Title: Information discovery methods using statistical processing to facilitate innovation creation

Authors: Takayuki Suzuki; Kiminori Gemba; Atsushi Aoyama

Addresses: Department of Business Administration, Kyoto University of Advanced Science, 18 Gotanda-cho, Yamanouchi, Ukyo-ku Kyoto, 615-8577, Japan ' Business School of Innovation Management, Hosei University, 3-3-9 Kudankita, Chiyoda-ku, Tokyo 102-0073, Japan ' Graduate School of Technology Management, Ritsumeikan University, 2-150 Iwakura, Ibaraki Osaka, 567-8570, Japan

Abstract: In the past, products were classified by hand, using qualitative judgement. This research proposes a method of discovering information that facilitates the creation of innovation using statistical processing, and an external theory targeting information associated with products. Word-of-mouth information concerning home appliances, aggregated during a specific period, was collectively analysed without dividing the total duration of the period. It was found that information lead to the emergence of innovation. This study aimed to discover information that supplements innovation through statistical processing (i.e., principal component analysis, non-negative matrix factorisation, and latent Dirichlet allocation). In the future, it will be crucial to re-examine various discussions that can be expressed in two dimensions or those that independently deal with product characteristics. Going forward, it will be essential to perform quantitative analysis of the relationship of product characteristics with each period of the product life cycle.

Keywords: information discovery methods; statistical processing; innovation creation; product characteristics; product life cycle; aesthetics; quantitative analysis; principal component analysis; latent Dirichlet allocation; non-negative matrix factorisation; NMF; Bass model.

DOI: 10.1504/IJBSR.2024.135778

International Journal of Business and Systems Research, 2024 Vol.18 No.1, pp.30 - 64

Received: 22 Mar 2021
Accepted: 07 Jun 2022
Published online: 05 Jan 2024 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article