Start Date

10-12-2017 12:00 AM

Description

Dictionaries have been used to analyze text even before the emergence of social media and the use of dictionaries for sentiment analysis there. While dictionaries have been used to understand the tonality of text, so far it has not been possible to automatically detect if the tonality refers to the present, past, or future. In this research, we develop a dictionary containing time-indicating words in a wordlist (T-wordlist). To test how the dictionary performs, we apply our T-wordlist on different disaster related social media datasets. Subsequently we will validate the wordlist and results by a manual content analysis. So far, in this research-in-progress, we were able to develop a first dictionary and will also provide some initial insight into the performance of our wordlist.

Share

COinS
 
Dec 10th, 12:00 AM

The Development of a Temporal Information Dictionary for Social Media Analytics

Dictionaries have been used to analyze text even before the emergence of social media and the use of dictionaries for sentiment analysis there. While dictionaries have been used to understand the tonality of text, so far it has not been possible to automatically detect if the tonality refers to the present, past, or future. In this research, we develop a dictionary containing time-indicating words in a wordlist (T-wordlist). To test how the dictionary performs, we apply our T-wordlist on different disaster related social media datasets. Subsequently we will validate the wordlist and results by a manual content analysis. So far, in this research-in-progress, we were able to develop a first dictionary and will also provide some initial insight into the performance of our wordlist.