The Virtual University of Pakistan has 6,000 hours of video lectures on Youtube

May 21, 2010, [MD]

I find distance universities very fascinating, and have learnt much from my time visiting the Open University of China and Indira Gandhi National Open University (IGNOU) in India. Distance universities have been in the forefront of researching alternative ways of delivering instruction and enabling learning, with radio and television, online classes etc. They are also uniquely poised to open up their materials, because their courses have been designed for distance/online learning from the outset. The Open University of the UK, the international pioneer of open universities, has been very innovative with its OpenLearn project.

I was extremely excited to find the amount of material published in India by both the Indian Institutes of Technology and IGNOU. IGNOU both uploaded a very large amount of materials to a DSpace repository called eGyanKosh, and also released thousands of hours of video on their YouTube channel. Being able to watch videos about rural planning, Hindi literature, economics, how to be a radio DJ, and many more topics — from an Indian point of view — is a very exciting opportunity.

This morning I had a wonderful conversation with Alastair Creelman, and he mentioned that the Virtual University of Pakistan (VU) had also put thousands of hours of video on YouTube. I immediately went to check, and found their channel with 5992 videos! (I wonder if there will be a cake served when they reach 6000?)

According to Wikipedia, the VU was founded in 2002, and is “the first distance learning university in Pakistan based entirely on modern information and communication technologies.” In that sense, it differs from IGNOU, which was established in 1985 (WP), and was initially based mainly on correspondence (although they also used satellite TV actively).

What was a positive surprise to me, when I began looking at some of the videos, is that they seem to all be in Urdu. Although Urdu is the national language of Pakistan, many of the universities teach largely in English (similar to the situation in India). The fact that these courses are in Urdu will obviously limit the utility of these courses to the international audience, but radically increase their value to the local population. (And let’s face it, it’s not like the English-speaking population is particularly underserved, globally speaking).

It’s very exciting to me, because I actually studied Hindi for a year (Hindi and Urdu are very similar) and spent three months in India. I am very interested in improving my Hindi, and also in learning Urdu. In the lectures I watched, I could mostly follow the dialogue, because the professors spoke very slowly (sometimes painfully so), and frequently mixed in English words, or used slides in English.

The one problem with the YouTube channel is that unlike the IGNOU account, which uses individual accounts for each faculty (which are all subscribed to b the main account), and playlists for individual courses, the VU account lumps every video into one big pool, and names them with cryptic codes like PSY032-Lecture34. It’s almost impossible to browse around, if you don’t know what you are looking for.

Luckily, they actually store the full title of the course in the description, but this does not come up unless you click on a video. I wanted to see if I could do something about this — mainly to scratch my own itch, and I found that there was a great Ruby library for accessing YouTube metadata programmatically. Using this, you can retrieve all the metadata for every video uploaded by a given user (or in a given category, with a different keyword, etc) using just a few lines of code. I downloaded all the metadata, used the descriptions as course titles, and sorted the filenames alphabetically, and generated an index page.

This page makes it much easier to see all courses on offer, find a sorted list of lectures, etc. Now I can really begin to explore! It’s not perfect, some of the lectures seem to be missing, and there is a course called “Click to add description…”, but overall this is very useful, in just about an hour and a half of hacking, having not even played with the YouTube API previously.

Here’s the code that I ended up with — this downloads all the metadata, and spits out the HTML file that I now host.

require 'rubygems'
require 'youtube_g'

client = YouTubeG::Client.new
allvids = Array.new
count = 1

# get metadata for all videos
begin
 while true
 videos = client.videos_by(:user => 'vu', :page => count)
 allvids = allvids + videos.videos
 count = videos.next_page
 $stderr << count
 break unless count
 end
rescue Exception=>e
end

# sort videos into courses based on descriptions
keywords = Hash.new
courses = Hash.new
allvids.each do |vid|
 unless courses[vid.description]
 courses[vid.description] = Array.new
 keywords[vid.description] = vid.keywords
 end
 courses[vid.description] << vid
end

# start generating the index
puts "<h1>Course titles</h1><ul>"
courses.sort.each do |element|
 puts "<li><a href=\##{element[0]}>#{element[0]}</a></li>"
end
puts "</li><p>"

puts "<h1>Individual courses</h1><ul>"
courses.each_pair do |desc,vids|
 uniq_ary=[]
 puts "<a name=#{desc}>"
 puts "<li><b>#{desc}</b> - <i>#{keywords[desc].join(", ")}</i></li><ul>"
 vids.sort! {|a,b| a.title.downcase <=> b.title.downcase}
 vids.each do |vid|
 next if uniq_ary.index(vid.title)
 uniq_ary << vid.title
 puts "<li><a href=#{vid.player_url}>#{vid.title}</a></li>"
 end
 uniq_ary = []
 puts "</ul><p>"
end
puts "</ul><p>"

Here's a small sample of the courses available:

So go check it out!

Stian

Stian Håklev May 21, 2010 Toronto, Canada
comments powered by Disqus