. OER Watson Watson Artificial Neural Networks ANN MP Rosenblatt BP IBM Deep Blue CNN RNN IBM Watson Thomas J. Watson Watson IBM 2014 CNN C

Size: px

Start display at page:

Download ". OER Watson Watson Artificial Neural Networks ANN MP Rosenblatt BP IBM Deep Blue CNN RNN IBM Watson Thomas J. Watson Watson IBM 2014 CNN C"

黏足束
5 years ago
Views:

1 Open Education Research Vol. 24 No. 1 Feb G434 A Hinton & Salakhutdinov Google Brain ImageNet 2016 Alpha Go 2006 Hinton DOI / j. cnki. kfjyyj JZD043 yusq@ bnu. edu. cn 16

2 . OER Watson Watson Artificial Neural Networks ANN MP Rosenblatt BP IBM Deep Blue CNN RNN IBM Watson Thomas J. Watson Watson IBM 2014 CNN CNN Chu & Krzyzak 2011 Watson RNN Britz 2015 RNN RNN Watson IBM IBM Watson RNN Liu et al Watson Graves & Jaitly Watson ImageNet Watson 17 Deng et al

3 . OER YouTube 1000 Vernon Vinge Le et al clips Tegmark Frank Wilczek AI YouTube Luckin et al The New Yorker Dark Factory BBC Frey & Osborne % 98. 5% 97. 6% 97. 0% 96. 8% 3. 8% 4. 5% 6. 2% 3. 5% 2. 1% 1. 8% 0. 4% Ashock Goel MOOC Jill Watson IBM AI Graaf

4 . OER BBC 0. 4% Ray Clifford 2 AI Teacher AI Teacher 19

5 . OER Find

6 . OER

7 . OER ABCD

8 . OER IBM Watson AI

9 . OER TPACK 12 24

10 . OER Self

11 . OER /

12 . OER Britz D Recurrent Neural Networks Tutorial Part 1-Introduction to RNNs EB/OL http / /www. wildml. com/ Publishing M. A The future of em- C Conference Frey C. B &. Osborne ployment How susceptible are jobs to computerisation J 2015 /09 / recurrent-neural-networks-tutorial-part-1-introduction-to-rnns. 2 Chu J. L. & Krzyzak A Analysis of feature maps selection in supervised learning using convolutional neural networks C. Canadian Conference on Artificial Intelligence. Springer International 3 Deng J. Dong W. Socher R. Li L. J. Li K. & Li F. F ImageNet A large-scale hierarchical image database. Computer Vision and Pattern Recognition. CVPR IEEE. Technological Forecasting and Social Change Graves A. & Jaitly N Towards end-to-end speech recognition with recurrent neural networks C. International Con- 27

13 . OER ference on Machine Learning Hinton G. E. & Salakhutdinov R. R Reducing the dimensionality of data with neural networks J. science Le Q. V. Ranzato M. Monga R. Devin M. Chen K. Corrado G. S. Dean J. & Ng A. Y Building highlevel features using large scale unsupervised learning EB /OL http / /arxiv. org /abs / Liu S. Yang N. Li M. & Zhou M A recursive recurrent neural network for statistical machine translation A. Meeting of the Association for Computational Linguistics Luckin R. Holmes W. Griffiths M. & Forcier L. B Intelligence unleashed. An argument for AI in Education. London Pearson. 10 Mouly F R. Kikuo Johnson s Tech Support EB /OL https / /www. newyorker. com /magazine / 2017 /10 / Graaf M. De A robot has been teaching grad students for 5 months... and NONE of them realized EB /OL http / /www. dailymail. co. uk /news /article /A-robotteaching-grad-students-5-months-NONE-realized. html. 12 Self J The defining characteristics of intelligent tutoring systems research ITSs care precisely J. International Journal of Artificial Intelligence in Education IJAIED J IBM Watson J J J J J The Future Roles of AI Teacher YU Shengquan School of Education Technology / Advanced Innovation Center for Future Education Beijing Normal University Beijing China Abstract In recent years with the leaps and bounds of microelectronics and the Internet the computing and storage capabilities have dramatically promoted the rapid growth of artificial intelligence AI. The breakthroughs and widespread applications of big data have driven the substantial progress of AI. Thus the application of AI in education has become a hot discussion topic. This paper introduces the three schools of AI and their typical cases discusses the attitude of mankind to deal with AI and specifically describes the possible twelve roles that AI teachers will assume in the future. The roles are as follows a teaching assistant who can automatically set a question and correct homework an analyst who can automatically diagnose learning disabilities and provide feedback a coach who can improve the quality of students problem solving ability a counselor who can evaluate students mental health and provide intervention measures a health physician who can monitor and improve students physical health a head teacher who can provide feedback on the report of students comprehensive quality evaluation a smart advisor who can offer personalized teaching service a smart tutor who can solve individual student s problems a career planner who can promote students growth a partner who can help each other during precision teaching research an intelligent a- gent which is able to automatically generate and gather personalized learning content an assistant who can help make education decisions driven by data. In this sense AI teachers will take a critical part in future schools and future education will be an era in which human teachers and AI teachers co-exist. Key words artificial intelligence application of artificial intelligence in education AI teacher co-existence future education 28

. 3. MOOC 2006 MOOC Automated Text Marker 2014 e-rater Yigal et al MOOC Coursera Edx 97

. 3. MOOC 2006 MOOC Automated Text Marker 2014 e-rater Yigal et al MOOC Coursera Edx 97 23 3 2017 6 Open Education Research Vol. 23 No. 3 Jun. 2017 1 2 1 2 1. 100875 2. 100875 PPT G434 A 1007-2179 2017 03-0096-09 Coursera Edx MOOC 100 2 3- social network services SNS 4 2008 2009 2017-03-