Retrieval: Robust Fuzzy Matching for Industrial Music Search
Greyson, P., Geng, Z., Zhang, W., & Yang, Y. (2026) · International ACM SIGIR Conference on Research and Development in Information Retrieval (SIGIR) · doi:10.1145/3805712.3808414
“Standard PLMs typically rely on fixed subword vocabularies (e.g., WordPiece), where minor perturbations can induce sharply different tokenizations and degrade robustness to severe query misspecifica-tions in search [7, 24].”
A State Space Model Approach to Filipino Sentiment Analysis
Villa, M. G., Fajardo, A. C., & Calanda, F. B. (2026) · IEEE International Colloquium on Signal Processing & Its Applications (CSPA) · doi:10.1109/cspa68262.2026.11517646
“…the original BERT architecture to support 104 languages through training on multilingual Wikipedia corpora and the use of a shared vocabulary based on WordPiece [4][5][6] , which is a subword tokenization algorithm that enables words to be represented as combinations of subword tokens [4][7] .”
Kashiwada, Y., Sakurai, R., Yokokawa, Y., Ando, K., Gocho, C., Murata, C., Tohda, K., Tanioka, M., Sonobe, S., Harada, T., et al. (2026) · Research Square · doi:10.21203/rs.3.rs-10026525/v1
Wangilisasi, E., Leo, J., & Sam, A. E. (2026) · Journal of Intelligent Learning Systems and Applications · doi:10.4236/jilsa.2026.183012
Tokenizations for Austronesian Language Models: Study on Languages in Indonesia Archipelago
Lumbantobing, A. B., Situngkir, H., & Ringo, K. S. (2026) · Lecture notes in computer science · doi:10.1007/978-3-032-29532-3_13
A Morphologically Labeled Dataset for Yakut: Design, Compilation, and Applications
Vissio, N. C. (2026) · Ural Symposium on Biomedical Engineering, Radioelectronics and Information Technology (USBEREIT) · doi:10.1109/usbereit70063.2026.11581110
Exploring speech technologies for the Kui language in a low-resource scenario
Pattanayak, S. S., Nayak, A. K., Mishra, S., Mohanty, P., Das, S. K., & Meher, A. A. (2026) · International Journal of Speech Technology · doi:10.1007/s10772-026-10278-5
“Graphemes, Phonemes, or byte-pair encoding can be used to handle out-of-vocabulary words in ASR (Schuster & Nakajima, 2012 ).”
Massively Multilingual Vision-Language Understanding
Geigle, G. (2026) · doi:10.25972/opus-44667
Towards robust and safe multilingual knowledge learning in Large Language Models
Wang, M. (2026) · Electronic Theses of LMU Munich (Ludwig-Maximilians-Universität München) · doi:10.5282/edoc.37293
The Words of Proteins: Motif-Level Language Modeling for Interpretable Protein Generation
Low, B. H., Goh, W. Y., Li, B., & He, D. (2026) · ACM SIGKDD Conference on Knowledge Discovery and Data Mining (KDD) · doi:10.1145/3770855.3819054
“WordPiece uses a similar approach optimized for likelihood [23].”
Liao, H., Kong, X., Mu, Y., Li, J., Junfeng, H., Hu, G., & Zhang, T. (2026) · Mathematics · doi:10.3390/math14162947
Building a Large-Scale Cross-Script Kazakh Parallel Corpus for Low-Resource Language Data Science
Gao, J., & Zhang, M. (2026) · Data Science & Big Data Technology · doi:10.63646/gtcu8198
WIP: LLM Odyssey: A Game-Based Platform for Teaching LLM Engineering ConceptsPreprint
Tripathi, P. (2026) · arXiv (Cornell University) · doi:10.48550/arxiv.2608.16924
“Students receive text samples spanning multilingual content, Python code, and legal documents, and select among four tokenizers: BPE [16], WordPiece [17], SentencePiece [18], and Unigram [19].”
Turkish Subword Segmentation: A Unified Intrinsic–Extrinsic Analysis
Gunes, U. K., Yıldız, O. T., & Karadenız, İ. (2026) · IEEE Access · doi:10.1109/access.2026.3725837
“WordPiece follows a similar frequency-driven philosophy, but is typically deployed with a fixed vocabulary and model-specific constraints [9].”
Sindane, T., Marivate, V. N., Modupe, A., & Combrink, H. (2026) · Intelligent Systems with Applications · doi:10.1016/j.iswa.2026.200723
Adkins, J., Walsh, A., Davis, B., & Dhonnchadha, E. U. (2026) · arXiv (Cornell University) · doi:10.48550/arxiv.2609.05022
“WordPiece (Schuster & Nakajima, 2012) is also a merge-based algorithm that selects the co-occuring characters that maximise the likelihood of the training data, i.e. it merges characters with maximum mutual information value, using an n-gram language model to score the likelihood of possible pairs.”
Computational approaches to semantic change: the case of Christian Latin
Lunardi, V. (2026) · eScholarship (California Digital Library)
Beyond Atomic Tokens: Factorizing Syllables for Language Model PretrainingPreprint
Nguyen, N. H., Huynh, T. B., Dinh-Le, B.-A., Hoang, P. G., Nguyen, D. T., Nguyen, K. V., & Nguyen, N. L.-T. (2026) · arXiv (Cornell University) · doi:10.48550/arxiv.2609.21362
“Most PLMs rely on corpus-induced algorithms such as Byte Pair Encoding (BPE) [1], WordPiece [2], or the Un-igram Language Model [3].”
Bayram, M. A. (2026) · Yildiz Technical University - AVESIS
Multi-Signal Learning Framework for Robust Detection of Visually Deceptive Text
Jovanovic-Hacon, A., Luo, S.-W., Cao, C., Wanderley, M., Boriel, H., Tenison, I., Kim, H., Beserra, F. S., Ferreira, M. A., Palacios, R., et al. (2026) · Electronics · doi:10.3390/electronics15194531
More Than Words: Compositional Tokenization for Efficient Language ModelsPreprint
Reif, Y., Kaplan, G., & Schwartz, R. (2026) · arXiv (Cornell University) · doi:10.48550/arxiv.2610.05597
“Standard BPE remains the dominant tokenizer in LMs (Sennrich et al., 2016), with WordPiece (Schuster & Nakajima, 2012; Devlin et al., 2019) and Unigram LM (Kudo, 2018) serving as established alternatives.”
Transformer-Based Large Language Models in Architectural Advances of Medical Image Analysis
Gutam, B. G., & Kumar, M. S. (2026) · International Conference on Natural Language Processing and Computer Vision (ICNPCV) · doi:10.1109/icnpcv69434.2026.11708856
“The token generation commonly used in LLMs includes WordPiece [12], Byte-Pair Encoding (BPE) [13], and Unigram LM tokenization algorithms.”
Evaluating Portuguese Tokenizers as Morpheme Sequence in Relation to LLM Downstream Performance
de Mello, G. L., & Finger, M. (2026) · Simpósio Brasileiro de Tecnologia da Informação e da Linguagem Humana (STIL) · doi:10.5753/stil.2026.26562
“WordPiece [Schuster and Nakajima 2012] is a well known tokenization algorithm, but was never openly published by Google.”
Improving Low-Resource Morpheme Segmentation via Cross-Lingual Sequential Fine-Tuning
Morozov, D. A., & Shaimukhamedov, A. (2026) · Lecture notes in computer science · doi:10.1007/978-3-032-41546-2_11
Bach, L. R., Gebeyehu, T. A., & Tashu, T. M. (2026) · Lecture notes in computer science · doi:10.1007/978-981-92-6910-5_24
Evaluating LLM Tokenization Efficiency on Kuwaiti Dialect, MSA and Other Arabic Dialects
Alostad, H., & Almutairi, Y. (2026) · Big Data and Cognitive Computing · doi:10.3390/bdcc10100356
Finitely Verifiable Segmentation: Decidability for Every Tokenizer with a Finite-State VerifierPreprint
Nidhögg, N. (2026) · arXiv (Cornell University) · doi:10.48550/arxiv.2610.30168 · Cited by 1
“WordPiece, the subword tokenization method of Schuster and Nakajima [65], has a real-vocabulary instance too, at two depths.”
Qiu, E., Chen, T., Xu, C., Tu, Y., & Gao, J. (2026) · ACM International Conference on Information and Knowledge Management (CIKM) · doi:10.1145/3799682.3840958
The New Perspectiva Artificialis: From the Mechanical Eye to the Algorithmic Symbolic Form
Hutson, J., & Plate, D. (2026) · Routledge eBooks · doi:10.4324/9781042013173
Tokenizer Myths and Where To Find ThemPreprint
Schaeffer, R., & Koyejo, S. (2026) · arXiv (Cornell University) · doi:10.48550/arxiv.2611.04215
Li, Y., Li, Z., Xiao, H., & Li, X. (2026) · Research Square · doi:10.21203/rs.3.rs-10260345/v1
Gmail Smart Compose: Real-Time Assisted Writing
Chen, M. X., Lee, B. N., Bansal, G., Cao, Y., Zhang, S., Lu, J., Tsay, J., Wang, Y., Dai, A. M., Chen, Z., et al. (2019) · arXiv · doi:10.48550/arxiv.1906.00080
“We use WordPiece [43] models and split the tokens into sub-word units (wordpieces).”
Method
Metadata
- Authors
- Schuster, M., & Nakajima, K.
- Affiliations
- Google (United States) (US)
- Venue
- 2012 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)
- Publisher
- IEEE
- Published
- 2012-03
- Pages
- 5149-5152
- ISSN
- 1520-6149 (ISSN-L)
- OpenAlex
- W2121879602
- Cited by
- 862
- References
- 8
- Citations by year
- 2026: 36 · 2025: 66 · 2024: 90 · 2023: 124 · 2022: 137 · 2021: 155 · 2020: 144 · 2019: 59 · 2018: 36 · 2017: 10 · 2016: 3 · 2013: 2
- FWCI
- 0.49
- Type
- conference paper
- Language
- en
- Open access
- closed
- Locations
- IEEE International Conference on Acoustics Speech and Signal Processing (publishedVersion)
- Topics
- Physical Sciences › Computer Science › Artificial Intelligence › Speech Recognition and Synthesis
- Physical Sciences › Computer Science › Artificial Intelligence › Speech and dialogue systems
- Physical Sciences › Computer Science › Artificial Intelligence › Natural Language Processing Techniques
- Keywords
- voice search, Japanese language, Korean language, acoustic modeling
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.