documentation

DOCUMENTATION GOAL: To build a learning platform for the students to get a better understanding of the subject and hence make the video lectures more effective. OBJECTIVE: To build a Design Course Ontology and to facilitate query in such a way that if a term is searched, all the logically related terms (deriving a path for the user to learn) along with relevant video links are displayed. The video links must guide the user to different places where that particular term has been discussed. APPROACH: 1. Building the Ontology • The Ontology 101 paper answers all the basic questions on how to build an ontology. Link: protege.stanford.edu/publications/ontology_development/ontology101.pdf

• An open source software, Protége 5.2 is used to build the ontology. Basic Tutorials can be found in YouTube as a series of 10 videos. All the basic features of creation of Ontology in Protege are well explained. You can found these at the attached link below: https://www.youtube.com/watch?v=R9ERIUgvgwM&list=PLea0WJq13cnAfCC0az rCyquCN_tPelJN1

The wide range of terms were classified into four main classes namely, Theories, Concepts, Methods and Techniques

. Then they were further classified into various sub-classes depending on the course to which they belong.

• Object Properties were chosen having in mind the requirements of the user and user-friendliness of the learning platform.

• Then the terms were connected to each other with the various object properties . 2. SPARQL Query • SPARQL query is the one which is used in Protege to traverse through the ontology. • The query for this objective was written to print the first circle of elements that are linked to the term that is searched for . • The results were presented in a readable table form .

3. Video Analytics • Obtained text of the speech as .srt file and classification of keywords and their respective timestamps from VoiceBase website. • Coded to combine all the video clips of particular concept from all the videos by providing time stamps of the particular video clip. (Default player – VLC) • Subtitles are made permanent to all the videos, so that even the modified video plays with subtitles. • So the playlist now can be viewed for the modified video and can skip to any particular classification (definition, example, …etc.) • The modified video is given as link and can be downloaded from the website. NOTE: The modified video and the original videos should be in the same folder.

4. Building the User Interface • An external online application is required to run the SPARQL query and access the ontology from Protege. Out of the many options, the Apache Jena Fuseki Server is the most popular and commonly used SPARQL endpoint. • The interface was built on a html, php and MySQL coding platform using the XAMPP server for Windows 7. • A data base(called “links”) containing all the terms and corresponding video links was created in the phpmyadmin of XAMPP server. • The SPARQL Query was parsed into the php code using ARC2 SPARQL Parser. • The integration of all the above points was done on the XAMPP server using php and MySQL.

Solution Architecture

Speech to text conversion , keywords with timestamps - Using VOICE BASE

Commands used to merge all the video clips into a single video.

Display of playlist on the screen which enables user to go to certain topic by clicking on it .

Work Done So Far 1. Front End

2. Study Help => Connecting to the Ontology

3. Displaying SPARQL Query results on the Webpage

4. Link should be clicked to access the videos :

The video will be downloaded on clicking the word !

WORK TO BE DONE IN THE NEXT MONTH:

REFINING THE ONTOLOGY: - Removing unwanted connections. - Making the ontology more user-friendly. CREATING DIGITAL STORIES WITH VIDEO CLIPS FOR THE REMAINING KEYWORDS AUTOMATION OF VIDEO ANNOTATION INTEGRATION WITH THE COURSE WEBSITE ADDITIONAL FEATURES FOR NAVIGATION