Ottawa County Innovation and Technology Forum 2016


Big Data Panel Discussion

Kevin Desouza, Rod Davenport, and Paul Stephenson, professor and chair of the GVSU department of statistics talk about the value and challenges of big data.



What is Big Data?


Big Data at GVSU

  • Real-world applications (link from the real world back to the classroom)
  • Big Data (real world context and data sets are available)
  • Complex Content
  • Internships and Jobs (students are interested)
  • Knowledgeable Community (educational institutions have a drive)
  • Students at universities have: tech skills, eager and creative minds, discretionary time.

Data Scientist Skills

  • Visualization, Communication, Storytelling
  • Basic statistics and computer programming.
  • Domain knowledge and teamwork
  • Sampling and data storage/retrieval
  • Statistical modeling and machine learning
  • Curious, evaluative (critical thinking), innovative, strategic

Challenges of Big Data

  • Data quality
  • Accessibility of data
  • Re-purposed data
  • Privacy and security
  • Complicated systems
  • Analysts that don’t understand the question or understand the solution
  • Inferential thinking (focus on error bars and intervals)
  • Differentiating “signals from noise”
  • Balancing time constraints

Realizing the Promise of Data and Technologies for Local Governments

Kevin Desouza, associate dean for research at the College of Public Service & Community Solutions and ASU Foundation professor in the School of Public Affairs at Arizona State University, presented at the Ottawa County Innovation and Technology Forum.


  • Complex platforms and governance now requires use of tech, data, mobile.
  • Data and technologies provide situational awareness, transparency, engagement, policy, innovation, and governance.
  • Open data includes many platforms, including crowdsourcing so that the data management and tool development occurs using the data.
  • Issues with open data include: limited tech talent, public/private partnerships, success metrics are not defined, and there is a transparency vs privacy concern.
  • Example: Arizona Budget Analysis Tool (AZBat) – Took for months to build and it was built with 8 undergrads for a reasonable cost vs hiring an outside firm.
  • Big data issues include: local governments lack IT infrastructure and talent to conduct large-scale predictive analytics projects. It also includes data that is “volunteered”. How can we link all of our separate databases? With the access to the data, what does that mean, ethically.  Should we/can we “discriminate” aka predicative policing, knowing that we have the data…
  • Mobile data includes Fitbit like devices that contribute data around health and activities, this data is given up by end users. Real-time data from phones, wearable tech, social networks, etc. is growing rapidly. Issues include byod, regulating apps, encryption, interoperability, video data processing and curation (police cameras).
  • Emerging tech such as autonomous vehicles will cost local governments big bucks. There are big data concerns and challenges.
  • We all make decisions based on data, once we made a decision we often stop processing data. People often have emotions, hopes, and instincts but without data you can’t align resources.

Top 10 Governing Data and Tech for Societal Value

  1. Start with a Goal in Mind – Evidence-based Decision Making. Knowing what the objective and outcome is needed.
  2. Explore Design Options – Designing for the customer vs designing with the customer. Needs are best met with working directly with the customer.
  3. Rapid Prototype Development – Open and frugal innovation.
  4. Manage Scope Creep – Bound the problem and hold.
  5. Build Partnerships – Leverage and connect to resources.
  6. Harness Collective Intelligence – Design civic labs and crowdsourcing platforms. Open and welcoming for people to experiment.
  7. Experiment Constantly – Test interventions, simulate intended and unintended consequences. Bring in end users to do actually test and use, simulate, experiment, try, provide recommendations.
  8. Release in Beta – Iterative and the project is really never done.
  9. Promote “Intrapreneurship” – Develop competencies from WITHIN and promote innovation inside of the organization to promote innovation.
  10. Outputs and Outcomes – Track both for evaluation and communicate the ROIs.