Toward automatic forecasts for diffusion of innovations

The paper presents an automated framework for forecasting the diffusion of innovations. The framework utilizes existing diffusion information from any market areas or similar products introduced to the markets earlier. The existing data, be it little, enormous, or not present at all, defines a corre...

Full description

Saved in:
Bibliographic Details
Published in:Technological forecasting & social change 2006-02, Vol.73 (2), p.182-198
Main Authors: Ilonen, Jarmo, Kamarainen, Joni-Kristian, Puumalainen, Kaisu, Sundqvist, Sanna, Kälviäinen, Heikki
Format: Article
Language:English
Subjects:
Clustering
Diffusion
Forecast
Forecasting
Information Dissemination
Innovation
Internet
Neural networks
Self-organizing map (SOM)
Technological Innovations
Technology Transfer
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:The paper presents an automated framework for forecasting the diffusion of innovations. The framework utilizes existing diffusion information from any market areas or similar products introduced to the markets earlier. The existing data, be it little, enormous, or not present at all, defines a corresponding decision path in the model, and following the path generates a forecast by maximizing the available information. An information-processing technique called a self-organizing map, SOM, was used to generate a map of the economic, technological and social market characteristics that have been found to affect diffusion. This map is used as a basis for finding suitable analogies for predicting the diffusion of an innovation in a specific market. The framework is applied in the context of predicting the diffusion of cellular subscriptions and Internet use worldwide and, separately, in the European Union, including the new member states. In the experiments the model yielded significantly better results than a regression using the Bass model. The method allows analysts to concentrate on more qualitative issues and the system to perform complicated computing operations. Furthermore, the system is self-refining since its accuracy continuously improves when new and more up-to-date information is added to the database. The proposed framework and methods aim to move present theory toward more practical and automatic prediction tools for company analysts and diffusion researchers.
ISSN:0040-1625
1873-5509
DOI:10.1016/j.techfore.2004.11.005