by, ramnath.k 2000b4a7504 shivakumar. g 2000a7ps047
TRANSCRIPT
By,By,
Ramnath.K 2000B4A7504Ramnath.K 2000B4A7504
Shivakumar. G 2000A7PS047Shivakumar. G 2000A7PS047
Part 1 :Part 1 :What is and why E Leaning?What is and why E Leaning?
Defn : Delivery of formal and informal learning Defn : Delivery of formal and informal learning via use of electronic media like internet, via use of electronic media like internet, intranet, cdrom, dvd, personal organizers intranet, cdrom, dvd, personal organizers etc.etc.
Uses of e-learning & application :Uses of e-learning & application :
• Anytime, anywhere reachAnytime, anywhere reach
• Speed, cost effectiveSpeed, cost effective
• Self-paced learningSelf-paced learning
• Room for innovative multimedia contentRoom for innovative multimedia content
Previous Work on Retrieval of Previous Work on Retrieval of E learning DocumentsE learning Documents
• Content-based retrieval systems for Content-based retrieval systems for personalization of educational videos – personalization of educational videos – Mittal et al.Mittal et al.
•Enhanced Understanding and Retrieval of E-learning Documents through Relational and Conceptual Graphs – Mittal et al.
• Logically characterizing Adaptive Logically characterizing Adaptive Hypermedia Systems – Henze et al.Hypermedia Systems – Henze et al.
Part 2: Personalization & Part 2: Personalization & SummarizationSummarizationSummarization : Automated culling of lecture Summarization : Automated culling of lecture
video clips based on rules strongly built on video clips based on rules strongly built on human perceivable events (like changes in human perceivable events (like changes in sounds and on screen) with the specific sounds and on screen) with the specific intention of reducing the time length of the intention of reducing the time length of the video but with retention of the meaning as a video but with retention of the meaning as a whole.whole.
Personalization : Automated culling of lecture Personalization : Automated culling of lecture video clips based on rules strongly built on video clips based on rules strongly built on human perceivable events (like changes in human perceivable events (like changes in sounds and on screen) with the specific sounds and on screen) with the specific intention of making adaptive lecture material intention of making adaptive lecture material based on the needs of the users from the based on the needs of the users from the same document set available.same document set available.
Uses/ Need of Personalization Uses/ Need of Personalization and Summarization:and Summarization:
•Amount of information available Amount of information available is enormousis enormous
•Users requirements and Users requirements and preferences are varied and preferences are varied and dynamic dynamic
•Faster retrieval of correct and Faster retrieval of correct and precise dataprecise data
Part 3 : The student Part 3 : The student categoriescategories
• Student taking up a courseStudent taking up a course
• Student doing a courseStudent doing a course
• Student before an examStudent before an exam
• Student doing a course revisionStudent doing a course revision
The OthersThe Others
• A Teacher taking the course for the A Teacher taking the course for the first timefirst time
• A Researcher A Researcher
• A casual SurferA casual Surfer
• A General Course SummaryA General Course Summary
Student Taking up a CourseStudent Taking up a Course- - The list of pre-requisite courses has to be highlighted to The list of pre-requisite courses has to be highlighted to
the user. the user.
- The course content, topics covered, lecture hour timings, - The course content, topics covered, lecture hour timings, etc from the introduction video should be coveredetc from the introduction video should be covered
- professor’s teaching skills, his verbal ability, mannerisms - professor’s teaching skills, his verbal ability, mannerisms etc Such a feature is well described by a “best clip”, the etc Such a feature is well described by a “best clip”, the rules for extracting a “best clip” can be used here. rules for extracting a “best clip” can be used here.
- A typical example should be covered (ref to difficult parts - A typical example should be covered (ref to difficult parts of a course) – shows how topics are elucidated in the of a course) – shows how topics are elucidated in the class.class.
- - Other things on the website :Other things on the website : question bank (prev question question bank (prev question papers, lab assignments etc), professor profile and papers, lab assignments etc), professor profile and background , feedback from previous students, background , feedback from previous students, research prospects, latest updates, relevance to other research prospects, latest updates, relevance to other coursescourses
Introductory clip :Introductory clip :Here we propose rules based on which an Here we propose rules based on which an
introductory clip of any video lecture series can be introductory clip of any video lecture series can be identified :identified :
i) Introductory clip always occurs in the first lecture i) Introductory clip always occurs in the first lecture video of any course. Such a lecture has to be video of any course. Such a lecture has to be identified.identified.
ii) Since the introductory part mainly focuses on the ii) Since the introductory part mainly focuses on the professor delineating various parts of the course, professor delineating various parts of the course, the clip mostly has a zoom in view of the professor.the clip mostly has a zoom in view of the professor.
iii) Portion at which the camera zooms out from the iii) Portion at which the camera zooms out from the professor and black board activity commences. The professor and black board activity commences. The professor also talks on the topic while professor also talks on the topic while simultaneously writing it on the black board.simultaneously writing it on the black board.
Best clip :Best clip :• The rules for identifying ‘Best Clip’ The rules for identifying ‘Best Clip’
are :are :Video feature Audio feature
More movement to and fro from the blackboard
More voice from the students
More zooming in/out than normal / Less black board work
Student’s hand raise or talk amongst them (topic introduction).
More student population
More explanation (topic explanation).
Student Taking up a CourseStudent Taking up a Course
Semantic information
Audio Feature Video Feature
Topic: Gives him the topics covered in the course
Keywords like “today’s topic”, “topic is” , “we discuss about” etc
Zoom in/stay zoomed in Slide title
Example: Shows how the professor explains the concepts
Keywords like “for example”, “paradigms”
“Ex:” on board, *.gif file + rules in diff part
Discussion: Shows how the professor interacts with the class
The professor asks questions, Keywords like “lets talk about” , “now we discuss” etc
Zoom out/ zoom in on professor + rules for selecting interactive clip + rules for best clip
Student before an examStudent before an exam • Since the student is already familiar with basic definitions in Since the student is already familiar with basic definitions in
the course, only definitions for difficult parts of the course the course, only definitions for difficult parts of the course should be shown.should be shown.
• Given that the student has little time to revise, examples Given that the student has little time to revise, examples from important as well as from difficult parts of the course from important as well as from difficult parts of the course may be presented to illustrate the main concepts taught in may be presented to illustrate the main concepts taught in the course.the course.
• Theorem proofs from the difficult areas.Theorem proofs from the difficult areas.
• Parts of the course which have more weightage in the tests Parts of the course which have more weightage in the tests have to be stressed.have to be stressed.
• Questions asked by the professor in the class during Questions asked by the professor in the class during discussion.discussion.
• Links to previous questions.Links to previous questions.
Student before an examStudent before an examSemantic information
Audio Feature Video Feature
Definitions: Provides definitions of important topics to the student
Keywords like “We define”, “important definition”, “more weightage”
Zoom in/ word “definition” + topics which student marks
‘difficult’ Examples: Ex of difficult parts are shown to student for review
Keywords like “for
example”, “paradigms” “Ex:” on board, *.gif file + rules of difficult
part
Theorems / proofs/ equations: Difficult proofs & eqs are shown to the student
Keywords like “theorem”, “lemma” , “proof of theorem” etc
Zoom in “theorem” “proof” *.gif file + Topics marked ‘difficult’
Teacher taking up a new Teacher taking up a new coursecourse• Course content is shown.Course content is shown.
- topics that need to be included/excluded- topics that need to be included/excluded
• Examples, to show how the topics are dealtExamples, to show how the topics are dealt
-choose which topic need examples-choose which topic need examples
• Interactive discussions are shown Interactive discussions are shown
-amount of teacher –student interaction -amount of teacher –student interaction neededneeded
-attentiveness of the students-attentiveness of the students
ResearcherResearcher
• Research topics to be shown in areas Research topics to be shown in areas of interestof interest
• Definitions, key terms for topic Definitions, key terms for topic selectedselected
• Links to research papers, ongoing Links to research papers, ongoing research projects in the courseresearch projects in the course
The Proposed E-Learning The Proposed E-Learning System ModelSystem Model
The inputs to the systemThe inputs to the system• Information about it’s users. This input will Information about it’s users. This input will
help the system to decide on the extent help the system to decide on the extent and amount of abridgement needed in the and amount of abridgement needed in the video lectures.video lectures.
• CGPA and grades in relevant courses, the CGPA and grades in relevant courses, the score in the pre-course test, the score in the pre-course test, the projects/reports done by the student in his projects/reports done by the student in his earlier course of study.earlier course of study.
• the Q&A system that shall get the student the Q&A system that shall get the student mark his weak areas are all considered mark his weak areas are all considered before deciding on the extent to which the before deciding on the extent to which the video lectures are cut short. video lectures are cut short.
Data Repository :Data Repository :
This module of our e-learning system is This module of our e-learning system is the usual database part of all such the usual database part of all such systems. systems.
This storehouse caches all the course This storehouse caches all the course material (lecture slides, videos) in its material (lecture slides, videos) in its raw, unprocessed form. raw, unprocessed form.
CGPA, grades in pre-requisite courses, CGPA, grades in pre-requisite courses, scores in previous tests, and faculty scores in previous tests, and faculty profiles also form an integral part of our profiles also form an integral part of our data-repository.data-repository.
The Interactive Response The Interactive Response SystemSystem
Conceptually, the IRS is link between various Conceptually, the IRS is link between various subsystems. subsystems.
• Closed loop Q&A systemClosed loop Q&A system• Type of personalization is decided based on Q&A Type of personalization is decided based on Q&A
and user infoand user info• Weighted user responses + user info from data Weighted user responses + user info from data
repository = extent of personalization repository = extent of personalization
or summarizationor summarization
Hence IRS customizes learning material Hence IRS customizes learning material according to user preferences and requirements.according to user preferences and requirements.
The User Input Interface: The User Input Interface: ConsoleConsole• The user inputs to the system are taken through the The user inputs to the system are taken through the
console, which in case of a personal computer console, which in case of a personal computer comprises of the input devices and the display unit.comprises of the input devices and the display unit.
• Questions like:Questions like:““type of student?” (before exam, researcher etc.)type of student?” (before exam, researcher etc.)““interested topics?” (sorting, trees etc)interested topics?” (sorting, trees etc)““weak areas?” (proofs, DNA decoding etc)weak areas?” (proofs, DNA decoding etc)““time available for this session?” (1hr session)time available for this session?” (1hr session)““research background?” (bayesian nets etc)research background?” (bayesian nets etc)
• Online testsOnline tests• Message boardsMessage boards• Chat sessions Chat sessions
The output The output
• Summarization of lectures – videos, Summarization of lectures – videos, slides , text slides , text
• Personalization of lecturesPersonalization of lectures
• Links to research topicsLinks to research topics
• Links to FAQsLinks to FAQs
Experimental ResultsExperimental Results
- Singapore – MIT lecture videos and Singapore – MIT lecture videos and on videos from Rensselear university. on videos from Rensselear university.
- The personalized videos thus formed The personalized videos thus formed were shown to students and their were shown to students and their feedback on the summary was feedback on the summary was obtained obtained
- Is the shown video continuous? (y/n)Is the shown video continuous? (y/n)- Is the video shown relevant to the Is the video shown relevant to the
category chosen ? category chosen ?
Work completed as part of Work completed as part of paper :paper :• E-learning, its uses, scope and areas of applications E-learning, its uses, scope and areas of applications
were analyzedwere analyzed• How personalization and summarization can help usersHow personalization and summarization can help users• Formed rule sets for different user categoriesFormed rule sets for different user categories• Came up with a model for the whole systemCame up with a model for the whole system• Cut videos and marked cues using the rules developedCut videos and marked cues using the rules developed• Familiarized with usage of video converters, editing Familiarized with usage of video converters, editing
tools like Vegas, Pro media, rm converter, Video cutter tools like Vegas, Pro media, rm converter, Video cutter etc. etc.
• Feedback from students was taken and correctness of Feedback from students was taken and correctness of the rules were verifiedthe rules were verified
• Created web pages for the interface partCreated web pages for the interface part• Came up with a paper on the rules and the system Came up with a paper on the rules and the system
developeddeveloped
Future scopeFuture scope
• Model various categoriesModel various categories
• Add more user typesAdd more user types
• Implementation of rules into a Implementation of rules into a working systemworking system