Companion site for: Mühlematter, M. (2023) Education.ai – Exploring the Influence of Artificial Intelligence Adoption on Investor Behavior in the EdTech Sector (Master's Thesis), University of St.Gallen
A dataset containing information about 14’046 businesses operating in the Education Technology (EdTech) market was investigated regarding the adoption of Artificial Intelligence (AI). The ventures contained in the dataset could be assigned to 23 topics, however, none is labeled “AI”. The keywords representing said topics mostly reflect the primary visions and customer segments. The analysis suggests that AI is only a tool to realize these visions, therefore about the operational dimension of the business model. The idea of AI-first EdTech companies does not appear in the self-description of ventures listed on CrunchBase. The logical conclusion of this first analysis is that AI is not the “hot” topic in EdTech, as it is not a topic at all. A binary classification was added to the dataset in the second analytical step. It confirmed that Artificial Intelligence (AI) is not the current “hot” topic in EdTech, as the distribution of AI companies regarding size, total funding, and age appears random. As the data revealed no trend, analyzing investor behavior regarding a trend-based bias linked to AI was made impossible. However, the temporal analysis using Kernel Density Estimates (KDE) on the topics resulting from the first step revealed additional insights linked to the wider field of emerging technologies: In the past Virtual Reality (VR) was a “hot” topic, explicitly referring to a specific technology, in the EdTech market. The association with this “hot” topic influenced investor behavior negatively, the ventures founded during the hype phase with growing interest in VR received less funding on average than both VR business across the entire dataset and the wider EdTech peer group founded within the same timeframe. This finding partly validates the concept of “smart money”, as professional investors do not overinvest based on a trend-driven bias.
This online repository contains the code needed to reconstruct the tables and figures from the paper. The following notebooks contain the code:
| Name | Link |
|---|---|
| Notebook Pre-Processing |  |
| Notebook Processing | |
| Notebook Analysis + Plots | |
The notebooks are saved to GitHub, so you can directly utilize them by cloning the repo.
You are welcome to use, or reuse the code as you wish. If you find this work helpful, please cite the thesis.
@thesis{Muehlematter2023,
author = {Muehlematter, Mario},
title = {Education.ai - Exploring the Influence of Artificial Intelligence Adoption on Investor Behavior in the EdTech Sector},
shorttitle = {Education.ai}
year = {2023},
institution = {University St. Gallen},
type = {Master's Thesis},
}