search

jawilliams3000 / DAT_SF_12

0 stars 0 forks source link

Project Proposal Feedback #2

Open ostegm opened 9 years ago

ostegm commented 9 years ago

@jawilliams3000

Thanks for the proposal. I'll do some inline comments below:

You wrote: I would like to examine how major events involving violence in the US impact the popularity of music genres. Comment: Great - what sort of questions are you trying to answer? Where will your initial exploratory analysis go? I like the idea, but tell me more about what you're looking for and how the user might find this helpful.

You wrote: I plan to get data from Spotify and/or Soundcloud to identify changes in popularity in musical genres around events such as the Boston Marathon bombings, ISIS executions, school shootings, and Ferguson riots,

Comment: This could be more difficult that you think. Have you contacted them about this or is there a dataset available? Another option might be to try to use the twitter API to pull tweets related to music and do something with this data?

Other notes: If you could have your ideal dataset, what would that look like?What indicators would you be looking for to identify those trends around major traumatic events? The reason I ask this, if to help brainstorm other routes to answering your questions.

Let's keep this conversation going here. Tag me in your reply (@ostegm)

ostegm commented 9 years ago

http://www.wired.com/2011/12/hit-potential-equation/

jawilliams3000 commented 9 years ago

@ostegm

I have been having trouble, as you pointed out, getting the data for my original project idea. I have been pursuing another concept and would like to change. The new idea is to look at how weather affects crime rates in the city of Oakland. I would pull the past 5 years of weather data from NOAA and crime data from the Oakland Crimespotter website (http://oakland.crimespotting.org/). From there I would try linear regression analysis to see how much things like rain, temperature effect the violent, property and quality of life crime rates. I'd also be curious to see how well the model I selected predicts crime for 2015.

I'd appreciate any feedback and additional questions you might have on this concept (including other ideas on how to analyze the data). If it's a go, I'll pull the data down and get the more formal proposal done.