2018 International Conference on Sensor Network and Computer Engineering (ICSNCE 2018) 

33 

 

Advanced Dynamic Autonomous Knowledge Learning Method for Distance 

Learning 

Yanfang Fu 

School of Computer Science and Engineering, 

Xi`an Technological University 

Xi’an, 710032, China 

e-mail: 149481698@qq.com 

Xing Li 

School of Computer Science and Engineering, 

Xi`an Technological University 

Xi’an, 710032, China 

e-mail: 825338005@qq.com 

Xueyao Feng 

School of Computer Science and Engineering, 

Xi`an Technological University 

Xi’an, 710032, China 

e-mail: 903840457@qq.com 

Jing Ma 

School of Computer Science and Engineering, 

Xi`an Technological University 

Xi’an, 710032, China 

e-mail: 410329074@qq.com 

 

Abstract—As the manufacturing of vast amount of information 

transmits at an unprecedented rate, traditional learning is no 

way to meet the needs of open learning and lifelong learning, 

how can distance learning be used to help us study become a 

problem to think about .Distance learning requires 

improvement of interactivity and initiative to achieve their 

aptitude. Using a variety of formats data, learners master the 

initiative in the learning process, it is not easy to lead to 

deviation from the learning goals phenomenon. To this end, in 

order to solve this problem, we use the basis of ILTMK theory 

and distributed interactive technology to introduce the active 

learning factors and propose a dynamic autonomous 

knowledge learning model and build its model’s simulation 

running environment and related mechanism. Through 

simulation experiment, we can verify the effectiveness of the 

autonomous learning methods. 

Keywords-Distance Learning; ILTMK theory; Distributed 

Interactive Technology; Dynamic Autonomous Knowledge 

Learning Model 

I. INTRODUCTION 

Looking at the source of knowledge view, Knowledge is 

that the knower grasps the essence of things through the 

phenomenon. So, Bystanders knowledge view derived from 

the ancient Greek natural philosopher, which hold by 

Socrates and others, lead to the “acceptance learning” long 

period domination of human way of learning[1-2]. In China, 

participant knowledge view of Basic Education Reform is ' 

inquiry learning'. The history of education practice shows, 

only participant knowledge view instead of bystander 

knowledge view, inquiry learning may be instead of 

reception learning. Knowledge contains a personal factor. 

Knowledge acquisition process is the learners to participate 

in the process of knowledge construction. Jone Deway 

would explain that participant knowledge view is interaction 

of the human and the environment [3]. Inquiry learning not 

only focuses on participant personal experiences process, 

more emphasis on personal understanding [4]. 

With the popularity of Internet, distance learning has 

gone from a letter-based, based on broadcast television 

systems "receiving learning" toward network-based (Baidu 

library and Blackboard) "research study" direction. Distance 

learning is a complement to a new independent study model. 

Now some intelligent distance learning increases chat rooms, 

electronic whiteboard and virtual classroom, and other 

functions to improve the learning effect, but it is only to 



2018 International Conference on Sensor Network and Computer Engineering (ICSNCE 2018) 

34 

 

provide support in the learning environment, did not achieve 

the custom teaching mode of learner’s active learning. 

Existing model commonly used model based on Agent 

technology or ILTMK (Intelligent Long-distance Teaching 

Model based on Knowledge)[5].However, at present, most 

distance learning model is a simple information-sharing 

individual learning styles. Just solved how to make use of 

the Internet technology, makes the geographical separation 

of some groups can share learning materials, learners are 

always in a passive learning mode. So far no one is satisfied 

with the distance learning model [6]. If the learners can 

interact with active learning, distance learning can produce 

good learning results. 

As information technology, especially with the rapid 

development of network technology and communication 

technology, distance learning and creates a rapidly improve 

our knowledge [9]. How to establish interactive mode 

between the learner and knowledge grantee, plays an 

important role in research and innovation. 

II. RELEVANT THEORIES OF THE DISTANCE 

LEARNING 

To make full use of the knowledge resources on the 

network, we must adopt a deeper sharing strategy that 

absorbed in the knowledge sharing or knowledge unit 

sharing[7]. A knowledge unit or learning entity is relatively 

independent, and it stands for one element of the knowledge 

unit, knowledge can be hierarchical and nested, and itself is 

an independent description metadata. LO can be divided into 

two parts, namely the metadata and content data. Metadata 

section describes the characteristics of the LO, content data 

part gives the knowledge that represented by LO and the 

information. In order to realize the intelligent reasoning of 

the system, we use the production system to represent 

knowledge and information [8]. The reason why we choose 

this knowledge representation is that it has many advantages, 

especially its easy understanding, availabilities and 

manageability. Meanwhile, today’s most intelligent 

application adopted this kind of knowledge representation. 

Production system contains a set of production(or 

production rules) and facts, production rules use IFTHEN to 

show the knowledge, “attribute/ value” pairs are used to 

means the facts. Production rules has the following form: 

If (<conditon1><condition2><condition 3><condition 

n>) 

Then 

(<conclusion1><conclusion2><conclusio3>---<conclusion 

n>) 

When satisfy the application condition, the data will be 

operated by the order which the conclusion talks. The form 

of the LO in the production can be constants, variables, unit 

groups or tree, figure. 

III. LEARNING MODEL BAESD ON  

INTERACTOIN 

A. Characteristics of active distance-learning model 

design 

Remote active learning model should have the following 

characteristic elements. (1)boundary and exclusivity, which 

means personnel composition in a learning subject is clear, 

whether to learn the subject or not needs to be 

identified;(2)the purpose, learning subject exists as a certain 

purpose, also as the learning interaction that sharing some 

knowledge.(3)Rules, each member must abide by certain 

interactive code of conduct; (4)the obligations, In the 

process of learning, each knowledge sender bears certain 

teaching responsibility;(5)Independence, that is ,every 

member can make choice about their own rights and 

awareness. 

The concept of the “interaction” firstly appeared in 

communication and sociology, in sociology, the definition of 

the interaction is ways and processes of interactive 

influences, effects produced between humans and 

groups.“Interactive” dynamically performed the person’s 

social relations, all kinds of complicated social phenomenon 

are caused by interaction. When thinking about interaction 

in the field of communication[9], we find that the 

interaction’s is more likely to focus on the flow and 

exchange of study information, or relationship among 

transmission elements. 

Simple “teaching” or “learning” in the distance learning 

is difficult to reach the expected goals, the entire learning 



2018 International Conference on Sensor Network and Computer Engineering (ICSNCE 2018) 

35 

 

process should be placed in “relationship” and “interactive” 

so that it can be better achieved. Therefore, interaction 

stands out of the virtual learning community. In the learning 

process, we call the interaction produced from learners 

interaction with others through internet or with their teachers 

the interactive spread[10]. Due to the nature of distance 

education, the teachers and students are often in different 

location ,and time and space are relatively separated in the 

whole education, Therefore,  purposeful and bilateral will 

be very obvious in the transmission of teaching , and 

interaction will be more important in teaching. Not free 

interactive communication often makes it more difficult for 

learners to improve themselves. 

So we can analyze that the process of the knowledge 

learning can be partitioned into several periodic learning 

interactive topic. Each topic still be divided into several 

learning content, send requests to knowledge giver, when 

satisfy the condition of subject and the synchronization time, 

knowledge giver can respectively receive the request signals 

of each learner and not confused. Meanwhile, knowledge 

which knowledge giver send to multiple learners are 

sequentially arranged in a specific time to teach, each learner 

just needs to received it within the specified time so that they 

can distinguish their intellectual theme from confused 

knowledge learning process and receive it down. In distance 

learning network, it is impossible for knowledge giver to 

master all the knowledge, he still needs to learn from the 

other knowledge giver in the virtual network, when two or 

more knowledge giver send knowledge data with same 

subject at the same time, or send data to same knowledge 

learner, then the conflict happens, so we should schedule the 

time of knowledge interaction. 

B. Autonomous learning measurement method 

We should give two definitions in facing with the 

knowledge interaction scheduling: 

 In the network, each learner exists at least one 

interactive subject in a scheduling cycle. 

 Network learners cannot receive knowledge and send 

new knowledge at the same time, and learners can’t 

receive more than one of the knowledge interaction 

either. 

Each learner is learners in certain knowledge subjects, 

However, sometimes they can be the knowledge giver, so it 

increases the complexity of the system modeling. Therefore, 

the problem of scheduling knowledge interaction is to 

assigned knowledge content to learning object according to 

different subjects[11]. According to the knowledge content’s 

interest, we can be divide into two groups, that’s learners 

and the grantors. Assuming that the entire network of 

interest have A1, A2…An, and their knowledge content is 

divide into L1, L2…Lm. We can see it as a graph G(V,E), of 

which the node V(1,2,3 and n) stands for knowledge interest, 

n is for the number of interest, and the edge of E is for the 

collection of the interactions. When node i and j can interact 

with each other, we can think that there exists a undirected 

edge ji
e

,  between two nodes, and
Ee

ji


, , then the node i, 

j is one jump neighbor same with the learners and the 

knowledge givers; if the node i , j cannot directly 

communicate with each other, it means that they don’t have 

the same content of the knowledge, then we need to find out 

the knowledge giver about this knowledge content, 

Therefore ,only through an intermediate node k , can it do 

the communication interaction, that is ,there exist e and e, so 

the node i , j are two jumps neighbor node. It is an adjacency 

matrix. We can describe the network with an symmetric 

binary n n matrix, as: 

)...2,1,(][][ njiCGC
nnij




 

(C
ij

=

edgeshavenode

else

.j.i,.1

0



) 

Another nn matrix: 
 

)...2,1,(][)( njidGD
nmij




 

(D ij =

neighborjumpstwoisnode

else

.,..j i.,1

,0





) 

Considering above two jumps node can realize the 

spatial reuse, it can finally obtain the optimal autonomous 

learning model. The network has a number of M of the 

knowledge content,we can use a M  N binary matrix 

T= NMmi
t


][

: 



2018 International Conference on Sensor Network and Computer Engineering (ICSNCE 2018) 

36 

 

ilearnertocontentknowledge

else

mi
t

....1

0




 

Interactive rate: 

numbertopicknowledge

ilearnertonumbercontentknowledge
i

..

.....
 =

M

t
M

m mi 1
 

The entire network’s interactive rate is: 





N

i

i
N 1

1
 =

MN

1

 

M

m

N

i

mi
t

1 1

 

By two limiting conditions, two learners can send the 

request of the knowledge data at the same time only on the 

condition that two learners are separated by more than two 

jumps, we summarized the above content as the following 

model: 

Min M and Max


 

Limiting conditions: 

C1:
 

M

m mi
t

1    Ni ,1  

C2: 
 ],1[,],,1[,,2 MmjiNjittc

mjmiij


 

C3:
]),1[,,,],,1[,,(,1 MmikkjjiNkjitctc

mjjkmiik


 

The main idea of the interactive process is: Interaction is 

that each learner interact with knowledge giver according to 

the fixed knowledge content point and the chronological 

order on condition that they have common learning subject. 

The learners active behavior allows them to get more study 

time. If there is no learning behavior to release the study 

time, then let the knowledge giver’s interaction have a free 

time, if other learners initiative is more active and don’t have 

the enough to study further, then he can dispute with free 

time and improve more. 

IV. BUILD SIMULATION MODEL BASED ON THE 

OPNET 

Data Packages in OPNET consist of one or more fields, 

package format editor is used to define the package format, 

data packages in each fields has name, size(which length is 

bit), color, type and default values. Using graphical way, it 

can directly display the package’s each fields, The size of 

the fields can exactly reflect the bit length in package. We 

designed three different types of data packages as three 

different types of the data grouping, that is:knowledge 

content data grouping, active application grouping and 

knowledge givers’s reply grouping. 

 

Figure 1.  Learners process model 

This process mainly includes the following states( fig.1), 

each state’s function is introduced as follows: 

 Init state: this state is mainly realized the initialization 

of the simulation data, initial object mainly includes: 

active interaction time slice, interactive time, 

number of the knowledge subject, number of the 

knowledge content etc. After completing the 

initialization of an object, call the interrupt function 

op_intrpt_schedule_self (op_sim_time (), 0), it will 

lead to process’s state transformation from 

producing interruption at the current time to the 

request state only to do the intern’s judgment. 

 Idle state: free wait state, after all the state in 

performing the exit code should them return to this 

state, waiting for the right conditions for the next 

jump, In the exit code of this state will determine 

the current simulation’s interruption type, and 

assign to variable current_intrpt_type, then 

determine whether produce the current node’s 

interactive interruption or not, according to the type 

of the variable current_intrpt_type. If it is,then 

turned to the request or state tx to send the request 

package or data package. 

 Fr_rx state: responsible for identifying the type of the 

data packages that from the network receiver. 



2018 International Conference on Sensor Network and Computer Engineering (ICSNCE 2018) 

37 

 

Extract the corresponding information in the field 

ACK_Data,judge whether current node’s 

interactive application is success or not, if yes, then 

extract the knowledge content’s information that 

contains in the ACK_Data field of the data 

packages, modify the if_need_request to 0, means 

application succeed. So there is no need to send the 

application grouping, then will cycle around in the 

state of idle, fr_rx, tx and fr_src_2, so the 

knowledge giver will send knowledge request in the 

assigned time slices until data queue is empty. 

 Fr_src_1 state: The condition that transform from idle 

state to this state is that source module in the Nodes 

model generates a package, and has not yet been 

assigned the interactive time slice. 

 Request state: The state of request is mainly 

responsible for calculating the current active 

application’s time slice position, determine whether 

to enter their own interaction time, if he entered 

into his interaction time , then fill in the fields of 

information, send secondary intellectual content 

after set up the fields information; if not into its 

time slice or the current time slice has exceeded its 

clock, then compute the next arrival time(absolute 

simulation time ), and set up a break at this time. 

 Fr_src_2 state: like the fr_src_1 status’s function, fill in 

each field of the data package which receives from 

the data source, and insert it into the queue. The 

difference is that when the data source produces 

one package, only when the site’s application for 

time slice succeed, can state transform from idle to 

fr_src_2. 

 Tx state: when the node’s application queue is not 

empty, sent data in this time slot. If the data 

grouping exists in the queue, you need to send the 

interaction request packet to the knowledge giver, 

apply for a number of time slice and use it to send 

data from a queue. Process at this time in idle, fr_rx, 

fr_src_1 and request jump between four states. 

After the success of the application, each frame of 

data within the time slot waiting for their own time 

slot to send data. This process will be idle, fr_rx, 

fr_src_2 and tx , judge the time slice and send the 

data of internship for data to cycle until the system 

closed. 

V. SIMULATION EXPERIMENT 

The scale of the personnel simulation is about 480 

people (Fig.2), the learners are set to the same request rate 

(intervalsetting function is negative expo-numberdistribution 

function),knowledge givers also bears for the same rate 

interact with learners. A learning subject forms one subnet, 

the learners interactwithknowledge giver to complete each 

content of the study, the learners may be as learners involved 

in the other subject learning subnet, a large number of this is 

nested. 

 

Figure 2.  Modeling of simulated network level 

We can see from the diagram that knowledge giver under 

the theme interact with a large number of students, and the 

learners complete the knowledge content learning. From the 

figure 3, we can see that the improvement of the new model 

reflects the active learning behavior of learners. When the 

model adopting the dynamic autonomous learning, learners 

in sending interactive data clearly in dynamic changes (as 

shown in figure 4). Can be seen from the figure b, learners 

after receiving the grantor knowledge content, knowledge 

acquisition of related request change significantly, the 

interesting thing is the simulation data curve downward 

point is far lower than the original model of rate, this should 

be thatthe increase of the initiative will also reduce the 

interest in learning theme. 



2018 International Conference on Sensor Network and Computer Engineering (ICSNCE 2018) 

38 

 

 

Figure 3.  Learners average interactive rate 

 

(a)Knowledge acquisition rate 

 

(b)Knowledge acquisition request rate 

Figure 4.  learners sending interactive data in dynamic changes 

From the above two groups of simulation data, we can 

see that the improvement of the new model not only reflects 

the active learning behavior of learners, but also increase the 

degree to which are not interested in learning content. 

VI. CONCLUSION 

The modern distance learning is bestowed with 

knowledge individuals, related to personal participation and 

learning knower. The enthusiasm of the individual, personal 

exploration, personal opinions are indispensable part of the 

knowledge. Knowledge itself is the intrinsic personal factors, 

the knowledge acquisition process is the knower individuals 

involved in the process of knowledge construction. Dewey 

explains the participants knowledge view as "interaction" of 

the human and the environment. Distance learning requires 

interactivity and initiative to improve, to achieve their 

aptitude, use now study in a variety of formats data, learners 

to master the initiative in the learning process, it is not easy 

to lead to deviation from the learning goals. To this end, on 

the basis of THLMT theory and distributed interactive 

technology, the introduction of active learning factors, put 

forward a kind of active learning rules of interaction, the 

interaction between learners and knowledge imparter is 

studied, under the condition of active learning, completely 

control rules generated by interactive knowledge subject 

data with method of study, established a model of 

autonomous learning. The model not only focus on learners 

"through" process, more emphasis on the learner in the 

experience, get personalized understanding in learning 

enthusiasm. This way of learning value is that learners in 

"facing the complex itself" keep some kind of "seek" 

uncertainty, in the process of "certainty for" construction 

"personal knowledge", in the process of "seeking" 

uncertainty "eagerly seeking knowledge", and will set up the 

learning process in the simulation software, in order to 

dynamically determine the distance learning program and 

related learning strategy, really improve the intelligent and 

personalized distance learning. 

ACKNOWLEDGMENT 

This research was supported by the industrial research 

project of Shaanxi science and technology department 

(nos.2016GY-088) and the educational reform fund of Xi'an 



2018 International Conference on Sensor Network and Computer Engineering (ICSNCE 2018) 

39 

 

Technological University (nos. 15JGZ05, 16JGY25, 

XAGDYJ160208). 

 

REFERENCES 

[1] Lee B, Yoon J,and Lee I,“Learners’ acceptance of e-learning in 
South Korea: Theories and results,” Computers & Education. 53rd 
ed,vol.3, 2009,pp.1320–1329. 

[2] Ahmed H M S. “Hybrid E-Learning Acceptance Model: Learner 
Perceptions,” Decision Sciences the Journal of Innovative Education. 
8th ed,vol.2, 2010,pp.313–346. 

[3] Ikpeze C H,and Boyd F B. “Web-Based Inquiry Learning: 
Facilitating Thoughtful Literacy with WebQuests,” Reading Teacher. 
60th ed,vol.7, 2007,pp.644–654. 

[4] Maudsley G, Williams E M I, and Taylor D C M. “Problem-based 
learning at the receiving end: A ‘mixed methods’ study of junior 
medical students’ perspectives,”Advances in Health Sciences 
Education. 13th ed,vol.4, 2008,pp.435-451. 

[5] Yun L X, sheng Z Y, and li Z S. “Intelligent Internet long-distance 
teaching model based on knowledge,”Journal of Hefei University of 
Technology. 2003,pp.514-519. 

[6] English T, Harrison A L, and Hart A L. “A Distance Learning Model 
in a Physical Therapy Curriculum,” Journal of Allied Health.1998, 
pp.27. 

[7] Jenkins J R P C G J M. “Special education and the regular education 
initiative: basic assumptions,” Exceptional Children. 56th ed,vol.6, 
1990,pp.479-491. 

[8] Harjumaa M, and Oinas-Kukkonen H. “TowardsDeeper 
Understanding of Persuasion in Software and Information Systems,” 
International Conference on Advances in Computer-human 
Interaction. IEEE Computer Society, 2008, pp.200-205. 

[9] Gellman-Danley B, and Fetzner M J. “Asking the really tough 
questions: Policy issues for distance learning.”Online Journal of 
Distance Learning Administration . 1998,pp.49-54 

[10] Qing P, Yun X, and Shou-hong W.“Research and Implementation of 
Animation Technology on Distance Teaching Courseware,” 
Application Research of Computers. 21st ed,vol.2, 2004,pp.144-146. 

[11] Boyle T. “Design principles for authoring dynamic reusable learning 
objects,” Australian Journal of Educational Technology. 19th 
ed,vol.1, 2003,pp.46--58.