Graph-theoretic autofill

TL;DR

This paper introduces two novel graph-theoretic methods for dynamically autofilling missing form input values based on predefined dependencies modeled as directed graphs, enhancing data completion accuracy.

Contribution

It proposes two new graph-based algorithms for real-time autofill of missing form data, improving upon fixed defaults and static methods.

Findings

Effective in handling missing data in web forms

Improves accuracy over default-based methods

Applicable to various dependency structures

Abstract

Imagine a website that asks the user to fill in a web form and -- based on the input values -- derives a relevant figure, for instance an expected salary, a medical diagnosis or the market value of a house. How to deal with missing input values at run-time? Besides using fixed defaults, a more sophisticated approach is to use predefined dependencies (logical or correlational) between different fields to autofill missing values in an iterative way. Directed loopless graphs (in which cycles are allowed) are the ideal mathematical model to formalize these dependencies. We present two new graph-theoretic approaches to filling missing values at run-time.