Exploring the Relationship Between Company Website Text and Innovation Investment
Keywords
No Thumbnail Available
Authors
Issue Date
2020-01-03
Language
en
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
We investigate the relationship between text on the website of companies and the investment
of these companies into various kinds of innovation, building on an existing study by the Dutch
national statistics institute CBS. To this end, we carry out three short studies. First, we try
to replicate the results of the original study. Here, we nd some inconsistencies for one of
the experiment's conditions. Second, we apply the methods of the original paper to a new
set of company websites from Sweden, nding that the classi cation accuracy is substantially
lower. Third, we produce a new set of Dutch websites with labels for individual types of
innovation, and measure the classi cation scores of two logistic regression classi ers and a set
of neural classi ers. We conclude that there is a correlation between company webpage text
and investment in product innovation, but not for investment in other types of innovation. We
also nd that the neural network models generally fail to surpass the scores of the regression
classi ers.
Description
Citation
Supervisor
Faculty
Faculteit der Sociale Wetenschappen