Measuring Intangible Assets Using Parametric and Machine Learning Approaches

TL;DR

This study evaluates how intangible assets like innovation and branding influence business performance, using parametric and machine learning methods on survey and Google Reviews data to improve measurement accuracy.

Contribution

It introduces a combined approach using parametric and machine learning techniques to measure and predict the impact of specific intangible assets on business performance.

Findings

Intangible capital significantly impacts business performance.

Google Reviews data effectively predict branding variables.

Machine learning models outperform traditional methods in prediction accuracy.

Abstract

Intangible capital as the result of digitalization and globalization has not been fully measured yet in the economy because of several challenges. The limitation of data sources and the methodological issue related to how to measure and capitalize intangible assets are some fundamental issues. This paper aims at studying the contribution of intangible capital to business performance. The specific intangible capital, such as innovation, intellectual property, and branding are explored using parametric and machine learning methods. There are two data sources utilized in this study: survey data and Google Reviews data. Some variables are utilized as predictors based on the data sources. The variable selection techniques are implemented, followed by applying parametric regression and machine learning methods to predict business performance based on intangible capital variables. The results show that the proxy of intangible capital used in this paper has a significant contribution to business performance. In addition, variables that are obtained from google reviews can be used to predict the use of branding with high accuracy.