Is your data aggregated and exposed like this?

Scary news just came across my feed. Another data exposure problem, with a marketing research firm that compiled a huge data set and did not secure it. Anyone with the link could access it. They don't *think* it was accessed maliciously by anyone.

The full story is here: http://www.bbc.com/news/technology-40331215

The data set included name, address, phone number, political affiliation, and likely views on any number of issues (was being used by GOP in the US). Estimated number of people affected: Nearly 200 million. The data were aggregated from multiple sources (including Reddit threads -- so yes, this is the spooky bit where everything you ever do online is somehow compiled and put together as a profile of you).

With this in mind, are you concerned? Do you think we need a law to govern the collection, storage, and security of data? Or is it too late or unwieldy to do anything in this arena?

Podcast metrics

Heard on NPR this morning ... apparently podcast metrics have been behind the times, but that's about the change! Follow the link to learn more.

https://www.marketplace.org/2017/06/16/business/podcast-data

The data and Internet club?

Found on my fridge tonight. Does my 8 year old know that I'm teaching analytics right now?

Automatically monitoring cities ... analytics alert us to security issues?

http://www.wbur.org/hereandnow/2017/06/06/london-attack-urban-security

The above link goes to a podcast of an NPR story I just heard. Fascinating stuff -- in short, it's an interview with an architect about security in big cities given recent attacks. What's it got to do with analytics? At one point, the architect discussed how they can use machines to monitor an area, and things like unusual sound patterns, movement patterns, etc. can be detected and can set off alerts for people to check. Pretty cool, huh?

Big data to increase graduation rates?

Here is a very good example of predictive analytics in a higher education institution. With the help of an outside consulting firm, Georgia State University used data from its past students to predict the factors that decrease a student’s chances for graduation. It is remarkable how the system worked as the system notified the academic adviser if a student was flagged for one of the 700 factors. The university increased the graduation rates 6 percent in last three years, and students are graduating half a semester earlier that saves a lot $ in tuition. It appears that the project was successful, however, what about privacy? Should educational institutions be able to flag their students for their possible future failure, although it is for a good cause? What if the information gets in the hands of people who are not authorized? We will explore more about privacy and ethics in the following weeks, but feel free to comment what you think.

http://www.npr.org/sections/ed/2016/10/30/499200614/how-one-university-used-big-data-to-boost-graduation-rates

Analytics of a Tweeting President

I found this story today:

http://www.npr.org/2017/02/04/513469456/when-trump-tweets-this-bot-makes-money

Basically, someone decided to see what happens to the stock market when Trump tweets about something.

A company called T3 created a bot -- "Trump and Dump Bot" -- to play the stock market when Trump tweets something negative about publicly traded companies. Read the story. Or read about how the bot works here:  https://www.t-3.com/works/the-trump-and-dump-bot/

Not to worry, it's all in fun, and they're not profiting. Proceeds to go ASPCA -- save the puppies!

Big Data: Biased?

Sam shared two articles with me and suggested you all might be interested. I concur! Thanks, Sam!

He states: I ran across an interesting article that you may want to share in your blog. It essentially explains that a lot of the big data samples are not representative of the truth and carry inaccuracies that may violate discrimination laws. They mention the Fair Credit Reporting Act, Equal Credit Opportunity Act, Americans with Disabilities Act, the Age Discrimination in Employment Act, the Fair Housing Act, the Genetic Information Nondiscrimination Act, the Federal Trade Commission Act, Etc. Some of this big data may have hidden biases within it and it is possible for accidental discrimination to happen. They talk about a smartphone app for reporting potholes in Boston and how the people in the lower income communities did not have a lot of smartphones - which meant that a lot of the pot holes in the lower income neighborhoods were not getting fixed for some time. They mention that these kinds of accidental discrimination slips could get people in trouble in the Big data world - people getting labeled by the data could end up causing all sorts of problems in the future. Thought that it was a thought provoking look at the future of Big Data.

https://www.datanami.com/2016/01/08/beware-of-bias-in-big-data-feds-warn/

There was one other article from the Harvard Business Review that was referenced within this article - it also stressed the importance of using caution when Big Data is linked to location or human culture. Interesting to learn that there are several cognitive biases that can be involved in the interpretation of this Big Data

Here is the other referenced article: https://hbr.org/2013/04/the-hidden-biases-in-big-data

So, what do you all think? 

Note that we'll be discussing ethical issues toward the end of the course, and we'll totally want to think about these articles then!

The Celiac-Banana Connection

Odds are you know at least one person with Celiac Disease who manages it by eating a gluten-free diet. Yesterday I stumbled upon this story about how kids with Celiac used to be put on a (temporary) banana diet. Why? Because someone made a connection between farmers eating a banana-rich diet not having the problem / people not eating bananas having the problem. It's a pretty quick read and super interesting:

http://www.npr.org/sections/thesalt/2017/05/24/529527564/doctors-once-thought-bananas-cured-celiac-disease-it-saved-kids-lives-at-a-cost

tl;dr? Folks felt better after eating the banana diet because they weren't eating wheat (gluten), and so bananas were hailed as the cure. They weren't, of course, but they probably saved some lives.

the data of going viral

Today on my morning drive I heard a story on the radio and made mental note to look it up and share it with all of you. The story? It was about crowdfunding -- kickstarter -- and the consulting industry that has grown up around it. In particular, the host was questioning if people engaged in crowdfunding should be spending money (on PR, flashy videos, etc.) based on promises that these folks can help their campaigns "go viral."

Well, the expert they had on the clip (from the Wharton crowdfunding study - https://crowdfunding.wharton.upenn.edu) said that no one can promise you'll go viral -- he's looked at the data and it just can't be predicted.

Read the story here: https://www.marketplace.org/2017/05/22/business/crowdfunding-has-created-whole-new-consulting-industry

or listen:
 

Your data is worth $10 million!

Yes, now companies are buying and selling datasets as data is of invaluable help to investors when deciding on which stocks to invest. Here is a very brief NPR podcast mentioning how these things work. I believe you will enjoy listening to it!

Algorithms: Friend of Foe of the Average Guy?

Thought some of you, if you have 15 minutes to spare (and it works well to listen while doing something else, the visuals don't add much; I folded laundry while listening/watching), might enjoy this video by Cathy O'Neil, author of Weapons of Math Destruction. She talks about algorithms and how they are used -- and how they hurt people, hurt a democracy.

Right around the 5 minute mark she discusses teacher accountability models, how they have accountability for the teachers, but there is no accountability for the model.

Could Germany's secret weapon be big data in 2014?

Any soccer fans? Back in Turkey, soccer, or the football in there, is a very popular game. I grew up playing soccer all the time and watching the legends of European and Latin American football. After I came to the States it has been difficult for me to catch up with the results as, between you and me, grad school doesn't let you have that kind of luxury, and of course the time difference.

You probably heard how FSU football team made use of the GPS technology to collect data on the players' heart rates, acceleration, and speed around 2013-2014 - good times (more on this). Well, Germany may have used a similar technology back in 2014 during the World Cup. It was a very successful tournament for Germany as they beat 5-time winner Brazil with a shocking score of 7-1. Of course the success of the team depends on hard work mostly, but it has been reported that the German Football Association was in collaboration with a software company called SAP AG in the beginning of the tournament. The company was using the data from previous games and reporting it to the German team so that the team can be best prepared against the rivals (more on this). It is not sure that it was the big data which helped Germany against Brazil or it was Brazil's bad performance but data is everywhere and it is possible to make use of it for decision making. What do you think? Details are here.

Correlate things ... because you can?

It's certainly possible to run -- and find! -- correlations between any number of things, but we all hopefully) know that correlation does not indicate causality, or really that one item has anything to do with the other. Sometimes correlations indicate things that co-occur in a meaningful way (they may not have a causal relationship but rather a common underlying cause), and other times correlations are just random.

Tyler Vigen has a blog -- and now a book -- that identifies all manner of spurious correlations. Y'know, like the correlation between Japanese passenger cars sold in the US and suicides by crashing motor vehicles. The correlation is real! The numbers don't lie! But we need the be careful about interpreting it.

Check it out! Which is your favorite?

http://www.tylervigen.com/spurious-correlations

You can even discover your own. For example, what correlates with sunlight in FL? Find the answers here: http://tylervigen.com/discover?type_select=sunlight&var_select=Florida&exclude_county=on

EME6356 Summer 2017 - Let the Analytics Fun Begin!

Hi Everyone!

Throughout the EME6356 course, I maintain this blog as a space for "extras." What does that mean? This is a place where I can share current events and other relevant things that weren't designed into our course, and you are free to comment and converse here. You would be surprised how often analytics come up in the news, for example. You just don't see it until you start really paying attention.

I'm happy to accept suggestions and submissions for this blog. Just email anything you find. Yasin says he'll join me as a blogger/contributor, too!

I think I'll also use this space to track some of my own analytic data (and maybe you'll encourage me and cheer me along?).

Vanessa