International Research Journal of Engineering and Technology (IRJET)
e-ISSN: 2395-0056
Volume: 09 Issue: 06 | June 2022
p-ISSN: 2395-0072
www.irjet.net
INAPPROPRIATE & ABUSIVE CONTENT CENSORSHIP Rachana P Bennur1, Samruddhi C Shetty2, Anagha P3, Hema G R4, Prof. Shalini K C5 1,2,3,4 B.E
Student, Dept. of Computer Science Engineering, JSS Science & Technology University, Mysore, India 570006 5 Faculty, Dept. of Computer Science Engineering, JSS Science & Technology University, Mysore, India - 570006 ---------------------------------------------------------------------***---------------------------------------------------------------------
Abstract – With the growth and easy access of internet, kids
User’s safe streaming is further ensured by inclusion of censorship over abusive comments, where all the comments of a video undergo sentiment analysis and if the comments turn out to be toxic or inappropriate, they are either blurred or are not displayed to the other users in the comments section.
today have access to content that are not appropriate for their age. Studies have shown that this can impact their mental development negatively. To prevent a kid from watching inappropriate content, while not affecting the eligible users, is important. The current solutions that determine the age of the user based on the login credentials is not efficient in combating the problem. This project is an attempt to provide safe streaming environment by classifying the videos as nsfw or sfw and to actively predicting the age of the user while he tries to watch a nsfw video. If the user is underaged, the video will be blocked. Further, every comment is checked for its toxicity before posting on the platform.
2. LITERATURE SURVEY 2.1 EXTRACTIVE TEXT SUMMARISATION USING DEEP NATURAL LANGUAGE FUZZY PROCESSING The review paper "Extractive Text Summarisation using Deep Natural Language Fuzzy Processing" by Neelima G, Veeramanickam M.R.M, Sergey Gorbachev, and Sandip A. Kale focuses on the summarising of a document into a smaller set of meaningful and usable sentences. It employs fuzzy logic, with the Naïve-Bayes technique being used to identify the most essential sentences in a given document.
Key Words: Age Prediction, Video Classification, Sentiment Analysis, Non-Safe For Work (NSFW), Safe For Work (SFW).
1. INTRODUCTION
2.2 UNSUPERVISED KEYPHRASE EXTRACTION USING MASKED DOCUMENT EMBEDDING
Social media today is one of the best ways to connect to people, learn and engage in communications. With several social media platforms available, streaming videos has become a primary source of entertainment on social media. However, sites across internet have content varying from age-appropriate to age-inappropriate videos. Such ageinappropriate videos can affect the mental development of kids.
Linhan Zhang, Qian Chen, Wen Wang, Chong Deng, Shiliang Zhang, Bing Li, Wei Wang, and Xin Cao's review paper "A Masked Document Embedding Rank Approach for Unsupervised Key-Extraction" focuses on extracting keywords from a huge document of text. The keywords are generated using a BERT-based model. It extracts essential phrases by selectively masking terms in the manuscript depending on their importance and use.
However, it is difficult for the parents to keep a check constantly on the content that children watch. Currently the streaming platforms blocks the NSFW content based on the age that the user has given in the login credentials. However, there is a loophole that the account belonging to an adult can be misused by a kid to watch NSFW content.
2.3 GENDER AND AGE DETECTION USING DEEP LEARNING Prof. Supriya Mandar Khatavkar, Abhinav Banerjee, Hemanka Sarma, and Anurag Sharma's review paper, "Gender and Age Detection Utilizing Deep Learning Techniques," focuses on predicting an individual's age by using Convolution Neural Network algorithms (CNN) to extract face data.
In order to combat this issue, we are building an age prediction module, which predicts the age of the user actively by capturing real time facial images when the user clicks on any video that is labelled as “age-inappropriate”. The content is blocked to the user if their age is determined as underaged (below 12). With the help of pre-trained machine learning modules, the video content uploaded on the streaming platform is labelled as “Safe for Work” and “Not Safe for Work”.
© 2022, IRJET
|
Impact Factor value: 7.529
2.4 SENTIMENT ANALYSIS IN SOCIAL MEDIA The review paper "Sentiment Analysis in Social Media and Its Application: Systematic Literature Review" by Zulfadzli Drus and Haliyana Khalid investigates how to use the high-
|
ISO 9001:2008 Certified Journal
|
Page 2783