ISIT219-Knowledge and Information Engineering: Assignment 2 Solved

Business Case
YouTube is one of the largest video-sharing websites worldwide, with an estimated monthly viewership of 1 billion and serves as an important source for analyzing online user activity. In this assignment, we are taking YouTube as the main resource. There is a great potential of using YouTube data in a wide range of real-life applications. As a group of knowledge engineers, your team is required to use knowledge creation and representation techniques to analysis available YouTube data, for gaining an in-depth knowledge of user online activity. You will need to decide one topic that is of your interest, and clearly state that in your report. The data structure from YouTube is shown as follows:

Table. 1 Data structure for harvested YouTube content

Columns/Attributes
Description 
Columns/Attributes
Description 
video_id 
ID for a video
channel_title 

 
Name of video channels
category_id 

 
Type of the video
trending_date 
Date of video trending
tags 

 
Tags for the comments/videos
views 

 
How many views of the video
likes      
The accumulated number of likes
dislikes  
The accumulated number of dislikes
comment_count 

 
The accumulated number of comments until the publish_time
description 

 
Comments content
Description of category_id:

 

1                    - Film & Animation  

2                    - Autos & Vehicles 10 - Music 

15 - Pets & Animals 

17   - Sports 

18   - Short Movies 

19   - Travel & Events 

20   - Gaming 

21   - Videoblogging 

22   - People & Blogs 

23   - Comedy 

24   - Entertainment 

25   - News & Politics 

26   - Howto & Style 

27   - Education 

28   - Science & Technology 

29   - Nonprofits & Activism 

30   - Movies 

31   - Anime/Animation 

32   - Action/Adventure 

33   - Classics 

34   - Comedy 

35   - Documentary 

36   - Drama 

37   - Family 

38   - Foreign 

39   - Horror 

40   - Sci-Fi/Fantasy 

41   - Thriller 

42   - Shorts 

43   - Shows 

44   - Trailers 
 

Your tasks:

 

Some related topics include, but not limited to:
the influence analysis from video channels (tips: identify popular video channels and explore
their influence in relation to type of video, likes/dislikes and received comments, etc., over the time span)

sentiment analysis of comments (tips: find out the relationship between “likes” (“dislikes”)
and “description”)

NLG (nature language generator) (tips: find out the relationship between “tags” and “description”)
categorising videos based on comments (tips: find out the relationship between “category_id” and “description”)
prediction of video popularity (tips: find out the relationship between “views” and “description, comment_count, category_id”, etc)
 

You need to choose a YouTube-related topic, and state it explicitly in your report.

 

Apart from the available datasets, it is expected that you collect other necessary information and/or existing case studies from academic resources (such as journal papers and books) to facilitate your research. This will be presented as the knowledge acquisition part in your project.
 

Various knowledge creation techniques can be employed including, but not limited to:
Classification (such as DT or ANN)
Clustering (such as SOM)
Association analysis (such as rule mining)
 

Finally, you need to write a report (maximum 2500 words) to elaborate on the following item:
Knowledge Acquisition or elicitation process
The techniques that you have employed for knowledge creation o You need to justify the choice of techniques 
Explain and justify the possible inconsistencies in the gathered knowledge