TwitAnalysis 1.0.19

Package for analyzing Twitter data

Purpose

Every individual user on Twitter has a customized, personal experience and only views a tiny portion of the actual conversations, content and opinions that are put out on the platform. The purpose of this project is to provide tools to allow people to use Twitter data to obtain a more holistic perspective of the social network landscape. The TwitAnalysis modules allow for the live processing of Tweet streams as well as processing mass amounts of posted content related to certain topics, trends or users. This allows for analysis of a much larger sample size of Twitter data, allowing us to estimate the impact/reach of Twitter content. Basically, this means that we can go beyond just seeing what our friends are thinking/saying on the platform, and see the opinions of Twitter at large.

Scope

The scope of the project is limited by a number of different factors which we will attempt to document to allow for transparency. While not necessarily all inclusive, hopefully this can serve as a foundation for Twitter analysis, and a starting point for more targeted projects in the future.

Functionality

Currently the project is split into two main modules. The TwitLive module is used for streaming/processing live Twitter data. The TwitProcess module is used for processing bulk Twitter data.

from TwitLive import TwitLive

# Make sure to have your keys/tokens defined in your '.config' file. See example file for details
live = TwitLive(a)
live.TrendAnalysis("United States", 3, False)

TODO:

• Calculate Impact (How many people are being reached or impacted by this topic/trend)
• Stream Tweet trend data
• Determine Tweet sentiment
• Calculate Sentiment (General sentiment surrounding topic/trend)
• Test/Verify Sentiment model
• Stream based on location
• Process search data

---

Documentation

https://developer.twitter.com/en/docs/tutorials/building-high-quality-filters

Research

https://www.sciencedirect.com/science/article/pii/S0268401218306005 https://epjdatascience.springeropen.com/articles/10.1140/epjds/s13688-018-0178-0