Monday, March 21, 2011

YouTube Traffic Experiment

YOTUBE TRAFFIC EXPERIMENT



______________



Submitted to
Mrs. Mavic Pineda
De La Salle University-Manila



______________




In Partial Fulfillment
of the Requirements for SOC-MED
Third Term AY 2010-2011




A Report by:
Africa, Maria Andrea Krisitiana M
Bernales, Bon Lester O.
Bool, Jamaica Jane S.
Reboredo, Ramon Luigi H.





March 21, 2011





I. Introduction

YouTube started when Steve Chen, Chad Hurley, and Hawed Karim made a website that would let users upload and share videos in 2005. The domain name ‘YouTube’ was registered on February 15, 2005 and the first video upload was on April 23, 2005. In order to make YouTube famous, they started a contest that promises to give an iPod Nano to any random member each day. YouTube continually became famous and was spread throughout the world.

II. The Experiment

Objectives:

In this experiment, the group monitored sets of most recent Youtube videos on a daily basis in order to achieve these following points:

·         To determine the level of popularity, confidence, and web host reliability of the said Web 2.0 application in terms of being a resource/multimedia facility
·         To produce information that will tell ‘what kind of videos are being uploaded’ and ‘who often uploads videos’ at a specific time of the day
·         To make an analysis about the trending that happened after the experiment by making an assessment about the processed data that will be gathered and produced and by making an actualization about Mike Wesch's lecture last 2008

Methodology:

1.     Each group member was assigned to monitor a set of the most recent 20 videos from seven o'clock to nine o'clock in the evening on a daily basis which was done from March 7 until March 11 of the year 2011. This method ensured the consistency and integrity of videos evaluation since an individual used a particular method in accomplishing the task.

2.     The group used a special site known as mibazaar.com - ( http://www.mibazaar.com/youtubevision/) to determine the most recently posted Youtube videos which made it easier for the group to determine what videos are to be evaluated.

3.     Every member watched each video to determine the language and the content of the post. The length of the each video was also recorded in order to determine the average time of the videos being posted online. Moreover, the channel of the uploader was also tracked to determine his/her age and origin country. Tag evaluation from www-958.ibm.com - (http://www-958.ibm.com/software/data/cognos/manyeyes/visualizations/89ade5ae18be22b20118beb016cb0109/comments/8d3e022cad9c11ddbb7a000255111976) was also done to determine the scale of popularity of the category at the specific time.

4.     After the entire experiment, the group tallied and produced graphs and tables about the corresponding data in terms of category, country uploader, language used, and tag cloud scale.

5.     Lastly, an analysis, conclusion, and assessment were made to answer the objectives of this study.

Scope and Limitations:

Since the group used mibazaar.com to track the most recently posted Youtube videos in order to focus on the first 20 videos, the group was not able to determine if there were a number of videos blocked during the time of monitoring. Moreover, the age of the uploaders was set aside since it was not accurate. Some uploaders have not specified their real ages and the group cannot actually determine the integrity of data. The group only referred on the thumbnail picture of the users but some of them only were actually animated pictures. This is why the group decided not to get and assess these data. In addition, tags used for all the videos have not been recorded since it is inconsistent and it contains different language characters that the group cannot understand.

III. Results and Interpretation

            a. Uploads by Country

Here are the top per country uploaders from the 35 countries found:

                   We can see the diversity of the sources of videos being posted having the United States as the top contributor with a percentage of 25% and next to it is the UK with 15% on their share of the top contributors. Asian nations like the Philippines and Thailand also made their way up in the rankings with 15% and 13% respectively. The data clearly shows that the uploading of videos are widespread and is not only concentrated in the areas of economic abundance.

            b. Uploads by Language

For the 19 Languages found excluding the audio and non-audio type, these are the main types of languages found:

In here, we can see that English is the most common language used since United States and United Kingdom was the top uploaders. Audio only means that the video may be a music video, singing video or a performing group, band, choir or orchestra. This result connects with the uploads by category since the music category is the second most popular. In the no audio, most of these uploads are home videos, so this is again supported by the uploads by category since people&blogs is the top category found.

 c. Uploads by Category


           

Tag cloud based on: http://www-958.ibm.com/software/data/cognos/manyeyes/visualizations/89ade5ae18be22b20118beb016cb0109/comments/8d3e022cad9c11ddbb7a000255111976

Based from the table, we have seen 16 video categories in youtube. The most occurred category is the People&Blogs, which appeared 22 times. This result can be supported by the tag cloud since it has also a high scale of 30. However, the other results did not match with the tag cloud scale by Mike Wesch. For example, the second most occurring category is Music, but it’s tag cloud scale is 0. Also, some of the least occurring categories such as Autos&Vehicles and Film&Animation, have a high tag cloud scale.

We believe that the mismatch of our results might be caused by the date the tag cloud was created. Since the results of the tag cloud were based on the youtube experiment conducted last March 17, 2008, the trending videos would have changed by now. In addition, some of our categories are not found in the tag cloud such as Gaming and Nonprofits&Activism.





            d. Length of Videos



Based on this table, the computed average time for the videos is 241 seconds or 4:01 minutes.

            e. Commonly Used Tags


The top tags include video, webcam, dance japan, music, 2010, 01 and the. These tags are heavily connected with what is popular (webcam and video for people & blogs,) what is current (japan for the recent japan tsunami and earthquake) and what is entertaining (dance and music). The trending of top categories is also a factor to the number of occurrences of each tag.

IV. Overall Analysis and Conclusion

There were different kinds of videos that was being posted between 7pm to 9 pm ranging from videos of the entertainment category to videos from the category of Nonprofit & Activism. Our group clearly saw the different categories of videos that were being posted by the hour with videos that entertains people and videos with informative point of views by regular people on the top of the scale, while videos of specific interests like pets & animals and the like are on the bottom part of the scale. Top categories during the time that the experiment was being conducted include People & Blogs, Music, Entertainment, Comedy, and Science & Technology.

People & Blogs, Music, Entertainment, Comedy, and Science & Technology are the categories that clearly stood up in terms of the number of uploads. This usually occurs because of the current interest of the citizens of the world. People & blogs topped the charts, it was made possible because it is in this that category that ordinary people can blurt out their opinions about everything and anything under the sun. Now, regular people can agree to these their kind who voices out their ideas and creates a mutual understanding about the things that directly affect them. Music, entertainment, and comedy also made a big impact on the trending because these categories are the source of joy and laughter of millions of people all over the world. It is through these videos that people can forget about their problems and listen to great music and laugh about simple things. The other category Science & technology, also made it to the top of the charts mainly because of the news regarding new technology, medical advancements, new discoveries, and of course gadgets that will surely help in maintaining or improving the quality of our lives and also educate us about the still undiscovered world of ours.
































Appendix


Per Country
Bulgaria
1
Costa Rica
1
Denmark
1
France
1
Ireland
1
Mexico
1
Nepal
1
Panama
1
Portugal
1
South Africa
1
Spain
1
Tajikistan
1
Turkey
1
Singapore
2
Brazil
2
Hongkong
2
India
2
Italy
2
Poland
2
Switzerland
2
Vietnam
2
Indonesia
3
Australia
3
Belgium
3
Taiwan
3
Canada
4
Germany
4
Ukraine
4
Netherlands
4
Japan
5
Russia
6
Thailand
6
Philippines
7
United Kingdom
7
United States
12


Per Language
Bulgarian
1
Dutch
1
Filipino
1
Irish
1
Mexican
1
African
1
Portugese
1
Swedish
1
Turkish
1
German
2
Thai
2
Vietnamese
2
French
3
Chinese
4
Japanese
4
Spanish
4
Ukranian
4
Russian
5
No Audio
13
Audio Only
22
English
24









Tags(Monday)
Tags(Tuesday)
Tags(Wednesday)
Tags(Thursday)
Tags(Friday)
webcam
dance
dj
amelie
VTS
video
art
2010
ann
01
12raunaq
list
uwa
het
1
Camera Plus Pro
safn
How
tekenen
belmont
(removed by user)
museum
To
YUI
springluncheon
(removed by youtube)
reykjavik
Do
simply
Parte
Fun
rvk
Riffle
white
El
silly
iceland
Shuffle
The secret law of attraction
Juico
mirrors
LHÍ
Tutorial
Law of attraction
Final
laughter
IAA
Bicycle
Positive thoughts
Arpel
squashed face
RAM
Playing
Negative thouhgts
Arredamenti
bed
LR
Cards
Abundance
Mobile
in bed
exercise
USPCC
Wealth
Letti
couple
pos353503152
USA
Thoughts
Armadi
ugartamendia
2011
thinking
Giapponesi
KSERA
This
naja0344
Falegnameria
Nao
Mickey
Unworn
skater
Moderni
Navigation
Fiddy
Mens
Campanha
Su
Person
Sandwiches
Audemars
Cadastramento
Misura
Localisation
FiddySandwiches
Piguet
Medulla ossea
Striperformance7777
Particle
World
Royal
Breno
Filter
Piece
Oak
Chilli design
1230101920
Mobile
sad
Peace
Offshore
Ameo
earthquake
song
Part
SURVIVOR
Medulla
tsunami
15’ of fame
PLAXIS
2 6165IO.OO.A002CA.01
Aplasia
8.9 earthquake
podiumkunstenaar
PLAXIS 3D
watch
Doe
Ichihara
aartselarr
importing
was
Doe vida em vidd
japanese
muziekacademie
structures
registered
Hackworth Reprographics
Honshu
Optreden
mesh
in
auditions
magnitude
singleproducing
volumes
And
spokesperson
·         關鍵時刻
cultureel centrum
preparing
has
Website
·         汪嘉伟排球教学片
aambeeld
converting
Donington
cute kids
06
WAK
3D faces
BTCC
funny kids
Clip0
week van de amateurkunsten
AutoCAD Civil 3D
2003
Printer
News
in hd
geotechnical
Team
Printing
flash
Cheap
software
Varta
Large format printing
·         Орлёнок1
Trick
analysis
shomi
123yoyofun
IES
the
Finite Element  method
Show me  love
rinorshqiptar00
Puerto
Flame
FEM
Stefan k
·         театр
Del
-Know
delafino
South Africa
Rosario
winxfan222
Guitar
stereo six
Choir
None
Power
sakso
National anthem
Copyright claim(removed)
bboy
antwerp
Choral
Final1
M16
powermoves
luikstraat 7
Lungile Jacobs
00
dancers
break
Your dose of house & happiness
Voices of cape town
TLM
dansarar
music
NORTHERN
International
Iyon
dans
b-boy
SOUL
Acclaimed capetown
Rhone

ess

Local
Rhone-ain



Nkosi sikelel iafrica
information



Volker
Infos



Ball
JT



Idiot
Actus



Dalig
AHE



Ljud
Uczelnia z pasja



Crap
Film promocyjny



Korv
Uczelni



bert
Slidwshow



idol
animoto



Oscar
Societe forever



Osvin
Societe forever living



mroscan
Nicy



disney
Keros-n



japan
Au



classic
Cabaret




bms




Quiquetron




DJ




proyect




80’s




VideoMix