|
Title:
|
AN APPROACH TO QUANTIFYING INFORMATION IN TWEETS |
|
Author(s):
|
Ruchishya Ramineni and K.M. George |
|
ISBN:
|
978-989-8533-82-1 |
|
Editors:
|
Pedro Isaías and Hans Weghorn |
|
Year:
|
2018 |
|
Edition:
|
Single |
|
Keywords:
|
Information Measure, Information Diffusion, Topic Modeling |
|
Type:
|
Full Paper |
|
First Page:
|
304 |
|
Last Page:
|
312 |
|
Language:
|
English |
|
Cover:
|
|
|
Full Contents:
|
click to dowload 
|
|
Paper Abstract:
|
A new method is proposed to quantify information presented by tweets. Two underlying assumptions are made to develop the method. They are 1) information is an unsigned quantity (i.e. nonnegative which is consistent with entropy) and 2) publicity or being in the news increases the information magnitude. A method based on topic modeling is proposed for quantifying information of lexical tokens. For analysis purpose, information is associated to a topic/question (for example, the weight of a topic, or an actions impact). We may also view our method as a generalization of count which is nonnegative and is a significant measure. The proposed method is used to quantify information of tweets in different domains. |
|
|
|
|
|
|
