Jump to content

 
Photo

Interactive Map Competition


  • Please log in to reply
5 replies to this topic

#1
ccantey

ccantey

    Newbie

  • New Member
  • Pip
  • 4 posts
  • United States

#Sandy created by Chris Cantey, Caroline Rose, and Morgan Jarocki, University of Wisconsin-Madison

 

http://geo-odyssey.com/links/sandy/

 

The results returned are randomly selected, but interesting topics and associated tweets/photographs include: "crane", "battery park", "water".  "Explosion" will usually return video of the ConEd plant explosion (just click submit until a link shows up).  You can also test the tweet geo-callibration by typing in the boroughs.

 

 



#2
Adam Wilbert

Adam Wilbert

    Legendary Contributor

  • Validated Member
  • PipPipPipPipPip
  • 275 posts
  • Gender:Male
  • Location:Bellingham, WA, USA
  • United States

one question I had on this map was if the data was normalized against the overall number of tweets in NYC. In other words, It looks like you show the % of how many tweets in any given hex area, but don't mention how many overall tweets there were there. Any time I see one of the bright blue "100% of the tweets in this area contain this keyword" hexagons, I immediately assume that there was only a single tweet in that area. 100% of 1 tweet is not very interesting. If it were 100% of 1000 tweets, that's much more powerful but I can't tell if this is the case or not.


Adam Wilbert
CartoGaia.com & AdamWilbert.com
Lynda.com author of "Up and Running with ArcGIS"


#3
ccantey

ccantey

    Newbie

  • New Member
  • Pip
  • 4 posts
  • United States

We randomly sample 50,000 tweets (for every query), then take the keyword tweets per hexagon normalized by total tweets per hexagon. I think there is something like 200 hexagons, so we feel like it is a fair sample size.  The database is too big to use all the tweets.  

 

I understand what you are saying, and I quickly changed the denominator to the whole population and I got some funky results back due to the classification scheme provided by D3js.  I think it would be a good idea to look into changing the functionality to represent a total NYC tweet normalization.... Or add a function to re-symbolize using raw point data and/or proportional symbols.


Edited by ccantey, 15 October 2013 - 02:11 PM.


#4
David Medeiros

David Medeiros

    Hall of Fame

  • Validated Member
  • PipPipPipPipPipPipPip
  • 1,085 posts
  • Gender:Male
  • Location:Redwood City CA
  • Interests:Cartography, wood working, wooden boats, fishing, camping, overland travel, exploring.
  • United States

So that means it is more likely that those 100% hexagons are out of a very small number of tweets overall rather than a statisticaly large number of tweets containing the keyword.

 

50,00 by 200 is only 250 tweets per hex if evenly distributed (not likely), so it wouldn't take much to create a number of hexs with very small tweet counts.

 

If you can't increase the sample size it would at least make sense to report the total number of tweets per hex, the map is not very informative otherwise. And the legend should probably read proportion of tweets with keyword, not frequency right?

 

We randomly sample 50,000 tweets (for every query), then take the keyword tweets per hexagon normalized by total tweets per hexagon. I think there is something like 200 hexagons, so we feel like it is a fair sample size.  The database is too big to use all the tweets.  

 

I understand what you are saying, and I quickly changed the denominator to the whole population and I got some funky results back due to the classification scheme provided by D3js.  I think it would be a good idea to look into changing the functionality to represent a total NYC tweet normalization.... Or add a function to re-symbolize using raw point data and/or proportional symbols.


GIS Reference and Instruction Specialist, Stanford Geospatial Center.

 

www.mapbliss.com

 


#5
ccantey

ccantey

    Newbie

  • New Member
  • Pip
  • 4 posts
  • United States

That makes sense... We could leave the percentage onHover and change the symbology to: TweetFreq.png



#6
ccantey

ccantey

    Newbie

  • New Member
  • Pip
  • 4 posts
  • United States

Or just throw in percentage of x tweets in hexagon on the onhover






0 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users

-->