1. The Center for Data Science and Public Policy at the University of Chicago
dsapp.uchicago.edu @datascifellows
The Center for Data Science and Public Policy at the University of Chicago
dsapp.uchicago.edu @datascifellows
Data Science for Social Good: How
predictive analytics can help governments
and non-profits
Lauren Haynes
@Lnhaynes
http://dsapp.uchicago.edu
@datascifellows
2. The Center for Data Science and Public Policy at the University of Chicago
dsapp.uchicago.edu @datascifellows
3. The Center for Data Science and Public Policy at the University of Chicago
dsapp.uchicago.edu @datascifellows
4. The Center for Data Science and Public Policy at the University of Chicago
dsapp.uchicago.edu @datascifellows
Data Science in the Social Sector
5. The Center for Data Science and Public Policy at the University of Chicago
dsapp.uchicago.edu @datascifellows
~20% of the inspected homes have
children who will get lead poisoning
within the next few months
6. The Center for Data Science and Public Policy at the University of Chicago
dsapp.uchicago.edu @datascifellows
50% fewer false positives
20% more officers identified correctly
7. The Center for Data Science and Public Policy at the University of Chicago
dsapp.uchicago.edu @datascifellows
2.4 billion people don’t have sanitation
facilities
946 million people defecate in the open.
2.5 million people die each year due to
sanitation-related diseases.
Operate 2.5x more toilets with the same
resources
Serve 45,000 more people
8. The Center for Data Science and Public Policy at the University of Chicago
dsapp.uchicago.edu @datascifellows
Identify children more accurately
and 4 years earlier
9. The Center for Data Science and Public Policy at the University of Chicago
dsapp.uchicago.edu @datascifellows
240,000 main breaks/yr in US
$13 billion in 2010 to repair
Expected $30 billion by 2040
180 breaks/yr in Syracuse, NY
62% of blocks in the top 1% of
predictions were correctly
predicted for last year
10. The Center for Data Science and Public Policy at the University of Chicago
dsapp.uchicago.edu @datascifellows
Before:
~400 Violations
per 1,000
Our Model:
~750 Violations
per 1,000
87%
Improvement
11. The Center for Data Science and Public Policy at the University of Chicago
dsapp.uchicago.edu @datascifellows
27% of incidents were under-dispatched
can get to the hospital faster
2440 yearly
12. The Center for Data Science and Public Policy at the University of Chicago
dsapp.uchicago.edu @datascifellows
11 million people move through 3,100 Jails
$22 Billion in costs
64 % suffer from mental illness,
68% have a substance abuse disorder
44 % suffer from chronic health problems
In the top 200 predictions
104 individuals went to jail in the next year
19 years total jail time
13. The Center for Data Science and Public Policy at the University of Chicago
dsapp.uchicago.edu @datascifellows
• Can I detect who’s going to get lead poisoning early?
• Can I determine which home inspections to prioritize?
• How do I improve the scheduling and assignment of my medics/ambulances/firetrucks?
• Can I route citizen requests more efficiently and effectively?
• Which policies do I modify to improve maternal mortality ?
• How much impact is my after-school program having?
• Can I get data that helps me match employers with employees ?
Data Science Problem Templates
X
X
X
X
X
X
X
14. The Center for Data Science and Public Policy at the University of Chicago
dsapp.uchicago.edu @datascifellows
“We are used to using data to justify
funding decisions. Now we can use
data to improve what we do.”
14
Center for Data Science and Public Policy
University of Chicago
dsapp.uchicago.edu @datascifellows
15. The Center for Data Science and Public Policy at the University of Chicago
dsapp.uchicago.edu @datascifellows
Data are
People
16. The Center for Data Science and Public Policy at the University of Chicago
dsapp.uchicago.edu @datascifellows
XKCD Knows a thing or two
16
Center for Data Science and Public Policy
University of Chicago
dsapp.uchicago.edu @datascifellows
17. The Center for Data Science and Public Policy at the University of Chicago
dsapp.uchicago.edu @datascifellows
Challenges
• Project Scoping
• Data Use Agreements
• Technical Capacity
• Resource Availability
• Funding
• Measuring Intervention Effectiveness
18. The Center for Data Science and Public Policy at the University of Chicago
dsapp.uchicago.edu @datascifellows18
Center for Data Science and Public Policy
University of Chicago
dsapp.uchicago.edu @datascifellows
19. The Center for Data Science and Public Policy at the University of Chicago
dsapp.uchicago.edu @datascifellows19
Center for Data Science and Public Policy
University of Chicago
dsapp.uchicago.edu @datascifellows
Project Scoping Guide
1. Goals: Define the goal(s) of the project
2. Actions: What actions/interventions will you inform?
3. Data: What data do you have internally? What data do you need? What can you augment from
external and public sources?
4. Analysis: What analysis needs to be done?
How will it be validated?
20. The Center for Data Science and Public Policy at the University of Chicago
dsapp.uchicago.edu @datascifellows
21. The Center for Data Science and Public Policy at the University of Chicago
dsapp.uchicago.edu @datascifellows
Relevance and Sufficiency
GOAL
Relevant but Insufficient
Relevant and Sufficient
Irrelevant and Insufficient
21
Center for Data Science and Public Policy
University of Chicago
dsapp.uchicago.edu @datascifellows
22. The Center for Data Science and Public Policy at the University of Chicago
dsapp.uchicago.edu @datascifellows
Lessons Learned
• Shift from “What can we do with the data we have?” to “What
problems are you trying to solve today and how can your data help
you?”
• Don’t shame anyone for the state of their data
• Often really simple methods and analysis will add value
• Open Source tools are often as good if not better than the
expensively licensed tools
• The social sector has a lot of really crappy UIs and limited software
(and often the organizations don’t own their own data)
23. The Center for Data Science and Public Policy at the University of Chicago
dsapp.uchicago.edu @datascifellows
Unicorns
• You don’t need to be a data science Unicorn to add
value in the social sector
• Train your people, hire people, or “rent” people
(temps/consultants) (or for the people here,
volunteer!)
24. The Center for Data Science and Public Policy at the University of Chicago
dsapp.uchicago.edu @datascifellows
Questions?
Lnhaynes@uchicago.edu
@Lnhaynes