الفهرس 
A model for enhancing cyber bullying detection for arabic language in social media
AbstractOnly 14 pages are availabe for public view
 |  | from | 72 |  |  |
| | | from | 72 | | |
Abstract
The social networking platform in Egypt is rapidly evolving, this rapid growth has made our lives easier and connected friends and families across the globe. As any tool, social networking is a double-edged weapon, and cyberbullying attacks have been proportionally increasing to this evolution.This research proposes an enhanced model to detect cyberbullying on social networks utilizing sentiment analysis for Arabic language. As there is a lack of existing research using Arabic sentiment analysis compared to English sentiment analysis, due to the nature and difficulty of the Arabic language, a deficiency in Arabic dataset used in sentiment analysis can be noticed. In the past few years, Arabic sentiment analysis has been put under a spotlight by the research community. However, most of the current published work focuses on English content, and not contributing much to the Arabic one. In this research we proposed an algorithm to detect cyberbullies on twitter using lexicon-based sentiment analysis specifically, Arabic tweets. The Proposed algorithm takes into consideration two main factors: the user history in terms of their past tweets, and the emoji{u2019}s included in the tweet. Our model starts with constructing our own dictionary for Egyptian dialect and then it includes three main stages which are Twitter Data Collection, Data Preprocessing and finally Data Classification and calculating the score.To confirm the feasibility of our model we ran a set of experiments and tested the model against multiple scenarios