Example analysis of tweets using Python text analytic tools
Text analytics with Python Preview of hands-on experience using Python and Python Libraries to analyze sample tweets Goal is not to teach programming, but to illustrate in a hands-on how these processes work Understand where and how they may be useful Potential use cases Roughly what is involved in analyzing
Supervised machine learning …
Can you spot the sea cow
Can you spot the sea cow https://blog.google/topics/machine-learning/could-machine-learning-save-sea-cow/
On to text… Can use similar approach to analyze text, where your expertise identifies the type of categorization OR one where you let the computer do more of the exploring—’unsupervised machine learning’ Haven’t predefined what is right—doing ‘unsupervised learning’ to see what groupings are But we still involve human judgment, as you will see Example analysis of tweets re Haiti and Cholera
Text analytics schematic #Haiti has begun a large-scale vaccination campaign to battle a #cholera haiti has begun a large-scale vaccination campaign to battle a cholera ‘haiti’ ‘ has’ ‘begun’ ‘a’ ‘large-scale’ ‘vaccination’ ‘haiti’ ‘ has’ ‘begun’ ‘a’ ‘large-scale’ ‘vaccination’ ‘haiti’ ‘ has’ ‘begun’ ‘a’ ‘large-scale’ ’vaccination’
Jupyter helps communicate
Clean the data…
Stop words and phrase length
Word clouds illustrate topics **Just the beginning…. Can trend and track over time, combine with other data
Python Notebook Breakout 12/14/2016