Comparing Boston vs. SiliconValley & NYC w/ Social Media APIs & "R" (NoFood)

Microsoft New England Research and Development Center 1 Memorial Drive, Cambridge, MA


The $2.50 fee (in place to reduce no-shows) will go to charity as described below (50%) and some O’Reilly books (50%) to be given away to the group.  Please note there is an event list with pizza for $5.


The O'Reilly Books have been ordered:

Machine Learning for Hackers

Social Network Analysis for Startups

R Cookbook


Please note if you want pizza to RSVP here at the $5 event.


Boston's Meetup Tech and Business scene has grown stronger; however Silicon Valley and New York City have both higher adoption and scale.  What might Boston do to improve its Tech and Biz community landscape?

Social Network Analysis and Machine Learning are used on data pulled from Meetup's API.  The talk will cover the application-side, analytics using R, and some on pulling data via an API.




This talk is a follow-up to the "Utilizing Meetup to Analyze Boston, Silicon Valley and NYC" presentation I gave back February.   This extension includes additional Social Network and Text Machine Learning analysis towards addressing questions such as "What specific Meetup Groups would be helpful for Boston towards improving its tech and biz scene relative to Silicon Valley and NYC?"

There are two main layers of the presentation:

(1) Regional Comparisons:  Silicon Valley, not surprisingly, has a higher percentage of tech and business groups than either NYC or Boston.  NYC is slightly above the national average; whereas Boston has an on-par tech scene, though lower than average percentage of business groups.  This will be explored in more detail.

(2) Analytics:  Social Network Analysis and Text Machine Learning as a means to describe different communities.  "R" (tnet and RTextTools libraries among others) was used with data originating primarily from the Meetup API)

Please note the talk is low on hardcore Computer Science, i.e. there is not going to be discussion of achieving 99.999% accuracy using 20 different models ensembled together ala Amazon or Netflix. Please refer to Stanford's Online Learning website ( which has a great ML course; and both Coursera and Udacity have upcoming Social Network Analysis courses.  It would be great to have a 2-hour ML event; however this is not the time for it.

The fee, which is to reduce the no-shows/waitlist, will partially be given to charities. One thought is to have the donations given to a local non-profit for whom Meetup could be a benefit (much as it has shown to be in tech and biz communities). Interestingly, this brings up a different set of questions including "What is the adoption rate of Meetup in inner cities?"   As of this time I don't know which organization; however there's time to explore this before the event, and I may be able to go back into the Meetup data and local economic data to see if those sources can be helpful.  If you are interested in helping out in identifying a non-profit organization, then please let me know.