Analyzing Submissions for a Waste Reduction Competition in Gunma Prefecture

Abstract

Many local governments in Japan are known to hold public idea competitions as a means of raising awareness about the issue of waste reduction. The study outlined in this paper tests whether the public majority opinion can be effectively determined by text-mining all submissions of a competition sponsored by Gunma Prefecture, rather than focusing only on the winning submissions. The analysis focuses on elementary school students who comprised over 80 % of the applicants. Results indicate that nouns such as ʻpaperʼ, ʻclothʼ, ʻPET bottlesʼ, and ʻskin (vegetable peelings)ʼ appeared frequently in the submissions among this age group. This suggests there is a strong awareness about garbage and which of the 3Rs (reduce, reuse, recycling) should be promoted with regard to each different waste group. Furthermore, the mined text revealed that the judging system was likely to award a high score to the minority of submissions that were linguistically more original than others. The methodology used in this study to demonstrate that it is possible to understand and analyze majority public opinion on an issue may offer a way to add value to the conventional ideas behind holding such competitions.