Difference between revisions of "ISSS608 2017-18 T1 Assign GONG QIANG"

From Visual Analytics and Applications
Jump to navigation Jump to search
Line 3: Line 3:
  
 
Analysis:
 
Analysis:
 +
First of all,by filtering key word : flu|soar|sneez|fever|cough in Text, then sort by blog time (Created_at,converted to Date type), I've managed to narrow down all blogs only to those flu-relevant blogs.
  
by filtering key word : flu|sort|cold|fever|cough in Text, then sort by blog time (Created_at,converted to Date type), I can narrow down all blogs only to those flu-relevant blogs.
+
Then load the sorted & filtered data into Tableau,and plot them by day, I can see almost every day people mention flu,fever etc.,but only from May 18, the total amount soared significantly comparing to previous days, after 8am it even hit 123 records which is evidently unusual.(See attached image : GongQiang_Image_1)
Then load the sorted & filtered data into Tableau,and plot them by day, I can see almost every day people mention flu,fever etc.,but only from May 18, the total amount rising signification comparing to previous days, after 8am it hit 123 records which is evidently not normal, hence I determine the first outbreak happens in May 18.
+
 
Moreover,by plotting all blogs into the map per latitude,longitude, put a time animation there to check hour by hour on May 18, I can see outbreak burst in Downtown area then move to Eastside
+
As such,I determine the first outbreak happens in May 18.
 +
 
 +
Moreover,by plotting all blogs into the map per latitude,longitude, put a time animation there to check hour by hour on May 18, I can see outbreak burst in Downtown area then moved to Eastside.

Revision as of 16:03, 15 October 2017

The first ourbreak looks to have happend in Downtown area on May 18. The affect area was from Downtown to Eastside as per the hourly trend.

Analysis: First of all,by filtering key word : flu|soar|sneez|fever|cough in Text, then sort by blog time (Created_at,converted to Date type), I've managed to narrow down all blogs only to those flu-relevant blogs.

Then load the sorted & filtered data into Tableau,and plot them by day, I can see almost every day people mention flu,fever etc.,but only from May 18, the total amount soared significantly comparing to previous days, after 8am it even hit 123 records which is evidently unusual.(See attached image : GongQiang_Image_1)

As such,I determine the first outbreak happens in May 18.

Moreover,by plotting all blogs into the map per latitude,longitude, put a time animation there to check hour by hour on May 18, I can see outbreak burst in Downtown area then moved to Eastside.