Knowledge Transfer Using User-Generated Data within Real-Time Cloud Services = Knowledge Transfer Using User-Generated Data within Real-Time Cloud Services

다국어 초록 (Multilingual Abstract)

When automatic speech recognition (ASR) is provided as a cloud service, it is easy to collect voice and application domain data from users. Harnessing these data will facilitate the provision of more personalized services. In this paper, we demonstrate our transfer learning-based knowledge service that built with the user-generated data collected through our novel system that deliveries personalized ASR service. First, we discuss the motivation, challenges, and prospects of building up such a knowledge-based service-oriented system. Second, we present a Quadruple Transfer Learning (QTL) method that can learn a classification model from a source domain and transfer it to a target domain. Third, we provide an overview architecture of our novel system that collects voice data from mobile users, labels the data via crowdsourcing, utilises these collected user-generated data to train different machine learning models, and delivers the personalised real-time cloud services. Finally, we use the E-Book data collected from our system to train classification models and apply them in the smart TV domain, and the experimental results show that our QTL method is effective in two classification tasks, which confirms that the knowledge transfer provides a value-added service for the upper-layer mobile applications in different domains.

참고문헌 (Reference)

1 T. Hofmann, "Unsupervised learning by probabilistic latent semantic analysis" 42 (42): 177-196, 2001

2 F. Zhuang, "Triplex transfer learning : exploiting both shared and distinct concepts for text classification" 44 (44): 1191-1203, 2014

3 O. V. Joldzic, "The impact of cluster characteristics on HiveQL query optimization" 837-840, 2013

4 S. Y. Ho, "The attraction of personalized service for users in mobile commerce : an empirical study" 3 (3): 10-18, 2002

5 T. Yoshioka, "The NTT CHiME-3 system : Advances in speech enhancement and recognition for mobile multi-microphone devices" 436-443, 2015

6 W. Xiong, "The Microsoft 2016 conversational speech recognition system" 5255-5259, 2017

7 A. Sheshadri, "SQUARE : A benchmark for research on computing crowd consensus" 156-164, 2013

8 J. Pan, "Quadruple Transfer Learning : Exploiting both shared and non-shared concepts for text classification" 90 : 199-210, 2015

9 M. Mehrabani, "Personalized speech recognition for Internet of Things" 369-374, 2015

10 J. Zhang, "Multi-Class Ground Truth Inference in Crowdsourcing with Clustering" 28 (28): 1080-1085, 2016

11 J. Dean, "MapReduce : simplified data processing on large clusters" 51 (51): 107-113, 2018

12 T. Condie, "Machine learning for big data" 939-942, 2013

13 D. M. Blei, "Latent Dirichlet allocation" 3 : 993-1022, 2003

14 J. Zhang, "Imbalanced multiple noisy labeling" 27 (27): 489-503, 2015

15 G. Chen, "Federation in cloud data management : Challenges and opportunities" 26 (26): 1670-1678, 2014

16 C. Wang, "Faster, larger, easier : reining real-time big data processing in cloud" 1-2, 2012

17 W. Galuba, "Encyclopedia of Database Systems" 903-904, 2009

18 M. Long, "Dual Transfer Learning" 540-551, 2012

19 X. Wu, "Data mining with big data" 26 (26): 97-107, 2014

20 F. Zhuang, "Concept learning for cross-domain text classification : A general probabilistic framework" 1960-1966, 2013

21 L. Besacier, "Automatic speech recognition for under-resourced languages : A survey" 56 : 85-100, 2014

22 S. J. Pan, "A survey on transfer learning" 22 (22): 1345-1359, 2010

23 J. Zhang, "A distributed cache for Hadoop distributed file system in real-time cloud services" 12-21, 2012