Paper Type

Complete

Paper Number

1588

Description

With the explosion of data, analytics and artificial intelligence, information systems research focuses on the use, management and consequences of algorithms. This far, only a handful of papers offer insights into how algorithmic solutions work. To address this gap, we studied the code making up 45 public data science Jupyter notebooks containing algorithmic solutions developed to predict customer churn in a credit card dataset on a data science platform Kaggle.com. We synthesized a process model of an algorithmic solution: preparing the environment, reading in data, cleaning data, exploratory data analysis, pre-processing the dataset, building and training the model, and testing and validating model. Unboxing the algorithm and investigating the process offers a more fine-tuned understanding and language to better conceptualize the use, management and consequences of algorithmic solutions. It also provides a scaffolding for research into the development of algorithmic solutions, highlighting their variability, experimentation and data scientist decisions.

Share

COinS
 
Aug 9th, 12:00 AM

Unboxing the Algorithm: A Process Model of an Algorithmic Solution

With the explosion of data, analytics and artificial intelligence, information systems research focuses on the use, management and consequences of algorithms. This far, only a handful of papers offer insights into how algorithmic solutions work. To address this gap, we studied the code making up 45 public data science Jupyter notebooks containing algorithmic solutions developed to predict customer churn in a credit card dataset on a data science platform Kaggle.com. We synthesized a process model of an algorithmic solution: preparing the environment, reading in data, cleaning data, exploratory data analysis, pre-processing the dataset, building and training the model, and testing and validating model. Unboxing the algorithm and investigating the process offers a more fine-tuned understanding and language to better conceptualize the use, management and consequences of algorithmic solutions. It also provides a scaffolding for research into the development of algorithmic solutions, highlighting their variability, experimentation and data scientist decisions.

When commenting on articles, please be friendly, welcoming, respectful and abide by the AIS eLibrary Discussion Thread Code of Conduct posted here.