Chance Discovery as Analogy Based Value Sensing

Chance Discovery as Analogy Based Value Sensing

Yukio Ohsawa (The University of Tokyo, Tokyo, Japan), Akinori Abe (ATR Knowledge Science Laboratories, Japan) and Jun Nakamura (The University of Tokyo, Tokyo, Japan)
DOI: 10.4018/978-1-4666-1577-9.ch003
OnDemand PDF Download:
$30.00
List Price: $37.50

Abstract

The authors are finding rising demands for sensing values in existing/new events and items in the real life. Chance discovery, focusing on new events significant for human decision making, can be positioned extensively as an approach to value sensing. This extension enables the innovation of various artificial systems, where human’s talent of analogical thinking comes to be the basic engine. Games for training and activating this talent are introduced, and it is clarified that these games train the an essential talent of human for chance discovery, by discussing the experimental results of these games on the logical framework of analogical abductive reasoning.
Chapter Preview
Top

Cases Using Scenario Maps As Basis Of Chance Discovery And Value Sensing

In projects we conducted with companies, the marketing teams acquired novel awareness of valuable parts of their market they had not taken into consideration so far. For acquiring this awareness, KeyGraph assisted business people by showing a diagram as a map of the market having (1) clusters of items frequently bought as a set, i.e., at the same time together, and (2) items bridging the clusters in (1), which may embrace a latent market coming up in the near future.

From the original algorithm (Ohsawa et al., 1998), we improved and extended KeyGraph in response to opinions of users working in real business. For example, let us show an example where a diagram obtained by KeyGraph assisted textile marketers seeking new hit products (Ohsawa & Usui, 2006). Although they already had popular products, they also desired to develop new markets from a niche product, i.e., a product which may be rare for the time being but may expand the company’s opportunity. For this purpose, they started from data which had been collected in exhibition events, where pieces of textile samples had been arranged on shelves for customers representing apparel companies to pick preferable samples. Previously, the list of picked-up samples had been used as an order card on which to send samples to customers. However, once the marketers came to aim at hit sales, the same list was put into an electronic dataset. In comparison with data on past sales, the exhibition data were expected to include customers’ preferences of products not yet displayed in stores.

Complete Chapter List

Search this Book:
Reset