Twitter Sentiment Analysis To Predict Humanitarian Issues During Wars

Speaker: Bahia Halawi, Lebanese American University

Abstract:

Today, twitter a popular social media platform, which offers users the ability to express their ideas and engage with others. As a result, many scientists believe that analyzing social data could be a very reflexive process that indicates many of the user's behaviors and traits and one of these methods is sentiment analysis.

That's why, this project will focus on analyzing tweets to come up with new facts regarding the Syrian refugees in the Middle East migrating to neighboring countries.. Basically, the paper will try to bring interesting analysis & visualizations.

The outcomes will be used to contrast what is being said in the countries witnessing the highest rates of reception for the Syrian refugees. Nonetheless, statistics about the situation of Syrians there along with the numbers will be provided to give a clearer identification about the topic. This can be a very helpful project; especially that sentiment analysis in languages different than English is still embryonic. It should also shed the light over the fact that the quality of services being received at the hosting country affects the choice of other refugees when deciding where to move to.

Data for Good Goal:
-To identify possible techniques or approaches that would assist scientists in inferring more about societies and people from tweets in Arabic.
-Focus on conversations and common phrases that can be directly associated with Syrian refugees and what media provides about them.
-The purpose of this project is humanitarian, aiming to shed light over problems and sufferings of innocent people by listening to tweets and conversations.

Data Available:
To be extracted from twitter:
Possible variables: (to be added into an Arabic dictionary)
-things related to food
-poverty
-weather
-bedsheets
-camps
-school