Journal of Intelligent & Fuzzy Systems 37 (2019) 3381–3391 3381

DOI:10.3233/JIFS-179141
IOS Press

English teaching practice based on artificial

intelligence technology
Yi Bina,∗ and Durbadal Mandalb
a School of Foreign Languages, Huizhou University, Huizhou, Guangdong, China
b National Institute of Technology, Durgapur

Abstract. The automatic scoring of English composition is an inevitable trend of the rapid development of computer technol-
ogy and artificial intelligence technology. This makes the research on the automatic scoring system of the composition more
theoretical and feasible. The integration of information technology and English curriculum has deepened the application of
artificial intelligence in the field of teaching, creating a new opportunity for the optimization of English teaching process, and
opening up a new space for the creation of intelligent and personalized English teaching environment. Based on the integration
of information technology and curriculum, aiming at improving the quality and effect of English teaching in middle schools,
this paper explores the application of artificial intelligence in English teaching in middle schools by using relevant theories of
curriculum theory, literature analysis and field investigation. An implementation plan of College English assisted instruction
system based on AI technology is proposed. Combined with English teaching, some functions of the English teaching system
are improved and humanized. The application of artificial intelligence technology in English teaching is explored in order to
improve the quality and effect of English teaching.

Keywords: artificial intelligence, teaching English, system construction

1. Introduction and realize human intelligent behavior [2]. In many

areas of computer science, artificial intelligence is
Since the reform and opening up, English majors one of the most challenging and creative areas [].
have received the attention of the education man- With the birth and development of artificial intelli-
agement department as a compulsory course in the gence, people began to use computers in the field
basic education stage. The scale of English teach- of teaching [3, 4]. Artificial intelligence (AI) is a
ing has been continuously expanded, the teaching challenging and creative field in English teaching. Its
level has been continuously improved, and the over- birth and development have enhanced the importance
all quality of English teaching in middle schools of information technology and curriculum integra-
has also improved [1]. Artificial intelligence is an tion. Artificial intelligence theory studies how to use
emerging edge discipline developed on the basis of artificial methods to simulate, extend and expand
mutual penetration of computer science, cybernetics, intelligence, mainly the work of science and engi-
information theory, neurophysiology, psychology, neering researchers, based on the theory of natural
philosophy, linguistics and other disciplines. How to intelligence [5]. Artificial intelligence theory studies
use machine computers is mainly studied to imitate how to use artificial methods to simulate, extend and
expand intelligence, mainly the work of science and
∗ Corresponding author. Yi Bin, School of Foreign Languages, engineering researchers, based on the theory of natu-
Huizhou University, Huizhou, Guangdong, 516007, China. ral intelligence [6]. If we can understand the working
E-mail: [email protected]. mechanism of various natural intelligences and the

ISSN 1064-1246/19/$35.00 © 2019 – IOS Press and the authors. All rights reserved
3382 Y. Bin and D. Mandal / English teaching practice based on artificial intelligence technology

structural relationship of its various functional com- in the research of artificial intelligence and the com-
ponents, we can simulate, extend and expand them by bination of English teaching, which makes some
constructing similar structures with highly developed functions of the expert system of English auxiliary
electronic, optical and biological devices, thus real- teaching is more perfect and humanized. Meanwhile,
izing artificial intelligence [7, 8]. In recent decades, it is important to apply the artificial intelligence
with the maturity of artificial intelligence technology, technology in teaching system. In other aspects of
some of its research results have been applied to the teaching, artificial intelligence technology can also
field of teaching, promoting the reform of educational establish human reasoning model learning tools and
development and the modernization of teaching pro- many other applications, showing that more and more
cess. The importance of AI in the teaching system good practicability. In summary, our contributions are
has also reached a consensus [9]. as follow:
At present, technology is still developing rapidly
in the US, Europe and Japan. In 2012, Sheikhan M 1. An artificial intelligence technology is pre-
put forward a [10] based on artificial intelligence. In sented. The integration of information technol-
2013, long Shu Le put forward [11] on the teaching ogy and English curriculum has deepened the
of AI for college students. Garca proposed a link- application of artificial intelligence in the field
based link approach in 2014 to compare links to of teaching, creating a new opportunity for the
Wikipedia [12]. Jain S proposed the dynamic char- optimization of English teaching process, and
acteristics of recommender systems in 2015. Zou S opening up a new space for the creation of
put forward the design and practice of the university intelligent and personalized English teaching
English teaching platform based on artificial intelli- environment.
gence in 2017 [13]. Foreign educational technology 2. An artificial intelligence technology is
researchers are also concerned about a wide range. designed, which can not only conducive to
Whether it is computer-aided teaching, intelligent adapting to English curriculum standards,
robots, intelligent evaluation technology, knowledge but also can realize the reform of English
engineering, artificial neural network and other artifi- classroom teaching mode and the improvement
cial intelligences related technology research, or the of students’ learning methods. The goal is to
introduction of artificial intelligence related technol- improve the visual and auditory perception
ogy achievements into the real process of teaching ability of the intelligent machine, so that
and learning to improve the teaching activities of the intelligent mechanism can understand
many links [14, 15]. People continue to introduce human daily written oral language and realize
techniques such as problem solving and knowledge human-computer interaction.
representation into assisted instruction. In recent 3. Artificial intelligence is implemented to assist
years, the maturity of artificial intelligence tech- teaching college English. The universal data
nology has strengthened the application of English transmission system designed with the above
teaching to its research results. The development of technology optimizes server performance. Sup-
English teaching reform and its teaching moderniza- ported by the embedded database HSQLDB
tion process have also been greatly promoted [16]. In is used for data integration between heteroge-
addition, artificial intelligence has also received great neous platforms. The designed English teaching
attention in the field of assisted English teaching. assistant expert system, using the combina-
The combination of language learning and artifi- tion of production and framework knowledge
cial intelligence applications has gradually developed expression method, using uncertainty reasoning
and improved from exploration. Through decades of technology, can complete the basic evaluation
development, a relatively systematic theoretical sys- of college English teachers’ knowledge points
tem has been formed, which contains extremely rich of student groups.
content, and has been widely used in practice and 4. A set of experimental evaluations are designed
played a significant role [17]. for the proposed method. Artificial intelli-
In this paper, an artificial intelligence technology is gence technology can also establish human
proposed, which can give full play to its own advan- reasoning model learning tools and many
tages in the field of language learning and make up for other applications, showing that more and
many shortcomings in the process of English teach- more good practicability. With the develop-
ing. The innovation of the paper is mainly embodied ment of Internet and the wide application of
 Y. Bin and D. Mandal / English teaching practice based on artificial intelligence technology 3383

virtual reality technology, ITS will be further 3. Expert system and its theoretical methods
improved.
Artificial intelligence is an emerging discipline.
The rest of the paper is organized as follows: Sec- There is no uniform definition yet, and there are
tion II demonstrates the practical significance of the different understandings and definitions of artificial
artificial intelligence technology in English teaching intelligence in different fields. Artificial intelligence
and discusses the motivations of this research. Then is a branch of computer science. It is a compre-
expert system and its theoretical methods is presented hensive discipline that studies computer simulation
in Section 3. In Section 4, experimental design and and extends human brain function. Its precise defi-
analysis are presented. The contents and methods nition is a computer system with human knowledge
used in this article is presented in Section 5. and behavior, and the ability to learn, infer, judge,
remember knowledge and understand human natural
language. Artificial intelligence is the most promi-
2. Related work nent feature in the simulation of human thinking
mode and the reorientation of thinking process [28].
Artificial intelligence as a crosscutting frontier The complexity of English teaching needs to use
is changing people’s traditional concepts, thinking artificial intelligence to give its unique solution to
and lifestyles, improving human knowledge and each student and provide appropriate assistance to
education [18, 19]. With English as the universal develop a dynamic teaching strategy. These unique
language, the country and the society urgently need thinking activities were only done by humans in the
compound talents with comprehensive qualities, and past. Nowadays, the artificial intelligence basic tech-
pay more attention to the practical use of English niques such as knowledge representation technology,
for college graduates. Under this circumstance, col- search technology, reasoning technology, inductive
lege English teaching reform came into being [20]. technology and association technology have realized
Trying to design an artificial intelligence system the simulation of human thinking activities [29–32].
module for middle school English teaching, to cre- Knowledge representation based on artificial intelli-
ate a good language communication environment, gence takes knowledge as the object, takes computer
strengthen the integration of information technol- software and hardware, computer science, artificial
ogy and English curriculum, the quality of middle intelligence and expert system technology as the
school English teaching demands improvement and tool, takes philosophy, psychology and logic as the
teaching effect demands optimization to some extent method and guidance, and expresses knowledge into
[21]. The system consists of two parts, one is a rule- a “knowledge base” which can be directly processed
based expert system, and the main user is an English by computer. Artificial intelligence technology can
teacher. Domain experts deposit domain knowledge give full play to its own advantages in the field of lan-
into knowledge base according to knowledge rules, guage learning and make up for many shortcomings
and English teachers diagnose the whole students in the process of English teaching. It is not only con-
through the system [22, 23]. Thus, students’ mas- ducive to adapting to English curriculum standards,
tery of basic knowledge can be obtained and teaching but also can realize the reform of English classroom
guidance can be given [24]. The other part is a student teaching mode and the improvement of students’
autonomous learning system based on neural network learning methods. Its research goal is to improve the
and memory psychology. The main users are students. visual and auditory perception ability of the intelli-
The innovation of this paper is mainly embodied in gent machine, so that the intelligent mechanism can
the research of artificial intelligence and the combina- understand human daily written oral language and
tion of English teaching, which makes some functions realize human-computer interaction.
of the expert system of English auxiliary teaching With the improvement of College English teach-
more perfect and humanized [25]. There is a consen- ing and freshmen’s English profi-ciency, the Ministry
sus on the importance of the application of artificial of Education has made new regulations on College
intelligence technology in teaching system. It is of syllabus and new requirements for college English
great practical significance to study the application teaching. Chinese English teaching has been con-
of artificial intelligence technology in English teach- fronted with the problems of practical application
ing under the guidance of information technology and and lack of communicative ability. The construc-
curriculum integration [26, 27]. tion of artificial intelligence English teaching system
3384 Y. Bin and D. Mandal / English teaching practice based on artificial intelligence technology

Table 1
Weighted U
U1 U2 U3
3.6 3.8 5.6
2.3 7.5 2.6
3.6 8.9 6.7

can be done. It is a breakthrough to promote the

reform of modern information technology in English
teaching, improve the quality of teachers’ English
teaching and students’ initiative in learning English.
Language is the most basic and important intelli-
gent behavior of human beings. Human beings use
language to communicate, reasoning and understand Fig. 1. Weighted U.
the laws of things. Natural language understand-
ing is also called natural language processing. It is
Table 2
the use of computers to understand human natu- Concept influence relationship table
ral language to achieve the purpose of simulating
Precondition Value
human language intelligent module. Chinese-style
C1 2
English teaching has always faced the problem of C2 6
lack of practical application and communication abil- C3 5
ity. The construction of artificial intelligence English
teaching system can be used as a breakthrough to
promote the transformation of modern information
technology to English teaching, improve the quality
of teachers’ English teaching and students’ initiative
in learning English. The achievements of artificial
intelligence in language translation and speech under-
standing programs have evolved into new concepts
for human language processing. Not only can you
find the student’s mistakes, but you can also point
out the students’ mistakes in targeted counseling or
study suggestions. Moreover, according to the char-
acteristics of the students, the teaching content is
automatically selected, the teaching progress is auto-
matically adjusted, and the teaching strategies and
methods are automatically selected. In the process
of social development, teaching plays a cornerstone
role. Therefore, the application field of artificial intel-
ligence is bound to expand first in education and
Fig. 2. Conceptual impact diagram.
teaching.
Combining the 25 students’ answer samples and
matching the rules in the knowledge base, the above The expert system is based on the knowledge
grammatical points are evaluated comprehensively, and experience provided by experts in a certain
and the quantified results are shown in Table 1 and field to make inferential judgments and simulate
Fig. 1 below. the decision process of human experts. It can be
Before constructing the concept impact diagram, regarded as a kind of computer intelligent program
the relationship between concepts should first be system with expertise and experience, and its abil-
represented by a two-dimensional table (called the ity comes from its expert knowledge. The two cores
concept impact relationship table CET), as shown in of artificial intelligence are “manual” and “smart.”
Table 2 and Fig. 2 below: Intelligence includes: information and knowledge
 Y. Bin and D. Mandal / English teaching practice based on artificial intelligence technology 3385

Table 3
Numerical representation of credibility fuzzy quantifiers
Fuzzy quantifier Numerical value
15 1.23
20 2.8
35 9.4

processing, symbol processing, problem solving, etc.

Intelligence can also be said to be a combination
of extensive information processing capabilities and
presentation information capabilities. Natural lan-
guage processing, speech recognition technology,
machine translation, etc. in artificial intelligence
research are closely related to language learning.
An intelligent system without any learning ability
cannot be called a real intelligent system, machine Fig. 3. Numerical representation of credibility fuzzy quantifiers.
learning has a profound impact on all walks of life,
especially in the field of education. In addition, the Table 4
main task of machine learning is to do data analy- P vector system space
sis, that is, data mining. It is necessary to adopt a P vector Grammar point Numerical value
planned or proficient learning method to urge students P1 Temporal 25
to arrange learning time regularly and effectively so P2 Noun 36
that students can achieve their learning goals through P3 Verb 23
self-monitoring, feedback and adjustment. An expert
system only has the knowledge of this special field, so
it can only solve the problems in this particular field,
which improves the possibility of system implemen-
tation and the effectiveness of solving problems. In
the ideal control of the teaching process, a ITS should
include a variety of teaching strategies. And what
strategy should be chosen depends on the specific sit-
uation of students in the student model. It is always
one of the most concerned contents for scientists and
scholars in the field of artificial intelligence to make
man-machine conversation possible. The application
of artificial intelligence in English teaching is likely
to be a breakthrough in this field.
For the convenience of teachers, combined with the
experience of domain experts, some fuzzy quantifiers
Fig. 4. P Vector system space.
and quantitative representations indicating credibility
are selected, as shown in Table 3 and Fig. 3:
Through systematic reasoning analysis, the knowl- penetrated into all areas of social life. Natural Lan-
edge points involved in the 15 selected questions are guage Understanding (NLP) technology provides a
distributed on three different grammatical points, and new human-computer interaction interface for com-
the three grammatical points are written as P vectors puters, and realizes the revolution from text interface,
as shown in Table 4 and Fig. 4: graphical interface to natural language interface.
Because of the continuous improvement of com- Its technical connotation mainly includes speech
puter speed and the expansion of storage capacity, recognition, character recognition and so on. The
computer hardware has made rapid progress, many advantages of using skeleton system are time-saving,
previously impossible works can be achieved, there- fast speed, high development efficiency, strong inher-
fore, artificial intelligence has a new climax. Some itance, but the disadvantages are poor flexibility and
of these technologies have been applied, and AI has versatility. The reasoning mechanism and knowledge
3386 Y. Bin and D. Mandal / English teaching practice based on artificial intelligence technology

Table 5
Training text set
Composition source Theme Number of articles
Pub med english composition 365 15 389
New TOEFL high score composition 23 246
New oriental classics 13 156

representation of the shell are fixed and unchanged.

It is only suitable for the development of the same
type of expert system as the original system, and
is technically limited by the original level of skele-
ton system. Once the content of the instruction is
selected, the system will generate questions to give
the student an exercise or test. In particular, the envi-
ronment simulation module needs to be emphasized. Fig. 5. Training text set.
The main part of the module is an artificial intelli-
gence chat robot. On its interface, the user can input
Table 6
voice or text information, and simulate English every-
Distribution of topics and scores in corpus
day through functions such as text, video, role playing
Fraction 7 8 9 10
and scene dialogue. Life scenes create a good and
relaxed language learning atmosphere. The machine My View on Job-Hopping 89 56 63 78
Haste Makes Waste 86 96 69 88
can analyze the total data of a student’s learning in My view on Fake Commodities 45 63 74 68
each subject in a year, and mine valuable data, and
can effectively analyze the reasons for the formation
of this student’s learning problems and propose cor-
responding learning suggestions, can give a sample
of the established format needed to train the neu-
ral network. The main privilege includes knowledge
base maintenance (adding, querying, modifying, and
deleting static knowledge base and dynamic knowl-
edge base), neural network sample maintenance (add,
query, modify, delete), password modification, and so
on.
The scoring process of this model is a simulation of
the person’s scoring process, so a sample essay refer-
ence is required for the essay. The selection principle
of the sample set is the model of CET-4 and CET-6,
that is, the score of 15 points, and the same as the
title of the essay to be tested, so that the score has a
Fig. 6. Training text set.
measure of scale. It is shown in Table 5 and Fig. 5:
Because these annotations are not required in the
scoring process, regular expressions are used to delete but also to develop international communication
the annotations in the text in order to extract feature and participate in social activities. The best way of
words and text scoring. The distribution of the three language learning is, of course, through communica-
headings in the ST4 corpus is shown in Table 6 and tive acquisition. Therefore, in the process of middle
Fig. 6. school English teaching, it is necessary to create such
a language atmosphere, but in practice there is one.
It’s difficult. Like other large-scale software devel-
4. Experimental design and analysis opment processes, the development process of expert
system should follow the steps and principles of soft-
The purpose of language learning is not only to ware engineering, including five stages: requirement
master English thinking and communicative tools, analysis, system design, system implementation,
 Y. Bin and D. Mandal / English teaching practice based on artificial intelligence technology 3387

testing and maintenance. Expert Module: Ideally, the courses and traditional classroom complementary.
expert module in the system should have the abil- It can provide technical support for the interactive
ity to answer questions. For the problems generated English teaching environment in middle schools. It
by the system and the questions raised by students, can process text, graphics, images, sounds and other
the correct answers can be generated through auto- media information synthetically through technical
matic reasoning. English teachers need to respond to computer, and establish a logical connection and
the intelligent operation of the SAIETS system based integration of these information into an intelligent
on specific curriculum requirements and content. system. In the process of teaching, the SAIETS sys-
Select appropriate teaching strategies by observing tem is used to encourage students to make full use of
students’ preferences and mastery of the system. the system in their English learning after class, and to
The basic knowledge hinders the improvement of use different system modules according to the course
students’ English level to a certain extent. There- schedule and teaching objectives. The new concepts
fore, it is necessary to help students to consolidate of human-computer interaction and human-computer
basic knowledge and further improve the quality of symbiosis have been created by this method, which
teaching. In order to retrieve and manage in the com- has expanded human capabilities and promoted all
puter, the knowledge base and the inference engine aspects of education reform.
are designed. Finally, the overall structure of the The formula of knowledge representation consid-
design system, including system architecture, knowl- ering weight is as follows:
edge base design, knowledge descriptive language
design, interface design, reasoning and interpretation |a − a | + |b − b | + |c − c | < ε (1)
mechanism design, and knowledge base management
design. The administrator is mainly responsible for Let the credibility of each sub-condition AI be
the management and maintenance of system infor- D, then the credibility of the combined evidence is
mation and the management of user information and calculated by Equation (1):
permissions. The teacher mainly manages the course-
ware and test question bank, analyzes and evaluates (It + Qt ) (It + Qt )
AIt = · (2)
the student’s learning behavior, and timely adjusts 2 Dt
and improves the teaching strategy and teaching
progress according to the results of data mining. If the normalization condition is not met, then W
Artificial intelligence technology has been grad- is calculated by the following formula:
ually applied to intelligent teaching systems and T −t
has a close relationship with teaching moderniza- w(t) = w2 + (w1 − w2 ) (3)
T
tion. The utility that may be applied in middle
school English teaching includes providing interac- W is the credibility of the rules and satisfies
tive English teaching environment and changing the 0 < X<1. The credibility of the conclusion is calcu-
current state of “Sound English”. Provide person- lated from the following formula:
alized teaching and optimize teaching results. The
 
development of AI artificial intelligence technology Xi
will also greatly promote the development of ITS. wij = wij + a − wij (4)
m
Students are grouped and taught on the basis of the
first step of work. Students with the same learning In this system, the rules in the knowledge base
ability are divided into the same group to teach stu- are based on credibility, so the uncertain reasoning
dents in accordance with their aptitude, and targeted method is used to realize the uncertain reasoning
and individualized teaching is given to students with engine. And the following fuzzy relations are estab-
different learning abilities. Any artificial intelligence lished:
system should focus on “intelligence.” An intelligent
English teaching aid system should attempt to predict FP = (PA − PB ) AP (5)
the mastery of student knowledge from observable
student test results and exercise results. The purpose Where W is the conclusion vector and K is the
of this teaching method is to combine traditional logical operator. Solve the above formula to get the
classroom teaching with modern information tech- inference model formula of the system:
nology, and implement the advantages of network
wij (k + 1) = wij (k) + ␩␦i xj (6)
3388 Y. Bin and D. Mandal / English teaching practice based on artificial intelligence technology

fi
Pi = (9)

N
fi
i=1

The credibility vector is the credibility of the rules

in the knowledge base. The weighting vector is the
weight vector Q given by the expert and passed down
by f. The weighted E, i.e. S, is given by the following
formula:
1 d 2 Sas Rf
qf = − πd 2 Ea 2
+ (10)
4 dx dx
Fig. 7. Forward reasoning graph.
Let the connection weight from the a layer neuron
to the d layer neuron x be S, p is the current learning
sample, and E is the output of the d layer neuron under
the p-sample. The transformation function takes the
following function, namely:

La −x
4P ch α · d −4Rf
Sas = ·  + · x (11)
πdEa α sh α · La πα2 Ea
d

For the first D sample, the output error of the net-

work is P:
D(p)
Dr(p) = (12)
min(PA , PB )
Fig. 8. Backward reasoning diagram.
The output of each neuron in the hidden and output
layers of the network is calculated by the following
Where CI represents the forward-inferential con- operation on each sample. The following formulas
clusion X-quantized vector: are given:
X+Y +Z min(AA , AB )
CI =  (7) RA (p) = (13)
X2 + Y2 + Z2 max(AA , AB )

P is a weight vector, which makes t = u suppose The output error is further calculated:
that y is a weighted combination logic formula: Kg
Gg (s) = (14)
  1 + Td0 s
Pt+t = pt (1 + tut (y)) + pt 1 + tut (y ) (8)
The system adopts B/S three tier architecture. In
There are two basic ways of reasoning in pro- the three-tier architecture, almost all the business pro-
duction systems: forward reasoning and backward cessing is done in the middleware. The client only
reasoning. Forward reasoning starts from known facts needs to install the browser, not need to do any other
and gradually derives the conclusion, also known as installation and configuration work. It really realizes
data driven. The reasoning process is shown in Fig. 7. “zero maintenance of the client”. In other aspects of
Backward reasoning first proposes hypotheses, teaching, artificial intelligence technology can also
then verifies the authenticity of these hypotheses, establish human reasoning model learning tools and
and finds all the evidence or facts that the hypoth- many other applications, showing that more and more
esis holds, also known as goal-driven. The reasoning good practicability. With the development of Internet
process is shown in Fig. 8. and the wide application of virtual reality technol-
Combining the 25 students’ answer samples and ogy, ITS will be further improved. The intelligent
matching the rules in the knowledge base, the above operation of SAIETS system is dealt with radially,
three grammatical points are evaluated comprehen- and appropriate teaching strategies are selected by
sively, and the P quantification formula is obtained. observing students’ preferences and mastery of the
 Y. Bin and D. Mandal / English teaching practice based on artificial intelligence technology 3389

Table 7
Weight table of indicators
Weight ratio Index
10% 23.5
25% 12.3
20% 45.8

system. Teachers can not only use the teaching system

to inject fresh vitality into the English classroom, but
also expand learners’ extracurricular English learn-
ing resources. In addition, the diversity of online
English education resources needs to play the “intelli-
gent” choice effect. The three-tier structure is a strict
layering method that delivers requests to the busi-
ness logic layer through the presentation layer. The
Fig. 9. Weight table of indicators.
business logic layer completes the relevant business
rules and logic and accesses the database through
the data access layer to obtain data. The intelligent Table 8
teaching system model based on Web and data min- Model performance evaluation
ing mainly uses data mining technology to analyze Number of Pearson Average
the large amount of information accumulated on the compositions coefficient deviation rate
site to discover patterns and rules of interest to users. 156 0.77 0.65
365 0.32 0.41
Through the systematic and efficient evaluation, the 100 0.78 0.12
teaching effect and the students’ language ability can
be reflected. In addition, the learners in the AI English
teaching system can gradually solve the problems
encountered in the English learning process through
the guidance of the artificial intelligence English
teaching system. According to the characteristics of
English teaching, the assessment and consolidation
of basic knowledge is mainly reflected in the analy-
sis of teachers’ questions on objective questions of
students. Each objective question has one or more
knowledge points included.
By referring to the materials and communicating
with the teachers of the Foreign Languages Institute,
the weight of the score of this indicator is reasonable.
The weights and descriptions of indicators as shown
in Table 7 and Fig. 9 are as follows:
In order to more clearly compare the results of the
machine score and the manual score, and in Table 8 Fig. 10. Model performance evaluation.
and Fig. 10, the Pearson coefficient and the average
deviation rate between the machine score and the system provide the basis and convenience for teach-
manual score are used as indicators for evaluating ers’ two teaching. At the same time, it also provides
the performance of the score model. teachers with information about students’ learning
Students preview and review the knowledge points courses, and provides them with different difficult
by video or text through the knowledge explain- learning content, so as to establish an intelligent and
ing module in the system, and apply what they personalized education environment. Expert learn-
have learned. Thirdly, students can understand their ing function in the system is independent of fault
English knowledge system and test their basic diagnosis function, not open to ordinary teachers.
knowledge through exercises module. The over- It only provides an external interface for English
all evaluation data and analysis generation of the experts to maintain knowledge rules. Through the
3390 Y. Bin and D. Mandal / English teaching practice based on artificial intelligence technology

interface, experts can modify or delete the contents shortcomings expected to be applied in middle school
of the knowledge base, which is conducive to the English teaching. For the fault diagnosis of the
expansion of the system and enhance the credibility. already formalized knowledge points, the system
In particular, the environment simulation module in only supports the entry and judgment of objective
the system should be emphasized. The module takes questions in the selection of the question bank and
the chat robot as the main part, and the user inputs the neural network sample library. Then the applica-
the voice or text information system intelligently to tion is more responsive. In all kinds of middle schools,
obtain the input information through the interface. the drawbacks of traditional English teaching meth-
The system should also have the level of assessment ods have not yet provided a good environment for
and the generation of corresponding learning sug- language communication, which leads to the lack of
gestions and guidance. Not only that, but teachers interest and motivation in English learning, and the
should also be able to make paperless assessments unsatisfactory teaching effect, resulting in the stu-
of students in this module and master the real situ- dents’ comprehensive language use ability cannot be
ation of students’ English learning. Most teaching improved. The uncertain reasoning technology can
systems use a graphical user interface (GUI) or a be used to complete the basic assessment of College
natural language communication interface, or both. English teachers’ mastery of students’ knowledge
However, this kind of autonomous learning is not a and improve the teaching quality. To enhance the
completely isolated learning. If in the independent understanding of artificial intelligence technology,
learning, learners can get personalized guidance and the implementation and application of advanced arti-
help from choosing the learning content, understand- ficial intelligence technology need continuing study
ing the learning objectives, evaluating the learning at home and abroad, to constantly improve the sys-
structure, and diagnosing the learning process to carry tem’s implementation capacity.
on effectively.
Acknowledgement
5. Conclusions
Fund Project: Foreign Language Informatization
Project of Guangdong Provincial Social Science
This paper introduces the content of artifi-
Planning in 2018 (GD18WXZ04); Special Inno-
cial intelligence. The implementation of rule-based
vation Project of Guangdong Province in 2017
uncertainty expert system is analyze, with an overall
(2017GXJK164); Innovation Team Development
implementation scheme of college English auxil-
Plan of Huizhou University.
iary teaching system proposed based on artificial
intelligence. With the rapid development of artifi-
cial intelligence and its wide application in learning References
and life, the universal data transmission system
designed with the above technology optimizes server [1] J. Xia and L. Longshu, The research on teaching of artificial
performance. Supported by the embedded database intelligence for undergraduate students, Science Education
HSQLDB is used for data integration between het- Article Collects, 2013.
[2] S. Zou, Designing and practice of a college english teaching
erogeneous platforms. The designed English teaching platform based on artificial intelligence, Journal of Compu-
assistant expert system, using the combination of tational & Theoretical Nanoscience 14(1) (2017), 104–108.
production and framework knowledge expression [3] Stanisław Duer, Applications of an artificial intelligence
method, using uncertainty reasoning technology, can for servicing of a technical object, Neural Computing &
Applications 22(5) (2013), 955–968.
complete the basic evaluation of college English [4] C. Rana and S. Kumar Jain, A study of the dynamic features
teachers’ knowledge points of student groups. With of recommender systems, Artificial Intelligence Review
the NIO mode, the communication server can still 43(1) (2015), 141–153.
perform smooth file reading and writing when the [5] M. Dzikovska, N. Steinhauser and E. Farrow, BEETLE II:
Deep natural language understanding and automatic feed-
load on the receiving end increases, which shortens back generation for intelligent tutoring in basic electricity
the response time of receiving data. Taking the com- and electronics, International Journal of Artificial Intelli-
prehensive development of middle school students gence in Education 24(3) (2014), 284–332.
[6] D.B. Lenat and P.J. Durlach, Reinforcing math knowledge
as the fundamental starting point and destination,
by immersing students in a simulated learning-by-teaching
the application of English teaching is explord in experience, International Journal of Artificial Intelligence
middle schools, including the possible effects and in Education 24(3) (2014), 216–250.
 Y. Bin and D. Mandal / English teaching practice based on artificial intelligence technology 3391

[7] G.P. Jain, V.P. Gurupur and J.L. Schroeder, Artificial [21] E. Saquete, J.L. Vicedo and Martı́nezBarco, Patricio,
intelligence-based student learning evaluation: A concept Enhancing QA systems with complex temporal question
map-based approach for analyzing a student’s understand- processing capabilities, Journal of Artificial Intelligence
ing of a topic, IEEE Transactions on Learning Technologies Research 35(1) (2014), 775–811.
7(3) (2014), 1–1. [22] I. Maglogiannis and K. Karpouzis, Artificial intelligence
[8] M.K. Najafabadi and Mohd Naz’ri Mahrin, A systematic applications and innovations, Artificial Intelligence Review
literature review on the state of research and practice of 42(3) (2013), 309–311.
collaborative filtering technique and implicit feedback, Arti- [23] I. Shahin, Speaker identification in emotional talking envi-
ficial Intelligence Review 45(2) (2015), 1–35. ronments based on CSPHMM2s, Engineering Applications
[9] S.Y. Lim, M.S. Mohamad and L.E. Chai, Distributed of Artificial Intelligence 26(7) (2013), 1652–1659.
computing and artificial intelligence, 13th International [24] A. Lorenzatti, M. Abel and S.R. Fiorini, AI*IA 2013:
Conference, Frontiers of Information and Electronic Engi- Advances in artificial intelligence, Lecture Notes in Com-
neering (English) 17(4) (2016), 281–282. puter Science 8077(4) (2013), 329–330, 371.
[10] K.D. Althoff, German Journal on Artificial Intelligence, KI [25] N. Matsuda and W.W. Cohen, Teaching the teacher: Tutor-
– Künstliche Intelligenz 27(2) (2013), 89–91. ing SimStudent leads to more effective cognitive tutor
[11] E.N. Benderskaya, Nonlinear trends in modern artificial authoring, International Journal of Artificial Intelligence
intelligence: A new perspective, BEYOND ARTIFI- in Education 25(1) (2015), 1–34.
CIAL INTELLIGENCE: CONTEMPLATIONS, EXPECTA- [26] E.G. Blanchard, Socio-cultural imbalances in AIED
TIONS, APPLICATIONS 4(1) (2013), 113–124. research: Investigations, implications and opportunities,
[12] M. Lauritsen, On balance, Artificial Intelligence & Law International Journal of Artificial Intelligence in Education
23(1) (2015), 1–20. 25(2) (2015), 204–228.
[13] M. Baldoni, C. Baroglio, G. Boella, AI*IA 2013: Advances [27] A. Charlesworth, The comprehensibility theorem and the
in artificial intelligence, Lecture Notes in Computer Science foundations of artificial intelligence, Minds & Machines
8249(4) (2013), 329–330, 371. 24(4) (2014), 439–476.
[14] Garcı́a, Norberto Fernández, Fisteus, Jesús Arias, [28] K. Shankar, Mohamed Elhoseny, E. Dhiravida chelvi,
Fernández and Luis Sánchez, Comparative evaluation S.K. Lakshmanaprabu and Wanqing Wu, An efficient
of link-based approaches for candidate ranking in link- optimal key based chaos function for medical image secu-
to-Wikipedia systems, Journal of Artificial Intelligence rity, IEEE Access, 2018. https://doi.org/10.1109/ACCESS.
Research 49(1) (2014), 733–773. 2018.2874026
[15] T. Dodson, N. Mattei and J.T. Guerin, An english-language [29] K. Shankar, Mohamed Elhoseny, R. Satheesh Kumar,
argumentation interface for explanation generation with S.K. Lakshmanaprabu and Xiaohui Yuan, Secret image
Markov decision processes in the domain of academic advis- sharing scheme with encrypted shadow images using
ing, Acm Transactions on Interactive Intelligent Systems optimal homomorphic encryption technique, Journal of
3(3) (2013), 1–30. Ambient Intelligence and Humanized Computing, 2018.
[16] Y.B. Zhong, Y. Xiang and H.L. Liu, A multi-objective artifi- https://doi.org/10.1007/s12652-018-1161-0
cial bee colony algorithm based on division of the searching [30] K. Karthikeyan, R. Sunder, K. Shankar, S.K. Laksh-
space, Applied Intelligence 41(4) (2014), 987–1011. manaprabu, V. Vijayakumar, Mohamed Elhoseny and
[17] L. Pareto, A teachable agent game engaging primary school Gunasekaran Manogaran, Energy consumption analysis of
children to learn arithmetic concepts and reasoning, Interna- Virtual Machine migration in cloud using hybrid swarm
tional Journal of Artificial Intelligence in Education 24(3) optimization (ABC–BA), The Journal of Supercomputing,
(2014), 251–283. 2018. https://doi.org/10.1007/s11227-018-2583-3
[18] Y. Guo, J. Hu and Y. Peng, Research of new strategies for [31] Noura Metawaa, M. Kabir Hassana and Mohamed Elhoseny,
improving CBR system, Artificial Intelligence Review 42(1) Genetic algorithm based model for optimizing bank lending
(2014), 1–20. decisions, Expert Systems with Applications 80 (2017), 75–
[19] M.G.B.A. Amsa, A.M. Aibinu and M.J.E. Salami, A novel 82.
hybrid artificial intelligence technique for Colpitts Oscil- [32] I.S. Farahat, A.S. Tolba, M. Elhoseny and W. Eladrosy,
lator design, Journal of Control Automation & Electrical Data security and challenges in smart cities. In: Hassanien
Systems 25(1) (2014), 10–21. A., Elhoseny M., Ahmed S., Singh A. (eds) Security in
[20] G.A. Khachatryan, A.V. Romashov and A.R. Khachatryan, Smart Cities: Models, Applications, and Challenges. Lec-
Reasoning mind genie 2: An intelligent tutoring system as a ture Notes in Intelligent Transportation and Infrastructure.
vehicle for international transfer of instructional methods in Springer, Cham, 2019. https://doi.org/10.1007/978-3-030-
mathematics, International Journal of Artificial Intelligence 01560-2 6
in Education 24(3) (2014), 333–382.
Copyright of Journal of Intelligent & Fuzzy Systems is the property of IOS Press and its
content may not be copied or emailed to multiple sites or posted to a listserv without the
copyright holder's express written permission. However, users may print, download, or email
articles for individual use.