In order to join virtual sessions, you must be registered and logged-in(Were you registered for the in-person meeting in Denver? if yes, just log in.) 
Note: All session times are in Mountain Daylight Time.

Point Pattern Analysis of Sentiment Classified Twitter Data

Authors: Kenneth Camacho*, Michigan State University
Topics: Spatial Analysis & Modeling, United States, Regional Geography
Keywords: twitter, sentiment, opinion, point, pattern, analysis, social, media
Session Type: Guided Poster
Presentation File: No File Uploaded


Varieties of sentiment analysis and point pattern analysis are being applied to social media data to address a broad range of questions, but they are rarely used in tandem. This study outlines a methodology that combines these two approaches to analyze the spatial distribution of sentiment classified opinions from social media data. Twitter postings on natural gas were downloaded and classified using a variety of sentiment analysis methods into positive, negative, and neutral categories. The classifications were then converted into spatial points using the location data associated with the tweets, whereby point pattern analysis techniques were applied to the points to examine the patterns of positive and negative tweet locations with respect to a background rate of neutral tweets across the contiguous Unites States. Considerations are discussed on the accuracy limitations of sentiment analysis and the uncertainty of point patterns when generating points from location data at different scales. With careful implementation, this methodology can open the door to a range of spatiotemporal analyses of social media opinions.

Abstract Information

This abstract is already part of a session. View the session here.

To access contact information login