Interplay Between Video Recommendations, Categories, and Popularity on YouTube using bigdata

ABSTRACT:

YouTube is one of the most popular video sharing sites, which allows users to upload, view, rate, share and comment on videos. Three of the many YouTube features, namely, tagging of videos into different categories, their popularity ratings, and video recommendations are notable to promoting users experience on this site. Therefore, this paper examines the interplay between these features through three research questions. We find that about 40% of the video recommendations come from categories other than that of the original video, with Entertainment being the most preferred cross-linked category; (ii) popularity measures including the number of views and comments strongly impact video recommendations; and (iii) video categorization has a higher influence on video recommendations compared to their popularity ratings. Taken together, these three findings suggest that users should carefully consider the tagging of their videos into the most suitable category when they upload to boost their videos popularity. This placement is especially critical because YouTube allows only category to be associated with each video.

EXISTING SYSTEM:
Several researches have been undertaken on different aspects of YouTube video features. Among them Views and uploaded are one of the important one to make a decision (rating, topic categories etc.,) about the particular video. These comments are also used to annotate the video object. Uplolads also reflect the user’s behavior and could use to find the troll users. Moreover, by analyzing the sentiment of comments it is possible to find the users positivity or negativity about video. Based on comments, researchers categorize the videos in several category. Furthermore, for improving video retrieval process  
In this section, we closely examine the You- Tube Insight’s views Data from the following three parts: 
1)  Popularity analysis that investigates the lifetime of one video; 
2)  Visitor trace that explains the viewers’ re-visiting behaviour;
3) users’ rate/favourite/comment data that discover the impacts of users’ engagement activities.

DISADVANTAGE:

Accuracy for finding rating is less because of less dataset used for classification
The ultimate goal of YouTube partners’ everyday operation is to increase views 
PROPOSED SYSTEM:

We proposed a method based on basic feature and social feature . Work on YouTube video comments, like, dislikes for showing that user’s perception (like/dislike) are influenced by valuable comments. 
These two methods worked to find the popularity of video using various features so that it could help to retrieve the useful video. 
Although these two proposed approaches showed impressive work for video retrieval process but they used and views. Sometimes which may lead to inaccurate result. On the contrary, we only analyze a large amount of comments instead of others features (like/views etc.) for finding relevant video which might be useful for YouTube users.
And we will find out what are the top 5 categories with the maximum number of videos uploaded. So we can find most number of rated videos while video retrieval process 

ADVANTAGE:

1.Here top 10 output is generated Where-else you cannot see other uploaded videos after these top ten which will be also been most streamed
2. Here output will not be generated based on keyword. Output will be generated based on query

SYSTEM ARCHITECTURE:
 
SYSTEM REQUIREMENTS:

HARDWARE REQUIREMENTS:

System : INTEL I3
Hard Disk            : 500 GB.
Mouse : Logitech.
Ram : 4GB.
Operating system             :          64-bit.

SOFTWARE REQUIREMENTS:

Operating system : Linux.
Coding Language : Java
Database : HDFS
TOOL                    :         Map-reduce                

BIBLIOGRAPHY:
 Wikipedia.org. 2016. Big Data. https://en.wikipedia.org/wiki/Big_data. [Online] February 2016. https://en.wikipedia.org/wiki/Big_data.