Concept mapping for Subject Linking in a WWW

Concept mapping for Subject Linking in a WWW Authoring Tool: My. English. Teacher: Teachers. Site ANNIE 2000 2022/1/16 AI Lab. , Japan Alexandra Cristea Toshio Okamoto 1 and Safia Belkada

Main research goals • free, adaptive, WWW, agent-based, longdistance teaching environment for academic English, for non-native English-speaking academician • => 2 environments: – student learning environment+ – teacher’s courseware design environment. 2022/1/16 AI Lab. , Japan 2

Main system modules - Story DB - Graphics DB - Audio DB - Video DB -Expression DB - Link DB - general student DB private student DBs - . . . Gl. A Story Editor Environment +display Learning Environment +display Teacher user 2022/1/16 PA Student user AI Lab. , Japan 3

Hypotheses: authoring • many course authors, representing different teaching strategies, with different amounts of time to spend in designing the course, different attention to details • ultimate goal is to find the optimal learning path in this heterogeneous net course • system helps teaching in adding semantics 2022/1/16 AI Lab. , Japan 4

Goals of authoring system • Low overhead • Great flexibility • To allow teachers to input as little or as much material as they want, within some reasonable limits 2022/1/16 AI Lab. , Japan 5

Main solutions for course authoring • For easy re-usage => course building bricks • For adding flexibility & semantics => manual & automatic concept mapping 2022/1/16 AI Lab. , Japan 6

course building bricks 2022/1/16 AI Lab. , Japan 7

Texts • smallest building block; • can have corresponding video/audio attached (of dialog, etc. ) • attributes: – main text, a short title, keywords, explanation, patterns to learn, conclusion, and exercises 2022/1/16 AI Lab. , Japan 8

Lessons • One or more texts build a LESSON • attributes (similar to text): – title, keywords, explanation, conclusion, combined exercises 2022/1/16 AI Lab. , Japan 9

Exercises • attributes: – title, keywords, patterns to test, explanation, conclusion This structure allows connecting exercises to rsp. texts and lessons automatically, via relatedness computations 2022/1/16 AI Lab. , Japan 10

Test Points • The teacher should mark TEST POINTS: a text/ lesson where it is necessary to pass a test in order to proceed (~ game theory). • if student wants to jump subjects => 1 test = combination of tests from current level • if student fails => another test generated 2022/1/16 AI Lab. , Japan 11

Course graph 2022/1/16 AI Lab. , Japan 12

Priority and Relatedness Lesson Text 1 Text 2. . . Text i. . . Text n 1 New Lesson Text 1 Text 2. . . Text i. . . Text n 2022/1/16 Lesson Text 1 Text 2. . . Text i. . . Text n w 2 Connections Lesson Text 1 Text 2. . . Text i. . . Text n AI Lab. , Japan Relatedness connections Priority connections Test point 13

Manual concept mapping 2022/1/16 AI Lab. , Japan 14

Method 1: Input of transversal links Current concept 2022/1/16 ? Keyword List: concept 1, concept 2, …. Title List: title 1; title 2; --AI Lab. , Japan 15

Concept mapping application • Linear courseware as opposed to n 2022/1/16 Transversal courseware AI Lab. , Japan 16

Method 2: labeling links subcategory: Simple words Listening practice Listening to sentences Audio material Listening to words Change focus more focus 2022/1/16 AI Lab. , Japan 17

Student advice for ex. Listening practice Audio material 2022/1/16 subcategory: Simple words Listening to words • You can go from here to: • · Subcategory: simple words “listening to words”, or to • · “Audio material”, etc. AI Lab. , Japan 18

Courseware Processing views for easy overview + gradual processing in labeling+linking • => partial views of whole graph, (bird) • 1 concept + its “star”-links (all concepts currently linked to it). (fish) • non-linked concepts: “floating”-concepts 2022/1/16 AI Lab. , Japan 19

Teacher’s responsibilities • only responsible to link own course to existing courses (priority links) • cannot link other teacher’s lessons • is free to link and label whatever subjects desired (relatedness links) • no compulsion to link all texts in the system 2022/1/16 AI Lab. , Japan 20

Automatic concept mapping 2022/1/16 AI Lab. , Japan 21

Subject relatedness weight computation w. A, B 0= 1: teacher’s selection; 0. 5: system’s generation; 0: rest; } (1) where: w. A, B>0: weight between subjects A and B; if w. A, B = 0, the relatedness connection disappears 2022/1/16 AI Lab. , Japan 22

w. A, B t+tconst = w. A, B t + f 1(no. of times connection A, B activated) + + f 2(no. of times connection A, B was accepted, when proposed in relation to unknown subject) + + f 3(no. of times connection A, B was accepted, when proposed in relation to query) + + f 4(no. of times tests related to connection A, B were solved satisfactorily or not) (2) where: (0, 1): forgetting rate; f 1~f 4: linear functions; 2022/1/16 AI Lab. , Japan 23

Conclusions • This paper presents a way of using concept mapping techniques (manual & automatic) and breaking of contents into highly structured, information rich pieces of low granularity, in order to support teachers in creating English language teaching material • The application field of these techniques is larger, could be used not only for other language teaching materials, but also for other subjects (restricted only by the presentation power of the current system version) • This research is just a part of a larger project for academic English teaching 2022/1/16 AI Lab. , Japan 24

Other features of authoring env. • at 1 st system entry, teacher user must register (next fig. ) • for confidentiality, teacher user (like student user) must choose – pseudonym (username) + – password ( to access own profile + courses) 2022/1/16 AI Lab. , Japan 25

2022/1/16 AI Lab. , Japan 26

2022/1/16 AI Lab. , Japan 27

2022/1/16 AI Lab. , Japan 28

• next : selection menu for extra lesson formatting, correction, and viewing. The teachers can change the lesson input as many times as they desire. 2022/1/16 AI Lab. , Japan 29

2022/1/16 AI Lab. , Japan 30

• All teachers can view each of the course items. However, modification is only allowed for the teacher who is the designer of that specific item. 2022/1/16 AI Lab. , Japan 31

2022/1/16 AI Lab. , Japan 32

Teacher/student users and their private screens 2022/1/16 AI Lab. , Japan 33