O n f e r e n c e

An Intelligent EFL Learning Platform f
or Cultivation of Students’ 
Autonomy 
Bio data 
Yan Tian
is Professor in the School of Foreign Languages at Shanghai Jiao Tong 
University (SJTU). She received her Ph.D. in Computational Linguistics from the 
Department of Computer Science & Engineering at SJTU. Her interests include Intelligent 
Computer-assisted Language Learning and Computational Linguistics. She is the director 
of SJTU Learning Center for Foreign Languages and developed “SJTU Online English 
Learning Center”
˄
http://english.sjtu.edu.cn
˅
. She is the member of both CALICO and 
IALLT as well as the associate editor of The IALLT Journal for Language Learning 
Technologies. She is also the member of Foreign Language Educational Technology 
Association of Shanghai.
Abstract 
This paper addresses t
he issue of cultivating EFL learners’ autonomy with the aid of an 
online intelligent learning platform. A self-developed intelligent EFL learning platform is 
described, followed by an experiment of using it. Questionnaire survey indicates that 
learners’ re
sponse to this platform is positive. Thus, it comes to the conclusion that the 
technology-based approach to fostering language learning autonomy is worthy of the 
efforts. 
Short paper 
Learning a foreign language only in class cannot meet the requirements of the syllabus 
for non-English majors at universities in China. Students have to develop a strong out-of-
class language learning autonomy, too. However, Chinese students often lack learning 
autonomy because of their learning motivation at high schools. It is, therefore, of great 
significance to cultivate their learning autonomy the moment they enter the university.
To do so, an intelligent EFL learning platform (http://202.120.60.18) was developed at 
Shanghai Jiao Tong University (SJTU). 

---

-336- 
2014 CALL Conference 
LINGUAPOLIS
www.antwerpcall.be 
This platfo
rm is able to obtain the initial status of a learner by a “Learning Style Test”, a 
“Motivation Test” and “P
roficiency Tests
”
of different language skills, such as listening 
comprehension, reading comprehension, writing, and translation, with which the system 
would provide the learner with the overall recommendations of his/her online study plans 
and content arrangements. It is also able to automatically guide the learner to learn the 
language according to the results of his/her learning style test, motivation test and
proficiency tests by intelligent management and classification of the learning materials. 
In addition, it can present instant feedback to the learner by regular email, reminding 
him/her of the contents to be covered and tests to be taken before deadlines. When 
finishing the online tests, the scores as well as keys to objective questions and answers 
to subjective ones can be presented to the learner instantly. 
Eighty non-English majors from two natural English classes at SJTU participated in the 
experiment. When they entered SJTU in September 2012, they were asked to make their 
own out-of-class English learning plan. Without any interference, they were asked to 
answer the questionnaires concerning the implementation of their autonomous learning 
plan at the end of the first semester. The results indicate that only 1.64% of the students 
are “very satisfied” with 
their autonomous learning plan
; 3.28% are “satisfied”; 36.89% 
are “not satisfied”; and 54.1% are “not satisfied at all”. When asked of the
reasons 
leading to their failure of carrying out their autonomous learning plan, 62.3% of the 
students said because there is
“no supervision”; 53.28% 
said they are 
“too busy”; 
18.03% said that their plan is 
“not reasonable”; 
and 10.66% listed other reasons. 
At the second semester which began in February 2013, the students were asked to use 
the intelligent EFL learning platform to learn English after class. At the end of the 
semester, a questionnaire survey was again conducted to obtain the feedback of using 
the learning platform in improving the implementation of their autonomous learning plan. 
The survey demonstrates that 40.16% of the students said that the implementation of 
their autonomous learning plan was “greatly improved”; 43.44% said that it is 
“improved”; 10.66% said that they are “not sure”; and 5.74% said that the 
implementation of their autonomous learning plan was “not improved”. Evidently, this 
intelligent EFL learning platform has helped most students satisfactorily implement their 
autonomous learning plan. 

---

-337- 
2014 CALL Conference 
LINGUAPOLIS
www.antwerpcall.be 
Therefore, a conclusion can be reached that modern information technology has provided 
teachers of English a new way to cultivate students’ learning autonomy. Although further 
researches have to be conducted, there are reasons to be optimistic that the intelligent 
EFL platform can bring students a lot of benefit. 
However, the following challenges 
have been encountered in cultivating students’ 
learning autonomy. First, it is difficult to measure students’ learning autonomy. The 
practice of measuring the ability through performance was adopted because it was 
believed that learners will demonstrate their ability in task performance. Presently, only 
the time students spent online, the columns they visited, and the exercises they did can 
be traced. It is still impossible to identify whether it is the student him/herself who left 
those traces. Another challenge going together with this one is that it is difficult to 
observe students’ online behaviors due to the multidimensional features of autonomy
. To 
be specific, it is hard to observe students’ progress so that their learning behavior can be 
controlled, because some behavior is implicit which may not leave any online traces. 
Furthermore, the lack of a “test” method to measure autonomy is also a pr
oblem.
Next, it is hard to reward learners’ online autonomous behaviors. Currently, students are 
given online credits according to the traces they have left. They are also ranked 
according to their credits and their ranking is posted on the homepage with their real 
names. However, it is not certain whether it is necessary to associate their online 
behaviors with the scores of the course or there are other effective means to motivate 
the students. 
Finally, it is not easy to provide instant feedback of subjective exercises for students 
because of the difficulty in the technology of natural language processing. The intelligent 
EFL platform was intended to be designed in such a way that it is able to provide instant 
feedback for all language skills, namely listening, reading, writing and translating. 
However, it is hard to develop the modules of the feedback for writing and translating 
due to the complex technology of natural language processing. At present, the platform 
cannot provide instant feedback of writing for students. Although efforts have been made 
to develop a T-rater which is able to provide instant feedback of online translation for 
students, the precision rate still needs further increasing.
In a word, the challenges come from approaches to cultivating autonomy as well as from 
the technology to support the accomplishment of these approaches.