The Mine Project

Goals

Quantify the influence between two users in a conversation

Method

Using Natural language processing (NLP) techniques to measure the influence within a conversation between two given users. We use existing NLP techniques as well as developing new ones in order to estimate the influence solely based on textual data

We use social media data for the project purpose. We focus on Twitter due to the high activity volume in the platform and the availability of the data. We focus on spoken Arabic language. We do no limit our research to a specific domain (e.g., politics, sports) within the Arabic corpus.

Project Milestones

	Topic modeling - A Natural Language Processing (NLP) algorithmic way to reveal the central topics within a given set of documents. We use this technique to explore the different topics discusses in our corpus and understand the different between them.
	Sentiment analysis (SA) - Systematically identify, extract, quantify, and study whether a written text is positive, negative or neutral. We use SA in order to understand how positive (or negative) a conversation between people is. We also use this tool to understand the polarity of a person compared to the overall polarity in a conversation.
	Emotion detection - Identifying human emotion in a given text. The emotion is usually categorized into a closed set of emotions (e.g., happy, sad, angry). Same as in the sentiment analysis, we use this tool to understand the emotion of a person in a conversation and compare it with the overall emotion of others in the conversation.
	Tweet based sentiment analysis (TBSA) – Building an algorithm in order to predict the sentiment of a user towards a given tweet. This way we can infer the sentiment of a response tweet to the main tweet in which it responded to. The sentiment towards the main tweet helps us understand whether the response person agrees or not with the main claim raised in a previous tweet.
	In-person sentiment analysis – Building a machine learning model to predict whether a response tweet refers to the person that wrote the main tweet. In many cases, a responder does not refer to the subject of the main tweet but rather attacks/speaks in favor of the main tweet's author. We aim to identify cases in which there is an explicit reference to the author and if so to extract the sentiment towards him/her.

Goals

Method

Project Milestones

Resources