{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,7,7]],"date-time":"2026-07-07T15:52:58Z","timestamp":1783439578449,"version":"3.54.6"},"reference-count":76,"publisher":"Association for Computing Machinery (ACM)","issue":"4","license":[{"start":{"date-parts":[[2018,12,27]],"date-time":"2018-12-27T00:00:00Z","timestamp":1545868800000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/www.acm.org\/publications\/policies\/copyright_policy#Background"}],"content-domain":{"domain":["dl.acm.org"],"crossmark-restriction":true},"short-container-title":["Proc. ACM Interact. Mob. Wearable Ubiquitous Technol."],"published-print":{"date-parts":[[2018,12,27]]},"abstract":"<jats:p>Mobile users spend an extensive amount of time on typing. A more efficient text input instrument brings a significant enhancement of user experience. Deep learning techniques have been recently applied to suggesting the next words of input, but to achieve more accurate predictions, these models should be customized for individual users. Personalization is often at the expense of privacy concerns. Existing solutions require users to upload the historical logs of their input text to the cloud so that a deep learning predictor can be trained. In this work, we propose a novel approach, called DeepType, to personalize text input with better privacy. The basic idea is intuitive: training deep learning predictors on the device instead of on the cloud, so that the model makes personalized and private data never leaves the device to externals. With DeepType, a global model is first trained on the cloud using massive public corpora, and our personalization is done by incrementally customizing the global model with data on individual devices. We further propose a set of techniques that effectively reduce the computation cost of training deep learning models on mobile devices at the cost of negligible accuracy loss. Experiments using real-world text input from millions of users demonstrate that DeepType significantly improves the input efficiency for individual users, and its incurred computation and energy costs are within the performance and battery restrictions of typical COTS mobile devices.<\/jats:p>","DOI":"10.1145\/3287075","type":"journal-article","created":{"date-parts":[[2018,12,27]],"date-time":"2018-12-27T19:28:03Z","timestamp":1545938883000},"page":"1-26","update-policy":"https:\/\/doi.org\/10.1145\/crossmark-policy","source":"Crossref","is-referenced-by-count":64,"title":["DeepType"],"prefix":"10.1145","volume":"2","author":[{"given":"Mengwei","family":"Xu","sequence":"first","affiliation":[{"name":"Key Laboratory on High-Confidence Software Technologies (Ministry of Education), School of Electronics Engineering and Computer Science, Peking University, Beijing, China"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Feng","family":"Qian","sequence":"additional","affiliation":[{"name":"University of Minnesota - Twin Cities, Minneapolis, MN, USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Qiaozhu","family":"Mei","sequence":"additional","affiliation":[{"name":"University of Michigan, Ann Arbor, MI, USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Kang","family":"Huang","sequence":"additional","affiliation":[{"name":"Kika Tech, Beijing, China"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Xuanzhe","family":"Liu","sequence":"additional","affiliation":[{"name":"Key Laboratory on High-Confidence Software Technologies (Ministry of Education), School of Electronics Engineering and Computer Science, Peking University, Beijing, China"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"320","published-online":{"date-parts":[[2018,12,27]]},"reference":[{"key":"e_1_2_2_1_1","volume-title":"https:\/\/www.radicati.com\/wp\/wp-content\/uploads\/2015\/02\/Email-Statistics-Report-2015-2019-Executive-Summary.pdf","author":"Report Email Statistics","year":"2015","unstructured":"Email Statistics Report , 2015-2019. https:\/\/www.radicati.com\/wp\/wp-content\/uploads\/2015\/02\/Email-Statistics-Report-2015-2019-Executive-Summary.pdf , 2015 . Email Statistics Report, 2015-2019. https:\/\/www.radicati.com\/wp\/wp-content\/uploads\/2015\/02\/Email-Statistics-Report-2015-2019-Executive-Summary.pdf, 2015."},{"key":"e_1_2_2_2_1","volume-title":"Collaborative Machine Learning without Centralized Training Data. https:\/\/research.googleblog.com\/2017\/04\/federated-learning-collaborative.html","author":"Learning Federated","year":"2016","unstructured":"Federated Learning : Collaborative Machine Learning without Centralized Training Data. https:\/\/research.googleblog.com\/2017\/04\/federated-learning-collaborative.html , 2016 . Federated Learning: Collaborative Machine Learning without Centralized Training Data. https:\/\/research.googleblog.com\/2017\/04\/federated-learning-collaborative.html, 2016."},{"key":"e_1_2_2_3_1","volume-title":"three times more than SMS. https:\/\/www.theverge.com\/2016\/4\/12\/11415198\/facebook-messenger-whatsapp-number-messages-vs-sms-f8-2016","author":"Messenger","year":"2016","unstructured":"Messenger and WhatsApp process 60 billion messages a day , three times more than SMS. https:\/\/www.theverge.com\/2016\/4\/12\/11415198\/facebook-messenger-whatsapp-number-messages-vs-sms-f8-2016 , 2016 . Messenger and WhatsApp process 60 billion messages a day, three times more than SMS. https:\/\/www.theverge.com\/2016\/4\/12\/11415198\/facebook-messenger-whatsapp-number-messages-vs-sms-f8-2016, 2016."},{"key":"e_1_2_2_4_1","volume-title":"https:\/\/research.googleblog.com\/2017\/05\/the-machine-intelligence-behind-gboard.html","author":"Intelligence Behind Gboard The Machine","year":"2016","unstructured":"The Machine Intelligence Behind Gboard . https:\/\/research.googleblog.com\/2017\/05\/the-machine-intelligence-behind-gboard.html , 2016 . The Machine Intelligence Behind Gboard. https:\/\/research.googleblog.com\/2017\/05\/the-machine-intelligence-behind-gboard.html, 2016."},{"key":"e_1_2_2_5_1","volume-title":"https:\/\/danijar.com\/variable-sequence-lengths-in-tensorflow\/","author":"le Variable Sequence Masks","year":"2017","unstructured":"Masks to hand le Variable Sequence Lengths input to RNN. https:\/\/danijar.com\/variable-sequence-lengths-in-tensorflow\/ , 2017 . Masks to handle Variable Sequence Lengths input to RNN. https:\/\/danijar.com\/variable-sequence-lengths-in-tensorflow\/, 2017."},{"key":"e_1_2_2_6_1","volume-title":"https:\/\/developer.qualcomm.com\/software\/snapdragon-profiler","author":"Profiler Snapdragon","year":"2017","unstructured":"Snapdragon Profiler . https:\/\/developer.qualcomm.com\/software\/snapdragon-profiler , 2017 . Snapdragon Profiler. https:\/\/developer.qualcomm.com\/software\/snapdragon-profiler, 2017."},{"key":"e_1_2_2_7_1","volume-title":"http:\/\/www.chinadaily.com.cn\/china\/2017-11\/10\/content_34372396.htm","author":"WeChat","year":"2017","unstructured":"WeChat users send 38 billion messages daily. http:\/\/www.chinadaily.com.cn\/china\/2017-11\/10\/content_34372396.htm , 2017 . WeChat users send 38 billion messages daily. http:\/\/www.chinadaily.com.cn\/china\/2017-11\/10\/content_34372396.htm, 2017."},{"key":"e_1_2_2_8_1","volume-title":"https:\/\/www.tensorflow.org\/versions\/master\/programmers_guide\/embedding","author":"Embedding Deep Learning","year":"2018","unstructured":"Deep Learning Embedding . https:\/\/www.tensorflow.org\/versions\/master\/programmers_guide\/embedding , 2018 . Deep Learning Embedding. https:\/\/www.tensorflow.org\/versions\/master\/programmers_guide\/embedding, 2018."},{"key":"e_1_2_2_9_1","volume-title":"http:\/\/www.internetlivestats.com\/google-search-statistics\/","author":"Google","year":"2018","unstructured":"Google search statistics. http:\/\/www.internetlivestats.com\/google-search-statistics\/ , 2018 . Google search statistics. http:\/\/www.internetlivestats.com\/google-search-statistics\/, 2018."},{"key":"e_1_2_2_10_1","volume-title":"https:\/\/www.quora.com\/How-many-comments-are-made-on-reddit-each-day","author":"How","year":"2018","unstructured":"How many comments are made on reddit each day. https:\/\/www.quora.com\/How-many-comments-are-made-on-reddit-each-day , 2018 . How many comments are made on reddit each day. https:\/\/www.quora.com\/How-many-comments-are-made-on-reddit-each-day, 2018."},{"key":"e_1_2_2_11_1","volume-title":"https:\/\/blog.swiftkey.com\/neural-networks-a-meaningful-leap-for-mobile-typing\/","author":"SwiftKey Neural Network","year":"2018","unstructured":"Neural Network in SwiftKey . https:\/\/blog.swiftkey.com\/neural-networks-a-meaningful-leap-for-mobile-typing\/ , 2018 . Neural Network in SwiftKey. https:\/\/blog.swiftkey.com\/neural-networks-a-meaningful-leap-for-mobile-typing\/, 2018."},{"key":"e_1_2_2_12_1","volume-title":"https:\/\/www.nvidia.com\/en-us\/data-center\/tesla-k80\/","author":"NVIDIA TESLA","year":"2018","unstructured":"NVIDIA TESLA K80. https:\/\/www.nvidia.com\/en-us\/data-center\/tesla-k80\/ , 2018 . NVIDIA TESLA K80. https:\/\/www.nvidia.com\/en-us\/data-center\/tesla-k80\/, 2018."},{"key":"e_1_2_2_13_1","volume-title":"https:\/\/catalog.ldc.upenn.edu\/ldc99t42","author":"Bank Dataset Penn Tree","year":"2018","unstructured":"Penn Tree Bank Dataset . https:\/\/catalog.ldc.upenn.edu\/ldc99t42 , 2018 . Penn Tree Bank Dataset. https:\/\/catalog.ldc.upenn.edu\/ldc99t42, 2018."},{"key":"e_1_2_2_14_1","volume-title":"https:\/\/goo.gl\/8XaKrK","author":"SwitfKey Personalization","year":"2018","unstructured":"Personalization in SwitfKey . https:\/\/goo.gl\/8XaKrK , 2018 . Personalization in SwitfKey. https:\/\/goo.gl\/8XaKrK, 2018."},{"key":"e_1_2_2_15_1","volume-title":"https:\/\/www.tensorflow.org\/api_docs\/python\/tf\/train\/GradientDescentOptimizer","author":"TensorFlow Selective Parameter","year":"2018","unstructured":"Selective Parameter Update in TensorFlow . https:\/\/www.tensorflow.org\/api_docs\/python\/tf\/train\/GradientDescentOptimizer , 2018 . Selective Parameter Update in TensorFlow. https:\/\/www.tensorflow.org\/api_docs\/python\/tf\/train\/GradientDescentOptimizer, 2018."},{"key":"e_1_2_2_16_1","volume-title":"https:\/\/www.tensorflow.org\/","year":"2018","unstructured":"TensorFlow. https:\/\/www.tensorflow.org\/ , 2018 . TensorFlow. https:\/\/www.tensorflow.org\/, 2018."},{"key":"e_1_2_2_17_1","volume-title":"https:\/\/github.com\/tensorflow\/tensorflow\/tree\/master\/tensorflow\/examples\/android","author":"Android Camera Demo TensorFlow","year":"2018","unstructured":"TensorFlow Android Camera Demo . https:\/\/github.com\/tensorflow\/tensorflow\/tree\/master\/tensorflow\/examples\/android , 2018 . TensorFlow Android Camera Demo. https:\/\/github.com\/tensorflow\/tensorflow\/tree\/master\/tensorflow\/examples\/android, 2018."},{"key":"e_1_2_2_18_1","volume-title":"http:\/\/www.internetlivestats.com\/twitter-statistics\/","author":"Statistics Twitter Usage","year":"2018","unstructured":"Twitter Usage Statistics . http:\/\/www.internetlivestats.com\/twitter-statistics\/ , 2018 . Twitter Usage Statistics. http:\/\/www.internetlivestats.com\/twitter-statistics\/, 2018."},{"key":"e_1_2_2_19_1","doi-asserted-by":"publisher","DOI":"10.1145\/2976749.2978318"},{"key":"e_1_2_2_20_1","volume-title":"Neural machine translation by jointly learning to align and translate. arXiv preprint arXiv:1409.0473","author":"Bahdanau D.","year":"2014","unstructured":"D. Bahdanau , K. Cho , and Y. Bengio . Neural machine translation by jointly learning to align and translate. arXiv preprint arXiv:1409.0473 , 2014 . D. Bahdanau, K. Cho, and Y. Bengio. Neural machine translation by jointly learning to align and translate. arXiv preprint arXiv:1409.0473, 2014."},{"key":"e_1_2_2_21_1","doi-asserted-by":"publisher","DOI":"10.1145\/2968219.2971389"},{"key":"e_1_2_2_22_1","doi-asserted-by":"publisher","DOI":"10.1145\/2994551.2994564"},{"key":"e_1_2_2_23_1","doi-asserted-by":"publisher","DOI":"10.5555\/176313.176316"},{"key":"e_1_2_2_24_1","doi-asserted-by":"publisher","DOI":"10.1561\/1500000055"},{"key":"e_1_2_2_25_1","volume-title":"Handbook for safeguarding sensitive personally identifiable information","author":"Callahan M.","year":"2014","unstructured":"M. Callahan . Handbook for safeguarding sensitive personally identifiable information . Washington, DC : Department of Homeland Security . Retrieved February , 23: 2014 , 2012. M. Callahan. Handbook for safeguarding sensitive personally identifiable information. Washington, DC: Department of Homeland Security. Retrieved February, 23:2014, 2012."},{"key":"e_1_2_2_26_1","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-642-31680-7_8"},{"key":"e_1_2_2_27_1","doi-asserted-by":"publisher","DOI":"10.1007\/s10799-005-5879-y"},{"key":"e_1_2_2_28_1","doi-asserted-by":"publisher","DOI":"10.1145\/2541940.2541967"},{"key":"e_1_2_2_29_1","doi-asserted-by":"publisher","DOI":"10.1109\/ISCA.2016.40"},{"key":"e_1_2_2_30_1","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/P16-1022"},{"key":"e_1_2_2_31_1","first-page":"1269","volume-title":"Advances in neural information processing systems","author":"Denton E. L.","year":"2014","unstructured":"E. L. Denton , W. Zaremba , J. Bruna , Y. LeCun , and R. Fergus . Exploiting linear structure within convolutional networks for efficient evaluation . In Advances in neural information processing systems , pages 1269 -- 1277 , 2014 . E. L. Denton, W. Zaremba, J. Bruna, Y. LeCun, and R. Fergus. Exploiting linear structure within convolutional networks for efficient evaluation. In Advances in neural information processing systems, pages 1269--1277, 2014."},{"key":"e_1_2_2_32_1","first-page":"1269","volume-title":"Proceedings of the Annual Conference on Neural Information Processing Systems (NIPS'14)","author":"Denton E. L.","year":"2014","unstructured":"E. L. Denton , W. Zaremba , J. Bruna , Y. LeCun , and R. Fergus . Exploiting linear structure within convolutional networks for efficient evaluation . In Proceedings of the Annual Conference on Neural Information Processing Systems (NIPS'14) , pages 1269 -- 1277 , 2014 . E. L. Denton, W. Zaremba, J. Bruna, Y. LeCun, and R. Fergus. Exploiting linear structure within convolutional networks for efficient evaluation. In Proceedings of the Annual Conference on Neural Information Processing Systems (NIPS'14), pages 1269--1277, 2014."},{"key":"e_1_2_2_33_1","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-540-25955-8_12"},{"key":"e_1_2_2_34_1","doi-asserted-by":"publisher","DOI":"10.1145\/2508859.2516758"},{"key":"e_1_2_2_35_1","doi-asserted-by":"publisher","DOI":"10.1145\/1966445.1966447"},{"key":"e_1_2_2_36_1","volume-title":"Neural networks for text correction and completion in keyboard decoding. arXiv preprint arXiv:1709.06429","author":"Ghosh S.","year":"2017","unstructured":"S. Ghosh and P. O. Kristensson . Neural networks for text correction and completion in keyboard decoding. arXiv preprint arXiv:1709.06429 , 2017 . S. Ghosh and P. O. Kristensson. Neural networks for text correction and completion in keyboard decoding. arXiv preprint arXiv:1709.06429, 2017."},{"key":"e_1_2_2_37_1","doi-asserted-by":"publisher","DOI":"10.1109\/ISCA.2016.30"},{"key":"e_1_2_2_38_1","doi-asserted-by":"publisher","DOI":"10.1145\/2906388.2906396"},{"key":"e_1_2_2_39_1","doi-asserted-by":"publisher","DOI":"10.1109\/ICASSP.2014.6853595"},{"key":"e_1_2_2_40_1","doi-asserted-by":"publisher","DOI":"10.1162\/neco.1997.9.8.1735"},{"key":"e_1_2_2_41_1","doi-asserted-by":"publisher","DOI":"10.1109\/SP.2013.31"},{"key":"e_1_2_2_42_1","doi-asserted-by":"publisher","DOI":"10.1145\/1321440.1321573"},{"key":"e_1_2_2_43_1","first-page":"100","volume-title":"IEEE First International Conference on","author":"Khomenko V.","year":"2016","unstructured":"V. Khomenko , O. Shyshkov , O. Radyvonenko , and K. Bokhan . Accelerating recurrent neural network training using sequence bucketing and multi-gpu data parallelization. In Data Stream Mining & Processing (DSMP) , IEEE First International Conference on , pages 100 -- 103 , 2016 . V. Khomenko, O. Shyshkov, O. Radyvonenko, and K. Bokhan. Accelerating recurrent neural network training using sequence bucketing and multi-gpu data parallelization. In Data Stream Mining & Processing (DSMP), IEEE First International Conference on, pages 100--103, 2016."},{"key":"e_1_2_2_44_1","first-page":"303","volume-title":"Proceedings of the section on survey research methods","author":"Kim J. J.","year":"1986","unstructured":"J. J. Kim . A method for limiting disclosure in microdata based on random noise and transformation . In Proceedings of the section on survey research methods , pages 303 -- 308 , 1986 . J. J. Kim. A method for limiting disclosure in microdata based on random noise and transformation. In Proceedings of the section on survey research methods, pages 303--308, 1986."},{"key":"e_1_2_2_45_1","volume-title":"Federated optimization: Distributed machine learning for on-device intelligence. arXiv preprint arXiv:1610.02527","author":"Kone\u010dn\u00fd J.","year":"2016","unstructured":"J. Kone\u010dn\u00fd , H. B. McMahan , D. Ramage , and P. Richt\u00e1rik . Federated optimization: Distributed machine learning for on-device intelligence. arXiv preprint arXiv:1610.02527 , 2016 . J. Kone\u010dn\u00fd, H. B. McMahan, D. Ramage, and P. Richt\u00e1rik. Federated optimization: Distributed machine learning for on-device intelligence. arXiv preprint arXiv:1610.02527, 2016."},{"key":"e_1_2_2_46_1","doi-asserted-by":"publisher","DOI":"10.5555\/2959355.2959378"},{"key":"e_1_2_2_47_1","doi-asserted-by":"publisher","DOI":"10.1145\/2750858.2804262"},{"key":"e_1_2_2_48_1","doi-asserted-by":"publisher","DOI":"10.1145\/2623330.2623612"},{"key":"e_1_2_2_49_1","doi-asserted-by":"publisher","DOI":"10.1145\/2637166.2637225"},{"key":"e_1_2_2_50_1","doi-asserted-by":"publisher","DOI":"10.1145\/3130941"},{"key":"e_1_2_2_51_1","doi-asserted-by":"publisher","DOI":"10.1145\/2884781.2884828"},{"key":"e_1_2_2_52_1","volume-title":"Thinet: A filter level pruning method for deep neural network compression. arXiv preprint arXiv:1707.06342","author":"Luo J.-H.","year":"2017","unstructured":"J.-H. Luo , J. Wu , and W. Lin . Thinet: A filter level pruning method for deep neural network compression. arXiv preprint arXiv:1707.06342 , 2017 . J.-H. Luo, J. Wu, and W. Lin. Thinet: A filter level pruning method for deep neural network compression. arXiv preprint arXiv:1707.06342, 2017."},{"key":"e_1_2_2_53_1","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/D15-1166"},{"key":"e_1_2_2_54_1","unstructured":"M. Malekzadeh R. G. Clegg A. Cavallaro and H. Haddadi. Privacy-preserving sensor data analysis for edge computing.  M. Malekzadeh R. G. Clegg A. Cavallaro and H. Haddadi. Privacy-preserving sensor data analysis for edge computing."},{"key":"e_1_2_2_55_1","doi-asserted-by":"publisher","DOI":"10.5555\/2206206"},{"key":"e_1_2_2_56_1","first-page":"0","volume-title":"Sixth International Conference on Learning Representations (ICLR'18)","author":"McMahan H. B.","year":"2018","unstructured":"H. B. McMahan , D. Ramage , K. Talwar , and L. Zhang . Learning differentially private recurrent language models . In Sixth International Conference on Learning Representations (ICLR'18) , pages 0 -- 0 , 2018 . H. B. McMahan, D. Ramage, K. Talwar, and L. Zhang. Learning differentially private recurrent language models. In Sixth International Conference on Learning Representations (ICLR'18), pages 0--0, 2018."},{"key":"e_1_2_2_57_1","doi-asserted-by":"publisher","DOI":"10.1145\/2971648.2971731"},{"key":"e_1_2_2_58_1","volume-title":"Utilizing noise addition for data privacy, an overview. arXiv preprint arXiv:1309.3958","author":"Mivule K.","year":"2013","unstructured":"K. Mivule . Utilizing noise addition for data privacy, an overview. arXiv preprint arXiv:1309.3958 , 2013 . K. Mivule. Utilizing noise addition for data privacy, an overview. arXiv preprint arXiv:1309.3958, 2013."},{"key":"e_1_2_2_59_1","first-page":"91","volume-title":"USENIX Security Symposium","author":"Oberheide J.","year":"2008","unstructured":"J. Oberheide , E. Cooke , and F. Jahanian . Cloudav: N-version antivirus in the network cloud . In USENIX Security Symposium , pages 91 -- 106 , 2008 . J. Oberheide, E. Cooke, and F. Jahanian. Cloudav: N-version antivirus in the network cloud. In USENIX Security Symposium, pages 91--106, 2008."},{"key":"e_1_2_2_60_1","doi-asserted-by":"publisher","DOI":"10.1109\/SP.2013.24"},{"key":"e_1_2_2_61_1","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2007.383157"},{"key":"e_1_2_2_62_1","doi-asserted-by":"publisher","DOI":"10.1109\/4236.968828"},{"key":"e_1_2_2_63_1","volume-title":"Aug. 8","author":"Robinson B. A.","year":"2006","unstructured":"B. A. Robinson and M. R. Long\u00e9 . Keyboard system with automatic correction , Aug. 8 2006 . US Patent 7,088,345. B. A. Robinson and M. R. Long\u00e9. Keyboard system with automatic correction, Aug. 8 2006. US Patent 7,088,345."},{"key":"e_1_2_2_64_1","doi-asserted-by":"publisher","DOI":"10.1145\/2654822.2541949"},{"key":"e_1_2_2_65_1","doi-asserted-by":"publisher","DOI":"10.1145\/1273221.1273222"},{"key":"e_1_2_2_66_1","volume-title":"International Conference on Artificial Intelligence and Statistics (AISTATS'17)","author":"Vanhaesebrouck P.","year":"2017","unstructured":"P. Vanhaesebrouck , A. Bellet , and M. Tommasi . Decentralized collaborative learning of personalized models over networks . In International Conference on Artificial Intelligence and Statistics (AISTATS'17) , 2017 . P. Vanhaesebrouck, A. Bellet, and M. Tommasi. Decentralized collaborative learning of personalized models over networks. In International Conference on Artificial Intelligence and Statistics (AISTATS'17), 2017."},{"key":"e_1_2_2_67_1","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2016.521"},{"key":"e_1_2_2_68_1","doi-asserted-by":"publisher","DOI":"10.1145\/2741948.2741977"},{"key":"e_1_2_2_69_1","first-page":"2365","volume-title":"INTERSPEECH 2013, 14th Annual Conference of the International Speech Communication Association","author":"Xue J.","year":"2013","unstructured":"J. Xue , J. Li , and Y. Gong . Restructuring of deep neural network acoustic models with singular value decomposition . In INTERSPEECH 2013, 14th Annual Conference of the International Speech Communication Association , Lyon, France , August 25-29, 2013 , pages 2365 -- 2369 , 2013. J. Xue, J. Li, and Y. Gong. Restructuring of deep neural network acoustic models with singular value decomposition. In INTERSPEECH 2013, 14th Annual Conference of the International Speech Communication Association, Lyon, France, August 25-29, 2013, pages 2365--2369, 2013."},{"key":"e_1_2_2_70_1","volume-title":"Designing energy-efficient convolutional neural networks using energy-aware pruning. arXiv preprint arXiv:1611.05128","author":"Yang T.","year":"2016","unstructured":"T. Yang , Y. Chen , and V. Sze . Designing energy-efficient convolutional neural networks using energy-aware pruning. arXiv preprint arXiv:1611.05128 , 2016 . T. Yang, Y. Chen, and V. Sze. Designing energy-efficient convolutional neural networks using energy-aware pruning. arXiv preprint arXiv:1611.05128, 2016."},{"key":"e_1_2_2_71_1","doi-asserted-by":"publisher","DOI":"10.1109\/ISDEA.2012.381"},{"key":"e_1_2_2_72_1","volume-title":"Advances in Neural Information Processing Systems 27: Annual Conference on Neural Information Processing Systems 2014","author":"Yosinski J.","year":"2014","unstructured":"J. Yosinski , J. Clune , Y. Bengio , and H. Lipson . How transferable are features in deep neural networks? In Advances in Neural Information Processing Systems 27: Annual Conference on Neural Information Processing Systems 2014 , December 8-13 2014 , Montreal, Quebec, Canada, pages 3320--3328 , 2014. J. Yosinski, J. Clune, Y. Bengio, and H. Lipson. How transferable are features in deep neural networks? In Advances in Neural Information Processing Systems 27: Annual Conference on Neural Information Processing Systems 2014, December 8-13 2014, Montreal, Quebec, Canada, pages 3320--3328, 2014."},{"key":"e_1_2_2_73_1","volume-title":"Recurrent neural network regularization. arXiv preprint arXiv:1409.2329","author":"Zaremba W.","year":"2014","unstructured":"W. Zaremba , I. Sutskever , and O. Vinyals . Recurrent neural network regularization. arXiv preprint arXiv:1409.2329 , 2014 . W. Zaremba, I. Sutskever, and O. Vinyals. Recurrent neural network regularization. arXiv preprint arXiv:1409.2329, 2014."},{"key":"e_1_2_2_74_1","doi-asserted-by":"publisher","DOI":"10.1145\/2684746.2689060"},{"key":"e_1_2_2_75_1","doi-asserted-by":"publisher","DOI":"10.1109\/TETC.2013.2283233"},{"key":"e_1_2_2_76_1","first-page":"685","volume-title":"Advances in Neural Information Processing Systems","author":"Zhang S.","year":"2015","unstructured":"S. Zhang , A. E. Choromanska , and Y. LeCun . Deep learning with elastic averaging sgd . In Advances in Neural Information Processing Systems , pages 685 -- 693 , 2015 . S. Zhang, A. E. Choromanska, and Y. LeCun. Deep learning with elastic averaging sgd. In Advances in Neural Information Processing Systems, pages 685--693, 2015."}],"container-title":["Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/dl.acm.org\/doi\/10.1145\/3287075","content-type":"unspecified","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/dl.acm.org\/doi\/pdf\/10.1145\/3287075","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,6,18]],"date-time":"2025-06-18T01:02:08Z","timestamp":1750208528000},"score":1,"resource":{"primary":{"URL":"https:\/\/dl.acm.org\/doi\/10.1145\/3287075"}},"subtitle":["On-Device Deep Learning for Input Personalization Service with Minimal Privacy Concern"],"short-title":[],"issued":{"date-parts":[[2018,12,27]]},"references-count":76,"journal-issue":{"issue":"4","published-print":{"date-parts":[[2018,12,27]]}},"alternative-id":["10.1145\/3287075"],"URL":"https:\/\/doi.org\/10.1145\/3287075","relation":{},"ISSN":["2474-9567"],"issn-type":[{"value":"2474-9567","type":"electronic"}],"subject":[],"published":{"date-parts":[[2018,12,27]]},"assertion":[{"value":"2018-05-01","order":0,"name":"received","label":"Received","group":{"name":"publication_history","label":"Publication History"}},{"value":"2018-10-01","order":1,"name":"accepted","label":"Accepted","group":{"name":"publication_history","label":"Publication History"}},{"value":"2018-12-27","order":2,"name":"published","label":"Published","group":{"name":"publication_history","label":"Publication History"}}]}}