Predictive Data Transformation Suggestions in Grafterizer Using Machine Learning

Abstract

Data preprocessing is a crucial step in data analysis. A substantial amount of time is spent on data transformation tasks such as data formatting, modification, extraction, and enrichment, typically making it more convenient for users to work with systems that can recommend most relevant transformations for a given dataset. In this paper, we propose an approach for generating relevant data transformation suggestions for tabular data preprocessing using machine learning (specifically, the Random Forest algorithm). The approach is implemented for Grafterizer, a Web-based framework for tabular data cleaning and transformation, and evaluated through a usability study.