Discussion #8: Orange Homework

Orange (https://orange.biolab.si) is a data mining tool that features a process and analysis workflow using widgets. The “connect the widgets” approach allows users to analyze data without requiring programming skills or in-depth knowledge of statistics. 

Orange workflow from data - corpus - preprocess text - word cloud

The goal of the Tweet Analysis Using Orange sessions will be to familiarize you with the process of gathering and analyzing the content of tweets. In order to make this learning experience focused and applicable to your research, this homework should be completed prior to the beginning of the Institute. 

  1. Priority – think about the criteria that would allow you to collect and examine a set of tweets that interests you. These could be hashtags such as #MeToo (see sample below), or general terms, such as “COVID-19,” or even specific Twitter users (or a combination of all). Choose an applicable time period (preferably focused) for the data you want to analyze, such as 16–22 October 2017 for the first week of #MeToo. Throughout this module, you will learn how to collect data on your own. 
  2. Install Orange (https://orange.biolab.si/download) 3. View the following short YouTube videos on Orange to get an idea of the basic functionality. Sample data tables are available with the Orange installation and are used in these videos. Even though the content is not focused on tweets, it will give you a good introduction to the user interface: 
    1. https://www.youtube.com/watch?v=HXjnDIgGDuI (Welcome to Orange) 
    2. https://www.youtube.com/watch?v=lb-x36xqJ-E (Data Workflows) 
    3. https://www.youtube.com/watch?v=2xS6QjnG714 (Widgets and Channels) 
    4. https://www.youtube.com/watch?v=MHcGdQeYCMg (Loading Your Data) 
    5. https://www.youtube.com/watch?v=V70UwJZWkZ8 (Text Preprocessing) – This video will have you install the Text Add-on that will be used during the institute. 
  3. Please skim “Emotion Recognition on Twitter: Comparative Study and Training a Unison Model.” The Introduction and Emotion Classifications are the two sections of importance here. We will be using the POMS (Profile of Mood States) when we analyze your Twitter data during the institute and the adjectives for the different categories will be helpful. The rest of the article, if you care to go through it, describes how the models that Orange uses were trained. 
  4. Write a short post for our discussion forum (about one paragraph – five sentences) explaining:
    • The hashtag you’ve chosen and why; think about what you’re hoping to learn and why this was the best hashtag for answering your research question
    • The time period you’ve chosen and why; 
    • 2-3 things you’ve learned about Orange through watching the tutorials and installing the program;
    • 1-2 things you’re hoping to learn more about and/or are concerned about moving forward

           Make sure to respond to at least two (2) of your peers.   

Sample of #MeToo tweets from October 2017:

download pdf