Leveraging Social Network For Hate Speech Detection And Offensive Language

  • Mehdi Ebady Manaa, Laith Abbas Abdallah

Abstract

As online content keeps improving, hate speech is also spreading. We understand and take a look at the ways in which we have been researched through online programmed methods to reject the speech's content site. Among these challenges are nuances of language, various definitions of what includes obnoxious rhetoric, and states of accessing information to prepare and test these frameworks. Besides, several notable methodologies test the pathogenic effects of the problem of decoding capacity - that is, it is very difficult to understand why frames are set to the options that in this paper we investigate the technology to reveal hate speech on social media. We aim to implement supervised classification technologies using a newly released data set described for this purpose. As features, our system uses n-grams (k-shingles) with the Minhash algorithm and Document Reverse Frequency (TFIDF) term. We get results (91-99)% accuracy in identifying posts in three categories. The results show that the main challenge lies in random profanity and hate speech from each other. Our proposed approach outperforms all modern methods with a significant increase in accuracy.

Published
2020-05-01
How to Cite
Mehdi Ebady Manaa, Laith Abbas Abdallah. (2020). Leveraging Social Network For Hate Speech Detection And Offensive Language. International Journal of Control and Automation, 13(03), 239-250. Retrieved from https://sersc.org/journals/index.php/IJCA/article/view/24991