group 1poster 05 - school of information technology · 2013-07-01 · 'cv_kappa',...

1
Powered by OpenedX Django cache management results Tool Used :- Parameters:- Results:- Jmeter (HTTP Request Sampler) ? Number of Users : 10 ? Ramp-up period : 20 ? Loop-Count : 5 Future Scope: ? User Access Control ? Refined Search ? Uploading srt files from cms to YouTube directly ? Course Feedback ? Chat ? Videos from own server rather than YouTube Conclusion:- The experimental results shows that Memcached perform better than other caching system and performance under Local Memory caching is least of all. The order is shown below: Memcached > Filesystem Caching > Database Caching > Local Memory Caching Type of Caching Average response time (in ms) Deviation Local Memory Caching 168 196 Memcached 93 8 Database Caching 136 32 Filesystem Caching 121 23 iitbombayX Ease (Enhanced AI Scoring Engine) Training and Prediction Introduction: A massive open online course (MOOC) is an online education system, providing various courses, aimed at large-scale interactive participation and open access, via the web. Apart from the traditional course materials such as videos, study materials, and problem sets, MOOCs also aims to provide interactive user forums, that helps to build a community for the students and professors. MOOC is a recent development in the education system for making distant education possible and effective. Development of edX aims to implement all the standard features of edX MOOC and also to enhance it. We have developed a MOOC platform for IIT Bombay based on edX platform. This MOOC is called iitbombayX. Features: ? Interactive video lectures with subtitles and indexing of subtitles. ? Study materials like books, notes, cheat sheets, etc. ? Online tests of different types like video embedded quiz, practice sessions, midterm exam, final exam, etc. ? Set grading policy for the course. ? Calendar based schedule. ? Discussion Forums ? Assignment upload option for students (in open response problems). ? Wiki edits for implementing collaborative learning. ? Contacting authors through email. ? Progress report and other kinds of embedded analytics. ? Different kinds of assessment systems for submitted assignments (subjective). This includes: ? Peer Grading. ? Self Grading. ? Staff Grading. ? Machine Grading. ? Email and Notification facilities for registered students ? Provision of certification. ? Registration/Dropping of a course. In addition to the above mentioned features already offered by edx, we have added the following new features: ? Upload videos directly on youtube. ? Search by keyword. ? Course Ratings. ? Recommended courses. ? Number of students enrolled in a course. Grading Model File with essays and scores ModelCreator. read_in_data Create Model Returns two lists: List of scores and corresponding essays Contribution: The team successfully integrated the edX-platform with other necessary modules for comments, forums and grading. We have created ER Diagram for edX database which could be used by others for future developments and enhancements. We have performed experiments for testing django cache management and ease (AI) grading. Ease repository allows anyone to use machine learning based automated classification which can work on both free text (essays, content, etc), and on numeric values. Dumps data to a json file Selects algorithm Creates an essay set object and add essays to it Extract features Get classifier Get cv error Essay and score lists Returns results = {'errors', 'success', 'cv_kappa', 'cv_mean_absolute_er ror', 'feature_ext', 'classifier', 'algorithm', 'score','text', ‘prompt'} *cv=cross validation Contributors: Udayan Jain Sonam Gupta Gagan Khanijau Suhanth Boddu Piyush Tiwari Rupal Sharma Surabhi Prasad Zeel Shah Summer Internship 2013(9th May to 7th July) email: [email protected] design by: Bhairav Fundamental Research Group, Dept of CSE, IIT Bombay NME-ICT MHRD

Upload: others

Post on 20-May-2020

1 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: group 1poster 05 - School of Information Technology · 2013-07-01 · 'cv_kappa', 'cv_mean_absolute_er ror', 'feature_ext', 'classifier', 'algorithm', 'score','text', ‘prompt'}

Powered by OpenedX

Django cache management results

Tool Used :- Parameters:-

Results:-

Jmeter (HTTP Request Sampler)

? Number of Users : 10? Ramp-up period : 20? Loop-Count : 5

Future Scope:

? User Access Control? Refined Search? Uploading srt files from cms to YouTube directly? Course Feedback ? Chat? Videos from own server rather than YouTube

Conclusion:- The experimental results shows that Memcached perform better than other caching system and performance under Local Memory caching is least of all. The order is shown below:Memcached > Filesystem Caching > Database Caching > Local Memory Caching

Type of Caching Average response time (in ms)

Deviation

Local Memory Caching 168 196

Memcached 93 8

Database Caching 136 32

Filesystem Caching 121 23

iitbombayX

Ease (Enhanced AI Scoring Engine)

Training and Prediction

Introduction:

A massive open online course (MOOC) is an online education system, providing various courses, aimed at large-scale interactive participation and open access, via the web. Apart from the traditional course materials such as videos, study materials, and problem sets, MOOCs also aims to provide interactive user forums, that helps to build a community for the students and professors. MOOC is a recent development in the education system for making distant education possible and effective.

Development of edX aims to implement all the standard features of edX MOOC and also to enhance it. We have d e v e l o p e d a M O O C platform for IIT Bombay based on edX platform. This MOOC is called iitbombayX.

Features:? Interactive video lectures with subtitles and indexing of subtitles.? Study materials like books, notes, cheat sheets, etc.

? Online tests of different types like video embedded quiz, practice sessions, midterm exam, final exam, etc.? Set grading policy for the course.? Calendar based schedule.? Discussion Forums? Assignment upload option for students (in open response problems).? Wiki edits for implementing collaborative learning.? Contacting authors through email.? Progress report and other kinds of embedded analytics.? Different kinds of assessment systems for submitted assignments

(subjective). This includes:?Peer Grading.

?Self Grading.?Staff Grading.?Machine Grading.?Email and Notification facilities for registered students?Provision of certification.?Registration/Dropping of a course.

In addition to the above mentioned features already offered by edx, we have

added the following new features: ? Upload videos directly on youtube.

? Search by keyword.? Course Ratings.? Recommended courses.? Number of students enrolled in a course.

Grading Model

File with essays and scores

ModelCreator.read_in_data

Create Model

Returns two lists: List of scores and corresponding essays

Contribution:The team successfully integrated the edX-platform with other necessary modules for comments, forums and grading. We have created ER Diagram for edX database which could be used by others for future developments and enhancements. We have performed experiments for testing django cache management and ease (AI) grading.

Ease repository allows anyone to use machine learning based automated classification which can work on both free text (essays, content, etc), and on numeric values.

Dumps data to a json file

Selects algorithm

Creates an essay set object and add

essays to it

Extract features

Get classifier

Get cv error

Essay and score lists

Returns results = {'errors','success','cv_kappa','cv_mean_absolute_error','feature_ext','classifier','algorithm','score','text',‘prompt'}

*cv=cross validation

Contributors: Udayan Jain Sonam Gupta Gagan KhanijauSuhanth Boddu

Piyush Tiwari Rupal Sharma

Surabhi PrasadZeel Shah

Summer Internship 2013(9th May to 7th July) email: [email protected] des

ign

by: B

hair

av

Fundamental Research Group, Dept of CSE, IIT Bombay

NME-ICTMHRD