828: Are “Citizen Data Scientists” A Myth? — with Keith McCormick

Low Code Tools

Keith McCormick shares insights into the evolution of low code tools like KNIME and IBM SPSS Modeler, highlighting their significance in data science projects. He notes that while IBM's acquisition of SPSS in 2009 shifted the community, KNIME has remained a versatile tool, especially in workshops and data manipulation tasks. Keith explains, "I've been using KNIME for about ten years now," emphasizing its open-source nature and commercial scalability 1. These tools are not just for citizen data scientists but are also valuable for expert data scientists to facilitate collaboration and rapid prototyping 2.

   

Collaboration

Low code and no code tools significantly enhance collaboration between data scientists and subject matter experts. Keith describes how these tools allow for real-time interaction and experimentation, which is less feasible with traditional coding methods like Python 2. He shares an example of a successful project using IBM SPSS Modeler to detect fraud in a Manhattan hospital, demonstrating the practical applications of these tools 2. "I've done insurance fraud projects... in low code, no code tools," he states, illustrating their effectiveness in complex scenarios.

   

Citizen Scientists

The concept of citizen data scientists sparks debate, with Keith McCormick questioning its practicality. He suggests that while subject matter experts can identify trends, the transition to production requires a data science team and proper governance 3. Keith remarks, "I've always been a little bit of a skeptic," about the citizen data scientist movement, emphasizing the need for collaboration between experts and data scientists 3. He also highlights the role of low code tools in bridging the gap between business insights and technical execution 4.