以模糊推論為基礎之英文測驗與評量系統

(1)

以模糊推論為基礎之英文測驗

與評量系統

國立臺南大學黃國禎

數位學習科技系教授

資訊教育研究所所長

理工學院院長

(2)

Introduction (1/3)



Background



English language has become a

international language



English learning has always been

substantially difficult for ESL/EFL students



English tenses

play an important part in

account for explaining the temporal

background of English sentence



Applying AI technologies

to assist English

learning attracts researchers from various

fields

(3)

Introduction (2/3)



Motivation



Conventional examinations usually give students

nothing but their scores, which is not really help

them to improve their studies



Hence, we propose

ET-DES

(

E

nglish

T

enses

D

iagnosis

E

xpert

S

ystem).



ET-DES is able to

identify English learners’deficiency of tenses on English

verb tenses

provide learners with individualized learning suggestions

(4)

Research Architecture and Design

Classification of Tenses in English Grammar



Present Simple



Past Simple



Future Simple



Present

Progressive



Past Progressive



Future Progressive



Present Perfect



Past Perfect



Future Perfect



Present Perfect Progressive



Past Perfect Progressive



Future Perfect Progressive



Time Adverbs

(5)

Research Architecture and Design

Design of Learning Portfolio

ET-DES ET-DES Learner (id) Learner (id) EnterTime EnterTime Category (id,name) Category

(id,name) ExamExam ExitTimeExitTime

Concept+ (name) Concept+ (name) Concept_EnterTime Concept_EnterTime Category_EnterTime Category_EnterTime Exam_EnterTime

Exam_EnterTime Exam_ExitTime_{Exam_ExitTime} TestItem+

(num) TestItem+ (num) ConcentrationWindow* (num) ConcentrationWindow* (num) PopTime PopTime Concept_ExitTime Concept_ExitTime Category_ExitTime Category_ExitTime RelevantConcept

RelevantConcept LearnerAnswer_{LearnerAnswer} StandardAnswer_{StandardAnswer}

ResponseTime

(6)

Research Architecture and Design

Identification of Learner’

s Performances

Map performance into MFs to get the membership degrees for linguistic terms

Performance (e.g. concept understand

degree, concentration, average page browsing time)

Fuzzicatioation

Inference Engine

Fuzzy Rule Base

Inference Engine

Fuzzy Rule Base

Learner’s relative rating of performance

Learner’s relative rating of performance Learning Suggestion Base Learning Suggestion Base match recommend as feedback input Expert System Learning Group Individual Learner (IF … THEN …)

(IF … THEN …)_{(IF … THEN …)} (IF … THEN …) Center Average Defuzzifier Center Average Defuzzifier Center Average Defuzzifier Center Average Defuzzifier result Fuzzy MFs Fuzzy MFs Fuzzy Rulebase Fuzzy Rulebase High Medium Low High Medium Low L M H L M H Fuzzy Inference Defuzzification

(7)

Identification of Learner’

s Performances

Definition of the Linguistic Variables (1/3)

Linguistic Variables Definitions Linguistic Terms Input

LAsc(Si,Cj) Learner Si’s individual learning achievement toward

a certain concept Cj.

Low, Medium, High ALAc(Cj) The learning group’s average learning achievement

toward a certain concept Cj.

Low, Medium, High concentration(Si) Learner Si’s response rating toward the concentration

windows during the learning progress.

Grade 1~5 pageBT(Si) Learner Si’s average browsing time to the learning

materials.

Short, Moderate, Long

Output

RLAsc(Si,Cj) The Si’s relative learning achievement toward a

certain concept Cjcompared with the learning group.

Grade 1~5 Patience(Si) Learner Si’s patience rating performed during the

learning progress.

(8)

Identification of Learner’

s

Performances

Relations among the Linguistic Variables (2/3)

LAsc(S_i,C_j) LAsc(S_i,C_j) ALAc(C_j) ALAc(C_j) RLAsc(S_i,C_j) RLAsc(S_i,C_j) concentration(S_i) concentration(S_i) pageBT(S_i) pageBT(S_i) patience(S_i) patience(S_i) (3) (3) (5) (5) (3) (5)

(9)

9

Identification of Learner’

s

Performances

Relations among the Linguistic Variables (3/3)



concentration(S

_i

)



Whether a CW is valid or not is defined

as:

CW

(S

_i

)

= 1

if the student responses to the pop-up

window on time



patience(S

_i

)

1 ( ) ( ) N j i j i cw S concentration S N  



nStu S pageBT pageABT nStu i i



  1 ) (

Total time for to browse the Learning materials ( )

Number of learning materials that browsed

i i i S pageBT S S 

(10)

Identification of Learner’

s

Performances

Definition of the Membership Functions (1/3)

(3.5) 

LAsc(S

_i

,C

_j

) / ALAc(C

_j

)

1.0 0.0 High x Medium Low 1.0 degree of compatibility 0.5 0.6 0.8 0.2 0.4 Z Tri S      High x S Medium x Tri Low x Z is Term Linguistic ; ) 8 . 0 , 6 . 0 ; ( is Term Linguistic ; ) 8 . 0 , 6 . 0 , 4 . 0 ; ( is Term Linguistic ; ) 6 . 0 , 4 . 0 ; (

(11)

Identification of Learner’

s

Performances

Definition of the Membership Functions (2/3)

(3.6)



RLAsc(S

_i

,C

_j

) / concentration(S

_i

) / patience(S

_i

)

1/6 1.0 0.0 Grade 5 x Grade 4 Grade 3 Grade 2 Grade 1 2/6 3/6 4/6 5/6 6/6 degree of compatibility

0.5 _Z _Tri _Tri _Tri _S

         5 is Term Linguistic ; ) 6 / 6 , 6 / 5 , 6 / 4 ; ( 4 is Term Linguistic ; ) 6 / 5 , 6 / 4 , 6 / 3 ; ( 3 is Term Linguistic ; ) 6 / 4 , 6 / 3 , 6 / 2 ; ( 2 is Term Linguistic ; ) 6 / 3 , 6 / 2 , 6 / 1 ; ( 1 is Term Linguistic ; ) 6 / 2 , 6 / 1 , 0 ; ( Grade x S Grade x Tri Grade x Tri Grade x Tri Grade x Z

(12)

Identification of Learner’

s

Performances

Definition of the Membership Functions (3/3)

(3.7) 

pageBT(S

_i

)

1.0 0.0 Long x Moderate Short degree of compatibility 0.5 2 pageABT 2 3 pageABT pageABT 2pageABT Z Tri S      Long pageABT pageABT x S Moderate pageABT pageABT pageABT x Tri Short pageABT pageABT x Z is Term Linguistic ; ) 2 / 3 , ; ( is Term Linguistic ; ) 2 / 3 , , 2 / ; ( is Term Linguistic ; ) , 2 / ; (

(13)

13

Identification of Learner’

s

Performances

Definition of the Fuzzy Rules (1/4)

R1: IF LAsc(Si,Cj)= Low AND ALAc(Cj)= Low

THEN RLAsc(Si,Cj)= Grade 3

R2: IF LAsc(Si,Cj)= Low AND ALAc(Cj)= Medium THEN RLAsc(Si,Cj)= Grade 2

R3: IF LAsc(Si,Cj)= Low AND ALAc(Cj)= High THEN RLAsc(Si,Cj)= Grade 1

R4: IF LAsc(Si,Cj)= Medium AND ALAc(Cj)= Low THEN RLAsc(Si,Cj)= Grade 4

R5: IF LAsc(Si,Cj)= Medium AND ALAc(Cj)= Medium THEN RLAsc(Si,Cj)= Grade 3

R6: IF LAsc(Si,Cj)= Medium AND ALAc(Cj)= High THEN RLAsc(Si,Cj)= Grade 2

R7: IF LAsc(Si,Cj)= High AND ALAc(Cj)= Low THEN RLAsc(Si,Cj)= Grade 5

R8: IF LAsc(Si,Cj)= High AND ALAc(Cj)= Medium THEN RLAsc(Si,Cj)= Grade 4

R9: IF LAsc(Si,Cj)= High AND ALAc(Cj)= High THEN RLAsc(Si,Cj)= Grade 3

(14)

Identification of Learner’

s

Performances

Definition of the Fuzzy Rules (2/4)

R10: IF concentration(Si)= Grade 1 AND pageBT(Si)= Short THEN patience(Si)= Grade 1

R11: IF concentration(Si)= Grade 1 AND pageBT(Si)= Moderate THEN patience(Si)= Grade 2

R12: IF concentration(Si)= Grade 1 AND pageBT(Si)= Long THEN patience(Si)= Grade 3

(15)

Identification of Learner’

s

Performances

Definition of the Fuzzy Rules (3/4)

(16)

Identification of Learner’

s

Performances

Definition of the Fuzzy Rules (4/4)

(17)

17

Identification of Learner’

s

Performances

Individual Learning Suggestions (1/3)



Concept relative learning achievement

Grade 1:

“You haven’t reached this concept’s studying goal, you do not understand the topic, please work harder on it.”

Grade 2:

“You have some understanding of the concept, but still not enough. Please work harder on it.”

Grade 3:

“You understand the concept, but you still are a distance from your goals. Please spend a little more time on the subject.” Grade 4:

“Congratulation! You have reached the concept’s studying goals.”

Grade 5:

“Congratulation! You have not only reached the studying goals but also entirely understood and used the subject.”

(18)

Identification of Learner’

s

Performances

Individual Learning Suggestions (2/3)



Concentration

Grade 1:

“You concentration level during the material learning was

mediocre. Please find the reasons and try to do better if you want better results.”

Grade 2:

“You did not reach the average concentration level, please try harder on the working to improve next time.”

Grade 3:

“You have reached the standard level of concentration. If you can try harder, the subjects won’t be that difficult anymore.”

Grade 4:

“Your concentration level was higher than usual. Is this your secret weapon of studying? Go on and use it well!”

Grade 5:

“You have shown to be perfectly concentrated during the material learning. Congratulations, you are a part of the intellectuals! Hold on to it!”

(19)

Identification of Learner’

s

Performances

Individual Learning Suggestions (3/3)



Patience

Grade 1:

“Your patience level needs to be improved. You can ask for help if needed.”

Grade 2:

“You patience level is lower than average.” Grade 3:

“Your patience level is around average.” Grade 4:

“Your patience level is above average.” Grade 5:

(20)

Identification of Learner’

s

Performances

Illustrative Example of the Fuzzy Inference (1/12)

Concept C_j hkj C1 C2 C3 C4 C5 C6 C7 C8 C9 C10 NCk Q1 0 0 1 0 1 0 0 1 0 0 3 Q₂ 1 0 1 0 1 1 0 0 1 0 5 Q3 1 0 0 1 0 0 1 0 0 1 4 Q₄ 1 0 0 0 1 0 0 1 0 1 4 Q₅ 0 1 1 1 1 0 0 1 1 0 6 Q6 0 0 0 1 0 0 0 1 0 1 3 Q₇ 1 0 0 0 0 1 1 1 0 0 4 Q₈ 0 0 0 1 1 0 0 1 0 0 3 Q₉ 1 0 1 0 0 1 0 0 1 1 5 Q₁₀ 0 1 1 0 0 0 1 0 1 1 5 Test item Q_k T(Cj) 5 2 5 4 5 3 3 6 4 5

(21)

Identification of Learner’

s

Performances

Illustrative Example of the Fuzzy Inference (2/12)

Test item Q_k eik Q1 Q2 Q3 Q4 Q5 Q6 Q7 Q8 Q9 Q10 NEi S1 0 1 0 1 0 0 0 1 0 0 3 S2 0 0 0 1 0 1 0 1 0 1 4 S3 0 0 1 0 0 0 0 0 1 0 2 S4 1 1 0 0 1 0 0 0 0 1 4 Learneri S₅ 0 0 0 0 0 0 0 1 0 0 1 T(Qk) 1 2 1 2 1 1 0 3 1 2

Learner-Test item Relationship Table (LTRT)

Learners’information during the learning progress

Learneri S1 S2 S3 S4 S5

Total pages browsed 9 13 9 10 11

Total browse time(sec.) 331 560 267 379 492

Total CWs 9 16 7 10 14

(22)

Identification of Learner’

s

Performances

Illustrative Example of the Fuzzy Inference (3/12)

(3.15) (3.16) (3.17) nE(Si,Cj)/LAsc(Si,Cj) 

LAsc(S

₁

, C

₈

) (1/6)

) 1 ( ) , ( 



nTI  _ik  k kij j i C h e S nE ) ( ) , ( 1 ) , ( j j i j i C T C S nE C S LAsc   Concept Cj C1 C2 C3 C4 C5 C6 C7 C8 C9 C10 S₁ 2/0.6 0/1.0 1/0.8 1/0.75 3/0.4 1/0.67 0/1.0 2/0.67 1/0.75 1/0.8 S2 1/0.8 1/0.5 1/0.8 2/0.5 2/0.6 0/1.0 1/0.67 3/0.5 1/0.75 3/0.4 S3 2/0.6 0/1.0 1/0.8 1/0.75 0/1.0 1/0.67 1/0.67 0/1.0 1/0.75 2/0.6 S₄ 1/0.8 2/0.0 4/0.2 1/0.75 3/0.4 1/0.67 1/0.67 2/0.67 3/0.25 1/0.8 Learner_i S₅ 0/1.0 0/1.0 0/1.0 1/0.75 1/0.8 0/1.0 0/1.0 1/0.83 0/1.0 0/1.0 ALAc(Cj) 0.76 0.7 0.72 0.7 0.64 0.8 0.8 0.73 0.7 0.72 ) ( ) , ( ) ( j j i j C T C S LAsc C ALAc 



(23)

Identification of Learner’

s

Performances

Illustrative Example of the Fuzzy Inference (4/12)

Take concept C₈ for example. If we would like to know the relative

learning achievement degree of leaner S₁, we can map the value of

LAsc(S₁,C₈) into the membership function:



LAsc(S

₁

, C

₈

) (2/6)













High

S

Medium

Tri

Low

Z

is

Term

Linguistic

;

35 .

0 )

8 .

0 ,

6 .

0 ;

67 .

0 (

is

Term

Linguistic

;

65 .

0 )

8 .

0 ,

6 .

0 ,

4 .

0 ;

67 .

0 (

is

Term

Linguistic

;

0 )

6 .

0 ,

4 .

0 ;

67 .

0 (

(24)

Identification of Learner’

s

Performances

Illustrative Example of the Fuzzy Inference (5/12)

For the learning group, the average concept achievement degree can be known by mapping ALAc(C₈) to the membership function:













"

;

65 .

0 )

8 .

0 ,

6 .

0 ;

73 .

0 (

"

;

35 .

0 )

8 .

0 ,

6 .

0 ,

4 .

0 ;

73 .

0 (

"

;

0 )

6 .

0 ,

4 .

0 ;

73 .

0 (

High

Term

Linguistic

S

Medium

Term

Linguistic

Tri

Low

Term

Linguistic

Z



ALAc(C

₈

) (3/6)

(25)

Identification of Learner’

s

Performances

Illustrative Example of the Fuzzy Inference (6/12)

According to the membership degrees, there are four rules to be triggered: R5: min{0.65, 0.35}= 0.35  Grade 3 R6: min{0.65, 0.65}= 0.65  Grade 2 R8: min{0.35, 0.35}= 0.35  Grade 4 R9: min{0.35, 0.65}= 0.35  Grade 3 

LAsc(S

₁

, C

₈

) (4/6)

(26)

Identification of Learner’

s

Performances

Illustrative Example of the Fuzzy Inference (7/12)

Consequently, the output of fuzzy inference can be

acquired –the membership degree of the relative

achievement degree of S

₁

toward concept C

₈

.

          " 5 " ; 0 " 4 " ; 35 . 0 " 3 " ; 35 . 0 } 35 . 0 , 35 . 0 max{ " 2 " ; 65 . 0 " 1 " ; 0 Grade is Term Linguistic Grade is Term Linguistic Grade is Term Linguistic Grade is Term Linguistic Grade is Term Linguistic 

LAsc(S

₁

, C

₈

) (5/6)

(27)

Identification of Learner’

s

Performances

Illustrative Example of the Fuzzy Inference (8/12)

For the output linguistic variable RLAsc(S₁,C₈), we take the linguistic term Grade 2 with the largest membership degree of .65

Render the learner with the learning suggestion “your relative achievement degree toward concept Past Perfect is Grade 2,

between the Grade 1 (the worst) and Grade 5 (the best).”



Defuzzification



LAsc(S

₁

, C

₈

) (6/6)

463 .

0

35 .

0

35 .

0

65 .

0

35 .

0

35 .

0

65 .

0 (

₆4 6 3 6 2 ~ ~ ~ ~ ~ 1211 ~ 6 4 ~ 6 3 ~ 6 2 ~ 121 ~





































l E l D l C l B l A l E l D l C l B l A



(28)

0.0 x x 0.0 x 1.0 0.0 x y y y y 1/6 z Grade 3 2/6 3/6 4/6 5/6 6/6 1/6 z Grade 2 2/6 3/6 4/6 5/6 6/6 1/6 z Grade 4 2/6 3/6 4/6 5/6 6/6 1/6 z Grade 3 2/6 3/6 4/6 5/6 6/6 1/6 1.0 z 2/6 3/6 4/6 5/6 6/6 0.6 0.8 0.2 0.4 0.6 0.2 0.4 0.8 0.6 0.2 0.4 0.8 0.2 0.4 0.6 0.8 0.6 0.8 0.2 0.4 0.2 0.4 0.6 0.8 0.6 0.8 0.2 0.4 0.2 0.4 0.6 0.8 0.65 LAsc(S₁,C₈)=0.67 LAc(C8)=0.73 0.35 Min Min Min Min 1.0 0.35 0.0 1.0 0.65 1.0 0.65 0.35 0.0 0.65 0.35

Crisp input values

Fuzzy output

(29)

Identification of Learner’

s

Performances

Illustrative Example of the Fuzzy Inference (9/12)

By mapping concentration(S₁) = 7/9 = 0.778 into the membership function, we have: 

concentration(S

₁

)

              5 is Term Linguistic ; 667 . 0 ) 6 / 6 , 6 / 5 , 6 / 4 ; 9 / 7 ( 4 is Term Linguistic ; 333 . 0 ) 6 / 5 , 6 / 4 , 6 / 3 ; 9 / 7 ( 3 is Term Linguistic ; 0 ) 6 / 4 , 6 / 3 , 6 / 2 ; 9 / 7 ( 2 is Term Linguistic ; 0 ) 6 / 3 , 6 / 2 , 6 / 1 ; 9 / 7 ( 1 is Term Linguistic ; 0 ) 6 / 2 , 6 / 1 , 0 ; 9 / 7 ( Grade S Grade Tri Grade Tri Grade Tri Grade Z

Learning Suggestion of Grade 5: “You did concentrate during the material learning. Congratulations! You have already possessed the prerequisite for learning. Hold on this and

(30)

Identification of Learner’

s Performances

Illustrative Example of the Fuzzy Inference (10/12)

pageABT= = 39.02

By mapping into the MF, we have:

11 10 9 13 9 492 379 267 560 331         78 . 36 9 331 ) (S₁   pageBT 

Patience(S

₁

)

        Long S Moderate Tri Short Z is Term Linguistic ; 0 ) 53 . 58 , 02 . 39 ; 78 . 36 ( is Term Linguistic ; 885 . 0 ) 53 . 58 , 02 . 39 , 51 . 19 ; 78 . 36 ( is Term Linguistic ; 115 . 0 ) 02 . 39 , 51 . 19 ; 78 . 36 (

(31)

Identification of Learner’

s

Performances

Illustrative Example of the Fuzzy Inference

(11/12)

According to the membership degree, there are four rules to be triggered: R₁₉: min{0.333, 0.115}= 0.115  Grade 2 R₂₀: min{0.333, 0.885 }= 0.333  Grade 3 R₂₂: min{0.667, 0.115}= 0.115  Grade 2 R₂₃: min{0.667, 0.885}= 0.667  Grade 3 

Patience(S

₁

)

(32)

Identification of Learner’

s Performances

Illustrative Example of the Fuzzy Inference (12/12)

The output of the fuzzy inference can be acquired, which represents the membership degree of the learner S₁’s patience during the on-line material learning process



Patience(S

₁

)

           5 is Term Linguistic ; 0 4 is Term Linguistic ; 0 3 is Term Linguistic ; 667 . 0 } 667 . 0 , 333 . 0 max{ 2 is Term Linguistic ; 115 . 0 } 115 . 0 , 115 . 0 max{ 1 is Term Linguistic ; 0 Grade Grade Grade Grade Grade

Learning Suggestion of Grade 3: “Your patience has already approached to the normal level, but you still need to enhance it

(33)

Implementation of ET-DES

System Architecture

System Parameter Database Browsing tense materials

Receiving on-line testing

Inspecting personal learning portfolio Inspecting fuzzy analysis result and

Individualized learning suggestions Learners

Teachers System Administrator

Setting the relations between concepts and test items Inspecting the relations between

concepts and test items Inspecting learners’learning

performances

Tense Materials

Test item Database Setting quiz parameters

Setting system parameters

Learning Portfolio XML User Database Fuzzy Interface In te rn e t N e tw o rk in g In te rfa ce

Setting the learning suggestions

KB of LSA Inference Engine KB of LOD KB of LSA Inference Engine KB of LOD DBMS (SQL Server)

Browsing the testing results

(34)

Implementation of ET-DES

Process flow of Learning Portfolio

Learner ET-DES ET-DES operates _{<?xml version=…} XML Document <?xml version=… XML Document generates combine <?xml version=… XML Document <?xml version=… XML Document JDOM

Parse and Compute

JDOM

Parse and Compute

be parsed by

Avg score=… Avg ConRR01=…

Avg score=…

Avg ConRR01=… DRAMADRAMA

results Individual’s learning view

<?xml version=… XML Document <?xml version=… XML Document <?xml version=… XML Document <?xml version=… XML Document <?xml version=… XML Document <?xml version=… XML Document XML document base feedback outputs inputs

Diagnosis result & Remedial Instruction

(35)

Implementation of ET-DES

(36)

Implementation of ET-DES

(37)

Implementation of ET-DES

(38)

Implementation of ET-DES

Teacher Interface (2/6) –Setting the

(39)

Implementation of ET-DES

Teacher Interface (3/6) –Inspecting the

Relationship between Concepts and Test Items

(40)

Implementation of ET-DES

Teacher Interface (4/6) –Inspecting the

Learners’Performances

Average performance of the overall learning

group

The inspected learner’s individual performance

(41)

Implementation of ET-DES

Teacher Interface (5/6) –Inspecting the

Learners’Performances

Average performance of the overall learning

group

The local view of the range of learners’

(42)

Implementation of ET-DES

Teacher Interface (6/6) –Setting the Learning

Suggestions

Press this button to recover the default setting

(43)

Implementation of ET-DES

Learner Interface (1/10)

Brief introduction to ET-DES

Operating flow of the system The Materials

(44)

Implementation of ET-DES

Learner Interface (2/10) –Browsing learning

materials of English Tenses

(45)

Implementation of ET-DES

Learner Interface (3/10) –Taking the Tense

Testing

(46)

Implementation of ET-DES

Learner Interface (4/10) –Browsing the Quiz

Result

The original question The explanation for the answer

(47)

Implementation of ET-DES

Learner Interface (5/10)–Browsing Personal

Learning Portfolio

(48)

Implementation of ET-DES

Learner Interface (6/10) –Inspecting the

(49)

Implementation of ET-DES

Learner Interface (7/10) –Inspecting the

(50)

Implementation of ET-DES

Learner Interface (8/10) –Inspecting the

(51)

Implementation of ET-DES

Learner Interface (9/10) –Inspecting the

(52)

Implementation of ET-DES

Learner Interface (10/10) –Inspecting the

Analysis Result and Learning Suggestions(5/5)

(53)

Experiment and Analysis

Experiment Design (1/4)



Participants



257 undergraduate students

of six Information

Network literacy classes of NCNU



223 (86.77%) data

of subjects were completed

successfully, legible and thus able to be used for

the analysis



Parameters involved



Interval time of pop-up concentration window is

set at

16 seconds



Auto-close time of pop-up concentration window is

set at

4 seconds

(54)

Experiment and Analysis

Experiment Design (2/4)



Procedure

Instructor explains the experiment purposes and the system functions

Instructor shows the system operation flow step by step

Subjects operate the system

Subjects are asked to fill in second instrument for gathering their comment

END

Subjects fill in individual data and pre-use instrument

START

(55)

Experiment and Analysis

Experiment Design (3/4)



Instruments –the first one

(Appendix C)

 subject’s basic individual data  investigation of the subjects’

English level and a survey on

whether subjects have really

understood the characteristics and functions of the system

(56)

Experiment and Analysis

Experiment Design (4/4)



Instruments –the second

(Appendix E)

 System (system usefulness and

ease-of-use testing)

 Material (readability test)

 Test items of the quiz (difficulty test)  Diagnosis result (helpfulness test)

(57)

Experiment and Analysis

Background of the Participants (1/2)

Opinion Number of Participants Percentage Male 100 44.84 Female 123 55.16 Sex Total 223 100 College of Humanities 112 -College of Management 28 -College* College of Technology 63 -Freshman 194 87 Sophomore 26 11.66 Grade Junior 3 1.35

(58)

Experiment and Analysis

Background of the Participants (2/2)

Fi (3, 1%) Ec (10, 4%) CSIE (2, 1%) EE (18, 8%) AC (27, 12%) CE (16, 7%) IB (15, 7%) IM ( 20, 9%) PPA (12, 5%) CE (8, 4%) Hi (9, 4%) SPSW (18, 8%) EPA (17, 8%) EL (28, 13%) CL (20, 9%) CL CL (Chinese Language) EL EL (English Language)

SPSW SPSW (Social Policy and Social Work) Hi Hi (History)

CE CE (Comparative Education)

PPA PPA (Public Policy and Administration) EPA EPA (Educational Policy and Administration)

IB IB (International Business) IM IM (Information Management) Ec Ec (Economics)

Fi Fi (Finance)

CSIE CSIE (Computer Science and Information Engineering) EE EE (Electrical Engineering)

CE CE (Civil Engineering) AC AC (Applied Chemistry)

(59)

Experiment and Analysis

Subject’

s Self-cognition (1/2)

Facet question Opinion Frequency Percentage Less than 3 years 41 18.39

3~6 years 55 24.66 6~9 years 101 45.29 9~12years 10 4.48 How long have you been learning

English?

More than 12 years 16 7.17 Poor 33 14.8 Not Good 62 27.8 Mediocre 82 36.77 Good 31 13.9 How would you rate your English

level? Excellent 15 6.73 Detest 26 11.66 Dislike 61 27.35 Neutral 59 26.46 Like 68 30.49 E ng lis h L ea rn in g E xp er ie nc e

How much interest do you have toward learning English?

(60)

Experiment and Analysis

Subject’

s Self-cognition (2/2)

Facet question Opinion Frequency Percentage Strongly Disagree 12 5.38

Disagree 9 4.04 Neutral 113 50.67

Agree 77 34.53

After the demonstration, I already understand the operating flow of the system and its functions?

Strongly Agree 12 5.38 Strongly Disagree 11 4.93 Disagree 6 2.69 Neutral 130 58.3 Agree 68 30.5 S ys te m F un cti on an d O bj ec tiv e

After the demonstration, I’m more interested to this to-be-used system?

(61)

Experiment and Analysis

Statistics of each Facet

Frequency* (Percentage**) 576 (64.57%) 1.05 47 0.7647 0.153 9 3.6872 146 (16.4% ) 430 (48.2%) 229 (25.7% ) 65 (7.3% ) 22 (2.5%) DIAGNO SIS RESULT 175 (39.24%) 0.98 27 0.8648 0.713 4 3.083 30 (6.7%) 145 (32.5%) 135 (30.3% ) 104 (23.3 %) 32 (7.2%) QUIZ 235 (52.69%) 0.93 1 0.7617 0.219 8 3.4662 38 (8.5%) 197 (44.2%) 155 (34.8% ) 48 (10.8 %) 8 (1.8%) MATERIA LS 614 (68.83%) 1.09 31 0.7605 0.186 7 3.704 123* (13.8% **) 491 (55%) 198 (22.2% ) 51 (5.7% ) 29 (3.3%) SYSTEM Satisfacti on(4+5) Max Min S.D. Mean 5 4 3 2 1 Option Facet

(62)

Experiment and Analysis

The Efficacy of the Diagnosis Results

0% 10% 20% 30% 40% 50% 60% 70% Strongly Disagree

Disagree Neutral Agree Strongly Agree

Learning Achievement Degree Concentration Degree

(63)

Directions for Future Research



Defining more precise parameters for the

Membership Functions



Providing more accurate concept relationship

descriptions



Adopting advanced algorithms and methodologies



Launching more complete categories of English

grammar



以模糊推論為基礎之英文測驗 與評量系統